W9418-PR.qxd
11/19/03 2:19 PM Page iv
Saunders An Imprint of Elsevier
The Curtis Center Independence Square West Philadelphia, Pennsylvania 19106
TEXTBOOK OF NEUROINTENSIVE CARE Copyright © 2004, Elsevier Inc. All rights reserved.
ISBN: 0-7216-9418-7
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. Permissions may be sought directly from Elsevier’s Health Sciences Rights Department in Philadelphia, PA, USA: phone: (+1) 215 238 7869, fax: (+1) 215 238 2239, e-mail:
[email protected]. You may also complete your request on-line via the Elsevier Science homepage (http://www.elsevier.com), by selecting “Customer Support” and then “Obtaining Permissions”.
NOTICE Medicine is an ever-changing field. Standard safety precautions must be followed, but as new research and clinical experience broaden our knowledge, changes in treatment and drug therapy may become necessary or appropriate. Readers are advised to check the most current product information provided by the manufacturer of each drug to be administered to verify the recommended dose, the method and duration of administration, and contraindications. It is the responsibility of the licensed prescriber, relying on experience and knowledge of the patient, to determine dosages and the best treatment for each individual patient. Neither the publisher nor the authors assume any liability for any injury and/or damage to persons or property arising from this publication.
Library of Congress Cataloging-in-Publication Data Textbook of neurointensive care / [edited by] A. Joseph Layon, Andrea Gabrielli, and William A. Friedman.—1st ed. p. ; cm. ISBN 0-7216-9418-7 1. Neurological intensive care. I. Layon, A. Joseph. II. Gabrielli, Andrea. III. Friedman, William A. (William Alan) [DNLM: 1. Central Nervous System Diseases. 2. Intensive Care. 3. Perioperative Care. WL 300 T355 2004] RC350.N49T49 2004 616.8¢0428—dc22 2003066791 Executive Editor: Allan Ross Senior Editor: Natasha Andjelkovic Assistant Editor: Peter McEllhenney Printed in the United States of America Last digit is the print number: 9 8 7
6
5
4
3
2
1
W9418-PR.qxd
11/3/03 2:42 PM Page v
To all who struggle for humanity, peace, and justice To our people and their heroic aspirations To our families To the memory of our indefatigable friend and colleague, Peter Safar Hay hombres que luchan un dia y son buenos. Hay otros que luchan un año y son mejores. Hay quienes luchan muchos años y son muy buenos. Pero hay los que luchan toda la vida: Esos son los imprescindibles Die Schwachen kämpfen nicht. Die Stärkeren Kämpfen vielleicht eine Stunde lang. Die noch stärker sind, kämpfen viele Jahre. Aber die Stärksten kämpfen ihre Leben lang. Diese Sind unentberlich. Bertolt Brecht Kantate zu Lenins Todestag, #7 Gesammelte Werke 9, Ffm, Suhrkamp, 1966, 691
W9418-PR.qxd
11/3/03 2:42 PM Page vii
Contributors
Bizhan Aarabi, MD, FACS, FRCSC Associate Professor of Neurosurgery, Department of Neurosurgery, University of Maryland School of Medicine; Director of Neurotrauma, University of Maryland Medical System Baltimore, MD Shahram Amini, MD Visiting Scholar, University of Florida College of Medicine Gainesville, FL Assistant Professor of Anesthesiology; Chief, Division of Critical Care; Associate Professor for Education; Khatam-al-anbia Hospital School of Medicine Zahedan, IRAN John L. D. Atkinson, MD, FACS Professor of Neurosurgery, Mayo School of Medicine; Consulting Neurosurgeon and Co-Director, Neurosurgery/Neurology Intensive Care Unit St. Mary’s Hospital Rochester, MN Issam A. Awad, MD, MSc, FACS The Ogsbury-Kindt Chair in Neurosurgery, Professor of Neurosurgery, Neurology, and Pathology, University of Colorado Health Sciences Center Denver, CO
Matthew V. Burry, MD Endovascular Fellow, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL Lawrence J. Caruso, MD Associate Professor of Anesthesiology and Critical Care, University of Florida College of Medicine Gainesville, FL Jie Deng, MD Department of Neuroscience, McKnight Brain Institute, Powell Gene Therapy Center, University of Florida College of Medicine Gainesville, FL Howard M. Eisenberg, MD Professor and Chair, Department of Neurosurgery, University of Maryland School of Medicine; Chief of Neurosurgery, University of Maryland Medical System Baltimore, MD Richard G. Fessler, MD, PhD Professor and Chief, Section of Neurosurgery, University of Chicago School of Medicine Chicago, IL
Joella Beard, MD Rehabilitation & Sports Medicine, LLC Anchorage, AK
Kelly D. Foote, MD Assistant Professor, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL
Wilhelm Behringer, MD Department of Emergency Medicine, Vienna General Hospital Vienna, AUSTRIA
Cory M. Franklin, MD Professor of Medicine, Director, Medical Intensive Care Unit, Chicago Medical School Chicago, IL
Corinna Burger, PhD Assistant Professor, Department of Molecular Genetics and Microbiology, University of Florida College of Medicine Gainesville, FL
William A. Friedman, MD Professor and Chair, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL vii
W9418-PR.qxd 11/3/03 2:42 PM Page viii
viii
Contributors
Andrea Gabrielli, MD Associate Professor of Anesthesiology and Surgery, Medical Director, Hyperbaric Center; Medical Director, Respiratory Care, University of Florida College of Medicine Gainesville, FL Robin L. Gilmore, MD Neurology Center of Middle Tennessee Columbia, TN Eric H. Gluck, MD, FCCP, FCCM Associate Chair of Medicine, Finch University of Health Sciences, Chicago Medical School; Chief, Pulmonary and Critical Care Medicine, North Chicago Veterans Affairs Medical Center North Chicago, IL
Ricardo Morales Laramendi, MD Professor of Internal Medicine, Division of Critical Care Medicine, Saturnino Lora Provincial Teaching Hospital, Instituto Superior de Ciencias Medicas de Santiago de Cuba Santiago de Cuba, CUBA A. Joseph Layon, MD, FACP Professor of Anesthesiology, Surgery, and Medicine, University of Florida College of Medicine; Medical Director, Gainesville Fire Rescue Service Gainesville, FL Dean Lin, MD, PhD Resident, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL
Dietrich Gravenstein, MD Associate Professor of Anesthesiology, Jerome H. Modell Professor and Chair, Department of Anesthesiology, University of Florida College of Medicine Gainesville, FL
Emilio B. Lobato, MD Associate Professor, Department of Anesthesiology, University of Florida College of Medicine; Chief, Cardiac Anesthesia, Shands Hospital Gainesville, FL
Nikolaus Gravenstein, MD Professor of Neurosurgery, University of Florida College of Medicine Gainesville, FL
Michael E. Mahla, MD Associate Professor of Anesthesiology and Neurosurgery, University of Florida College of Medicine Gainesville, FL
David M. Greer, MD, MA Assistant Professor, Harvard Medical School; Instructor in Neurology, Massachusetts General Hospital Boston, MA
Ronald J. Mandel, PhD Department of Neuroscience, McKnight Brain Institute, Powell Gene Therapy Center, University of Florida College of Medicine Gainesville, FL
Bernard H. Guiot, MD, FRCSC Attending Neurosurgeon, Colorado Neurological Institute Denver, CO Hugh C. Hemmings, Jr., MD, PhD Professor of Anesthesiology and Pharmacology, Vice Chair for Research in Anesthesiology, Weill Medical College of Cornell University New York, NY Daniel Huddle, DO Associate Professor of Radiology and Neurosurgery, University of Colorado Health Sciences Center Denver, CO
Richard J. Melker, MD, PhD Professor of Anesthesiology, Pediatrics, and Biomedical Engineering, University of Florida College of Medicine Gainesville, FL Ehud Mendel, MD Associate Professor, Department of Neurosurgery, University of Texas, MD Anderson Cancer Center Houston, TX Robert A. Mericle, MD Assistant Professor, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL
Ahamed H. Idris, MD Professor of Emergency Medicine, University of Texas, Southwestern Parkland Medical Center Dallas, TX
Chet Morrison, MD Fellow, Department of Critical Care, University of Maryland Shock Trauma Center, University of Maryland School of Medicine Baltimore, MD
Pascal M. Jabbour, MD Resident in Neurosurgery, University of Colorado Health Sciences Center Denver, CO
Lorenzo F. R. Muñoz, MD Section Chief, Pediatric Neurosurgery, Rush Presbyterian–St. Luke’s Medical Center Chicago, IL
W9418-PR.qxd
11/18/03 5:59 PM Page ix
Contributors
ix
Katrina Murphy, MD, PhD Chief Resident, Department of Neurosurgery, University of Maryland Medical System Baltimore, MD
Albert L. Rhoton, Jr., MD Professor of Neurosurgery, University of Florida College of Medicine Gainesville, FL
Antoˆnio C. M. Mussi, MD Clinical Instructor, Department of Neurology, University of Sao Paulo Sao Paulo, BRAZIL
Steven A. Robicsek, MD, PhD Assistant Professor, Department of Anesthesiology, University of Florida College of Medicine Gainesville, FL
David A. Peace Medical Investigator, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL
Edgardo Rodriguez, PhD Graduate Assistant Research, Department of Neuroscience, McKnight Brain Institute, University of Florida College of Medicine Gainesville, FL
Carmen E. S. Peden, BS Department of Neuroscience, McKnight Brain Institute, Powell Gene Therapy Center, University of Florida College of Medicine Gainesville, FL Emily Piercefield, MD Department of Neuroscience, McKnight Brain Institute, Powell Gene Therapy Center, University of Florida College of Medicine Gainesville, FL David W. Pincus, MD, PhD Assistant Professor, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL Ronald G. Quisling, MD Professor, Department of Radiology; Director, Section of Neuroradiology, University of Florida College of Medicine Gainesville, FL Alejandro A. Rabinstein, MD Assistant Professor of Neurology, University of Miami School of Medicine Miami, FL Kenneth H. Rand, MD Professor of Pathology and Medicine, University of Florida College of Medicine; Director of Clinical Pathology, Director of Clinical Microbiology, Shands Hospital at the University of Florida Gainesville, FL Paul J. Reier, PhD Eminent Scholar, Department of Neuroscience, McKnight Brain Institute, University of Florida College of Medicine Gainesville, FL
Richard J. Rogers, MD, PhD Assistant Professor, Department of Anesthesiology, University of Florida College of Medicine Gainesville, FL Peter Safar, MD (Deceased 2003) Distinguished Service Professor, Safar Center for Resuscitation Research, University of Pittsburgh Pittsburgh, PA Jacob Samuel, MD Assistant Professor of Medicine, Rush Medical College Chicago, IL Andreas Sarrigiannidis, MD, FCCP Fellow, Division of Pulmonary and Critical Care Medicine, Finch University of Health Sciences, Chicago Medical School North Chicago, IL Christoph N. Seubert, MD, PhD Assistant Professor of Anesthesiolgy; Director, Intraoperative Neurophysiologic Monitoring Laboratory, University of Florida College of Medicine Gainesville, FL David H. Shafron, MD Pediatric Neurosurgeon, Phoenix Children’s Hospital Phoenix, AZ James W. Simpkins, PhD Professor and Chair, Department of Pharmacology/ Neuroscience; Director, Institute of Aging & Alzheimers Research University of North Texas Fort Worth, TX Lorna Sohn-Williams, MD Assistant Professor of Neuroradiology, Department of Radiology, University of Florida College of Medicine Gainesville, FL
W9418-PR.qxd
11/3/03 2:42 PM Page x
x
Contributors
Cheri A. Sulek, MD Associate Professor of Anesthesiology, University of Florida College of Medicine; Associate Professor of Anesthesiology, Malcolm Randall Veterans Affairs Medical Center Gainesville, FL Sean Michael Sullivan, PhD Associate Professor, Department of Pharmaceutics, University of Florida College of Pharmacy; Shands Cancer Institute Gainesville, FL Research Director, Let There Be Hope Research Institute Beverley Hills, CA Trent L. Tredway, MD Chief Resident, Department of Neurological Surgery, Rush Presbyterian–St. Luke’s Medical Center Chicago, IL Arthur J. Ulm, MD Resident, Department of Neurological Surgery, University of Florida College of Medicine Gainesville, FL Margaret J. Velardo, PhD Research Assistant Professor, Department of Neuroscience, McKnight Brain Institute, University of Florida College of Medicine Gainesville, FL Jian Wang, PhD Department of Anesthesia, Stanford University School of Medicine Palo Alto, CA
Max Weinmann, MBBS, MD, FRACP, DipMBM Associate Professor, Medical Director, Critical Care, Department of Anesthesiology, Medical College of Virginia Richmond, VA Robin L. Wellington, PhD Visiting Assistant Professor, Department of Neurosurgery, Rush Presbyterian–St. Luke’s Medical Center Chicago, IL Hung Tzu Wen, MD Clinical Instructor, Division of Neurosurgery, University of Sao Paulo, Hosptial des Clinicas Sao Paulo, BRAZIL Eelco F. M. Wijdicks, MD Professor of Neurology, Mayo Medical School; Chair, Division of Critical Care Neurology, Mayo Clinic, Saint Mary’s Hospital Rochester, MN Jack E. Wilberger, Jr., MD Allegheny University Pittsburgh, PA Shao-Hua Yang, MD Department of Pharmacology and Neuroscience, University of North Texas Health Science Center Fort Worth, TX
W9418-PR.qxd
11/3/03 2:42 PM Page xi
Preface
Whether apocryphal or not, it is said that in the near future, hospitals will be composed of three areas: the emergency department, the operating rooms, and the intensive care unit (ICU). The rationale for such a statement is that managed care is driving medicine in the United States toward outpatient care except in the cases of very ill patients, who are admitted into the ICU. Our experience is that the severity of illness of the patients we care for is greater every year. This is as true in the general ICU population as it is in those individuals with neurologic disease. Partly because of this increased severity of illness in patients with neurologic injury, we conceived the project that led to this book. The book before you is unusual in several respects. It is a textbook, rather than a monograph, of neurointensive care. We initiate the book with a solid review of neurophysiology and neuroanatomy, including anatomy as seen through the “eyes” of our radiology colleagues. We remind the reader of the problems that our neurosurgical colleagues expect to see, even in a well-performed procedure. The body of the book then follows, first with general topics and then with specific disease states. Difficult ethical issues, including topics such as access to health care, alterations of a do not resuscitate order in patients going to the operating room, withdrawal and withholding of therapy, physician-assisted suicide, and brain death are embraced and discussed. We finish the book with clinically relevant research issues that are present on the horizon, beckoning us forward with the unfulfilled promises that make up their potential. The use of evidence-based medicine when such data exist, provision of protocols and algorithms, and honesty when our best approximations and biases are the only data available has served as our credo.
As any authors should, we undertook this book with some hesitation. To write a book—any book—means laying open, for the world to see, one’s biases, flaws, and inadequacies. This is especially true when dealing with an area as broad and complex as treatment of the critically ill patient with neurologic injuries. While others might have written a different book, we undertook this project and offer it, with humility, to our colleagues. Although we live in a society that lionizes—at least rhetorically—the individual and individual exploits, work of any quality is of necessity the culmination of a collective effort. This is true in the case of our textbook. Our coauthors are dedicated clinicians and scientists with whom we are honored to be associated. They have worked diligently in the process of creation of this work. The publishers and printers are remarkable people and true professionals who put up with our foibles and ideas of cover art (we lost on that one). To Allan Ross, Executive Editor, Natasha Andjelkovic, Senior Editor, and Peter McEllhenney, Assistant Editor at Elsevier; Jesamyn Angelica; and Nancy Lombardi at PM Gordon Associates, we offer our heartfelt thanks and appreciation. To Poppy Meehan, the hand that guided the entire project, we can only say thank you. While this is a work of many, we are responsible for any errors or other flaws. We hope you find this text useful. Let us know what you think. There should, after all, be a second edition. A. Joseph Layon, MD, FACP Andrea Gabrielli, MD William A. Friedman, MD
xi
W9418-PR.qxd 11/3/03 2:42 PM Page xiii
Foreword
This textbook represents an enormously thorough effort at not only addressing the issue of neurointensive care in the specific treatment mode but also providing the reader with a comprehensive body of knowledge in regard to the anatomy and physiologic function, both normal and abnormal, of the patient suffering from disorders of neurologic systems. Although the editors have authored a number of the chapters themselves, they also have called on other international experts to contribute to this work. The result is a publication that is not only current in its contents, but also contains material presented by experts in the field. Many of the authors have dedicated their lifetime to the study of neurologic systems in health and disease. Others are relatively new to the field and add a breath of enthusiasm for the future. If I were to have one criticism of this publication, it would be that the title is much narrower than the content of the book itself. Thus, in seeing the title, prospective readers may
not appreciate the enormous amount of comprehensive material that is available to them for study and use in treating their patients. As someone who had the distinct pleasure of working with the three editors in their formative years in residency and/or fellowship, I read the prepublication manuscript with enormous pride. There are many who say that the success of educators can best be appreciated by the accomplishments of their students. The supreme compliment is when the students actually surpass their teachers by their deeds and accomplishments. I feel that through the Textbook of Neurointensive Care, Doctors Friedman, Gabrielli, and Layon have, indeed, paid the supreme compliment to Doctor Albert Rhoton, myself, and others who had a significant role in their growth and development as physicians, scientists, and academicians. Jerome H. Modell, MD
xiii
W9418-01.qxd 10/31/03 3:33 PM Page 3
Chapter 1 Basic Neuroanatomy for Neurointensive Care Unit Hung Tzu Wen, MD, Antônio C. M. Mussi, MD, and Albert L. Rhoton, Jr., MD
Introduction The goal of this chapter is to provide the necessary information on the neuroanatomy that enables the neurointensive care unit staff to: (1) perform a concise but precise neurologic examination and be able to establish the location of the event (anatomic diagnosis), (2) understand the major vascular (arterial and venous) territories of the brain and correlate them with the findings of the neurologic examination and the radiologic findings (computed tomography [CT] or magnetic resonance imaging [MRI]), (3) understand the risks and the potential neurologic complications of the most commonly used neurosurgical procedures, and (4) establish the prevention or early detection and treatment of those complications. The adult central nervous system can be divided into eight major components: (1) cerebral hemisphere, (2) basal ganglia, (3) diencephalon, (4) midbrain, (5) pons, (6) medulla, (7) cerebellum, and (8) spinal cord. The cerebral hemispheres plus basal ganglia and the thalamus are collectively called forebrain. In describing the anatomy of the central nervous system of the brain, some confusion can arise when terms such as rostral, caudal, or both, are used instead of anterior or superior, inferior or posterior. The term “rostral” means nose or mouth region and “caudal” means tail. As for spinal cord and the brainstem, ventral means anterior, dorsal means posterior, rostral means superior, and caudal means inferior. However, because of the 110-degree flexure that the human brain undergoes during development, for the cerebrum and diencephalon, rostral means anterior, caudal means
posterior, ventral means inferior, and dorsal means superior (Fig. 1-1).
Cerebrum Lateral Surface: Neural Structures The cerebrum is arbitrarily divided into five lobes: frontal, temporal, parietal, occipital, and the hidden insula.1 On the lateral surface, the central sulcus and the posterior ramus of the sylvian fissure separate the frontal lobe from the parietal and temporal lobes. Posteriorly, the lateral parietotemporal line, which runs from the impression of the parietooccipital sulcus on the lateral surface to the preoccipital notch, separates the occipital lobe from the parietal and temporal lobes. The parietal and the temporal lobes are separated by the posterior ramus of the sylvian fissure and by the temporo-occipital line, which runs from the posterior end of the posterior ramus of the sylvian fissure to the midpoint of the lateral parietotemporal line. The central sulcus starts from the medial surface of the hemisphere and extends on the lateral surface of the hemisphere from medial to lateral, superior to inferior, and posterior to anterior. It ends adjacent to the posterior ramus of the sylvian fissure approximately 2.5 cm behind the anterior ascending ramus of the sylvian fissure.1 As a characteristic of its trajectory, the central sulcus presents a sinuous silhouette, forming a well-defined superior knee with its convexity directed posteriorly “…” and a not constant inferior knee with its convexity directed anteriorly “Ã.” Together they resemble 3
W9418-01.qxd 10/31/03 3:33 PM Page 4
4
Section I
Introduction ral gyri. The inferior temporal gyrus occupies not only the lateral surface of the cerebrum, but it is also the most laterally placed gyrus on the basal surface of the cerebrum (Fig. 1-2A).
Figure 1-1. Midsagittal section displaying the differences between the spatial orientation of the cerebrum and the brainstem.
the shape of an inverted letter “S” that is best identified near the midline. Frontal Lobe The two main sulci are the superior and inferior frontal sulci, which are anteroposteriorly oriented and extend from the precentral sulcus to the frontal pole. These two frontal sulci divide the lateral surface of the frontal lobe into three gyri: the superior, the middle, and the inferior frontal gyri. The inferior frontal gyrus is divided into three parts by the anterior horizontal, the anterior ascending and the posterior rami of the sylvian fissure: pars orbitalis, triangularis, and opercularis. The apex of the pars triangularis is directed inferiorly toward the junction of three rami of the sylvian fissure; this junctional point coincides with the anterior limiting sulcus of the insula in the depth of the sylvian fissure. It also marks the anterior limit of the basal ganglia and the location of the anterior horn of the lateral ventricle. Temporal Lobe The temporal lobe is limited superiorly by the posterior ramus of the sylvian fissure and posteriorly by the temporooccipital and the lateral parietotemporal lines. It presents two main sulci: the superior and the inferior temporal sulci, which divide the lateral surface of the temporal lobe into three gyri, the superior, the middle, and the inferior tempo-
Parietal Lobe The parietal lobe is limited anteriorly by the central sulcus, medially by the interhemispheric fissure, inferolaterally by the sylvian fissure and the temporo-occipital line, and posteriorly by the lateral parietotemporal line. Its two main sulci are the postcentral and the intraparietal sulci. The postcentral sulcus is very similar to the central sulcus, except for the “inverted S” morphology and for its variable continuity. It is the posterior limit of the postcentral gyrus. The intraparietal sulcus starts at the postcentral sulcus and is directed posteriorly and inferiorly toward the occipital pole; its direction is often parallel and 2 to 3 cm lateral to the midline. The intraparietal sulcus divides the lateral surface of the parietal lobe into two parts: the superior parietal lobule, which is the superomedial and the smaller part, and the inferior parietal lobule, which is the inferolateral and the larger part. The inferior parietal lobule is constituted by the supramarginal and the angular gyri. The supramarginal gyrus is the posterior continuation of the superior temporal gyrus that turns around the posterior ascending ramus of the sylvian fissure to become the most posterior operculum of the sylvian fissure on the parietal side. The angular gyrus is the posterior continuation of the middle temporal gyrus and turns superiorly and medially up to the intraparietal sulcus; it is sometimes limited between the two posterior terminations of the superior temporal sulcus, the angular and the anterior occipital rami (Fig. 1-2B). Occipital Lobe The occipital lobe is located behind the lateral parietotemporal line and is composed of a number of irregular convolutions that are divided by a short horizontal sulcus, the lateral occipital sulcus, into the superior and inferior occipital gyri. As important as the knowledge of the superficial anatomy of the cerebrum is, its correlation to the neural structures located in the depth of the cerebrum is equally as important. The precentral gyrus begins at the medial surface of the cerebrum, just above the level of the splenium of the corpus callosum; it then runs from medial to laterally, and from posterior to anteriorly to pass above the body of the lateral ventricle, thalamus, posterior limb of the internal capsule, and the posterior part of the lentiform nucleus to finally reach the sylvian fissure midway between the anterior and the posterior limits of the insula (Fig. 1-2C). The functional map of the lateral surface of the hemisphere is shown in Figure 1-2D. Sylvian Fissure The sylvian fissure is not merely a complex fissure that carries the middle cerebral artery and its branches, and sep-
W9418-01.qxd 10/31/03 3:33 PM Page 5
Chapter 1
A
Basic Neuroanatomy for Neurointensive Care Unit
B
D C Figure 1-2. A, Lateral surface of the right hemisphere: 1 = central sulcus; 2 = superior precentral sulcus; 3 = superior frontal sulcus; 4 = intraparietal sulcus; 5 = postcentral sulcus; 6 = precentral sulcus; 7 = postcentral sulcus; 8 = inferior precentral sulcus; 9 = superior frontal gyrus; 10 = angular gyrus; 11 = supramarginal gyrus; 12 = middle frontal gyrus; 13 = inferior frontal sulcus; 14 = posterior ramus of the sylvian fissure; 15 = pars opercularis; 16 = Heschl’s gyrus; 17 = superior temporal gyrus; 18 = pars triangularis; 19 = pars orbitalis; 20 = middle temporal gyrus; 21 = inferior temporal gyrus. B, Posterolateral view of the right hemisphere: 1 = central sulcus; 2 = intraparietal sulcus; 3 = postcentral gyrus; 4 = superior parietal lobule; 5 = supramarginal gyrus; 6 = angular ramus; 7 = posterior ramus of the sylvian fissure; 8 = angular gyrus; 9 = anterior occipital ramus; 10 = superior temporal gyrus; 11 = superior temporal sulcus; 12 = middle temporal gyrus. C, Anterosuperior view of the precentral gyrus with its functional mapping. D, The functional mapping of the lateral surface of the right hemisphere.
5
W9418-01.qxd 10/31/03 3:33 PM Page 6
6
Section I
Introduction
arates the frontal and the parietal lobes from the temporal lobe. From a neurosurgical viewpoint, the sylvian fissure can be considered as the gateway connecting the surface of the anterior part of the brain to its depth with all the neural and vascular components along the way. The extensive spectrum of the neural and the vascular structures within the reach of the transylvian approach includes insula, basal ganglia, the lateral ventricle, middle cerebral artery, temporal operculum, frontal and parietal opercula, uncus, orbit, anterior cranial fossa, optic nerve, internal carotid artery and branches, lamina terminalis, and interpeduncular fossa. The sylvian fissure starts on the basal surfaces and extends to the lateral surface of the cerebrum. On both surfaces, the sylvian fissure presents a superficial and a deep part.2 The superficial part of the sylvian fissure presents a stem and three rami. The stem extends medially from the uncus, between the frontal and temporal lobes to the lateral end of the sphenoid ridge, where the stem divides itself into anterior horizontal, anterior ascending, and the posterior rami. The deep (cisternal) part is divided into an anterior part, “sphenoidal compartment,” and a posterior part, “operculoinsular compartment,” The sphenoidal compartment arises in the region of the limen insulae, and extends posteriorly to the sphenoid ridge between the basal frontal and the temporal lobes (Fig. 1-3A). The operculoinsular compartment is formed by two narrow clefts, the opercular cleft between the opposing lips of the frontoparietal and the temporal opercula, and the insular cleft, which has a superior limb located between the insula and the frontoparietal opercula, and an inferior limb between the insula and the temporal operculum (Fig. 1-3B). The opercular cleft—the gyri that constitute the frontal and parietal opercula of the sylvian fissure—are (from posterior to anterior): the supramarginal, the postcentral, and the precentral gyri, pars opercularis, triangularis, and orbitalis The gyri that constitute the temporal operculum of the sylvian fissure are (from posterior to anterior): planum temporale, Heschl’s gyrus, and the planum polare.3 Each gyrus of the frontoparietal opercula is closely related to its counterpart on the temporal side; the supramarginal gyrus is in close contact with the planum temporale; the postcentral gyrus to Heschl’s gyrus, and the precentral gyrus, pars opercularis, triangularis, and orbitalis are related to the planum polare. The site on the posterior ramus of the sylvian fissure where the postcentral gyrus meets Heschl’s gyrus is projected in the same coronal plane of the external acoustic meatus. The medial wall of the sylvian fissure is the insula or island of Reil, which can only be seen when the lips of the sylvian fissure are widely separated. The insula has the shape of a pyramid with its apex directed inferiorly, and it connects the temporal lobe to the posterior orbital gyrus via limen insulae. The limen insulae
is composed of fibers of the uncinate fasciculus covered by a thin layer of gray matter. “Limen,” meaning threshold, was introduced to indicate that the limen insulae serves as threshold between the carotid cistern medially and the sylvian fissure laterally.4 The insula is encircled and separated from the opercula by a deep furrow called the circular or limiting sulcus of the insula, which presents the superior, anterior, and inferior parts. From the limen insulae, the sulci and gyri of the insula are directed superiorly in a radial manner. The deepest sulcus, the central sulcus of insula, is a constant sulcus that extends upward and backward across the insula, in the general line of the central sulcus of the cerebrum. It divides the insula into a large anterior part that is divided by several shallow sulci into three to five short gyri, and a posterior part that is formed by anterior and posterior long gyri. From microsurgical and radiologic viewpoints, the insula represents the external covering of a mass constituted by the extreme, external, and internal capsules, claustrum, basal ganglia, and thalamus. The superior, anterior, inferior, and posterior limits of the insula on the lateral projection correspond to superior, anterior, inferior, and posterior limits of the previously mentioned mass (Fig. 1-3C). Lateral Ventricles Wrapping the previously described mass are the lateral ventricles. These are C-shaped cavities located close to the midline, one on each side of the hemisphere. Each ventricle has five components: frontal horn, body, atrium, and occipital and temporal horns.5 The frontal horn is located in front of the foramen of Monro, and presents roof; floor; and anterior, lateral, medial, and posterior walls. The roof is constituted by the transition between the genu and the body of the corpus callosum, the narrow floor by the rostrum of the corpus callosum, the medial and the posterior walls by septum pellucidum and the thalamus, respectively. The majority of the lateral wall of the frontal horn is represented by the head of the caudate nucleus, except for its most anterior part, constituted by the most anterior part of the anterior limb of the internal capsule, and it is in close relation to the anterior limiting sulcus of the insula. The body of the lateral ventricle is located behind the foramen of Monro, and extends to the point where the septum pellucidum disappears and the corpus callosum and fornix meet. It presents roof, floor, and lateral and medial walls. The roof is formed by the body of the corpus callosum, the medial wall by the septum pellucidum above and the body of the fornix below, the lateral wall by the body of the caudate nucleus, and the floor by the thalamus. The caudate nucleus and the thalamus are separated by the striothalamic sulcus, the groove in which the stria terminalis and the thalamostriate vein course.
W9418-01.qxd 11/21/03 2:34 PM Page 7
B
A
C
D
Figure 1-3. A, Basal view: 1 = lateral orbital gyrus; 2 = anterior orbital gyrus; 3 = medial orbital gyrus; 4 = rectus gyrus; 5 = olfactory tract; 6 = posterior orbital gyrus; 7 = temporal pole; 8 = lateral olfactory striae; 9 = medial olfactory striae; 10 = optic chiasm and infundibulum; 11 = anterior perforated substance; 12 = tuber cinereum and mamillary bodies; 13 = uncus; 14 = limen insulae; 15 = amygdala; 16 = rhinal sulcus; 17 = head of hippocampus; 18 = posterior perforated substance; 19 = crus cerebri; 20 = inferior temporal gyrus; 21 = lateral mesencephalic sulcus; 22 = medial geniculate body; 23 = tegmentum of the midbrain; 24 = parahippocampal gyrus; 25 = collateral sulcus; 26 = pineal and the splenium of the corpus callosum; 27 = fusiform gyrus; 28 = occipitotemporal gyrus; 29 = atrium; 30 = basal parietotemporal line. B, Frontal view: 1 = corpus callosum; 2 = body of the caudate nucleus; 3 = septum pellucidum; 4 = superior limiting sulcus of insula; 5 = frontal operculum; 6 = internal capsule; 7 = thalamostriate vein; 8 = putamen; 9 = superior cleft of the insular compartment; 10 = opercular compartment; 11 = third ventricle; 12 = globus pallidus; 13 = amygdala; 14 = inferior cleft of the insular compartment (inferior limiting sulcus of insula); 15 = head of the hippocampus. C, Lateral view of the right insula and lateral ventricle: 1 = cingulate gyrus; 2 = corpus callosum; 3 = septum pellucidum; 4 = bulb of the callosum; 5 = superior limiting sulcus of the insula; 6 = calcar avis; 7 = glomus; 8 = last short gyrus of insula and the central sulcus of insula. D, Superior view: 1 = frontal lobe; 2 = genu of the corpus callosum; 3 = frontal horn; 4 = septum pellucidum; 5 = head of the caudate nucleus; 6 = anterior limb of the internal capsule; 7 = body of the caudate nucleus; 8 = foramen of Monro; 9 = genu of the internal capsule; 10 = lentiform nucleus; 11 = striothalamic sulcus; 12 = posterior limb of the internal capsule; 13 = thalamus; 14 = internal cerebral vein; 15 = collateral eminence; 16 = glomus; 17 = pineal gland; 18 = collateral trigone; 19 = straight sinus.
W9418-01.qxd 10/31/03 3:33 PM Page 8
8
Section I
Introduction
The atrium and the occipital horn together form a roughly triangular cavity, with the apex pointing posteriorly in the occipital lobe and the base anteriorly on the pulvinar. The atrium has roof; floor; and anterior, medial, and lateral walls. The roof is formed by the body, splenium, and the tapetum of the corpus callosum; the floor by the collateral trigone, a triangular area that bulges upward over the posterior end of the collateral sulcus. The medial wall is formed by two roughly horizontal prominences: the upper prominence is the bulb of the callosum, which is formed by the large bundle of fibers called the forceps major, and it connects the two occipital lobes; the lower prominence is the calcar avis, which overlies the deepest part of the calcarine sulcus; the lateral wall has an anterior part, formed by the caudate nucleus as it wraps the lateral margin of the pulvinar, and a posterior part, formed by the fibers of the tapetum as they sweep anteroinferiorly along the lateral margin of the ventricle. At this part of the lateral ventricle, the tapetum separates the ventricular cavity from the optic radiation; the anterior wall has a medial part composed of the crus of the fornix as it wraps the posterior part of the pulvinar, and a lateral part formed by the pulvinar of the thalamus. The occipital horn extends posteriorly into the occipital lobe from the atrium. It varies in size from being absent to extending far posteriorly in the occipital lobe. Its medial wall is formed by the bulb of the callosum and the calcar avis, the roof and lateral wall are formed by the tapetum, and the floor by the collateral trigone. The temporal horn extends forward and inferiorly from the atrium into the medial part of the temporal lobe, and presents roof; floor; and anterior, lateral, and medial walls. The roof is formed by tapetum, the tail of the caudate nucleus, part of the retrolentiform and sublentiform components of the internal capsule, and the amygdaloid nucleus. The retrolentiform component in the roof of the temporal horn is the posterior thalamic radiation that includes the optic radiation; the part of the sublentiform component in the roof of the temporal horn is formed mainly by the acoustic radiation. The amygdaloid nucleus constitutes the most anterior part of the roof of the temporal horn, and is located just above the head of the hippocampus, anterior to the inferior choroidal point, which is the most anterior site of attachment of the choroid plexus in the temporal horn.6 There is no clear separation between the roof of the temporal horn and the thalamus because all fibers of the optic radiation come from the lateral geniculate body. The lateral wall is formed by the tapetum and the optic radiation, and the anterior wall by the amygdaloid body, the posterior two thirds of the medial wall by the choroidal fissure, and the anterior one third of the medial wall by the head of the hippocampus.6 The floor is formed medially by the hippocampus and laterally by the collateral eminence, which is the prominence overlying the collateral sulcus (Fig. 1-3C and D).
On discussing the lateral ventricle, several related elements can be mentioned: foramen of Monro, internal capsule, corpus callosum, fornix, thalamus, caudate nucleus, hippocampus, temporal amygdala, and choroidal fissure. Foramen of Monro. The foramen of Monro communicates
the lateral ventricle to the third ventricle. It is bounded anteriorly and superiorly by the fornix and posteriorly by the thalamus; the elements that run close to the foramen of Monro are the anterior septal vein superior and medially, choroidal plexus posterior and medially, and the thalamostriate vein lateral and posteriorly (Fig. 1-4A). Internal Capsule. The internal capsule has five parts:7,8 ante-
rior and posterior limbs, genu, retrolentiform, and the sublentiform parts. The anterior limb is located between the head of the caudate nucleus and the lentiform nucleus, it contains frontopontine fibers; the posterior limb is located between the thalamus and the lentiform nucleus, and contains corticospinal tract, frontopontine, corticorubral fibers, and fibers of the superior thalamic radiation (somaesthetic radiation) (Fig. 1-4B). The genu comes directly to the ventricular surface and touches the wall of the lateral ventricle immediately lateral to the foramen of Monro in the interval between the caudate nucleus and the thalamus, where the thalamostriate vein usually drains into the internal cerebral vein; the genu contains corticonuclear fibers and anterior fibers of the superior thalamic radiation. The retrolentiform part is located posteriorly to the lentiform nucleus and contains mainly parietopontine, occipitopontine, occipitocollicular, and occipitotectal fibers and the posterior thalamic radiation that includes the optic radiation. The sublentiform part is located below the lentiform nucleus and contains temporopontine, parietopontine fibers, acoustic radiation from the medial geniculate body to the superior temporal, and transverse temporal gyri. Corpus Callosum. The corpus callosum is the largest transverse commissure connecting the cerebral hemispheres. It contributes to the wall of each of the five parts of the lateral ventricle. The corpus callosum has two anterior parts, the rostrum and genu; a central part, the body; and a posterior part, the splenium. The rostrum is located below and forms the floor of the frontal horn. The genu gives rise to a large fiber tract, the forceps minor, which forms the anterior wall of the frontal horn as it sweeps obliquely forward and laterally to connect the frontal lobes. The genu and the body of the corpus callosum form the roof of both the frontal horn and the body of the lateral ventricle. The splenium gives rise to a large tract, the forceps major, which forms a prominence called bulb in the upper part of the medial wall of the atrium and occipital horn as it sweeps posteriorly to connect the occipital lobes. Another fiber tract, the tapetum, which arises in the posterior part of the body and splenium, sweeps
W9418-01.qxd 11/21/03 2:36 PM Page 9
A
B
C
D
Figure 1-4. A, Anterolateral view of the right foramen of Monro: 1 = anterior septal vein; 2 = choroid plexus; 3 = thalamus; 4 = thalamostriate vein; 5 = fornix. B, Superior view displaying the distribution of the fibers in the internal capsule; the fibers from the corticospinal tract occupy approximately the anterior half of the posterior limb of the internal capsule: 1 = frontal horn; 2 = anterior limiting sulcus of insula; 3 = head of the caudate nucleus; 4 = claustrum; 5 = lentiform nucleus; 6 = insula; 7 = foramen of Monro; 8 = thalamostriate vein; 9 = thalamus; FT = frontothalamic fibers; FP = frontopontine fibers; S = sensory fibers; TP = temporopontine fibers; V&A = visual and auditory fibers. C, Superior view of the right temporal lobe and temporal horn: 1 = planum temporale; 2 = rhinal sulcus; 3 = anterior segment of the uncus; 4 = apex of the uncus; 5 = posterior segment of the uncus; 6 = head of the hippocampus; 7 = collateral eminence; 8 = Heschl’s gyrus; 9 = inferior choroidal point; 10 = parahippocampal gyrus; 11 = body of the corpus callosum; 12 = tail of the hippocampus; 13 = collateral trigone; 14 = planum temporale. D, Superolateral view of the right lateral ventricle: 1 = left head of the caudate nucleus; 2 = splenium of the corpus callosum; 3 = body of fornix; 4 = thalamus; 5 = foramen of Monro; 6 = bulb of the callosum; 7 = crus of fornix; 8 = lateral geniculate body; 9 = lentiform nucleus; 10 = occipital horn; 11 = calcar avis; 12 = fimbria of fornix and tail of the hippocampus; 13 = collateral trigone; 14 = fimbria and the body of the hippocampus; 15 = head of the hippocampus.
W9418-01.qxd 10/31/03 3:33 PM Page 10
10
Section I
Introduction
laterally and inferiorly to form the roof and lateral wall of the atrium and the temporal and occipital horns. The tapetum separates the fibers of the optic radiations from the temporal horn and the atrium. Basal Ganglia. Although macroscopically fused and gathered into a “mass” that is involved laterally by insula, the basal ganglia and the thalamus are embryologically and functionally distinct structures. The basal ganglia are telencephalic structures and the thalamus is a diencephalic structure. The basal ganglia consist of four nuclei: (1) striatum (caudate nucleus, putamen and the nucleus accumbens), (2) globus pallidus, (3) substantia nigra, and (4) subthalamic nucleus. The basal ganglia play a major role in voluntary motor movement; however, they do not have direct input or output with the spinal cord. They receive their primary input from the cerebral cortex and send their output to the brainstem, and, via the thalamus, back to the prefrontal, premotor, and motor cortices. The motor activity of the basal ganglia is therefore mostly mediated by motor areas of the frontal lobe. Disturbance of the basal ganglia is usually characterized by: (1) tremor and other involuntary movements, (2) changes in posture and muscle tone, and (3) poverty and slowness of movement without paralysis. The caudate nucleus is another C-shaped structure that wraps around the thalamus; it has a head, body, and tail. The head and the body are lateral walls of the frontal horn and the body of the lateral ventricle. The tail extends from the atrium into the roof of the temporal horn and is continuous with the amygdaloid nucleus. Thalamus. The thalamus is located in the center of the
lateral ventricle. Each lateral ventricle wraps around the superior, inferior, and posterior surfaces of the thalamus. The anterior tubercle of the thalamus is the posterior limit of the foramen of Monro; the posterior part, called pulvinar (pillow) of the thalamus is the wall of three different compartments in the cerebrum: the posterolateral part of the pulvinar is the lateral half of the anterior wall of the atrium; the posteromedial part is covered by the crus of the fornix, and is part of the superolateral wall of the quadrigeminal cistern; the inferolateral part of the pulvinar is the roof of the wing of the ambient cistern. The medial part of the thalamus is the lateral wall of the third ventricle. The thalamus is not a relay station where information is simply passed on to the neocortex—the thalamus acts as a gatekeeper for information to the cerebral cortex, preventing or enhancing the passage of specific information depending on the behavioral state of the person. The thalamus is composed of more than 50 nuclei, which can be divided into specific or relay and nonspecific or diffusely projecting nuclei. The relay nuclei have a specific relationship with a particular region of the neocortex, and are clas-
sically divided into four groups, depending on their position in relation to the internal medullar lamina. The anterior group receives input from the mamillary bodies and from the subiculum of the hippocampal formation. The medial group receives input from the basal ganglia, amygdala, and midbrain, and has been implicated in memory; its major output is to the frontal cortex. The nuclei from ventral group are named according to their position within the thalamus. The ventral anterior and ventral lateral nuclei are important for motor control and carry information from the basal ganglia and cerebellum to the motor cortex. The ventral posterior lateral conveys somatosensory information to the neocortex. The posterior group includes the medial and lateral geniculate nucles, lateral posterior nucleus, and the pulvinar. The medial geniculate nucleus is a component of the auditory system; the lateral geniculate nucleus receives information from the retina and conveys it to the primary visual cortex; the pulvinar seems to be interconnected with parietal, temporal, and occipital lobes. The nonspecific or diffusely projecting nuclei are either located in the midline (midline nuclei) or within the internal medullary lamina (intralaminar nuclei). The largest intralaminar nucleus is the centromedian nucleus, and it projects to amygdala, hippocampus, and basal ganglia. These nuclei are also thought to mediate cortical arousal. Hippocampus. The hippocampus occupies the medial part of the floor of the temporal horn and is divided into three parts: head, body, and tail. The head of the hippocampus, the anterior and the largest part, is directed anterior and inferiorly, and then medially, and is characterized by three or four hippocampal digitations; its overall shape resembles a feline paw.9 Its posterior limit is characterized by the initial segment of the fimbria and the choroidal fissure. Superiorly, the head of the hippocampus is related to the posteroinferior portion of the amygdala, which bulges from the most anterior part of the roof of the temporal horn into the ventricular cavity. The body of the hippocampus has an anteroposterior and inferosuperior direction in the medial part of the floor of the temporal horn, and narrows as it approaches the atrium of the lateral ventricle. At the atrium of the lateral ventricle, the body of the hippocampus changes direction and has its longitudinal axis oriented transversely to become the tail of the hippocampus. The tail of the hippocampus is even more slender and constitutes the medial part of the floor of the atrium; medially the tail of the hippocampus fuses with the calcar avis. Macroscopically, the tail of the hippocampus ends when it meets the medial wall of the atrium, although histologically the terminal segment of the hippocampal tail continues as the subsplenial gyrus, which covers the inferior splenial surface (Fig. 1-4C and D). Amygdala. The amygdala, along with the hippocampus, constitute the core of the limbic system.10 The temporal amyg-
W9418-01.qxd 10/31/03 3:33 PM Page 11
Chapter 1
dala is composed of a series of gray matter nuclei classified into three main groups: basolateral, corticomedial, and central. From a neurosurgical viewpoint, the temporal amygdala can be considered as being entirely located within the boundaries of the uncus: superiorly, the amygdala blends into the globus pallidus without any clear demarcation; inferiorly, the posterior portion of the temporal amygdala bulges inferiorly into the most anterior portion of the roof of the temporal horn above the hippocampal head and the uncal recess; medially, it is related to the anterior and posterior segments of the uncus; it also constitutes the anterior wall of the temporal horn (Fig. 1-5A). Choroidal Fissure. The choroidal fissure is one of the most
important intraventricular surgical landmarks for neurosurgeons. The choroidal fissure is a cleft located between the thalamus and the fornix. It is the site of attachment of the choroid plexus in the lateral ventricle. It is a C-shaped arc that extends from the foramen of Monro through the body, atrium, and temporal horn of the lateral ventricle.11 The choroidal fissure is divided into three parts: (1) the body part between the body of the fornix and the thalamus, (2) the atrial part between the crus of the fornix and the pulvinar of the thalamus, and (3) the temporal part between the fimbria of the fornix and the stria terminalis of the thalamus. The choroid plexus of the lateral ventricle is attached to the fornix and to the thalamus via an ependymal covering called taenia; in the body and the atrial parts, the taenia fornicis attaches the choroid plexus to the body of the fornix, and the taenia choroidea attaches the choroid plexus to the thalamus. In the temporal part, the choroidal plexus is attached to the fimbria via taenia fimbriae and to the stria terminalis via taenia choroidea. The choroidal fissure is one of the most important landmarks in microneurosurgeries involving the temporal lobe: it separates those structures located laterally that can be removed (temporal structures) from those structures located medially that should be preserved (thalamic structures). Third Ventricle The third ventricle is a narrow, funnel-shaped, unilocular, midline cavity (Fig. 1-5B). It communicates at its anterosuperior margin with each lateral ventricle through the foramen of Monro and posteriorly with the fourth ventricle through the aqueduct of Sylvius. It has a roof; a floor; and an anterior, posterior, and two lateral walls.12,13 The roof extends from the foramen of Monro anteriorly to the suprapineal recess posteriorly and is constituted from superior to inferior by five layers of neural, vascular, and pial structures:14 the first layer is the fornix; the second layer is the superior membrane of the tela coroidea. The third layer is the vascular layer located in a space between the superior and the inferior membranes of the tela choroidea called velum interpositum, which contains the internal cerebral veins and branches of the medial posterior choroidal arter-
Basic Neuroanatomy for Neurointensive Care Unit
11
ies. The fourth layer is the inferior membrane of the tela choroidea, which forms the floor of the velum interpositum. The fifth layer is the choroidal plexus of the third ventricle, usually represented by two parallel strands of choroid plexus projecting downward on each side of the midline (Fig. 15C). The floor of the third ventricle extends from the optic chiasm anteriorly to the orifice of the aqueduct of Sylvius posteriorly, and it is constituted from anterior to posterior by the optic and infundibular recesses, tuber cinereum, mamillary bodies, posterior perforated substance, midbrain, and the aqueduct. The anterior wall is constituted by the lamina terminalis and the posterior wall is represented from inferior to superior by the posterior commissure, pineal recess, habenular commissure, pineal gland, and the suprapineal recess (Fig. 1-5D). The anterior wall and the floor form an acute angle that resembles the shape of the beak of a bird. At the inner angle formed by the roof and the anterior wall is the anterior commissure. Frequently there is another commissure located in the cavity of the third ventricle connecting both thalami called massa intermedia, which is located posterior to the foramen of Monro. The lateral wall of the third ventricle is constituted by thalamus above and by the hypothalamus below; the hypothalamus is separated from the thalamus by hypothalamic sulcus, a shallow groove extending from the foramen of Monro anteriorly to the aqueduct posteriorly. The hypothalamic sulcus is the cephalic continuation of the central canal in the spinal cord and the sulcus limitans in the brainstem. During the development of the central nervous system, the neural tube is divided by the sulcus limitans into two plates: dorsal to the sulcus limitans is the alar plate, and ventral to the sulcus limitans is the basal plate. In the spinal cord and brainstem, the structures evolved from the alar plate bear sensory and coordination functions; the structures evolved from the basal plate bear motor function. However, only the alar plate is evolved in the development of the telencephalon and diencephalon; in the diencephalon, the alar plate is further divided by the hypothalamic sulcus into a ventral and a dorsal part: the dorsal part becomes the thalamus (sensory and coordination) and the ventral part becomes hypothalamus (motor). Even though the neural control of emotion involves several regions, including the amygdala and the limbic association areas of the cerebral cortex, they all work through the hypothalamus to control the autonomic nervous system. The hypothalamus coordinates behavioral response to ensure bodily homeostasis, the constancy of the internal environment, by working through three major systems: the autonomic nervous system, the endocrine system, and an ill-defined neural system concerned with motivation. The third ventricle can be approached from the front, through the lamina terminalis via interhemispheric or pterional approaches; from behind, through the velum interpositum via supracerebellar infratentorial approach; or from
W9418-01.qxd 10/31/03 3:33 PM Page 12
12
Section I
Introduction
A
B
C
D
W9418-01.qxd 10/31/03 3:33 PM Page 13
Chapter 1 Figure 1-5. A, Coronal view: the left thalamus has been removed. 1 = bulb of the callosum; 2 = calcar avis; 3 = choroid plexus; 4 = Heschl’s gyrus; 5 = internal capsule; 6 = thalamus; 7 = tentorial edge and the dentate gyrus; 8 = fimbria of fornix and the tail of the caudate nucleus; 9 = globus pallidus and the anterior commissure; 10 = mamillary bodies (floor of the third ventricle); 11 = crus cerebri and the substantia nigra (midbrain); 12 = limen insulae; 13 = substantia inominata; 14 = amygdala; 15 = head of the hippocampus; 16 = optic chiasm. B, Left anterolateral view of the third ventricle: 1 = corpus callosum; 2 = thalamus; 3 = septum pellucidum; 4 = posterior commissure; 5 = thalamus; 6 = medial and lateral geniculate bodies; 7 = tail of the caudate nucleus; 8 = limen insulae; 9 = pons; 10 = optic nerve; 11 = internal carotid artery; 12 = uncus; 13 = left orbit; 3V = third ventricle; III = oculomotor nerve. C, Coronal section through the body of the lateral ventricle and the third ventricle, to display the five components of the roof of the third ventricle. D, Superior view of the third ventricle: the fornix has been reflected posteriorly. 1 = columns of fornix (cut); 2 = head of the right caudate nucleus; 3 = foramen of Monro; 4 = thalamostriate vein; 5 = choroid plexus; 6 = massa intermedia; 7 = right internal cerebral vein; 8 = superior choroidal vein; 9 = thalamus; 10 = midbrain; 11 = posterior commissure; 12 = pineal; 13 = fornix.
䉳
above, through its roof as in transcallosal interforniceal15,16 and transcallosal transchoroidal approaches.14 Lateral Surface: Venous Relationships The superficial venous system drains the superficial one fifth of the thickness of the cerebrum, while the deep venous system drains the remaining four fifths.17 On the lateral surface of the cerebrum, the superficial venous drainage system is accomplished via venous channels adjacent to the lobes. In the frontal and parietal lobes, the venous drainage can direct superiorly toward the superior sagittal sinus or inferiorly toward the superficial sylvian vein; in the temporal lobe, venous drainage can be superiorly toward the superficial sylvian vein or inferiorly toward the dural sinuses below the temporal lobe. The lateral surface of the occipital lobe is drained by the lateral occipital vein into the superior sagittal sinus. There are no large veins entering the superior sagittal sinus for a distance of 4 to 5 cm proximal to the torcula.18 There are three main anastomotic veins on the lateral surface of the cerebrum. 1. The superficial sylvian vein begins in the surface of the posterior part of the posterior ramus of the sylvian fissure, and runs inferior and anteriorly along the fissure. Along its trajectory, the superficial sylvian vein receives the frontosylvian, parietosylvian, and temporosylvian veins, and commonly anastomoses with the veins of Trolard and Labbé. In the region of the pterion, the superficial sylvian vein enters the dura and runs in the sphenoparietal sinus or sinus of the lesser wing of
Basic Neuroanatomy for Neurointensive Care Unit
13
the sphenoid19 to enter the anterior end of the cavernous sinus via the medial end of the superior orbital fissure, then drains into the basilar and the inferior petrosal sinuses. It can also drain into the tentorial sinus located in the dura of the middle fossa and then drains into the tentorium toward the transverse sinus. 2. The vein of Trolard, also called the superior anastomotic vein, is the largest anastomotic vein crossing the lateral surface of the frontal and parietal lobes between the superior sagittal sinus and the sylvian fissure. It is more frequently located at the parietal lobe. 3. The vein of Labbé, also called the inferior anastomotic vein, is the largest anastomotic vein that crosses the temporal lobe between the sylvian fissure and the transverse sinus. It usually arises from the middle portion of the sylvian fissure and is directed posterior and inferiorly toward the anterior part of the transverse sinus, at the level of the preoccipital notch (Fig. 1-6A). The deep part of the sylvian fissure is related to the deep sylvian or middle cerebral vein and its tributaries. The tributaries of the deep sylvian vein come mainly from the sulci on surface of the insula. The deep middle cerebral vein begins as a vein in the central sulcus of the insula, and runs anterior and inferior toward the limen insulae where it joins other insular veins to form a common trunk. Normally the deep middle cerebral vein courses medially into the carotid cistern, under the anterior perforated substance to form the first segment of the basal vein. The deep venous system of the cerebrum is divided into ventricular and cisternal groups; the ventricular group will be discussed here and the cisternal group will be discussed under the basal surface of the cerebrum. The ventricular veins are named mainly according to the location they course. Frontal horn: anterior caudate and anterior septal veins. Body of the lateral ventricle: thalamostriate, thalamocaudate veins, posterior caudate, and posterior septal veins. Atrium and the occipital horn: medial and lateral atrial veins. Temporal horn: inferior ventricular, amygdalar, and transverse hippocampal veins. Deep thalamic veins: anterior thalamic and superior thalamic veins. Superficial thalamic veins: anterior superficial thalamic, superior superficial thalamic, and posterior superficial thalamic veins. Choroidal veins: superior choroidal, and inferior choroidal veins.20 Lateral Surface: Arterial Relationships Most of the lateral surface of the cerebral hemisphere is supplied by the middle cerebral artery. The middle cerebral artery21,22 is divided into four segments. 1. The M1 or sphenoidal segment extends from the bifurcation of the internal carotid artery to the limen insulae. The M1 segment presents two types of
W9418-01.qxd 10/31/03 3:33 PM Page 14
Section I
14
Introduction
B
A
C
D
W9418-01.qxd 10/31/03 3:33 PM Page 15
Chapter 1 Figure 1-6. A, Lateral view of the right hemisphere: 1 = superior sagittal sinus; 2 = vein of Trolard; 3 = frontosylvian vein; 4 = superficial sylvian vein; 5 = vein of Labbé; 6 = Heschl’s gyrus; 7 = frontopolar vein; 8 = deep middle cerebral vein and the limen insulae; 9 = transverse sinus; 10 = right orbit. B, Superior view: 1 = right olfactory tract; 2 = genu of the middle cerebral artery; 3 = anterior cerebral artery; 4 = optic nerve; 5 = anterior clinoid process; 6 = limen insulae; 7 = internal carotid artery; 8 = uncus; 9 = P1 (posterior cerebral artery); 10 = anterior choroidal artery and the P2A (posterior cerebral artery); 11 = head of the hippocampus; 12 = midbrain; 13 = P2P (posterior cerebral artery); 14 = calcarine artery; M1 = sphenoid segment of the middle cerebral artery; M2 = insular segment of the middle cerebral artery; M3 = opercular segment of the middle cerebral artery. C, Basal view, displaying the sulci and gyri and the functional mapping of the basal surface of the cerebrum. 1 = olfactory tract; 2 = rectus gyrus; 3 = medial orbital gyrus; 4 = temporal pole; 5 = optic nerve; 6 = pituitary stalk; 7 = rhinal sulcus; 8 = optic tract; 9 = parahippocampal gyrus; 10 = inferior temporal gyrus; 11 = mamillary body and the posterior perforated substance; 12 = crus cerebri; 13 = medial geniculate body; 14 = tegmentum of the midbrain; 15 = collateral sulcus; 16 = pineal gland and the splenium of the corpus callosum; 17 = fusiform gyrus; 18 = occipitotemporal sulcus; 19 = lingual gyrus; 20 = preoccipital notch; III = oculomotor nerve. D, Basal view: 1 = deep middle cerebral artery; 2 = olfactory vein; 3 = anterior communicating artery; 4 = anterior cerebral vein; 5 = striate segment of the basal vein; 6 = chiasm; 7 = optic tract; 8 = crus cerebri; 9 = interpeduncular fossa and the peduncular vein; 10 = inferior ventricular vein; 11 = pulvinar of the thalamus; 12 = posterior mesencephalic segment of the basal vein; 13 = anterior choroidal artery (plexal segment); 14 = vein of Galen. Small arrows, Heubner’s artery. Arrowhead, peduncular vein joining the striate segment of the basal vein to form the peduncular segment of the basal vein.
䉳 branches: the lateral lenticulostriate arteries that arise mostly from the superior or posterosuperior aspect of the M1 and penetrate the anterior perforated substance to supply the basal ganglia, and early branches that course toward the temporal lobe to supply the temporal pole. 2. The M2 or insular segment extends from the limen insulae to the superior or inferior circular sulcus of insula; it runs in the insular compartment of the sylvian fissure, and is constituted by superior and inferior trunks and their branches. After reaching the superior or inferior circular sulcus of insula, the M2 branches enter the opercular compartment, and are called M3 segment. 3. The M3 or opercular segment runs in the opercular compartment and is related to the frontal and parietal opercula superiorly and to the temporal operculum inferiorly, and it depicts exactly the morphology of the opercula to which it is related. The morphology of the sylvian fissure is mainly determined by the operculum of the temporal lobe. The loop of the most posterior M3
Basic Neuroanatomy for Neurointensive Care Unit
15
segment branch that exits from the sylvian fissure is called “M point” or “sylvian point” (Fig. 1-6B).23 Anatomically the sylvian point is located behind the posterior part of the insula, above the medial end of Heschl’s gyrus. The angiographic sylvian point or “M point” not only displays the location of the medial end of the Heschl’s gyrus, but also depicts the posterior end of the insula, and consequently the posterior end of the mass described earlier constituted by basal ganglia, internal capsule, and thalamus. Medially it points toward the atrium of the lateral ventricle, and consequently it points to the posterior end of the thalamus (pulvinar of the thalamus), which is the anterior wall of the atrium. Therefore, the location of the medial end of Heschl’s gyrus, atrium of the lateral ventricle, posterior end of the insula, basal ganglia, and the thalamus can be determined on an anteroposterior angiogram by following the sylvian point. On lateral projection, the M2 and M3 segments form the “sylvian triangle” that depicts the shape of the insula; as the insula is the outer covering of the “mass” that comprises the basal ganglia, thalamus, and the internal capsule, the sylvian triangle shows the anterior, superior, and the inferior limits of this “mass.” 4. The M4 or cortical segment extends from the sylvian fissure to the lateral surface of the cerebrum. Basal Surface: Neural Relationships The basal surface comprises the basal surface of the frontal, temporal, and occipital lobes. The basal surface of the frontal lobe is divided by the olfactory tract and sulcus in two uneven parts, a smaller and medial part is the rectus gyrus, lodged within the olfactory groove; and a larger and lateral part, the orbital surface of the frontal lobe, located above the orbit and composed of orbital gyri. The orbital surface is divided by the orbital sulcus, a complex sulcus that presents a rough configuration of the letter “H,” into quadrants: the anterior orbital, medial orbital, posterior orbital, and lateral orbital gyri. The pars orbitalis of the inferior frontal gyrus is continuous with the posterior part of the lateral orbital gyrus and the lateral part of the posterior orbital gyrus. The posterior part of the rectus, medial orbital, and the posterior orbital gyri, along with the medial and lateral olfactory stria, constitute the anterior limit of the anterior perforated substance. The posterior orbital gyrus also continues posterior and inferiorly into the temporal lobe as limen insulae. The temporal lobe is separated posteriorly from the occipital lobe by the basal parietotemporal line, which extends from the preoccipital notch to the junction between the parieto-occipital and calcarine fissures and presents from lateral to medially the inferior temporal gyrus, occipitotemporal sulcus, fusiform gyrus, collateral sulcus, and parahippocampal gyrus. The inferior temporal gyrus runs
W9418-01.qxd 10/31/03 3:33 PM Page 16
16
Section I
Introduction
from the temporal pole anteriorly to the occipital lobe posteriorly. It is the most laterally located gyrus on the basal surface; medial to it is the occipitotemporal sulcus, a lateralto-medial, inferior to superiorly oriented sulcus that points toward the collateral eminence. The occipitotemporal sulcus has an anteroposterior orientation, and frequently fuses anteriorly and posteriorly with the collateral sulcus, delimiting the fusiform gyrus between them. Medial to the fusiform gyrus is the collateral sulcus. The collateral sulcus is an inferior-to-superior, medial-to-laterally oriented sulcus that bulges into the lateral part of the floor of the temporal horn anteriorly (collateral eminence) and the atrium posteriorly (collateral trigone). The collateral sulcus separates medially the allocortical parahippocampal gyrus from the mesocortical fusiform laterally. These gyri are separated anteriorly by the rhinal sulcus, which separates the uncus medially from the temporal pole laterally. The rhinal sulcus is the lateral limit of the uncus. Part of the floor of the third ventricle can also be seen from below: the optic chiasm, infundibulum, tuber cinereum, mamillary bodies, and the posterior perforated substance form the two-thirds anterior part of the floor of the third ventricle. The interpeduncular region, where the basilar artery bifurcates, is determined by two oculomotor nerves, the anteromedial surface, and the apex of the uncus laterally; diencephalic membrane of the Liliequist membrane (the membrane that goes from the dorsum sellae to the mamillary bodies), pituitary stalk and dorsum sellae anteriorly; tuber cinereum, mamillary bodies, and posterior perforated substance superiorly; inner surface of both crura cerebri posteriorly; the prepontine cistern forms the inferior limit of the interpeduncular fossa (Fig. 1-6C). The functional map of the basal surface is also shown in Figure 1-6C. Anterior Perforated Substance The anterior perforated substance (APS) is the entry site for perforating arteries from internal carotid, anterior choroidal, and anterior and middle cerebral arteries to the basal ganglia, the anterior limb, genu, and posterior limb of the internal capsule. It also is the exit site for the inferior striate veins. The APS is a rhomboid-shaped area buried deep in the sylvian fissure, bounded anteriorly by the lateral and medial olfactory striae; posteromedially by the optic tract and posterolaterally by the anteromedial surface of the uncus; laterally by the limen insulae. Medially the APS extends above the optic chiasm to the interhemispheric fissure.24 Intraoperatively, the APS and the carotid bifurcation can be identified by following the olfactory tract posteriorly. The head of the caudate nucleus, the anterior part of the lentiform nucleus and the anterior limb of the internal capsule are located immediately above the boundaries of the anterior perforated substance. Just like the insula can be understood as the outer covering of the basal ganglia and thalamus, the anterior perforated substance can be seen as
the “floor” of the anterior half of the basal ganglia. The anterior perforated substance can be considered the site where the anterior basal ganglia come to the surface extraventricularly (the caudate nucleus comes to the surface intraventricularly in the frontal horn and in the body of the lateral ventricle). Basal Surface: Venous Relationships The basal surface of the frontal lobe is drained by inferior frontal veins. The inferior frontal veins either drain anteriorly to the superior sagittal sinus (anterior group) or posteriorly to join the deep sylvian vein in the sylvian fissure (posterior group). The temporal lobe is drained by inferior temporal veins. The inferior temporal veins are divided into a lateral group that drains into the sinuses in the anterolateral part of the tentorium, and a medial group that empties into the basal vein as it courses along the medial edge of the temporal lobe. The occipital lobe is drained by the occipitobasal vein, which arises from tributaries that drain the inferolateral part of the lingual and adjacent part of the occipitotemporal and inferior temporal gyri. The occipitobasal vein courses anterolaterally toward the preoccipital notch and frequently joins the posterior temporobasal vein before emptying into the lateral tentorial sinus. The most important deep venous channel on the basal surface is the basal vein of Rosenthal.25 The basal vein originates below the APS by the union of the deep middle cerebral, inferior striate, olfactory, fronto-orbital, and anterior cerebral veins, and it usually drains into the vein of Galen after passing around the midbrain. The basal vein is divided into three segments. The first, also called anterior or striate segment, originates from the junction of the anterior cerebral, inferior striate, olfactory, fronto-orbital, and deep middle cerebral veins under the APS and runs posteriorly under the optic tract, medially to the anterior portion of the crus cerebri. This point denotes laterally the location of the apex of the uncus. The main tributaries of the first segment are the fronto-orbital, the olfactory, the inferior striate, the anterior cerebral, the deep middle cerebral, and the anterior pericallosal veins. The second, also called middle or peduncular segment, starts from this most medial point in the course of the basal vein, usually correspondent to the site where the peduncular vein joins the basal vein. It runs laterally between the upper part of the posteromedial surface of the uncus and the upper part of the crus cerebri, and under the optic tract to reach the most lateral part of the crus cerebri, which corresponds to the most lateral point of the vein as it turns around the crus cerebri, usually where the inferior ventricular vein joins the basal vein. This is called the anterior peduncular segment by Huang and Wolf 25; it then turns medially, superiorly, and posteriorly to the plane of the lateral mesencephalic sulcus behind the crus cerebri to constitute the posterior peduncular segment. The main
W9418-01.qxd 10/31/03 3:33 PM Page 17
Chapter 1
tributaries of the second segment are the peduncular or interpeduncular vein, the inferior ventricular, the inferior choroidal, the hippocampal, and the anterior hippocampal veins. The third, also called posterior or posterior mesencephalic segment, runs medially, superiorly, and posteriorly from the lateral mesencephalic sulcus, and under the pulvinar of the thalamus to penetrate the quadrigeminal cistern and generally drains into the vein of Galen. The main tributaries of the third segment are the lateral mesencephalic vein, the posterior thalamic, the posterior longitudinal hippocampal, the medial temporal, and the medial occipital veins. Sometimes, the precentral cerebellar, superior vermian, internal occipital, splenial, medial atrial, and direct lateral and lateral atrial subependymal veins may drain into the third segment of the basal vein (Fig. 1-6D). Basal Surface: Arterial Relationships The supraclinoid portion of the internal carotid artery, anterior choroidal artery, anterior perforating and the posterior cerebral arteries are better visualized from this surface. The internal carotid artery is divided into five parts: cervical, petrous, cavernous, clinoid, and the supraclinoid portions. Yet recent evidence has shown that the clinoid segment is also located inside the cavernous sinus.26 The supraclinoid portion of the internal carotid artery has been divided into three segments based on the origin of its major branches:27 the ophthalmic segment extends from the origin of the ophthalmic artery to the origin of the posterior communicating artery (PCom); the communicating segment extends from the origin of the PCom to the origin of the anterior choroidal artery (AChA), and the choroidal segment extends from the origin of the AChA to the bifurcation of the internal carotid artery. Each segment gives off a series of perforating branches with a relatively constant site of termination. The ophthalmic artery arises under the optic nerve, usually from the medial one third of the superior surface of the carotid artery, then it passes anteriorly and laterally to become superolateral to the carotid to enter the optic canal and the orbit.28 The perforating arteries from this segment arise from the posterior or medial or posteromedial aspect of the carotid artery and are distributed to the stalk of the pituitary gland, the optic chiasm, and less commonly to the optic nerve, premamillary portion of the floor of the third ventricle, and the optic tract. The superior hypophyseal arteries, which can range from one to five in number, pass medially across the ventral surface of the chiasm to reach the pituitary area, where they intermingle and constitute a fine anastomotic plexus around the pituitary stalk called the “circuminfundibular anastomosis,” which supplies the pituitary stalk and the anterior lobe of the pituitary gland. The posterior lobe is supplied by the inferior hypophyseal artery originated from the meningohypophyseal trunk of the intracavernous carotid artery. The infundibular arteries are
Basic Neuroanatomy for Neurointensive Care Unit
17
another group of arteries that arise from the PCom and supply the same area as the superior hypophyseal artery. The PCom arises from posteromedial or posterior or posterolateral aspect of the internal carotid artery and passes posteromedially below the tuber cinereum and above the sella turcica and oculomotor nerve to join the posterior cerebral artery (PCA). In the embryo, the PCom continues as the PCA, but in the adult the PCA becomes part of the basilar system. If the PCom remains the major origin of the PCA, the configuration is termed “fetal.” The perforating arteries from PCom range four to 14 in number, arising predominantly from the proximal half of the artery, course superiorly and terminate in the premamillary area, posterior perforated substance, interpeduncular fossa, optic tract, pituitary stalk, and the optic chiasm. They then reach the thalamus, the hypothalamus, the subthalamus, and internal capsule. The largest branch from the PCom is the premamillary artery or “anterior thalamoperforating artery,” which enters the floor of the third ventricle in front of or beside the mamillary body. The AChA arises either from the posterolateral or from the posterior aspect of the internal carotid artery. The AChA courses posteriorly below the optic tract and above the PCom toward the temporal horn by passing through the choroidal fissure. The AChA sends off branches to the optic tract, the crus cerebri, the lateral geniculate body, and the uncus, and supplies the optic radiation, the globus pallidus, the midbrain, the thalamus, and the retrolenticular and posterior portion of the posterior limb of the internal capsule.29,30 The choroidal segment of the internal carotid artery is the most frequent site of perforating arteries (range, 1 to 9), they arise either from the posterior, posterolateral, or posteromedial surfaces of the internal carotid artery. They terminate in the APS, optic tract, and the uncus. The anterior perforating arteries are those arising from the internal carotid, middle and anterior cerebral, and the anterior choroidal arteries, which enter the brain through the APS. The anterior perforating arteries of the internal carotid artery arise from the choroidal segment (Fig. 1-7A and B). Embryologically, the posterior cerebral artery (PCA)31–35 arises as a branch of the internal carotid artery, but up to birth its most common origin is the basilar artery. The PCA is classified, according to Yasargil and Rhoton, into four segments: the P1 segment extends from the basilar bifurcation to the site where the PCom joins the PCA. The P2 segment extends from the PCom to the posterior aspect of the midbrain. The P2 segment is further divided into P2A (anterior) and P2P (posterior) segments. P2A segment begins at the PCom and courses around the crus cerebri; inferiorly to the optic tract, AchA, and basal vein; and medially to the posteromedial surface of the uncus, up to the posterior margin of the crus cerebri. The P2P segment begins at the posterior margin of the crus cerebri and runs laterally to
W9418-01.qxd 10/31/03 3:33 PM Page 18
18
Section I
Introduction
A
B
C
D
Figure 1-7. A, Basal view: 1 = internal carotid artery; 2 = M1 (middle cerebral artery); 3 = posterior communicating artery; 4 = anterior choroidal artery; 5 = P1 (posterior cerebral artery); 6 = P2A (posterior cerebral artery); 7 = anterior inferior temporal artery; 8 = interpeduncular fossa; 9 = short circumflex arteries; 10 = P2P (posterior cerebral artery); 11 = middle inferior temporal artery; 12 = posterior inferior temporal artery. B, Basal view: 1 = internal carotid artery; 2 = middle cerebral artery; 3 = A1 (anterior cerebral artery); 4 = anterior communicating artery; 5 = anterior choroidal artery; 6 = hypothalamic arteries from anterior communicating artery; 7 = posterior communicating artery; 8 = lateral lenticulostriate arteries; 9 = P2P (posterior cerebral artery); 10 = interpeduncular fossa; 11 = anterior inferior temporal artery; 12 = short circumflex arteries; 13 = P2P (posterior cerebral artery). C, Inferomedial view of the left hemisphere: 1 = parieto-occipital sulcus; 2 = isthmus of the cingulate gyrus; 3 = posterior communicating artery; 4 = collicular point; 5 = P3 (posterior cerebral artery); 6 = P2A (posterior cerebral artery); 7 = internal carotid artery; 8 = posterior inferior temporal artery; 9 = P2P (posterior cerebral artery); 10 = calcarine sulcus; 11 = middle inferior temporal artery. D, Medial view: the midbrain has been removed. 1 = anterior choroidal artery entering the temporal horn (inferior choroidal point); 2 = P2A (posterior cerebral artery); 3 = posterior segment of the uncus; 4 = perforating branches from the anterior choroidal artery; 5 = middle cerebral artery; 6 = cisternal segment of the anterior choroidal artery; 7 = anterior cerebral artery; 8 = pons; 9 = P1 (posterior cerebral artery); 10 = posterior communicating artery; 11 = internal carotid artery.
W9418-01.qxd 10/31/03 3:33 PM Page 19
Chapter 1
the tegmentum of the midbrain within the ambient cistern, parallel and inferiorly to the basal vein, inferolaterally to the geniculate bodies and pulvinar, and medially to the parahippocampal gyrus to enter the quadrigeminal cistern. The P3 segment begins under the posterior part of the pulvinar in the lateral aspect of the quadrigeminal cistern and ends at the anterior limit of the anterior calcarine sulcus. P3 is often divided into its major terminal branches: the calcarine and the parieto-occipital arteries before reaching the anterior limit of the anterior calcarine sulcus. The point where the PCAs from each side are closer to each other is called the collicular or quadrigeminal point. It marks the posterior limit of the midbrain on angiogram. The P4 segment is the cortical branches of the PCA (Fig. 1-7C and D). The main branches arising from the PCA are the posterior thalamoperforating, the direct perforating, the short and long circumflex, the thalamogeniculate, the medial and the lateral posterior choroidal, the inferior temporal, the parieto-occipital, the calcarine, and the posterior pericallosal arteries. The posterior thalamoperforating arteries, which arise from P1 segment and enter the brain through the posterior perforated substance, interpeduncular fossa, and medial crus cerebri, supply the anterior and part of the posterior thalamus, hypothalamus, subthalamus, substantia nigra, red nucleus, oculomotor and trochlear nuclei, oculomotor nerve, mesencephalic reticular formation, pretectum, rostromedial floor of the third ventricle, and the posterior portion of the internal capsule.36 The direct perforating arteries to crus cerebri arise mainly from the P2A segment and supply the crus cerebri. The short and long circumflex arteries to the brainstem arise mainly from P1 segment, and less frequently from P2A segments; the short circumflex artery courses around the midbrain and terminates at the geniculate bodies; the long circumflex artery courses around the midbrain and reaches the colliculi. The thalamogeniculate arteries arise equally from P2A or P2P segments, perforate the inferior surface of the geniculate bodies, and supply the posterior half of the lateral thalamus, posterior limb of the internal capsule, and the optic tract. The medial posterior choroidal arteries arise mainly from P2A and less frequently from P2P and P1 segments, and course around the midbrain, medial to the main trunk of the PCA, turn around the pulvinar of the thalamus to proceed superiorly at the lateral side of colliculi and pineal gland, to enter the roof of the third ventricle through the velum interpositum, and finally course through the foramen of Monro to enter the choroid plexus in the lateral ventricle. The medial posterior choroidal arteries supply the crus cerebri, tegmentum, geniculate bodies (mainly the medial), the colliculi, pulvinar, pineal gland, and medial thalamus. Angiographically on lateral projection, the medial posterior choroidal artery describes the shape of the number “3.” The inferior curve of the “3” is when it turns around the pulvinar, and the superior curve is when it contours the colliculi before
Basic Neuroanatomy for Neurointensive Care Unit
19
entering the roof of the third ventricle. Lateral posterior choroidal arteries arise mainly from P2P, and less frequently from P2A segment, and pass laterally to enter the ventricular cavity directly through the choroidal fissure, to supply the choroid plexus in the atrium and the temporal horn. It anastomoses with the AChA and also supplies the crus cerebri, posterior commissure, part of the body and anterior portion of the column of the fornix, the lateral geniculate body, pulvinar, dorsomedial thalami nucleus, and the body of the caudate nucleus. Inferior temporal arteries are distributed to the basal surface of the temporal and occipital lobes. They include the hippocampal artery and three groups of temporal arteries, namely, anterior, middle, and posterior temporal arteries. The anterior temporal artery arises mainly from the P2A segment, while the middle and posterior temporal arteries arise mainly from the P2P segment. Parieto-occipital and calcarine arteries are usually terminal branches of the PCA; they arise predominantly from P3 segment; however, sometimes they may also arise from the P2P segment and course respectively in the parietooccipital fissure and calcarine fissure. As the calcarine fissure reaches laterally to bulge into the medial wall of the atrium and the occipital horn, the calcarine artery also follows laterally into the depth of the calcarine fissure. The splenial or posterior pericallosal artery supplies the splenium of the corpus callosum, and arises from the parieto-occipital artery in 62% of cases, from the calcarine artery in 12%, medial posterior choroidal artery in 8%, posterior temporal in 6%, P2P in 4%, P3 in 4%, and lateral posterior choroidal artery in 4%. Medial Surface: Neural Relationships The medial surface of the cerebrum comprises the sulci and gyri of the frontal, parietal, occipital, and temporal lobes. The general organization of the gyri of the frontal, parietal, and occipital lobes on this surface can be compared to that of a three-layer roll; the inner layer is represented by corpus callosum, the intermediate layer by cingulate gyrus, and the outer layer by the medial frontal gyrus, paracentral lobule, precuneus, cuneus, and the lingual gyrus. The cingulate gyrus is separated inferiorly from the corpus callosum by the callosal sulcus, and superiorly from the outer layer by cingulate sulcus. Several secondary rami ascend from the cingulate sulcus in a radiate pattern and divide the outer layer into several sections; there are two secondary rami of particular importance: the paracentral ramus, which ascends from the cingulate sulcus at the level of the midpoint of the corpus callosum, and separates the medial frontal gyrus anteriorly from the paracentral lobule posteriorly and the marginal ramus, which ascends from the cingulate sulcus at the level of the splenium of the corpus callosum, and separates the paracentral lobule anteriorly from the precuneus posteriorly.
W9418-01.qxd 10/31/03 3:33 PM Page 20
20
Section I
Introduction
The parieto-occipital and the calcarine sulci also have the same radiate pattern; the parieto-occipital sulcus separates the precuneus superiorly from the cuneus inferiorly, and the calcarine sulcus separates the cuneus superiorly from the lingual gyrus inferiorly. The marginal ramus of the cingulate sulcus intercepts the postcentral gyrus in almost 100% of the cases.1 The marginal ramus is an important landmark to determine the location of the sensory or motor areas in the lateral convexity through a midsagittal MRI. The paracentral ramus along with the marginal ramus determines the paracentral lobule, which is concerned with movements of the contralateral lower limb and perineal region, and is involved in voluntary control over defecation and micturition. The precuneus along with the part of the paracentral lobule behind the central sulcus forms the medial part of the parietal lobe. The precuneus presents the subparietal sulcus on its surface. This sulcus behaves as a posterior continuation of the cingulate sulcus and separates the precuneus above from the cingulate gyrus below. It has a variable vaguely H-shape where the vertical arms of the H tend to align with the marginal ramus and the parietooccipital sulcus. The parieto-occipital and the calcarine sulci determine the cuneus; the cuneus along with the medial part of the lingual gyrus are the medial portion of the occipital lobe. The calcarine sulcus starts at the occipital pole and directs anteriorly, presenting a slightly curved course with its characteristic upward convexity. The calcarine sulcus joins the parieto-occipital sulcus (only superficially) at an acute angle behind the isthmus of the cingulate gyrus, then continues anteriorly to intercept the isthmus of the cingulate gyrus. The portion of the calcarine sulcus anterior to the junction with the parieto-occipital sulcus is called anterior calcarine sulcus, which is crossed by a buried anterior cuneolingual gyrus and bulges into the medial wall of the atrium of the lateral ventricle as calcar avis. The part of the calcarine posterior to the union is called posterior calcarine sulcus and presents the striate (visual) cortex on its upper and lower lips, and the anterior calcarine sulcus presents the visual cortex only on its lower lip. Anteriorly, the cingulate and the medial frontal gyri wrap around the genu and the rostrum of the corpus callosum. At the inferior end of these two gyri, under the rostrum of the corpus callosum and in front of the lamina terminalis is a narrow triangle of gray matter, the paraterminal gyrus, separated from the rest of the cortex by a shallow posterior paraolfactory sulcus. Slightly anterior to this sulcus, the anterior paraolfactory sulcus, a short vertical sulcus may occur; the cortex between the posterior and anterior paraolfactory sulci is the subcallosal area or paraolfactory gyrus. Frequently two anteroposteriorly directed sulci, the superior and inferior rostral sulci, which are parallel to the floor of the anterior fossa, divide the inferior portion of the medial frontal gyrus into three parts. Posteriorly the cingulate gyrus continues inferiorly with the parahip-
pocampal gyrus through the isthmus of the cingulate gyrus (Fig. 1-8A). The mesial portion of the temporal lobe presents intraventricular and extraventricular elements.6 The intraventricular elements are the hippocampus, fimbria, amygdala, and the choroidal fissure; the extraventricular elements are the parahippocampal gyrus, uncus, and dentate gyrus. The parahippocampal gyrus extends from anterior to posterior, and at its anterior extremity it deviates medially and bends posteriorly to constitute the uncus. Posteriorly, just below the splenium of the corpus callosum, the parahippocampal gyrus is often intersected by the anterior calcarine sulcus, which divides the posterior portion of the parahippocampal gyrus into the isthmus of the cingulate gyrus superiorly, and the parahippocampal gyrus inferiorly, which continues posteriorly as the lingual gyrus. Superiorly the parahippocampal gyrus is separated from the dentate gyrus by the hippocampal sulcus. Laterally, the parahippocampal gyrus is limited by the collateral sulcus posteriorly and by the rhinal sulcus anteriorly. The rhinal sulcus marks the lateral limit of the entorhinal area of the parahippocampal gyrus. Medially the parahippocampal gyrus is related to the free edge of the tentorium and to the contents of the ambient cistern. The various components of the parahippocampal gyrus are the subiculum, presubiculum, parasubiculum, and entorhinal area, with the subiculum as its medial round edge. Uncus, meaning hook, is formed by the anterior portion of the parahippocampal gyrus, which has deviated medially and folded posteriorly. Inferiorly, the uncus is separated from the parahippocampal gyrus by the uncal notch (Fig. 1-8A). Anteriorly, the uncus continues with the anterior portion of the parahippocampal gyrus without a sharp limit; superiorly, the uncus is continuous with the globus pallidus. At the basal surface, the uncus is separated laterally from the temporal pole by the rhinal sulcus, and its medial part is normally herniated medially to the tentorial edge. When viewed from its basal surface, the uncus presents the shape of an arrowhead with its apex pointing medially, featuring an apex, an anterior segment, and a posterior segment. The anterior segment of the uncus is continuous with the parahippocampal gyrus and presents one surface, the anteromedial, which is related to the proximal sylvian fissure and carotid cistern, and is the posterolateral limit of the APS. The posterior segment is related to the hippocampus and has two surfaces: a posteromedial and an inferior surface. The posterior segment is occupied by three small gyri; from anterior to posterior, they are the uncinate gyrus, the band of Giacomini, and the intralimbic gyrus. The superior and the inferior portions of the posteromedial surface of the uncus are related, respectively, to the crural and ambient cisterns. Posteriorly and superiorly to the uncus is the inferior choroidal point, where the choroid plexus of the temporal horn begins. The inferior choroidal point usually corresponds to the site where the AChA enters the
W9418-01.qxd 10/31/03 3:33 PM Page 21
Chapter 1
A
B
C
D
Basic Neuroanatomy for Neurointensive Care Unit
21
Figure 1-8. A, Medial view of the medial surface of the left hemisphere: 1 = central sulcus; 2 = paracentral lobule; 3 = marginal ramus of the cingulate sulcus; 4 = cingulate sulcus; 5 = medial frontal gyrus; 6 = cingulate gyrus; 7 = precuneus; 8 = subparietal sulcus; 9 = body of the corpus callosum; 10 = genu of the corpus callosum; 11 = parieto-occipital sulcus; 12 = rostrum of the corpus callosum; 13 = cuneus; 14 = isthmus of the cingulate gyrus; 15 = splenium of the corpus callosum; 16 = calcarine sulcus; 17 = lingual gyrus; 18 = anterior calcarine sulcus; 19 = uncus; 20 = parahippocampal gyrus; 21 = temporal pole. B, The functional mapping of the medial surface of the left hemisphere. C, Medial view of the medial surface of the left hemisphere: 1 = postcentral gyrus; 2 = central sulcus; 3 = precentral sulcus; 4 = marginal ramus of the cingulate sulcus; 5 = paracentral lobule; 6 = paracentral ramus of the cingulate sulcus; 7 = medial frontal gyrus; 8 = cingulate gyrus; 9 = fornix; 10 = straight sinus; 11 = hypothalamic sulcus; 12 = subcallosal area; 13 = medial frontal gyrus; 14 = vein of Galen; 15 = posterior perforated substance; 16 = rectus gyrus and the olfactory tract; 17 = tentorium edge; 18 = cerebellum; 19 = superior petrosal sinus; 20 = middle fossa; II = optic nerve; III = oculomotor nerve. Arrow, foramen of Monro; A2, A3, A4, and A5 = segments of the anterior cerebral artery. D, Vascularization of the basal ganglia and the thalamus: APS = anterior perforated substance A1; A2 = segments of the anterior cerebral artery; M1 = sphenoid segment of the middle cerebral artery; AchA = anterior choroidal artery; Pcom = posterior communicating artery; PPS = posterior perforated substance.
temporal horn through the choroidal fissure. The inferior surface is the superior lip of the uncal notch, and it is visible only from below when the parahippocampal gyrus is removed. The dentate gyrus bears this name because of its characteristic toothlike elevations. The dentate gyrus continues
anteriorly with the band of Giacomini, also called the tail of the dentate gyrus, and continues posteriorly with the fasciolar gyrus, a smooth grayish band that is located posteriorly to the splenium of the corpus callosum; the fasciolar gyrus continues above the corpus callosum as the indusium griseum to finally end as the paraterminal gyrus. The
W9418-01.qxd 10/31/03 3:33 PM Page 22
22
Section I
Introduction
fimbrodentate and hippocampal sulci separate the dentate gyrus, respectively, from the fimbria superiorly and the parahippocampal gyrus inferiorly. The extraventricular and the intraventricular structures of the mesial temporal lobe are intimately related. The uncus is related medially to cisternal elements and laterally to intraventricular elements. The anterior segment is related to the proximal sylvian fissure, the lateral portion of the carotid cistern, and the amygdala. The apex is related to the oculomotor nerve medially, and to the uncal recess, and the amygdala laterally; the posterior segment is related to the head of the hippocampus and the amygdala laterally, to the P2A segment of PCA inferomedially, and to the AChA superomedially. The functional map of the mesial surface is shown in Figure 1-8B.
Medial Surface: Venous Relationships The medial surface of the frontal lobe is drained by the medial frontal veins. They can either empty superiorly into the superior sagittal sinus, or inferiorly either into the inferior sagittal sinus or into the veins that pass around the corpus callosum to drain into the anterior end of the basal vein. The medial surface of the parietal lobe is formed by part of the paracentral lobule located behind the central sulcus and the precuneus, and it is drained by the medial parietal veins. They can either empty superiorly into the superior sagittal sinus or course around the splenium of the corpus callosum (posterior pericallosal vein) and drain inferiorly into the vein of Galen or its tributaries. The occipital lobe is drained by the anterior and the posterior calcarine veins. The anterior calcarine, also called internal occipital vein, arises from tributaries that drain the anterior portion of the cuneus and lingual gyrus, and passes forward toward the vein of Galen. It frequently joins the posterior pericallosal vein near the splenium before terminating in either the internal cerebral vein or in the vein of Galen. The posterior calcarine vein arises from tributaries that drain the area bordering the posterior part of the calcarine fissure. The initial part of this vein may course posteriorly along the fissure, and then curve sharply upward on the cuneus to reach the superior sagittal sinus. The deep venous system of the mesial temporal region drains into the basal vein of Rosenthal. The cisternal group in the mesial temporal region comprises the basal vein; the anterior, middle, and the posterior temporal cortical veins; and the anterior longitudinal hippocampal, the anterior hippocampal, the lateral mesencephalic, the posterior mesencephalic, and the posterior longitudinal hippocampal veins. The ventricular group comprises those veins located in the temporal horn that ultimately drain into the second or peduncular segment of the basal vein of Rosenthal, usually via inferior ventricular vein. These include the amygdalar
vein, the transverse hippocampal veins, the inferior choroidal vein, and the inferior ventricular vein.
Medial Surface: Arterial Relationships The anterior cerebral artery (ACA) is classified according to Fisher into five segments:37 the A1 segment extends from the bifurcation of the internal carotid artery to the anterior communicating artery (ACom); the A2 segment extends from the ACom to the junction between the rostrum and the genu of the corpus callosum; the A3 segment extends from the genu of the corpus callosum to the point where the artery turns sharply and posteriorly above the genu of the corpus callosum (the A2 and A3 segments together are also called ascending segment); and the A4 and A5 segments extend above the corpus callosum, from the genu to the splenium. The combination of the A4 and A5 segments are also called horizontal segment. The separation between these two segments is the point bisected in the lateral view close behind the coronal suture. The segment of the ACA distal to the ACom (A2 to A5) has also been called the pericallosal artery. The A1 segment arises from the carotid bifurcation, in the carotid cistern and courses preferably above either the optic chiasm or the optic nerve to enter the lamina terminalis cistern. The medial lenticulostriate perforators, ranging from 1 to 11 branches (average of 6.4), arise from the superior, the posterior, or the posterior-superior aspect of the proximal half of A1 segment and pursue a direct posterior and superior course to the APS to supply the optic chiasm, the optic tract, the genu of the internal capsule, and the anterior part of the globus pallidus. They may extend to the adjacent part of the posterior limb of the internal capsule and, less commonly, to the thalamus. The ACom unites the paired anterior cerebral arteries in the lamina terminalis cistern to provide an anastomotic channel between the anterior circulation on both hemispheres. Embryologically the ACom develops from a multichanneled vascular network that coalesces to a variable degree by the time of birth. The ACom complex probably exists as a single channel in approximately 75% of the cases. In other cases, a spectrum of anomalies exists between the multichanneled network of the embryo and the single ACom, which include duplications and triplications, fenestrations, reticular patterns, and loops and bridges. The perforators from the ACom, ranging from 0 to 4 (average 1.6), usually arise from its postero-inferior aspect to supply the infundibulum, the APS, the optic chiasm, the subcallosal area, and the preoptic areas of the hypothalamus.38,39 The recurrent artery of Heubner of the ACA arises in 78% of the cases from the proximal A2 and it doubles back on its parent vessel, courses anterior to the A1 segment in 60% of the cases, and can be seen upon elevating the frontal lobe
W9418-01.qxd 10/31/03 3:33 PM Page 23
Chapter 1
before the visualization of the A1 segment; it is the largest and longest branch directed to the APS. After its origin, it passes above the carotid bifurcation and accompanies the middle cerebral artery into the medial part of the sylvian fissure before entering the anterior and middle portions of the full mediolateral extent of the APS (from above the optic chiasm, close to the interhemispheric fissure, to the limen insulae) to supply the most anterior and inferior part of the head of the caudate nucleus and putamen, and the adjacent part of the anterior limb of the internal capsule.40 The A2 segment is also the source of the central or the basal perforating arteries, which pass posteriorly to enter the optic chiasm, the lamina terminalis, and the anterior forebrain, below the corpus callosum, to supply the anterior hypothalamus, the septum pellucidum, the medial portion of the anterior commissure, the columns of the fornix, and the anterior-inferior part of the striatum. The two first cortical branches of the anterior cerebral artery supplying the medial surface, the orbitofrontal and the frontopolar arteries, also usually arise from the A2 segment. The segments A3 to A5 give rise to other cortical branches to supply the medial surface of the hemisphere. The A3 segment is a frequent site of origin for the anterior and the middle internal frontal and the callosomarginal arteries. The A4 segment frequently gives rise to the paracentral artery. The A5 segment gives rise to the superior and the inferior parietal arteries. All the cortical branches arise more frequently from the pericallosal than from the callosomarginal artery (Fig. 1-8C). The anterior cerebral artery syndromes include:41 (1) paracentral lobule syndrome, (2) supplementary motor area (SMA) syndrome, (3) anterior cingulate syndrome, (4) callosal syndrome, (5) basal forebrain syndrome, and (6) total ACA territory infarction. Paracentral syndrome is characterized by weakness of the contralateral lower limb, most intense in the foot and ankle, with or without sensory loss. The transient or permanent incontinence of urine can also be present. The SMA occupies the mesial surface of the superior frontal gyrus immediately anterior to paracentral lobule. SMA syndrome can be characterized by dysphasia (when the dominant hemisphere is affected), akinesia in the contralateral limb, contralateral hand grasping or groping, contralateral alien hand signs (when dominant hemisphere is affected, the right hand consistently interrupts manual tasks performed by the left hand), and dyspraxia. Anterior cingulate syndrome is more evident when the cingulate cortex is bilaterally and extensively affected; this might cause akinetic mutism, complex behavioral changes, loss of sphincter control, and autonomic disfunctions (temperature, cardiac, and respiratory irregularities). The callosal syndromes can be characterized by “split brain” signs and symptoms: left hand apraxia (inability to perform actions with left hand on verbal command), alien hand syndrome (left hand behaving like a foreigner or an alien, and acts uncooperatively), and left
Basic Neuroanatomy for Neurointensive Care Unit
23
hand agraphia. Basal forebrain syndrome occurs when the territories of the orbitofrontal and frontopolar arteries and the septal region are affected, and the signs and symptoms include amnesia, emotional disinhibition, inappropriate social conduct, and autonomic disturbances. Total ACA infarction syndrome is a combination of the previously mentioned syndromes. Vascularization of the Basal Ganglia and the Thalamus The “mass” that constitutes the core of the cerebrum comprises basal ganglia, internal capsule, and the thalamus. The vessels that supply this “mass” encircle it from below, behind, above, and from its lateral aspect. The external covering of this “mass” up to the claustrum, namely insula and extreme capsule, are supplied by the branches from M2 segment of the middle cerebral artery. The rest of the “mass” constituted by the basal ganglia is supplied by perforator branches from anterior cerebral, middle cerebral, internal carotid, and anterior choroidal arteries (Fig. 1-8D). The thalamus is supplied by four vascular groups that surround the thalamus: (1) anterior thalamoperforating arteries, coming from PCom artery, coming from inferior and anteriorly to supply the anterior thalamus; (2) posterior thalamoperforating arteries from P1 or basilar top, coming from inferiorly through the posterior perforated substance to supply the medial ventral part of the thalamus, which is related to alertness, memory, and emotion; (3) thalamogeniculate arteries arising from P2P segment of the PCA, coming from inferolaterally; and (4) medial and lateral posterior choroidal arteries from P2A or P2P segment of the PCA, coming from behind and above (Fig. 1-8D).42,43
Posterior Fossa The posterior fossa is the largest and deepest of the three cranial fossae. It comprises one eighth the intracranial space and contains the pathways regulating consciousness, vital autonomic functions, and motor activities, in addition to the centers for controlling balance and gait. Only two of the 12 pairs of cranial nerves are located entirely outside the posterior fossa. The posterior fossa extends from the tentorial incisura, through which it communicates with the supratentorial space, to the foramem magnum, through which it communicates with the spinal cord. The posterior fossa is separated from the supratentorial space by the tentorium cerebelli.44 The intracranial surface of the posterior fossa presents jugular foramen, internal acoustic meatus, hypoglossal canal, the vestibular and cochlear aqueducts, and several venous emissary foramina. The posterior fossa also presents neural (cerebellum, the brainstem, and the cranial nerves) and vascular elements (arteries and veins), which can be
W9418-01.qxd 10/31/03 3:33 PM Page 24
24
Section I
Introduction
characterized by the “rule of three”: the brainstem presents three parts (midbrain, pons, and medulla), the cerebellum presents three surfaces (petrosal, tentorial, and suboccipital), three cerebellar peduncles (superior, middle, and inferior), three fissures (cerebellomesencephalic, cerebellopontine, and cerebellomedullary), three main arteries (superior cerebellar artery [SCA], anterior inferior cerebellar artery [AICA], and posterior inferior cerebellar artery [PICA]), and three main venous draining groups (petrosal, galenic, and tentorial).
Brainstem The brainstem cannot be considered simply a connecting structure between the diencephalon and the spinal cord. Through the long projection systems of its reticular formation, the brainstem modulates sensory and motor pathways, and modulates arousal and conscious states (ascending projections to the diencephalon and the cerebrum); the brainstem presents the nuclei of ten cranial nerves that supply the sensory and motor functions of the face and head and the autonomic functions of the body; the brainstem also coordinates reflexes and simple behaviors mediated by the cranial nerves. As a general rule, the descending motor system occupies the anterior portion of the brainstem, while the long ascending and descending sensory tracts (the medial lemniscus; spinothalamic tract; and auditory, vestibular, and visceral sensory pathways) run within the reticular formation, which is located at the core (tegmentum) of the brainstem. The brainstem is divided into three portions: midbrain, pons, and the medulla. The midbrain is the upper part of the brainstem, connecting the diencephalon superiorly to the pons inferiorly. The midbrain is separated superiorly from the diencephalon by the optic tract, lateral geniculate body, and the pulvinar of the thalamus. Inferiorly the midbrain is separated from the pons by the pontomesencephalic sulcus and by the emergence of the trochlear nerve, which is a mesencephalic structure (Fig. 1-9A). The midbrain is divided by a midline sagittal plane into two halves that are called cerebral peduncles. Each peduncle is further divided into three parts: an anterior part, crus cerebri or basis pedunculi; an intermediate part, the tegmentum; and a posterior part located behind the aqueduct that is the tectum. The substantia nigra and the lateral mesencephalic sulcus separate the crus cerebri from the tegmentum. The oculomotor nerves emerge from the medial side of the crura cerebri in the interpeduncular fossa (Fig. 1-9B). The pontomesencephalic sulcus, which separates the midbrain superiorly from the pons inferiorly, originates in the depth of the interpeduncular fossa and runs around the inferior margin of the crus cerebri to join the lateral mesencephalic sulcus behind the crus cerebri. The posterior or dorsal aspect of the midbrain is characterized by the
superior and inferior colliculi (quadrigeminal plate). The superior colliculi are connected to the lateral geniculate bodies via brachium of the superior colliculus, and the inferior colliculi are connected to the medial geniculate bodies via brachium of the inferior colliculus. The pons presents a prominent anterior surface that is considerably convex from side to side, and it consists of transverse fibers that cross the median plane and converge on each side to form the middle cerebellar peduncles. The basilar sulcus is a shallow median groove on the anterior surface of the pons and usually lodges the basilar artery; this sulcus is bounded on each side by an eminence caused by the descent of the corticospinal fibers through the substance of the pons. The middle cerebellar peduncle is separated from the belly of the pons by a vertical shallow groove, the lateral pontine sulcus. Just lateral to the lateral pontine sulcus is the emergence of the trigeminal nerve, with its smaller superomedial motor root and a larger inferolateral sensory root (Fig. 1-9A). Posteriorly the pons constitutes the upper portion of the floor of the fourth ventricle. The medulla presents at its anterior aspect three longitudinal fissures, one median and two paramedian; the median one is the anterior median fissure, which continues inferiorly as the anterior median fissure of the spinal cord. The paramedian sulci of the anterior aspect of the medulla are the anterolateral sulci. At the medulla, the anterolateral sulcus is located medially to the olive; because of that it is also called preolivary sulcus. The preolivary sulcus is the upper continuation of the anterolateral sulcus of the spinal cord. The rootlets of the hypoglossal nerve that exit from the preolivary sulcus are analogous to the ventral motor rootlets that exit from the anterolateral sulcus of the spinal cord. The anterior region, located between the anterior median fissure and the preolivary sulcus, is characterized by the pyramid. The olives are located laterally to the preolivary sulcus; behind the olive, the rootlets of the accessory, the vagus, and the glossopharyngeal nerves exit from the postolivary sulcus. The postolivary sulcus is the continuation of the posterolateral sulcus of the spinal cord in the medulla oblongata; therefore, these cranial nerve rootlets are analogous to the dorsal spinal rootlets. Those rootlets emerge from the brainstem and extend almost straight laterally to the jugular foramen. The pontomedullary sulcus separates the pons from the medulla, and its junction with the preolivary sulcus marks the apparent origin of the abducent nerve (Fig. 1-9A). When viewed obliquely, the brainstem presents a triangular depression located behind and above the olive, anteromedial to the flocculus, that corresponds to the junction among the pons, the medulla, the middle and the inferior cerebellar peduncles; this area is called supraolivary fossette and is limited superiorly by the inferior aspect of the pons and the middle cerebellar peduncle, and posteriorly by the inferior cerebellar peduncle. The fossette resembles a rightangled triangle with its right angle located between the supe-
W9418-01.qxd 10/31/03 3:33 PM Page 25
Chapter 1
A
Basic Neuroanatomy for Neurointensive Care Unit
25
B
C D Figure 1-9. A, Frontal view: 1 = oculomotor nerve; 2 = crus cerebri; 3 = interpeduncular fossa; 4 = pontomesencephalic sulcus; 5 = pons; 6 = lateral pontine sulcus; 7 = pontomedullary sulcus; 8 = pyramid; 9 = flocculus; 10 = petrosal or great horizontal or cerebellopontine fissure; 11 = olive. V = trigeminal nerve; VI = abducent nerve; VII = facial nerve; VIII = vestibulocochlear nerve; IX = glossopharyngeal nerve; X = vagus nerve; XI = accessory nerve; XII = hypoglossal nerve. B, Basal view of the midbrain; the location of the fibers of the corticospinal tract in the crus cerebri is also shown. 1 = apex of the uncus; 2 = P2A (posterior cerebral artery); 3 = mamillary body; 4 = crus cerebri; 5 = substantia nigra; 6 = P2P (posterior cerebral artery); 7 = lateral mesencephalic sulcus; 8 = tegmentum of the midbrain; 9 = aqueduct; 10 = tectum of the midbrain; III = oculomotor nerve. C, Anterolateral view of the supraolivary fossette: 1 = pyramid; 2 = olive; * = supraolivary fossette. D, Midsagittal section of the posterior fossa: 1 = caudate nucleus; 2 = parieto-occipital sulcus; 3 = thalamus; 4 = fornix; 5 = anterior calcarine sulcus; 6 = massa intermedia and the hypothalamic sulcus; 7 = calcarine sulcus; 8 = superior colliculus; 9 = mamillary body; 10 = optic nerve; 11 = midbrain; 12 = lingual gyrus; 13 = pons; 14 = tentorium cerebelli; 15 = medulla; 4V = fourth ventricle. Ce = central lobule; Cu = culmen; De = declive; Fo = folium; Tu = tuber; Pi = pyramid; Uv = uvula; No = nodule; To = tonsil.
rior pole of the olive and the inferior aspect of the pons; the superior catheti corresponds to the inferior border of the pons and the middle cerebellar peduncle; the vertical catheti corresponds to the posterior border of the olive, and the hypotenuse corresponds to the inferior cerebellar peduncle. Cranial nerves VI, VII, and VIII exit from the brainstem at
the superior catheti of the supraolivary fossette; cranial nerves VII and VIII then pass above the flocculus to the internal acoustic meatus. Cranial nerves IX, X, and XI exit from the brainstem at the hypotenuse of the supraolivary fossette, and cranial nerves IX and X pass below the flocculus to the jugular foramen (Fig. 1-9C).
W9418-01.qxd 10/31/03 3:33 PM Page 26
26
Section I
Introduction
Among all structures located in the brainstem, the reticular formation deserves special consideration. The reticular formation of the brainstem is located at the tegmentum (core) of the brainstem and it modulates sensation, movement, consciousness, reflexive behavior, and the activities of ten of the 12 cranial nerves. By convention, the reticular formation is defined only for the brainstem and it contains specific groups of cells extending from the upper pons to the hypothalamus that are responsable for “activating” the cerebral cortex and thalamus, increasing wakefulness, vigilance, and the responsiveness of cortical and thalamic neurons to sensory stimuli, a state known as arousal; those cell groups constitute the ascending reticular activating system. The ascending reticular activating system reaches the cerebral cortex through two major branches at the junction of the midbrain and diencephalon. One is through the thalamus, where it activates and modulates thalamic relay nuclei as well as intralaminar and related nuclei with extensive diffuse cortical projections. The second branch is through the lateral hypothalamic area and is joined by the ascending output from the hypothalamic and basal forebrain cell groups that diffusely innervate the cerebral cortex. Damage to either branch of the ascending reticular activating system and or its projections to the cerebral cortex or bilateral damage of the cerebral cortex can impair consciousness. The reticular formation also contains the cranial nerve nuclei that are organized in longitudinal columns. As a general rule, the sensory columns are located dorsal to the sulcus limitans, and the motor columns are located ventrally to the sulcus limitans. The sulcus limitans extends along the brainstem, through the cerebral aqueduct into the diencephalon as hypothalamic sulcus. The sulcus limitans can be easily identified on the surface of the floor of the fourth ventricle.
Cerebellum The cerebellum (from the Latin for little brain) is made up primarily of white matter covered with a thin layer of gray matter (cerebellar cortex), and three pairs of deep nuclei: the fastigial, the interposed (composed by two nuclei, the globose and emboliform), and the dentate. The cerebellum plays an important role in motor function by evaluating disparities between the intended movement and the actual action, and by adjusting the operation of motor centers in the cortex and brainstem while a movement is in progress, as well as during repetition of the same movement. The cerebellum is provided with extensive information about the goals, commands, and feedback signals associated with the programming and execution of movement. There are therefore 40 times more axons projecting into the cerebellum than exit from it.45 The cerebellum is also involved in the control of the balance, eye movements, and the movements of the body and the limb.
From a morphologic viewpoint, the cerebellum is composed of three parts: a small, median portion called the vermis; and two large, lateral cerebellar hemispheres. Both the vermis and the hemispheres are divided, by fissures and sulci, into lobules. The cerebellum is connected to the brainstem through the three cerebellar peduncles, and through the brainstem, the cerebellum establishes its connections with the cerebrum and the spinal cord. However, at its central portion, the cerebellum is separated from the brainstem by the fourth ventricle. From a functional viewpoint, the cerebellum presents three distinct regions: one is the vermis, and other two regions are located in the intermediate and in the lateral parts of the cerebellar hemispheres. These three regions and the flocculonodular lobe receive different afferent inputs, project to different parts of the motor systems, and represent distinct functional subdivisions. The flocculonodular lobule or vestibulocerebellum, also called the archicerebellum, is the most primitive part of the cerebellum. It receives input directly from the primary vestibular afferents and projects to the lateral vestibular nuclei through cranial nerve VIII, and it is related to controlling eye movements and balance. The vermis and the intermediate part of the cerebellar hemisphere constitute the spinocerebellum. Both superior and the inferior vermis receive vestibular input and somatic sensory inputs from the head and proximal parts of the body. It then projects through the fastigial nucleus to cortical and brainstem regions that give rise to the medial descending systems that control proximal muscles of the body and limbs. Some specific parts of the vermis (declive, folium, tuber, and pyramid) also receive visual and auditory inputs. The intermediate part of the cerebellar hemisphere also receives somatic sensory inputs from the limbs, and then projects through the interposed nucleus to lateral corticospinal and rubrospinal systems to control the more distal muscles of the limbs. The lateral part of the cerebellar hemisphere is called the cerebrocerebellum, as it receives input exclusively from the cerebral cortex. It projects via the dentate nucleus to motor, premotor, and prefrontal areas, and is involved in planning and mental rehersal of complex motor actions and in the conscious assessment of movement errors. All the cerebellar output comes from the deep nuclei (fastigial, globose, emboliform, and dentate) and from the flocculonodular lobule. The superior cerebellar peduncle contains most of the cerebellar efferent projections. The cerebellum presents three surfaces: petrosal, tentorial, and the suboccipital surfaces. The petrosal surface of cerebellum is related anteriorly to the petrous part of the temporal bone: the tentorial surface is related superiorly to the tentorium cerebelli and inferiorly to the upper part of the roof of the fourth ventricle; the suboccipital surface in its anatomic position is related inferiorly to the squamosal part of the occipital bone, and anteriorly to the inferior part of the roof of the fourth ventricle. Because the fourth
W9418-01.qxd 10/31/03 3:33 PM Page 27
Chapter 1
ventricle and the cerebellum are intimately related, their anatomy will be considered together. The fourth ventricle is often described as a tent-shaped midline structure surrounded mainly by the vermian components of the cerebellum (Fig. 1-9D).46,47 A normal tent has a roof that is divided into two halves, a floor and two lateral walls; however, the actual overall shape of the fourth ventricle resembles that of a turned-over tent with its base facing forward and two open lateral walls (Fig. 1-10A): the floor is represented by the pons and medulla; the superior cerebellar peduncles, the superior medullary velum, and the adjacent lingula constitute the superior part of the roof of the fourth ventricle; the inferior part of the roof is composed by the inferior medullary velum, the tela choroidea, the choroid plexus, the uvula, and the nodule; the two open lateral walls of the fourth ventricle are open corridors represented by lateral recesses that communicate the fourth ventricle with the cerebellopontine angle.
Frontal View: Petrosal Surface of the Cerebellum and the Fourth Ventricle The two halves of the petrosal surface of the cerebellum are separated because of the interposition of the brainstem; each half of the petrosal surface is intersected by the great horizontal fissure that circumscribes the cerebellum; at the petrosal surface the great horizontal fissure is called the petrosal or cerebellopontine fissure. The petrosal fissure runs from lateral to medial and it presents, as its posterior wall, the white matter of the cerebellum. However, at the level of the flocculus, the petrosal fissure bifurcates into a larger superior portion and a smaller inferior portion. The superior portion or superior limb is the suprafloccular portion of the petrosal fissure. The inferior portion is the infrafloccular portion of the posterolateral fissure, which separates the flocculo-nodule lobule from the rest of the cerebellum, and communicates with the cerebellomedullary fissure at the cerebellopontine angle. The folia that constitute the upper half of the petrosal surface of the cerebellum are those folia of the tentorial surface that have folded over the middle cerebellar peduncle and over the core of the cerebellum. These folia are, respectively, the wing of the central lobule, the quadrangular, the simple, and the superior semilunar lobules. The folia that constitute the lower half of the petrosal surface of the cerebellum originate from the suboccipital surface of the cerebellum that have folded over the inferior cerebellar peduncle and over the core of the cerebellum, and correspond to the inferior semilunar and biventral lobules. The cerebellopontine (CP) angle is a triangular area on the petrosal surface of the cerebellum comprising the flocculus laterally, the supraflocular portion of the great horizontal fissure and the emergence of the trigeminal nerve superiorly, the infraflocular portion of the posterolateral
Basic Neuroanatomy for Neurointensive Care Unit
27
fissure and the emergence of the glossopharyngeal nerve inferiorly, and the supraolivary fossette medially. When the brainstem, which is the floor of the fourth ventricle is removed from the cerebellum (by sectioning the three cerebellar peduncles), the cavity and the two parts of the roof of the fourth ventricle are exposed. The upper half of the roof of the fourth ventricle comprises neural elements: the superior cerebellar peduncles, the superior medullary velum, and the lingula. The lingula can be visualized through transparency behind the superior medullary velum. The lower half of the roof comprises nonneural elements, and presents a horizontal portion and a vertical portion. The horizontal portion is constituted by the inferior medullary velum, which covers the nodule at the midline and the superior pole of the tonsils laterally; the vertical portion is constituted by the tela choroidea and the choroid plexus, covering the anterior aspect of the nodule, uvula, and partly the tonsils. The two portions of the lower half of the roof of the fourth ventricle unite at the telovelar junction and continue laterally as the floor of the lateral recess. At the midline, the upper and the lower halves of the roof of the fourth ventricle converge at the fastigium (Fig. 1-10B). The cavity of the fourth ventricle communicates with the CP angle cistern through the lateral recess of the fourth ventricle. The lateral recess is the lateral extension of the fourth ventricle and connects the fourth ventricle to the CP angle.48 It is directed from medial to laterally, slightly from superior to inferior and from posterior to anterior, forming an angle of approximately 45 degrees with the sagittal plane. The lateral recess presents an anterior, superior, and posterior wall; and a floor. The anterior and superior walls are constituted by the inferior cerebellar peduncle as it runs upward and then turns backward toward the white matter of the cerebellum. The floor of the lateral recess is constituted by the tela choroidea anteriorly, the choroid plexus in the middle, and the inferior medullary velum posteriorly; at the foramen of Luschka, the inferior medullary velum becomes thicker and is called the peduncle of the flocculus. The peduncle of the flocculus constitutes the posterior wall of the foramen of Luschka, and connects the nodule to the flocculus (flocculonodular lobule). The morphology of the choroid plexus of the fourth ventricle resembles the letter “T” with two vertical bars. The horizontal part of the choroid plexus that starts from the fourth ventricle and protrudes into the CP angle resembles the horns of a bull. The vertical part and the proximal half of the horizontal part of the choroid plexus of the fourth ventricle is usually supplied by PICA; the lateral half of the horizontal part and the choroid plexus located at the cerebellopontine angle are generally supplied by the AICA (Fig. 1-10A and B).49 The inferior medullary velum separates the tonsil inferiorly from the superolateral recess superiorly. The tonsils are two riniform structures that are hemispheric components of the uvula and are attached to the cerebellum through the peduncles of the tonsil, located at the supero-
W9418-01.qxd 10/31/03 3:33 PM Page 28
28
Section I
Introduction
F
F
A
B
D C Figure 1-10. A, A normal tent (left). The actual shape of the fourth ventricle resembles a turned over tent with its floor facing forward (right). F = floor. B, Frontal view: the pons and the medulla have been removed to display the roof of the fourth ventricle. 1 = culmen; 2 = central lobule; 3 = superior cerebellar peduncle; 4 = superior medullary velum; 5 = middle cerebellar peduncle; 6 = nodule; 7 = choroid plexus; 8 = tela choroidea. C, Right supero-anterolateral view of the fourth ventricle and the lateral recess. 1 = culmen; 2 = central lobule; 3 = superior cerebellar peduncle; 4 = superior medullary velum; 5 = middle cerebellar peduncle; 6 = nodule; 7 = superior pole of the tonsil (covered by the inferior medullary velum); 8 = peduncle of the flocculus; 9 = choroid plexus (exiting from the foramen of Luschka); 10 = choroid plexus; 11 = foramen of Magendie. D, Frontal view: the choroid plexus, the tela choroidea, and the right tonsil have been removed. 1 = superolateral recess; 2 = nodule; 3 = furrowed band of Reil; 4 = uvula; 5 = left tonsil; 6 = copula pyramidis; 7 = pyramid.
lateral aspect of each tonsil. The superior, the medial, the anterior, the posterior, and most of the lateral surfaces of the tonsils are free and can be separated easily from the adjacent structures. The tonsils, along with surrounding neural structures, determine important spaces: between its superior pole and the inferior medullary velum is the supratonsillar space; between the medial surfaces of the two tonsils is the vallecula; between the anterior surface of the tonsil and the medulla is the cerebellomedullary fissure; between the posterior surface of the tonsil and the adjacent vermis is the
retrotonsilar space where the inferior vermian veins originate (Fig. 1-10D).
Superior View: Tentorial Surface of the Cerebellum and the Fourth Ventricle The tentorial surface faces the tentorium and presents two cerebellar incisurae and three margins. The cerebellar
W9418-01.qxd 10/31/03 3:33 PM Page 29
Chapter 1
incisurae are the anterior and posterior cerebellar incisurae; the brainstem fits into the anterior cerebellar incisura and the falx cerebelli fits into the posterior cerebellar incisura. The margins are the anterosuperior margin, which is the posterior wall of the cerebellomesencephalic fissure; the anterolateral margin, which separates the tentorial from the petrosal surfaces of the cerebellum; and the posterolateral margin, which separates the tentorial from the suboccipital surfaces. The folia of the tentorial surface are represented by the superior vermis and its hemispheric counterpart; from anterior to posterior: lingula (without hemispheric correspondent), central lobule (wing of the central lobule), culmen (quadrangular lobule), declive (simple lobule), and folium (part of the superior semilunar lobule), being the primary fissure located between the quadrangular and simple lobules, and the most prominent one, the postclival fissure located between the simple and the superior semilunar lobules. Most of the lobules of the cerebellum occupy more than one surface (except the lingula, the tonsil, and the nodule, which occupy one surface; and the superior semilunar lobule, which occupies three surfaces of the cerebellum). The tentorial surface presents the cerebelomesencephalic or the precentral cerebellar fissure. This fissure is located between the cerebellum posteriorly and the midbrain anteriorly. Posteriorly in the midline, the cerebellomesencephalic fissure is bounded by the anterior part of the culmen above and the central lobule below; posterolaterally, it is limited by the anterior surface of the quadrangular lobule above and the wing of the central lobule below. Anteriorly it is limited from the midline to laterally by the lingula, and the superior and the middle cerebellar peduncles. Among the cerebellar nuclei, the dentate nucleus is the most laterally located and the largest one. Because the majority of the fibers that constitute the superior cerebellar peduncle arise from the dentate nucleus, this nucleus is located at the posterior projection of the superior cerebellar peduncle. The lateral limit of the dentate nucleus extends from 0.5 to 2.0 cm from the midline,50 but the lateral limit of the dentate nucleus can be considered as the posterior continuation of the interpeduncular sulcus. Located between the postclival sulcus and the bottom of the cerebellomesencephalic fissure, the dentate nucleus can be considered as the roof of the superolateral recess of the fourth ventricle (the space in the fourth ventricle lateral to the nodule and above the superior pole of the tonsils) (Fig. 1-11A). The interpeduncular or interbrachial sulcus, which separates the superior from the middle cerebellar peduncles, ascends from the bottom of the cerebellomesencephalic fissure toward the lateral aspect of the pons, where it is joined by the ponto-mesencephalic sulcus to proceed superiorly as the lateral mesencephalic sulcus to the medial geniculate body; the lateral mesencephalic sulcus separates the crus cerebri anteriorly from the tegmentum posteriorly (Fig. 1-11B).
Basic Neuroanatomy for Neurointensive Care Unit
29
Inferior View: Suboccipital Surface of the Cerebellum and the Fourth Ventricle This area of the cerebellum is located below the transverse and the sigmoid sinuses, and its surface faces inferiorly, almost parallel to the ground; therefore, for a better visualization of this surface either in surgery or for anatomic studies, the head or the cerebellum has to be bent forward; consequently, the cerebellum will not be in its anatomic position. The suboccipital surface presents the posterior cerebellar incisura, and the prominent vermohemispheric or paravermian fissure, which separates the inferior vermis from the cerebellar hemisphere. The components of the inferior vermis and its hemispheric correspondents are folium (superior semilunar lobule), tuber (inferior semilunar lobule), pyramid (biventral lobule), uvula (tonsil), and nodule (flocculus). In the anatomic position, the most inferior part of the inferior vermis is the pyramid. The most prominent fissure on the suboccipital surface is the great horizontal fissure, which is a circumferential fissure that begins in the posterior cerebellar notch between the folium and the tuber and runs forward and slightly downward on the suboccipital surface, between the superior and the inferior semilunar lobules, and then onto the petrosal surface. The secondary fissure is the one located between the tonsils and the biventral lobule (Fig. 1-11C). After the removal of the tonsils in the surgical position, the inferior portion of the roof of the fourth ventricle, namely tela choroidea and the inferior medullary velum, come to the view. After the removal of the inferior portion of the roof of the fourth ventricle, the floor of the fourth ventricle comes to the view (Fig. 1-11D). The floor of the fourth ventricle has a rhomboid shape and presents a strip between the lower margin of the cerebellar peduncles and the site of attachment of the tela choroidea; this strip is called the junctional part, and is characterized by the striae medullary that extends into the lateral recesses. The striae medullary are external arcuate fibers of the corticopontocerebellar afferents coming from the arcuate nuclei, located at the pyramid of the medulla that enter the cerebellum through the inferior cerebellar peduncle. The junctional part divides the floor of the fourth ventricle into two unequal triangles; the superior and larger one with its apex directed toward the aqueduct is the pontine part, and the inferior and smaller one with its apex directed toward the obex is the medullary part of the floor. These three parts of the floor are also divided longitudinally into two symmetrical halves by the median sulcus. The sulcus limitans, another longitudinal sulcus, divides each half of the floor into a raised median strip called the median eminence, and a lateral strip called area vestibular. The superior or pontine part is characterized by two rounded prominences called facial colliculi located on the median eminence, one on each side of the median sulcus. The facial colliculi are limited laterally by the
W9418-01.qxd 10/31/03 3:33 PM Page 30
30
A
Section I
Introduction
B
C D Figure 1-11. A, Superior view of the cerebellum: 1 = quadrigeminal plate; 2 = wing of the central lobule; 3 = AICA and the internal acoustic meatus; 4 = superior petrosal sinus; 5 = quadrangular lobule; 6 = interpeduncular sulcus and the precentral branches of the superior cerebellar artery to the dentate nucleus; 7 = middle cerebellar peduncle; 8 = primary fissure; 9 = simple lobule; 10 = paramedian nuclei of the cerebellum; 11 = straight sinus; 12 = superior semilunar lobule; 13 = postclival fissure; 14 = transverse sinus; IV = trochlear nerve; 4V = fourth ventricle; DN = dentate nucleus. B, Posterosuperior view of the brainstem: 1 = right frontal horn; 2 = thalamus (floor of the body of the lateral ventricle); 3 = fornix; 4 = choroidal fissure; 5 = pineal gland; 6 = thalamus (anterior wall of the atrium); 7 = thalamus (roof of the wing of the ambient cistern); 8 = quadrigeminal plate; 9 = brachium of the superior and the inferior colliculi, and the lateral mesencephalic sulcus and the crus cerebri; 10 = branches of the superior cerebellar artery; 11 = superior cerebellar peduncle and the interpeduncular sulcus; 12 = middle cerebellar peduncle; 13 = floor of the fourth ventricle; 14 = superior pole of the left tonsil; 15 = nodule. C, Suboccipital view of the cerebellum: 1 = great horizontal fissure; 2 = inferior semilunar lobule; 3 = posterior cerebellar incisura; 4 = tuber; 5 = prepyramidal fissure; 6 = pyramid; 7 = biventral lobule; 8 = uvula; 9 = tonsil; 10 = flocculus; 4V = fourth ventricle. D, Suboccipital view: part of the biventral lobule and both tonsils have been removed to display the inferior portion of the roof of the fourth ventricle and the floor of the lateral recess. 1 = pyramid; 2 = peduncle of the tonsil; 3 = uvula; 4 = inferior medullary velum; 5 = tela choroidea; 6 = peduncle of the flocculus; 7 = rhomboid lip; 8 = flocculus; 9 = spinal cord; 4V = fourth ventricle.
W9418-01.qxd 10/31/03 3:33 PM Page 31
Chapter 1
superior fovea, a dimple formed by the sulcus limitans. The inferior or medullary part presents the configuration of a feather, or pen nib, called calamus scriptorius, and it is characterized by three triangular areas overlying the hypoglossal and vagal nuclei (hypoglossal and vagal trigones), and the area postrema; just lateral to the hypoglossal trigone, the limitans sulcus presents another dimple called inferior fovea. At the junctional part, the sulcus limitans is discontinuous (Fig. 1-12A). As a general rule, the motor nuclei of the cranial nerves are located medially to the sulcus limitans, and the sensory nuclei are located laterally to the sulcus limitans. Their respective locations are shown in Figure 1-12B.
The Cranial Nerves Olfactory Nerve (Cranial Nerve I) The olfactory nerve is concerned with the sense of smell. The olfactory nerves are located in the olfactory epithelium, which covers most of the superior nasal conchae and the opposed surfaces of the nasal septum. The approximately 20 olfactory nerves traverse the ethmoidal cribriform plate to end in the “glomeruli”of the homolateral olfactory bulb. The input then goes along the olfactory tract up to the olfactory trigone, where the olfactory tract divides into two striae: medial and lateral olfactory striae. The medial olfactory striae terminate in the paraolfactory area, subcallosal gyrus, and in the inferior part of the cingulate gyrus. The lateral olfactory striae terminate in the uncus, anterior portion of the hippocampus, and the amygdaloid nucleus. The tertiary olfactory cortex occupies the posterior portion of the orbital gyri (medial orbital, posterior orbital, and the lateral orbital gyri). Optic Nerve (Cranial Nerve II) Two important aspects regarding the optic nerve are the vision and the pupillary light reflex. The central visual pathway is best summarized in Figure 1-12C. The main structures related to the vision and its pathways are lens, retina, optic nerve and chiasm, lateral geniculate body, pretectum area and the superior colliculi of the midbrain, optic radiation, and the primary visual cortex. The surface of the retina is divided in two parts with respect to the midline (where the fovea is located): the nasal and temporal hemiretinae. The former is responsible for the temporal side of the visual field, and the latter is responsible for the nasal visual field. The fovea is the area of the retina with the highest density of ganglion cells, and is responsible for the central region of the visual field. The superimposed area of the visual fields of both eyes is called the binocular zone, and the lateral part, or nonsuperimposed areas, of the visual field are called the monoc-
Basic Neuroanatomy for Neurointensive Care Unit
31
ular zone. There is a normal blind spot in our visual field that corresponds to the optic disc, which is the exit site of the ganglion cell axons, and contains no photoreceptors; because the optic disc is located medially to the fovea, the natural blind spot of each eye is located in the temporal visual field of that eye. When an image passes through the eye to project onto the retina, the lens inverts the visual image: the upper half of the visual field projects onto the inferior half of the retina, while the lower half of the visual field projects onto the superior half of the retina. The visual information is then carried by the axons from the ganglion cells that exit the retina from the optic disc and, at the optic chiasm, the fibers from the nasal half of each retina cross to the opposite side of the brain. Therefore axons from the right half of each retina project in the right optic tract, which carries the complete representation of the left hemifield of vision. From the optic tract, the visual information will project to three subcortical regions: the lateral geniculate body (thalamus), the pretectum, and the superior colliculus (midbrain). Ninety percent of the retinal axons terminate in the lateral geniculate body. From the lateral geniculate body, the visual information is projected to the visual cortex via optic radiation. The optic radiation is a bundle of fibers that extends from the lateral geniculate body to the visual area in the occipital lobe. According to the direction of its fibers, the optic radiation may be divided in three parts: anterior, middle, and posterior. In the anterior part, the fibers initially take an anterior direction along the roof of the temporal horn, usually reaching as far anteriorly as the tip of the temporal horn and then loop backward in the lateral and inferior aspects of the atrium and occipital horn to end in the lower lip of the calcarine fissure; the anterior loop is called Meyer’s loop. The anterior part represents the upper quadrants of the visual field. In the middle part, the fibers take a lateral direction initially, coursing along the roof of the temporal horn, and then proceed posteriorly along the lateral wall of the atrium and the occipital horn; the middle part contains the macular fibers. The fibers of the posterior part course directly posteriorly along the lateral wall of the atrium and the occipital horn to end in the upper lip of the calcarine fissure; these fibers are responsible for the lower quadrants of the visual field (Fig. 1-12C).51–53 Each half of the visual field is represented in the contralateral primary visual cortex that is located along the lips of the posterior calcarine fissure, mainly at the medial surface. The visual cortex might extend into the lateral surface of the occipital lobe. The upper visual fields are mapped below the calcarine fissure, and the lower fields above it. Most of the visual cortex is devoted to representation of the fovea and region around the fovea. Pupillary Reflexes Light shining in one eye causes constriction of the pupil in that eye (the direct response) and in the other eye
W9418-01.qxd 10/31/03 3:33 PM Page 32
Section I
32
Introduction
A
B
C
D
Figure 1-12. A, Posterior view of the floor of the fourth ventricle: 1 = superior cerebellar peduncle; 2 = median sulcus and the median eminence; 3 = sulcus limitans; 4 = middle cerebellar peduncle; 5 = facial colliculus; 6 = superior fovea and the vestibular area; 7 = inferior cerebellar peduncle; 8 = striae medullary; 9 = hypoglossal trigone; 10 = inferior fovea; 11 = vagal trigone; 12 = area postrema; 13 = obex; IV = fourth nerve. B, Posterior view: the motor nuclei of the cranial nerves are located medially to the sulcus limitans and the sensory nuclei are located laterally to the sulcus limitans. C, Superior view of the visual pathways and the pupillary reflex: Cilliary gg = cilliary ganglion; III = oculomotor nerve; E.W. = Edinger-Westphal nucleus. D, Superolateral view of the right cavernous sinus: 1 = cerebellomesencephalic segment of the superior cerebellar artery; 2 = lower midbrain; 3 = basilar artery; 4 = pituitary stalk; 5 = anterior pontomesencephalic segment of the superior cerebellar artery; 6 = internal carotid artery; 7 = tentorium edge; 8 = lateral pontomesencephalic segment of the superior cerebellar artery; 9 = cerebellum; 10 = superior petrosal sinus; 11 = facial and the superior vestibular nerves; 12 = greater petrosal nerve; 13 = middle meningeal artery; 14 = geniculate ganglion and the superior semicircular canal; V1, V2, V3 = ophthalmic, maxillary, and the mandibular divisions of the trigeminal nerve; GG = gasserian ganglion.
W9418-01.qxd 10/31/03 3:33 PM Page 33
Chapter 1
(consensual response). These reflexes are mediated by retinal ganglion cells that project from the optic tract to the pretectal area of the midbrain. The cells in the pretectal area project bilaterallly to preganglionic parasympathetic neurons in the Edinger-Westphal nucleus. From the Edinger-Westphal nuclei, the preganglionic parasympathetic fibers travel in the oculomotor nerve (inferior ramus) to the ciliary ganglion to innervate the pupillary sphincter that constricts the pupil (Fig. 1-12C). The retinal projection to the superior colliculus controls the saccadic (visually guided) eye movement. Oculomotor Nerve (Cranial Nerve III) The oculomotor nerve innervates striated muscles of the eyelid and all extraocular muscles except the superior oblique and lateral rectus muscles. Autonomic function mediates pupillary constriction and accommodation of the lens for near vision. The oculomotor nerve emerges from the interpeduncular fossa and constitutes the lateral limit of the interpeduncular cistern and passes between the posterior cerebral (above) and superior cerebellar (below) arteries; it then runs forward and laterally, below the apex of the uncus toward the cavernous sinus. During its subarachnoid course, the parasympathetic pupillary fibers lie peripherally in the dorsomedial part of the nerve. An uncal herniation will compress the oculomotor nerve from its lateral aspect, causing the dilation of the pupil. The nerve then perforates the posterior portion of the roof of the cavernous sinus, through the center of a triangular area called oculomotor trigone to course on the lateral wall of the cavernous sinus, where it lies above the trochlear nerve (Fig. 1-12D). It divides into superior and inferior rami, which enter the orbit by the superior orbital fissure within the annulus of Zinn (annulus tendineus communis), with the nasociliary and abducent nerves between them. The smaller superior ramus supplies the superior rectus and the levator palpebrae muscles; the inferior ramus divides into three branches: one passes under the optic nerve to the medial rectus muscle, another to the inferior rectus, the third and longest passes forward between the inferior rectus and lateral rectus muscles to the inferior oblique muscle. From the nerve to obliquus inferior muscle, a short branch passes to the lower part of the ciliary ganglion as its motor, parasympathetic root to innervate the sphincter pupillae and ciliaris muscles (Fig. 1-13A). The sympathetic and sensory fibers merely pass through the ciliary ganglion; however, the parasympathetic root, derived from the branch of the oculomotor nerve to the inferior oblique, consists of preganglionic fibers from the Edinger-Westphal nucleus, which relay in the ganglion, and have their postganglionic fibers traveling in the short ciliary nerves to the sphincter pupillae and ciliaris muscle. The sympathetic root consists of postganglionic fibers from the superior cervical ganglion, which traverse the ganglion without
Basic Neuroanatomy for Neurointensive Care Unit
33
synapsing, to emerge into the short ciliary nerve. They are distributed to blood vessels of the eyeball but may include axons supplying the dilator pupillae when these do not follow their usual course in the ophthalmic, nasociliary, and long ciliary nerves. The sensory root is a ramus communicans of the nasociliary nerve, containing sensory fibers from the eyeball that reach the ganglion in short ciliary nerves and traverse it without synapsing. The ramus leaves the ganglion posteriorly and runs back to join the nasociliary nerve near its orbital entry.
Trochlear Nerve (Cranial Nerve IV) The trochlear nerve is the thinnest and the only cranial nerve to emerge dorsally from the brainstem. Its nucleus supplies the contralateral superior oblique muscle. The trochlear nucleus lies in the grey matter in the floor of the cerebral aqueduct, at the level of the upper part of the inferior colliculus. After leaving its nucleus, the trochlear nerve passes laterally through the tegmentum and then curves dorsocaudally around the aqueduct into the superior medullary velum; then it decussates with its mate, crossing the midline to emerge as one or more rootlets below the inferior colliculus. The nerve crosses to the lateral aspect of the superior cerebellar peduncle and winds around the cerebral peduncle just above the pons, between the posterior cerebral and superior cerebellar arteries. It pierces the dura below the free edge of the tentorium cerebelli, just behind the posterior clinoid process, and then passes forward in the lateral wall of the cavernous sinus, inferior to the oculomotor nerve and above cranial nerve V1. Near the anterior end of the cavernous sinus it crosses the oculomotor nerve, entering the orbit via superior orbital fissure, above the annulus of Zinn and medial to the frontal nerve. In the orbit it inclines medially, above the origin of the levator palpebrae muscle to enter the orbital surface of the superior oblique muscle (Figs. 1-12D; 1-13A and B). Trigeminal Nerve (Cranial Nerve V) The sensory function mediates sensation from the skin of the external ear canal and taste from the anterior two thirds of the tongue. The motor function innervates muscles of mastication, plus tensor tympani, tensor veli palatini, mylohyoid muscle, and anterior belly of digastric muscle (see Figs. 1-12D; 1-13A and B). Ophthalmic Nerve (Cranial Nerve V, Segment 1) The ophthalmic nerve is the smallest division of the trigeminal nerve. It is totally sensory supplying the eyeball, lacrimal gland and conjuntiva, part of the nasal mucosa and the skin of the nose, eyelids, forehead, and part of the scalp. It passes in the lateral wall of the cavernous sinus below the
W9418-01.qxd 10/31/03 3:33 PM Page 34
34
Section I
Introduction
A
B
C
D
Figure 1-13. A, Anterolateral view of the right orbit and the annulus of Zinn: 1 = basilar artery; 2 = nasociliary nerve; 3 = superior ophthalmic vein; 4 = ophthalmic artery and the ciliary ganglion; 5 = frontal nerve; SR = superior rectus muscle; Sup. div. = superior division; LP = levator palpebrae muscle; IR = inferior rectus muscle; LR = lateral rectus muscle; Inf. div. = inferior division; IO = inferior oblique muscle. B, Superior view of the right orbit: 1 = trochlea; 2 = supraorbital nerve; 3 = supratrochlear nerve; 4 = superior oblique muscle; 5 = anterior ethmoidal nerve and artery; 6 = frontal nerve; 7 = nasociliary; 8 = medial rectus muscle; 9 = superior rectus muscle; 10 = superior ophthalmic vein; II = optic nerve; III = oculomotor nerve; IV = trochlear nerve; V1 = ophthalmic division of the trigeminal nerve. C, Anterior view of the veins of the posterior fossa: 1 = crus cerebri; 2 = transverse pontine vein; 3 = vein of the cerebellopontine fissure; 4 = vein of the middle cerebellar peduncle; 5 = vein of the cerebellomedullary fissure or vein of the lateral recess of the fourth ventricle; FL = flocculus. D, Suboccipital view of the veins of the posterior fossa: 1 = inferior hemispheric vein; 2 = inferior vermian vein; 3 = inferior hemispheric vein; 4 = retrotonsillar veins; 5 = superior cerebellar peduncle; 6 = vein of the lateral recess of the fourth ventricle; 7 = tela choroidea; 8 = flocculus; VA = vertebral artery.
W9418-01.qxd 10/31/03 3:33 PM Page 35
Chapter 1
oculomotor and trochlear nerves. Just before entering the orbit via the superior orbital fissure, it divides into lacrimal, frontal, and nasociliary branches. The lacrimal nerve is the smallest of the main ophthalmic branches, and enters the orbit through the lateral part of the superior orbital fissure and supplies the lacrimal gland and the adjoining conjunctiva. The frontal nerve is the largest branch of the ophthalmic division and enters the orbit via the superior orbital fissure above the annular tendon of Zinn to divide into a small supratrochlear and a large supraorbital branch. The nasociliary nerve is more deeply located in the orbit, which it enters through the annular tendon, lying between the two rami of the oculomotor nerve. It crosses the optic nerve with the ophthalmic artery and runs obliquely below the superior rectus and superior oblique muscles to the medial orbital wall. Here, as the anterior ethmoidal nerve, it traverses the anterior ethmoidal foramen to enter the cribriform plate and then into the nasal cavity. The nasociliary nerve connects with the ciliary ganglion and has long ciliary, infratrochlear, and posterior ethmoidal branches. The posterior ethmoidal nerve leaves the orbit via the posterior ethmoidal foramen and supplies the ethmoidal and sphenoidal sinuses.
Basic Neuroanatomy for Neurointensive Care Unit
35
temporal region; part of the auricle, including the external meatus and tympanum; the lower lip; the lower part of the face; the muscles of mastication; the mucosa of the anterior two thirds of the tongue; and the mucosa of the floor of the oral cavity. Cranial nerve V3 has a large sensory and a small motor root. The sensory root exits from the trigeminal ganglion and leaves the skull base through the foramen ovale in the floor of the middle fossa. The small motor root passes under the ganglion and unites with the sensory root just outside the skull. The nerve immediately passes between the tensor veli palatini muscle and the lateral pterygoid muscle. Just beyond this junction, a meningeal branch and the nerve to the medial pterygoid muscle leaves the medial side of the nerve, which then divides into a small anterior and large posterior trunk. The anterior trunk presents the following branches: sensory (buccal nerve), and motor (masseteric, deep temporal, and nerve to lateral pterygoid). The posterior trunk presents the auriculotemporal, lingual, and the inferior alveolar nerves. Abducent Nerve (Cranial Nerve VI)
The maxillary nerve is totally sensory and, after leaving the trigeminal ganglion, it passes horizontally forward and medially in the lateral wall of the cavernous sinus to traverse the foramen rotundum. Through the foramen rotundum it courses directly into the posterior wall of the pterygopalatine fossa, where it sends off two large ganglionic branches containing fibers to the nose, palate, and pharynx. It then inclines laterally on the posterior surface of the orbital process of the palatine bone and on the upper part of the posterior surface of the maxilla in the inferior orbital fissure (which is continuous posteriorly with the pterygopalatine fossa) outside the orbital periosteum giving off its zygomatic branch, and then posterior superior alveolar branches. About halfway between the orbital apex and the orbital rim the nerve turns medially to enter the infraorbital canal as the infraorbital nerve through which it is conveyed progressively further below the orbital floor, in the roof of the maxillary antrum, until it emerges onto the face through the infraorbital foramen approximately 1 cm below the inferior orbital rim (in line with the pupil). The branches of the maxillary nerve are: in the cranial cavity, meningeal; in the pterygopalatine fossa, ganglionic, zygomatic, and posterior superior alveolar; in the infraorbital canal, middle superior alveolar and anterior superior alveolar; and on the face, palpebral, nasal, and superior labial.
This innervates the lateral rectus muscle (see Figs. 1-12D and 1-13A). The fibers arise from a small nucleus in the superior part of the floor of the fourth ventricle near the midline and beneath the facial colliculus. They descend ventrally through the pons, emerging in the junction of the pontomedullary sulcus and the anterolateral sulcus. After leaving the brainstem, the abducent nerve ascends anterolaterally through the prepontine cistern along the clivus, usually dorsal to the AICA. It pierces the dura mater lateral to the dorsum sellae and bends sharply forward over the superior border of the apex of the petrous temporal bone, medial to the trigeminal nerve. Here it enters a fibro-osseous canal (Dorello’s canal) formed by the apex of the petrous temporal bone and the petrosphenoidal ligament. The latter is a fibrous band connecting the lateral margin of the dorsum sellae to the upper border of the petrous temporal bone near its medial end. The nerve then traverses the cavernous sinus, lying laterally to the intracavernous internal carotid artery. Among those cranial nerves crossing through the cavernous sinus, the abducent nerve is the only one that traverses inside the cavernous sinus. All the rest course on the lateral wall of the cavernous sinus. Cranial nerve VI enters the orbital cavity via the medial end of the superior orbital fissure within the annulus of Zinn, between the superior and the inferior divisions of cranial nerve III and inferolateral to the nasociliary nerve, to finally enter the lateral rectus muscle through its ocular surface.
Mandibular Nerve (Cranial Nerve V, Segment 3)
Facial Nerve (Cranial Nerve VII)
The mandibular nerve is the largest trigeminal division. It supplies the teeth and gums of the mandible; the skin in the
The sensory function mediates sensation from the skin of the external ear canal and taste from the anterior two thirds of
Maxillary Nerve (Cranial Nerve V, Segment 2)
W9418-01.qxd 10/31/03 3:33 PM Page 36
36
Section I
Introduction
the tongue. The motor function is responsible for innervating muscles of facial expression, plus stylohyoid, stapedius, and posterior belly of the digastric muscle. The autonomic function mediates all salivary glands except the parotid gland, as well as lacrimal glands and cerebral vasculare. The facial nerve has a motor root and a sensory root (nervus intermedius). The nervus intermedius gains its name from its position between the facial and vestibulocochlear nerves at the CP angle. The two roots arise from the pons at the superior catheti of the supraolivary fossette, slightly medial and anterior to the vestibulocochlear nerves. The facial nerve is divided into five segments: (1) cisternal, (2) labyrinthine, (3) tympanic or horizontal, (4) mastoid, and (5) extratemporal. The cisternal segment extends from its origin from the brainstem to the internal acoustic meatus. From their emergence from the brainstem, the two roots pass anterolaterally with the vestibulocochlear nerve toward the internal acoustic meatus; in this location, the motor root is in an anterosuperior groove on the vestibulocolchlear nerve, with the sensory root between them. At the lateral end of the internal acoustic meatus, the Bill’s bar and the transverse crest divide the meatus into quadrants: the facial nerve is located at the anterosuperior quadrant, the cochlear nerve at the anteroinferior quadrant, the superior vestibular nerve at the posterosuperior quandrant, and the inferior vestibular nerve at the posteroinferior quandrant. In its intratemporal course, its bony canal is narrowest at the meatal foramen. It is accompanied by the labyrinthine branch of the AICA or from the basilar artery. The labyrinthine segment of the nerve runs across the axis of the petrous pyramid to the geniculate ganglion. It is just medial to the tip of the cochleariform process. After the geniculate ganglion, the facial nerve turns 130 degrees and forms the tympanic segment, which is approximately 10 to 12 mm long, and passes lateral to the vestibule, above the oval window, and below the horizontal semicircular canal. In the medial wall of the middle ear, it slopes down from anterior to posterior forming an angle of approximately 10 degrees with the horizontal semicircular canal. It lies medial to the malleus head anteriorly, the incudomalleolar joint; incus and attic posteriorly. After the horizontal or tympanic segment, the facial nerve angles inferiorly within the mastoid to constitute the mastoid segment, and here it gives off the nerve to the stapedius muscle. Aproximately 5 mm before exiting from the stylomastoid foramen, the facial nerve gives off the corda tympani nerve. After exiting from the stylomastoid foramen, the nerve runs forward in the parotid gland, lateral to the styloid process, retromandibular vein, and external carotid artery and divides behind the neck of the mandible into branches that pierce the anteromedial surface of the parotid gland to innervate the face through five branches. The branches of the facial nerve are: from geniculate ganglion, greater petrosal
nerve; from facial canal, auricular branch to vagus, nerve to stapedius, and chorda tympani nerve; at the exit from the stylomastoid foramen, posterior auricular branch and branches to posterior belly of digastric and stylohyoid muscles; on the face, temporal, zygomatic, buccal, marginal mandibular, and cervical branches. Vestibulocochlear Nerve (Cranial Nerve VIII) This nerve is responsible for hearing and sense of motion (angular and linear acceleration). The vestibulocochlear nerve exits from the superior catheti of the supraolivary fossette and courses along with the facial and the intermedius nerves toward the internal acoustic meatus. At the fundus of the internal acoustic meatus, the vestibular part of the vestibulocochlear nerve expands to form the vestibular ganglion and then divides into superior vestibular and inferior vestibular nerves. The superior and inferior vestibular nerves are connected to the ampulle of the semicircular canals and the maculae of the saccule and utricle. Glossopharyngeal Nerve (Cranial Nerve IX) Cranial nerve IX is both motor and sensory, supplying motor fibers to the stylopharyngeus, parasympathetic secretomotor fibers to the parotid gland via the otic ganglion, and sensory fibers to the tympanic cavity, Eustachian tube, tonsils, nasopharynx, uvula, inferior surface of the soft palate and posterior third of the tongue; it is also the gustatory nerve for this part of the tongue. Cranial nerve IX is the uppermost single nerve that exits from the hypotenuse of the supraolivary fossette, anteriorly to the choroid plexus of the foramen of Luschka, and courses anterolaterally toward the jugular foramen. After exiting the jugular foramen, the glossopharyngeal nerve passes forward between the internal jugular vein and internal carotid to reach the posterior border of the stylopharyngeus. It curves forward on the stylopharyngeus and either pierces the lower fibers of the superior phryngeal constrictor or passes between it and the middle constrictor to be distributed to the tonsil, the mucosa of the pharynx, and the posterior part of the tongue. Vagus Nerve (Cranial Nerve X) The vagus nerve is also mixed in function. Its sensory function mediates sensation from the posterior pharynx, visceral sensation from pharynx, larynx, thoracic, and abdominal organs; and taste from posterior tongue and oral cavity. Its motor function innervates striated muscles of larynx and pharynx. The autonomic function innervates smooth muscle and glands of gastrointestinal, pulmonary, and cardiovascular systems in neck, thorax, and abdomen. The vagus nerve emerges as eight or ten rootlets from the hypotenuse of the supraolivary fossette, below cranial nerve
W9418-01.qxd 10/31/03 3:33 PM Page 37
Chapter 1
IX. The vagal rootlets unite to form a flat cord that passes below the flocculus to the jugular foramen, where the vagus acompanies cranial nerve XI, sharing an arachnoid and dural sheath; both lie anteriorly to a fibrous septum that separates them from cranial nerve IX. The vagus has a more extensive course and distribution than any other cranial nerve, traversing the neck, thorax, and abdomen. Accessory Nerve (Cranial Nerve XI) The accessory nerve has two components: the cranial root, which joins the vagus, has been considered a special visceral efferent nerve. The spinal root is considered a motor nerve that innervates the trapezius and the sternocleidomastoid muscles. The cranial root is the smaller one, and emerges as four or five fine rootlets from the dorsolateral surface of the caudal medulla oblongata below cranial nerve X. It then runs laterally toward the jugular foramen, where it becomes separated from its spinal mate and joins cranial nerve X. These fibers are distributed in the pharyngeal branches of the vagus nerve. The spinal root arises from the lateral aspect of the ventral horn. However it does not exit either from the anterolateral nor from the posterolateral sulci; it exits anteriorly to the posterolateral sulcus and behind the dentate ligament. Its origin can extend from the junction of the spinal cord and medulla to the sixth cervical segment. It then ascends through the foramen magnum toward the jugular foramen. After exiting from the jugular foramen, the spinal root runs posterolaterally passing either medial or lateral to the internal jugular vein. The nerve then crosses the transverse process of the C1 and is itself crossed by the occipital artery. It descends obliquely, medial to the styloid process, stylohyoid, and the posterior belly of the digastric muscles. With the superior sternocleidomastoid branch of the occipital artery, it reaches the upper part of the sternocleidomastoid and enters its deep surface. It crosses the posterior triangle on the neck, on the levator scapulae muscle, and it is separated from that muscle by the prevertebral layer of deep cervical fascia and adipose tissue. Hypoglossal Nerve (Cranial Nerve XII) The hypoglossal nerve innervates intrinsic muscles of the tongue except the palatoglossus. Cranial nerve XII rootlets exit from the preolivary sulcus and run laterally behind the vertebral artery, then they are collected into two bundles that perforate the dura mater separately to enter the hypoglossal canal in the occipital condyle. The two bundles become united at the lateral end of the hypoglossal canal. This single bundle exits from the canal medially to the internal jugular vein, internal carotid artery, cranial nerves IX to XI, and passes inferolaterally behind the internal carotid artery and
Basic Neuroanatomy for Neurointensive Care Unit
37
cranial nerves IX and X to the interval between the artery and the internal jugular vein. Here it makes a half-spiral turn around the inferior vagal ganglion, being united with it by connective tissue. The hypoglossal nerve then descends almost vertically between the vessels and anterior to cranial nerve X to a point level with the mandibular angle, becoming superficial below the posterior belly of the digastric and emerges between the internal jugular vein and internal carotid artery to supply the tongue. Veins of the Posterior Fossa The posterior fossa venous system can be divided into three groups: the anterior or petrosal group that drains into the superior and inferior petrosal sinuses, the superior or Galenic group that drains into the vein of Galen, and the posterior or tentorial group that drains into the sinuses near the torcula.54 There is a tendency for the veins to empty into the nearest draining system. The veins running on the petrosal surface of the cerebellum (anterior hemispheric veins, inferior and superior groups, vein of the great horizontal fissure) and the anterior surface of the brainstem tend to drain into the petrosal sinuses via superior petrosal veins, except those veins running on the surface of the midbrain that drain to the vein of Galen (Fig. 1-13C).55 The tentorial surface of the cerebellum and the posterior aspect of the brainstem are enclosed within the three draining systems; the midline portion of the cerebellomesencephalic fissure is closer to the vein of Galen. Therefore, the veins around the central lobule and culmen (superior vermian vein), and the veins draining the intermediate portion of the wing of the central lobule and the quadrangular lobule (superior hemipheric veins, anterior group) tend to drain into the vein of Galen. As the superior petrosal sinus runs along the anterolateral margin of the cerebellum, all the veins draining the lateral portion of the wing of the central lobule, quadrangular lobule, simple lobule, and tentorial part of the superior semilunar lobule (superior hemispheric veins, lateral group) tend to drain into the superior petrosal sinus. The tentorial surface of the cerebellum is enclosed posteriorly by the transverse sinuses and the torcula; therefore, the veins draining the declive, folium (declival vein), and the intermediate portion of the simple and the superior semilunar lobules (superior hemispheric veins, posterior group) tend to drain into those sinuses or into the tentorial sinus located in the tentorium cerebelli.56,57 The veins draining the suboccipital surface of the cerebellum tend to drain into the torcula or into the transverse sinus or into the sinuses located in the tentorium cerbelli. The cerebellar hemispheres are therefore drained by the posterior inferior hemipheric veins. The inferior vermis, comprising the tuber, pyramid, and uvula, is drained by the inferior vermian veins, which are formed by the junction of
W9418-01.qxd 10/31/03 3:33 PM Page 38
38
Section I
Introduction
the superior and the inferior retrotonsilar veins that run in the retrotonsillar space.58,59 The inferior portion of the roof of the fourth ventricle and the lateral recess are drained by the vein of the lateral recess of the fourth ventricle, also called vein of the cerebellomedullary fissure. Although the vein of the lateral recess of the fourth ventricle is best seen from the suboccipital view, it courses laterally under the lateral recess toward the CP angle, then it passes above or below the flocculus and joins the vein of the middle cerebellar peduncle or joins the vein of the cerebellopontine fissure to finally empty into the superior petrosal sinus via the superior petrosal vein. The vein of the lateral recess of the fourth ventricle also can anastomose with the retrotonsilar veins at the retrotonsillar space, establishing therefore a communication between the petrosal and the tentorial groups of venous drainage (Fig. 1-13D). The brachial veins running in the cerebellomesencephalic fissure can also establish a communication between the petrosal and the galenic groups of venous drainage via the pontotrigeminal and precentral cerebellar veins, respectively (Fig. 1-14A). The veins of the posterior fossa can be summarized as follows: • The petrosal group may be divided into (1) veins related to the anterior aspect of the brain stem; (2) veins in the wing of the precentral cerebellar fissure (the brachial veins); (3) veins on the tentorial and mainly on the petrosal surfaces of the cerebellar hemispheres—superior and inferior hemispheric veins, including the veins of the great horizontal fissure; (4) veins on the cerebellar side (the medial tonsillar vein) and medullary side of the cerebellomedullary fissure (the retro-olivary vein and vein of the inferior cerebellar peduncle); and (5) the vein of the lateral recess of the fourth ventricle. • The superior or galenic group drains into the vein of Galen and includes (1) mesencephalic tributaries—the median anterior pontomesencephalic, the lateral anterior pontomesencephalic, the lateral pontomesencephalic, the lateral mesencephalic, the peduncular, the posterior mesencephalic, and the tectal veins; and (2) cerebellar tributaries—the precentral cerebellar vein and its variants and the superior vermian vein. • The posterior or tentorial group drains into the tentorial sinuses near the torcula and includes basically the veins that drain the suboccipital surface of the cerebellum: the inferior vermian vein and its superior and inferior retrotonsillar tributaries, and the superior and inferior hemispheric veins. Arteries of the Posterior Fossa The vertebral artery arises on each side from the subclavian artery, then enters the transverse foramen of the C6 and ascends through the transverse foramina of the upper cervi-
cal vertebrae up to the C2. After exiting from the transverse foramen of the C2, the vertebral artery deviates laterally to enter the laterally placed transverse foramen of C1. The vertebral artery then turns behind the lateral mass of the C1, above the posterior arch of the C1 to course medially and superiorly to pierce the dura at the foramen magnum. At this level, the vertebral artery usually gives off the posterior spinal and the posterior meningeal arteries. The intradural segment of the vertebral artery is divided into lateral medullary and anterior medullary segments before joining its contralateral mate to form the basilar artery. The lateral medullary segment of the vertebral artery extends from its entrance into the posterior fossa to the preolivary sulcus. From its entrance, the vertebral artery courses anterior, medial, and superiorly through the lower cranial nerve rootlets and laterally to the medulla to reach the preolivary sulcus. The anterior medullary segment begins at the preolivary sulcus, courses in front of, or between, the hypoglossal rootlets, and crosses the pyramid to join with the other vertebral artery at or near the pontomedullary sulcus to form the basilar artery. The main branches of the vertebral artery are the posterior spinal, anterior spinal, PICA, and anterior and posterior meningeal arteries. The vertebral artery also sends off branches to supply the lateral and anterior parts of the medulla along its way around the medulla (Fig. 1-14B). The PICA supplies the medulla, the inferior vermis, the inferior portion of the fourth ventricle, the tonsils, and the inferior aspect of the cerebellum.60–62 It is the most important branch of the vertebral artery, from which it usually arises at the anterolateral aspect of the brainstem near the inferior olive. It often has a tortuous course, and its area of supply is the most variable of the cerebellar arteries. The PICA gives off perforating, choroidal, and cortical arteries; when PICA is absent, the AICA usually supplies this area. The entire inferior cerebellar hemispheres may be supplied by the contralateral PICA. The “normal” PICA has the most complex and variable course of the cerebellar arteries, and is divided into five segments: 1. The anterior medullary segment, which lies on the front of the medulla and extends from the origin to the level of the inferior olive 2. The lateral medullary segment, which courses beside the medulla, and extends from the inferior olive to the origin of the glossopharyngeal, the vagus, and the accessory nerves 3. The tonsillomedullary or posterior medullary segment, which courses around the caudal half of the cerebellar tonsil and begins at the level of the nerves and loops below the inferior pole of the cerebellar tonsil and upward along the medial surface of the tonsil toward the inferior medullary velum (caudal loop) 4. The telovelotonsillar or supratonsillar segment, which courses in the cleft between the tela choroidea and the
W9418-01.qxd 10/31/03 3:33 PM Page 39
Chapter 1
Basic Neuroanatomy for Neurointensive Care Unit
B
A
C
D
Figure 1-14. A, Superolateral view: 1 = transverse sinus; 2 = tentorial sinus; 3 = internal cerebral vein; 4 = anterior septal vein; 5 = declival veins; 6 = superior vermian vein; 7 = vein of Galen; 8 = pineal gland; 9 = precentral cerebellar vein; 10 = red nucleus; 11 = anterior cerebral vein; 12 = pontotrigeminal vein; 13 = posterior cerebral vein; 14 = internal carotid artery; 15 = superior petrosal vein; 16 = superior petrosal sinus. B, Posterior view: 1 = optic nerve; 2 = internal carotid artery; 3 = oculomotor nerve; 4 = posterior cerebral artery; 5 = trochlear nerve; 6 = superior cerebellar artery; 7 = trigeminal nerve; 8 = basilar artery; 9 = AICA; 10 = PICA; 11 = anterior spinal artery; 12 = first triangular process of the dentate ligament; 13 = second triangular process of the dentate nucleus. IAM, internal acoustic meatus; VA, vertebral artery; Jug. for., jugular foramen. C, Posterolateral view: 1 = thalamus; 2 = choroidal fissure; 3 = insula; 4 = fornix; 5 = thalamus; 6 = pineal gland; 7 = superior colliculus; 8 = inferior colliculus; 9 = branches from the superior cerebellar artery; 10 = middle cerebellar peduncle; 11 = anterior medullary segment of the PICA; 12 = lateral medullary segment of PICA; 13 = caudal loop of PICA; 14 = posterior medullary segment of PICA; 15 = supratonsillar segment of PICA; 16 = pyramidal loop of PICA; IAM = internal acoustic meatus; FL = flocculus; Jug. for. = jugular foramen; VA = vertebral artery. D, Frontal view: 1 = right posterior cerebral artery; 2 = left posterior cerebral artery; 3 = right superior cerebellar artery; 4 = left superior cerebellar artery; 5 = basilar artery; 6 = AICA; 7 = PICA; VA = vertebral artery.
39
W9418-01.qxd 10/31/03 3:33 PM Page 40
40
Section I
Introduction
inferior medullary velum rostrally, and the superior pole of the cerebellar tonsil caudally, and begins below the fastigium, where the PICA turns posteriorly over the medial side of the superior pole of the tonsil. This segment forms the “cranial loop.” It sometimes passes posteriorly before reaching the superior pole of the tonsil, thus giving the cranial loop a variable relationship to the fastigium. The junction of the posterior medullary segment and the supratonsillar segment is called the choroidal point. 5. The cortical segment: after a short distance distal to the apex of the cranial loop, the PICA continues posteriorly downward in the retrotonsillar fissure where it usually bifurcates into two terminal trunks, the tonsillohemispheric and the inferior vermian branches. The inferior vermian branch lies on the lower aspect of the inferior vermis, and forms a loop convex around the copula pyramidis (pyramidal loop). The most anterior point of the curve of the pyramidal loop is also called the copular point. The terminal portion of the vermian branch curves around the tuber in the posterior cerebellar notch. The tonsillohemispheric branch runs inferiorly near the prepyramidal sulcus and gives off anterior or tonsillar branches and posterior or hemispheric branches that curve downward and backward around the biventral lobule to the under aspect of the cerebellar hemisphere (Fig. 1-14C). The AICA and PICA are defined according to their origin rather than by the portions of cerebellum that they supply. The AICA supplies the lateral pontine tegmentum and base, and the most anterior undersurface of the cerebellum.63,64 The AICA arises from the basilar artery and the PICA from the vertebral artery. The AICA arises more frequently from the lower third of the basilar artery and courses posteriorly, laterally, and usually downward on the belly of the pons, in contact with either the superior or inferior aspect of the abducent nerve. In this course, it supplies the lateral aspect of the lower two thirds of the pons and the upper medulla. Either immediately before or after crossing the roots of the facial, the intermedius, and the vestibulocochlear nerves within the cerebellopontine angle, the AICA bifurcates into its two major branches, the rostrolateral and the caudomedial arteries. The rostrolateral (RL) trunk has been divided into three segments—premeatal, meatal, and postmeatal—according to their relationship to cranial nerves VII and VIII. The premeatal segment begins at the basilar artery and courses around the brainstem to reach cranial nerves VII and VIII and the region of the meatus, usually anteroinferiorly to the nerves. Seventy-seven percent of the internal auditory arteries and 49% of the recurrent perforating arteries, which course medially from their origin to supply the brainstem, arise from this segment. The meatal segment is located in the vicinity of the internal auditory meatus, where the nerve-related vessels turn toward the brainstem. This segment often forms a later-
ally convex loop, the meatal loop, directed toward or through the meatus. It usually stays medial to the meatus, but sometimes it protrudes into the canal. The postmeatal segment begins distally to the nerves and courses medially to supply the brainstem and the cerebellum. The subarcuate artery to the subarcuate fossa usually arises from this segment. The caudomedial (CM) artery courses medially and downward toward the medial and anterior border of the cerebellum, supplying the biventral lobule and the middle cerebellar peduncle; the medial branch also has abundant anastomoses with branches of the PICA. From its origin on the lateral aspect of the pons in the vicinity of the sixth cranial nerve, the CM passes posterosuperiorly, toward the pontomedullary sulcus to describe its own caudal loop on the lateral aspect of the pons and medulla. This lateral loop can lie on the antero-inferolateral aspect of the flocculus, or on the petrosal aspect of the biventral lobule, or on the petrosal aspect of the undersurface of the biventral lobule. Multiple small arteries to the choroid plexus of the lateral recess often arise from the inner aspect of this lateral loop. Distal to the loop, the biventral segment turns postero-inferiorly on the lateral aspect of the biventral lobule, on the biventral ridge, or within the cerebellomedullary fissure to reach the posterior surface of the cerebellum. In a manner analogous to the descending branch of the RL, the CM may also anastomose with PICA or give rise to ascending hemispheric branches that supplement or replace the hemispheric branches of PICA (Figs. 1-14D and 1-15A). The superior cerebellar artery (SCA) is the most rostral of the infratentorial vessels, and it arises near the apex of the basilar artery and encircles the pons and the lower midbrain. It supplies the tentorial surface of the cerebellum, the midbrain tegmentum, the deep cerebellar nuclei, and the inferior colliculi.65–67 The SCA is divided in four segments: 1. The anterior pontomesencephalic segment extends from its origin to the anterolateral margin of the brainstem. It courses laterally on the anterior aspect of the upper pons, often in an arcuate curve convex inferiorly; at the anterolateral margin of the brainstem it lies inferior to the third nerve. 2. The lateral pontomesencephalic segment begins at the anterolateral margin of the brainstem and follows caudally onto the lateral side of the upper pons in the infratentorial portion of the ambient cistern to terminate at the anterior margin of the cerebellomesencephalic fissure. This segment is related medially to the brainstem, laterally to the wing of the central lobule, and to the middle cerebellar peduncle inferiorly. The anterior part of this segment is often visible above the free edge of the tentorium, whereas its caudal loop projects toward and often reaches the root entry zone of the trigeminal nerve. The bifurcation of the SCA into its rostral and caudal trunks often occurs in this segment, while the rostral trunk supplies the vermis and a
W9418-01.qxd 10/31/03 3:33 PM Page 41
Chapter 1
Basic Neuroanatomy for Neurointensive Care Unit
A
B
C
D Figure 1-15. A, Frontal view: 1 = posterior cerebral artery; 2 = superior cerebellar artery; 3 = superior petrosal sinus; 4 = AICA; 5 = facial nerve and the meatal loop of AICA; 6 = origin of PICA; 7 = anterior spinal artery; 8 = vertebral artery between C1 and C2; VA = intradural vertebral artery. B, Posterior view: 1 = perforating branches from basilar tip and P1; 2 = medial posterior choroidal artery; 3 = superior cerebellar artery; 4 = perforating branches from midbasilar artery; 5 = PICA; BA = basilar artery; VA = vertebral artery. C, Midsagittal view of the interpeduncular region: 1 = thalamus; 2 = aqueduct; 3 = foramen of Monro; 4 = fornix; 5 = midbrain; 6 = hypothalamus; 7 = posterior perforated substance region; 8 = lamina terminalis; 9 = mamillary body; 10 = tuber cinereum; 11 = pituitary stalk; 12 = optic chiasm; 13 = pons and the posterior thalamoperforating artery; 14 = posterior communicating artery; 15 = P1; 16 = internal carotid artery; 17 = parahippocampal gyrus; III = oculomotor nerve. D, Coronal view: General pattern of the vascularization of the central nervous system (left); the approximate vascular territories of the major arteries (right).
41
W9418-01.qxd 10/31/03 3:33 PM Page 42
42
Section I
Introduction
variable portion of the adjacent tentorial surface, the caudal trunk supplies the surface lateral to the area supplied by the rostral trunk. 3. The cerebellomesencephalic segment courses in the cerebellomesencephalic fissure through a series of hairpin-like curves; it then passes upward to reach the anterosuperior margin of the cerebellum. Inside the cerebellomesencephalic fissure, the cortical branches from the rostral and caudal trunks send off small arterial twig branches called precentral branches. Those precentral branches arising from the rostral trunk supply the inferior colliculi (the superior colliculi are supplied by PCA) and superior medullary velum, and those arising from the caudal trunk supply the deep cerebellar nuclei. 4. The cortical segment is represented by the hemispheric and the vermian branches to supply the tentorial surface of the cerebellum. Among these cortical branches, the marginal or lateral branch deserves a special attention: it was present in 62% of the cases; it is the first large cortical branch of the SCA, and arises from the lateral pontomesencephalic segment to course anteriorly and laterally to reach the anterolateral margin of the cerebellum, and then extends posterolaterally in the region of the petrosal fissure to supply the adjacent areas. Its area of supply is inversely related to the area supplied by AICA (Figs. 1-15A and B). The basilar artery averages approximately 3 cm and is formed at the level of the pontomedullary junction by the union of the two vertebral arteries. The basilar artery then courses superiorly in front of the pons toward the interpeduncular fossa to divide into the two posterior cerebral arteries. The main branches of the basilar artery are AICA, SCA, and short and long segment circumflex perforating arteries. These perforating arteries arise along the entire length of the basilar artery and supply the ventral pons and rostral brainstem. The branches arising from the top of the basilar artery have been discussed under the posterior cerebral artery (Fig. 1-15B and C).
The Brainstem Syndromes The general pattern of the vascularization of the central nervous system is applied for the whole central nervous system, from the telencephalon to the spinal cord. Following this pattern, the “tube-shaped” central nervous system is completely circumscribed by the major arteries of the cerebrum and the posterior fossa. These major arteries not only circumscribe the neural tube but also give off central perforators (paramedian arteries) to the central base region of the neural tube, lateral perforators to the lateral part of the neural tube, and finally perforators to the dorsal aspect of the neural tube. This pattern is still evident in the mature central nervous system (Fig. 1-15D and 1-16A).
The brainstem syndromes of vascular origin can be divided into midbrain, pontine, and medullary syndromes.68,69 Midbrain syndrome can be further divided into ventral midbrain, tegmental, and dorsal midbrain syndromes. Ventral midbrain syndrome, also called Weber’s syndrome, usually is characterized by diplopia, with homolateral dilated pupil (oculomotor nerve), occasional ataxia, and contralateral motor deficit, mainly upper limb and face (pyramidal tract) (Fig. 1-16B). Tegmental midbrain syndrome, also called Claude’s syndrome, is usually characterized by diplopia, with homolateral dilated pupil (oculomotor nerve), homolateral Horner’s syndrome, contralateral loss of pain and temperature sensation from trunk, limbs (spinothalamic tract), face (trigeminothalamic tract), contralateral loss of position sense in the limbs (medial lemniscus), contralateral resting tremor (red nucleus), contralateral ataxia (dentatothalamic tract after crossing), and contralateral motor deficit affecting mainly the leg (leg fibers from the pyramidal tract) (Fig. 1-16B). The combination of the ventral and the tegmental midbrain syndromes is called Benedikt’s syndrome. Tectal midbrain syndrome, also called Parinaud syndrome, is characterized by deficit of upward conjugate gaze (type 1) or deficit of downward conjugate gaze (type 2) (Fig. 1-16B). Pontine syndrome can be divided in medial and lateral pontine syndromes. The medial pontine syndrome, also called Millard-Gubler syndrome, is characterized by homolateral abducent nerve palsy, homolateral facial paralysis, homolateral ataxia (transverse fibers of pons), and contralateral loss of pain and temperature sensation from the face (trigeminothalamic tract), contralateral loss of position sense of limbs (medial lemniscus), and contralateral motor deficit (pyramidal tract) (Fig. 1-16C). Lateral pontine syndrome, also called Foville’s syndrome, is characterized by homolateral tinnitus or deafness (cochlear nerve), homolateral ataxia (inferior cerebellar peduncle), paralysis of homolateral conjugate gaze (paraabducent nucleus), vomiting, nystagmus, vertigo (vestibular nerve), homolateral Horner’s syndrome, homolateral anesthesia of the face (sensory root of the trigeminal nerve), weakness of jaw muscles (motor root of trigeminal nerve), homolateral facial paralysis, and contralateral loss of pain and temperature sensation from trunk and limbs (spinothalamic tract) (Fig. 1-16C). Medullary syndromes can be divided into medial and lateral medullary syndromes. Medial medullary syndrome is characterized by weakness of tongue (hypoglossal nerve), contralateral loss of position sense in limbs (medial lemniscus), and contralateral motor deficit sparing the face (pyramid) (Fig. 1-16D). Lateral medullary syndrome, also called Wallenberg’s syndrome, is characterized by vertigo, vomiting, nystagmus (vestibular nucleus), homolateral ataxia of limbs (inferior cerebellar peduncle), homolateral loss of pain and temperature sensation from face (spinal tract of trigeminal nerve), homolateral Horner’s syndrome, dysphagia, hoarseness (nucleus ambiguus),
W9418-01.qxd 10/31/03 3:33 PM Page 43
B
A
C
Figure 1-16. A, Axial view: the approximate vascular territories of the major arteries (right). B, The midbrain syndromes: 1 = ventral midbrain syndrome; 2 = tegmental midbrain syndrome; 3 = tectal midbrain syndrome. C, The pontine syndromes: 1 = medial pontine syndrome; 2 = lateral pontine syndrome. D, The medial and lateral medullary syndromes (shaded areas).
D
W9418-01.qxd 10/31/03 3:33 PM Page 44
44
Section I
Introduction
A
B
Figure 1-17. A, The location of the pyramidal and the lateral spinothalamic tracts in the spinal cord. B, Frontal view of the dermatomes.
and contralateral loss of pain and temperature sensation from trunk and limbs (lateral spinothalamic tract) (Fig. 1-16D).
The Spinal Cord For a quick and concise neurologic examination of the spinal cord in an intensive care unit, two basic aspects of spinal cord
organization should be kept in mind: (1) the topographical organization of the main fiber tract in the spinal cord (the motor and the superficial sensory tracts; Fig. 1-17A), and (2) the distribution of the dermatomes—the strips of skin supplied by the posterior nerve roots. The clinical importance of the dermatomes is evident in determining the level of the spinal cord lesion (for instance in spinal cord injuries; Fig. 1-17B).
W9418-01.qxd 10/31/03 3:33 PM Page 45
Chapter 1
Basic Neuroanatomy for Neurointensive Care Unit
45
P earls 1. The cerebrum is arbitrarily divided into five lobes: frontal, parietal, temporal, occipital, and the hidden insula. 2. From the neurosurgical viewpoint, the sylvian fissure can be considered as the gateway connecting the surface of the anterior part of the brain to its depth with all the neural and vascular components along the way. 3. The corpus callosum is the largest transverse commissure connecting the cerebral hemispheres. 4. The basal ganglia play a major role in voluntary motor movement; however, they do not have direct input or output with the spinal cord. 5. The disturbance of the basal ganglia is usually characterized by (1) tremor and other involuntary movements, (2) changes in posture and muscle tone, and (3) poverty and slowness of movement without paralysis. 6. The thalamus is not a relay station where information is simply passed on to the neocortex; the thalamus acts as a gatekeeper for information to the cerebral cortex, preventing or enhancing the passage of specific information depending on the behavioral state of the person. 7. The choroidal fissure is one of the most important landmarks in microneurosurgeries involving the tem-
References 1. Ono M, Kubik S, Abernathey CD: Atlas of the Cerebral Sulci. Stuttgart, Georg Thieme Verlag, 1990. 2. Gibo H, Caarver CC, Rhoton AL Jr: Microsurgical anatomy of the middle cerebral artery. J Neurosurg 1981;54:151. 3. Szikla G, Bouvier T, Hori T, et al: Angiography of the Human Brain Cortex. Berlin, Springer, 1977. 4. Wolf BS, Huang YP: The insula and deep middle cerebral venous drainage system: Normal anatomy and angiography. Am J Roentgenol Radium Ther Nucl Med 1963;90:472. 5. Timurkaynak E, Rhoton AL Jr, Barry M: Microsurgical anatomy and operative approaches to the lateral ventricles. Neurosurgery 1986;19:685. 6. Wen HT, Rhoton AL Jr, de Oliveira E, et al: Microsurgical anatomy of the temporal lobe: Part 1. Mesial temporal lobe and its vascular relationship as applied to amygdalohippocampectomy. Neurosurgery 1999;45:549. 7. Warwick R, Williams PL: Gray’s Anatomy, 35th ed. Philadelphia, WB Saunders, 1973. 8. Williams PL: Gray’s Anatomy, 38th ed. London, Churchill Livingstone, 1995. 9. Duvernoy HM: The Human Hippocampus: An Atlas of Applied Anatomy. Munich, JF Bergmann Verlag, 1988. 10. Gloor P: The temporal Lobe and Limbic System. New York, Oxford University Press, 1997. 11. Nagata S, Rhoton AL Jr, Barry M: Microsurgical anatomy of the choroidal fissure. Surg Neurol 1988;30:3.
8.
9.
10.
11.
poral lobe; it separates those structures located laterally that can be removed (temporal structures) from those structures located medially that should be preserved (thalamic structures). The internal carotid artery is divided into five parts: cervical, petrous, cavernous, clinoid, and supraclinoid portions. Paracentral syndrome can be characterized by weakness of the contralateral lower limb, most intense in the foot and ankle, with or without sensory loss. The transient or permanent incontinence of urine can also be present. SMA syndrome can be characterized by dysphasia (when the dominant hemisphere is affected), akinesia in the contralateral limb, contralateral hand grasping or groping, contralateral alien hand signs (when dominant hemisphere is affected, the right hand consistently interrupts manual tasks performed by the left hand), and dyspraxia. Anterior cingulate syndrome is more evident when the cingulate cortex is bilaterally and extensively affected; this might cause akinetic mutism, complex behavioral changes, loss of sphincter control, and autonomic dysfunctions (temperature; cardiac and respiratory irregularities).
12. Rhoton AL Jr: Microsurgical anatomy of the third ventricular region. In Apuzzo MLJ (ed): Surgery of the Third Ventricle Region. Baltimore, Williams & Wilkins, 1987, pp 570–590. 13. Yamamoto I, Rhoton AL Jr, Peace D: Microsurgery of the third ventricle: Part 1. Microsurgical anatomy. Neurosurgery 1981; 8:334. 14. Wen HT, Rhoton AL Jr, de Oliveira E: Transchoroidal approach to the third ventricle: An anatomic study of the choroidal fissure and its clinical application. Neurosurgery 1998;42:1205. 15. Apuzzo MLJ, Giannotta SL: Transcallosal interforniceal approach. In Apuzzo MLJ (ed): Surgery of the Third Ventricle Region. Baltimore: Williams & Wilkins, 1987, pp 354–380. 16. Apuzzo MLJ: Transcallosal interforniceal exposure of lesions of the third ventricle. In Schmidek HH, Sweet WH (eds): Operative Neurosurgical Techniques. Indications, Methods, and Results, vol 1. Orlando, Grune & Stratton, 1988, pp 389–395. 17. Yasargil MG: Microsneurosurgery: AVM of the Brain. Stuttgart, Georg Thieme Verlag, 1988, vol III-B. 18. Oka K, Rhoton AL Jr, Barry M, et al: Microsurgical anatomy of the superficial veins of the cerebrum. Neurosurgery 1985;17:711. 19. Wolf BS, Huang YP, Newman CM: The superficial sylvian venous drainage system. Am J Roentgenol Radium Ther Nucl Med 1963;89:398. 20. Ono M, Rhoton AL Jr, Peace D, et al: Microsurgical anatomy of the deep venous system of the brain. Neurosurgery 1984;15: 621. 21. Osborn AG: Normal vascular anatomy. In Osborn AG (ed): Diagnostic Neuroradiology. St. Louis, Mosby, 1994, pp 117–153.
W9418-01.qxd 10/31/03 3:33 PM Page 46
46
Section I
Introduction
22. Ture U, Yasargil MG, Al-Mefty O, et al: Arteries of the insula. J Neurosurg 2000;92:676. 23. Taveras JM, Wood EH: Diagnotic Neuroradiology. Baltimore, Williams & Wilkins, 1964. 24. Rosner SS, Rhoton AL Jr, Ono M, et al: Microsurgical anatomy of the anterior perforating arteries. J Neurosurg 1984;61:468. 25. Huang YP, Wolf BS: The basal cerebral vein and its tributaries. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol 2, book 3. St. Louis, CV Mosby, 1974, pp 2111–2154. 26. Seoane E, Rhoton AL Jr, de Oliveira E: Microsurgical anatomy of the dural collar (carotid collar) and rings around the clinoid segment of the internal carotid artery. Neurosurgery 1998;42:869. 27. Gibo H, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the supraclinoid portion of the internal carotid artery. J Neurosurg 1981;55:560. 28. Hayreh SS: The ophthalmic artery. In Newton TH and Potts DG (eds): Radiology of the Skull and Brain. Angiography, vol 2, book 2. St. Louis, CV Mosby, 1974, pp 1333–1350. 29. Rhoton AL Jr, Fujii K, Fradd B: Microsurgical anatomy of the anterior choroidal artery. Surg Neurol 1979;12:171. 30. Theron J, Newton TH: Anterior choroidal artery: 1—Anatomic and radiographic study. J Neuroradiol 1976;3:5. 31. Hayman LA, Berman SA, Hinck VC: Correlation of CT cerebral vascular territories with function: II. Posterior cerebral artery. AJR 1981;137:13–19. 32. Margolis MT, Newton TP, Hoyt WF: The posterior cerebral artery II. Gross and roentgenographic anatomy. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol 2, book 2. St. Louis, CV Mosby, 1974, pp 1551–1576. 33. Saeki N, Rhoton AL Jr: Microsurgical anatomy of the upper basilar artery and the posterior circle of Willis. J Neurosurg 1977;46:563. 34. Yasargil MG: Microneurosurgery: Microsurgical Anatomy of the Basal Cisterns and Vessels of the Brain. Stuttgart, Georg Thieme Verlag, 1984, vol I. 35. Zeal AA, Rhoton AL Jr: Microsurgical anatomy of the posterior cerebral artery. J Neurosurg 1978;48:534. 36. George AE, Raybaud CH, Salamon G, et al: Anatomy of the thalamoperforating arteries with special emphasis on arteriography of the third ventricle: Part 1. AJR 1975;124:220–230. 37. Perlmutter D, Rhoton AL Jr: Microsurgical anatomy of the distal anterior cerebral artery. J Neurosurg 1978;49:204. 38. Dunker RO, Harris B: Surgical anatomy of the proximal anterior cerebral artery. J Neurosurg 1976;44:359. 39. Perlmutter D, Rhoton AL Jr: Microsurgical anatomy of the anterior cerebral-anterior communicating-recurrent artery complex. J Neurosurg 1976;45:259. 40. Gomes F, Dunovny M, Umansky F, et al: Microsurgical anatomy of the recurrent artery of Heubner. J Neurosurg 1984;60:130. 41. Hung TP, Ryu SJ: Anterior cerebral artery syndromes. In Vinken PJ, Bruyn GW, Klawans HL (eds): Handbook of Clinical Neurology, vol 53, part 1. Amsterdam, Elsevier Science, 1988, pp 339–352. 42. Salamon G, Lazorthes G: Tumours of the basal ganglia: An angiographic study. Neuroradiology 1971;2:80–89. 43. Westberg G: Arteries of the basal ganglia. Acta Radiol Diagn 1966;N.S.5:581–596. 44. Rhoton AL Jr: The Posterior Cranial Fossa: Microsurgical Anatomy and Surgical Approaches. Neurosurgery 2000;47:S1. 45. Ghez C, Thach WT: The cerebellum. In Kandel ER, Schwartz JH, Jessel TM (eds): Principles of Neuroscience, 4th ed. New York, McGraw Hill, 2000, pp 832–852.
46. Matsushima T, Rhoton AL Jr, Lenkey C: Microsurgery of the fourth ventricle: Part 1: Microsurgical anatomy. Neurosurgery 1982;11:631. 47. Corrales M, Greitz T: Fourth ventricle: I. A morphologic and radiologic investigation of the normal anatomy. Acta Radiol 1972;12:113. 48. Bentson JR, Alberti JB: Lateral recess of the fourth ventricle. Radiology 1972;104:593. 49. Fujii K, Lenkey C, Rhoton AL Jr: Microsurgical anatomy of the choroidal arteries: fourth ventricle and cerebellopontine angles. J Neurosurg 1980;52:504. 50. Gortvai P, Teruchkin S: The position and extent of the human dentate nucleus. Acta Neurochir 1974;21:101. 51. Buren van JM, Baldwin M: The architecture of the optic radiation in the temporal lobe of man. Brain 1958;81:2. 52. Ebeling U, Reulen HJ: Neurosurgical topography of the optic radiation in the temporal lobe. Acta Neurochir 1988;92:29. 53. Rasmussen AT: The extent of recurrent geniculo-calcarine fibers (loop of Archambault and Meyer) as demonstrated by gross brain dissection. Anat Rec 1943;85:277. 54. Matsushima T, Rhoton AL Jr, de Oliveira E: Microsurgical anatomy of the veins of the posterior fossa. J Neurosurg 1983;59:63. 55. Huang YP, Wolf BS, Antin SP, et al: The veins of the posterior fossaanterior or petrosal draining group. AJR 1968;104:36. 56. Huang YP, Wolf BS: The veins of the posterior fossa-superior or galenic draining group. AJR 1965;95:808. 57. Huang YP, Wolf BS: Precentral cerebellar vein in angiography. Acta Radiol 1966;5:250. 58. Huang YP, Wolf BS, Okudera T: Angiographic anatomy of the inferior vermian vein of the cerebellum. Acta Radiol 1969;9:327. 59. Huang YP, Wolf BS: Veins of the posterior fossa. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol II, book 3. St Louis: CV Mosby, 1974, pp 2155–2216. 60. Lister JR, Rhoton AL Jr, Matsushima T, et al: Microsurgical anatomy of the posterior inferior cerebellar artery. Neurosurgery 1982; 10:170. 61. Takahashi M, Okudera T, Fukui M, et al: The choroidal and nodular branches of the posterior inferior cerebellar artery. Radiology 1972;103:347. 62. Wolf BS, Newman CM, Khilnani MT: The posterior inferior cerebellar artery on vertebral angiography. AJR 1962;87:322. 63. Martin RG, Grant JL, Peace D, et al: Microsurgical relationships of the anterior inferior cerebellar artery and the facial-vestibulocochlear nerve complex. Neurosurgery 1980;6:483. 64. Naidich TP, Kricheff II, George AE, et al: The normal anterior inferior cerebellar artery. Anatomic-radiographic correlation with emphasis on the lateral projection. Radiology 1976;119:355. 65. Hardy DG, Peace D, Rhoton AL Jr: Microsurgical anatomy of the superior cerebellar artery. Neurosurgery 1980;6:10. 66. Hoffman HB, Margolis MT, Newton TH: The superior cerebellar artery: I. Normal gross and radiographic anatomy. In Newton TH, Potts DG (eds): Radiology of the Skull and Brain, vol 2, book 2. St. Louis, CV Mosby, 1974, pp 1809–1830. 67. Mani RL, Newton TH, Glickman MG: The superior cerebellar artery: An anatomic-roentgenographic correlation. Radiology 1968;91:1102. 68. Brust JCM: Circulation of the brain. In Kandell ER, Schwartz JH, Gessell TM (eds): Principles of Neuroscience, 4th ed. New York, McGraw-Hill, 2000, pp 1302–1316. 69. Fitzgerald MJT: Brain stem. In Fitzgerald MJT (ed): Neuroanatomy Basic and Applied. London, Bailliére Tindall, 1985, pp 84–93.
W9418-02.qxd 10/31/03 3:35 PM Page 47
Chapter 2 Neuroradiologic Imaging Ronald G. Quisling, MD and Lorna Sohn-Williams, MD
Understanding Magnetic Resonance Imaging The Physical Basis for Magnetic Resonance Imaging Diagnostic imaging is often critical in the treatment of patients undergoing acute neurological decline. Brain imaging in this context typically relies on magnetic resonance imaging (MRI) and computed tomography (CT) studies, although occasionally angiography and nuclear medicine testing is also valuable. However, only brain MRI and CT modalities will be discussed in this chapter. Fundamental to providing adequate sectional imaging is access to sufficient clinical information to appropriately protocol the examination. No single test fits all, and not all patients can undergo all varieties of imaging studies. When an especially complex diagnostic dilemma arises, direct discussion with the neuroradiologist is necessary to (1) select all areas of clinical interest, (2) choose the most appropriate imaging modality, (3) optimize the imaging parameters, and (4) exclude patients unsuitable for imaging, especially with MRI. There is a wide array of technical choices for how an imaging procedure(s) can be performed, but usually only a finite time in which to acquire the data. Imaging protocols must include the most appropriate imaging sequences, the best plane of section to achieve true longitudinal or transverse sectional anatomy (unplanned obliquity often causes ambiguity or worse), and a decision on whether to use contrast medium. The delivery of the contrast material then requires decisions on the type and amount of contrast material, and how to
time the injection relative to image acquisition. Choices must also be made concerning physical factors related to section thickness, region of interest, magnification, postprocessing algorithms, and actual CT energy or MRI parameter settings. Each procedure must be tailored to individual patient requirements; it is rare that unfocused imaging will substantially contribute to treatment decisions. The primary objective for any diagnostic MRI or CT study is to distinguish between normality and abnormality (“lesion detection”). Establishing normality is typically more difficult than recognizing obvious pathology. To declare that an imaging procedure reveals either “no apparent and/or no significant disease” requires the examination to have been performed appropriately with the most sensitive imaging modality and scanned in the right location using parameters capable of providing the optimum spatial and contrast resolution. Once a lesion is observed the role of imaging shifts toward more specialized tasks, including predicting the type of lesion, establishing the spatial relationship between the lesion and eloquent areas, and judging the acuteness of the changes (“time course”). These tasks require sophisticated imaging and experienced interpretation. In essence, the neuroradiologic imager should be provided with the presenting symptom complex and the main clinical considerations before the examination to enable the imaging to guide an appropriate treatment plan. The following discussion will provide an overview of the basic principles of MRI and CT; a summary of the strengths, weaknesses, and restrictions of special imaging for both MRI and CT, and to provide examples of key neuropathology for the neurointensivist. 47
W9418-02.qxd 10/31/03 3:35 PM Page 48
48
Section I
Introduction
Understanding the Standard Magnetic Resonance Imaging Sequences Magnetic resonance imaging compares the relative intensity differences between anatomic subunits of tissue that are exposed to both a constant main magnetic field, and intermittently to a changing set of secondary gradients (to give spatial resolution) and to an external radiofrequency (RF) pulse(s) to energize the nuclear spin systems.1–4 Proton MR, in essence, is a map of hydrogen nuclei bound to oxygen, as water, or to carbon, as organic compounds. It is obvious that protons must be present for any signal to be detected at all. Hence, air generates no signal. Signal intensity is a measure of how all protons within any small block of tissue (volume element or “voxel”) respond to the RF signal. This signal is then compared to its neighbors and represented as a shade on a gray scale. What is critical in generation of MR signal is that all the tissue within a voxel responds more or less similarly magnetically; all the tissue in the voxel need not be organically the same, but all needs to be of similar magnetic susceptibility. If tissue within a voxel behaves discordantly, then the overall signal of that voxel is diminished or lost (i.e., “susceptibility artifact”). These susceptibility artifacts can actually be emphasized with certain sequence to detect, for example, hemosiderin. It is also imperative that tissue remain stationary during the standard MR experiment; otherwise, “flow-related artifacts” occur. These flow effects can actually be emphasized and stationary brain de-emphasized in the process of acquiring MR angiography (MRA). The process of deriving signal from the magnetic response of most stationary brain tissue placed under the conditions of the MR experiment is referred to as “relaxivity.” The actual measurable response time is called the “relaxation rate.” Types of tissue are referenced by their relative relaxation rates; some are faster and some are slower. These differences are portrayed on MR images as differences in shade-of-gray. Although technically complex and beyond the scope of this chapter, all MR experiments require a refocusing of signal before the read-out (signal detection phase). This refocusing is obtained either by using a second RF pulse (i.e., the spin-echo experiment), or by using magnet gradients (i.e., gradient-echo experiment or GRE). The former have better contrast resolution; the latter are substantially faster. Both are described in the following sections along with a general overview of basic MR signal characteristics. Standard Spin-Echo Sequence with T1 Weighting The T1 experiment evaluates a progressively increasing signal over time (i.e., a constructive effect), while the T2 experiment measures a declining signal over time (i.e., a destructive effect). In essence, the tissue with quicker recovery (i.e., more rapid T1 relaxation rates) will appear relatively brighter than their slower neighbors on a T1-weighted
MR sequence. The terms, signal hyperintensity and hypointensity, are descriptive and based on the signal characteristics of the object in question compared to normal brain, which is assigned the term isointensity. Thus, inherent T1 hyperintensity (compared to isointense normal brain) will appear “brighter” (approaching a shade toward white) and implies the presence of substances that naturally exhibit rapid proton relaxation (a more rapid return to baseline), such as structures with high lipid content. Other substances with faster T1 relaxivity include methemoglobin (subacute residua of hemorrhage), melanin, dilute amounts of intracellular calcium, and hyperconcentrated proteins (as might occur in a colloid containing mass or very hypercellular tumors). Inherent T1 hypointensity (compared to normal brain signal) is displayed as a darker shade of gray, and indicates that tissue in these voxels are not like fat but is either rigidly bound (an anisotrophic effect), as seen in fibrosis or matrix calcification, or virtually unbound (an isotropic effect), as seen in edema, necrosis, or cyst formation. T1 hypointensity reflects both extremes of hydrogen states: too tightly bound or too loosely bound compared to lipid or lipid-like protons. T1 weighting assigns a bright signal (tending toward white) to tissues containing higher inherent lipid concentration but also to those affected by substances that cause proton relaxation enhancement. As mentioned, compounds that quickly return to baseline (and appear hyperintense to brain), include methemoglobin, melanin, concentrated protein, and increased intracellular calcium ions, and intravenous gadolinium injected at the correct concentration. Gadolinium will normally cause proton relaxation “enhancement” in tissues that lack endothelial tight junctions and an intact blood-brain barrier (pineal gland, pituitary gland, choroid plexus) normally. Gadolinium causes abnormal enhancement in the context of an altered bloodbrain barrier or an increased regional capillary blood volume (Fig. 2-1). Standard Spin-Echo Sequence with T2 Weighting T2-weighted sequences provide different information than T1-weighted data. In physical terms T2 rate reflects the time required for the signal to degrade or to lose signal intensity. It is a measure of loss of phase coherence of the precessing magnetic vectors, which had been refocused by either the second RF pulse for spin-echo sequences or by the secondary gradients in gradient echo (GRE) imaging. Tissue with more rapid T2 relaxation rates (T2 hypointensity) destroy signal more quickly and appear darker than their neighbors. Abnormal T2 hypointensity implies a more restricted bulk water pool. The latter occurs in fibrosis, ossification, and dehydrated tissues. Tissues that possess more unbound water molecules (like cerebrospinal fluid [CSF] or tissue edema) will preserve phase coherence longer, which translates into a slower T2 rate, and a brighter signal or T2 hyperintensity (Fig. 2-2).
W9418-02.qxd 10/31/03 3:35 PM Page 49
Chapter 2
Figure 2-1. Arachnoid cyst on T1 sequence. An axial, nonenhanced, T1-weighted MR image through the posterior fossa demonstrates the features of a large arachnoid cyst residing in the right cerebellopontine angle cistern. The fluid within the cyst matches cerebrospinal fluid (hypointense to normal brain) consistent with either an arachnoid cyst or an epidermoid cyst. Differentiation between these entities requires additional information, which is provided in Figures 2-2, 2-6, and 2-7.
Magnetic Resonance Imaging Contrast Medium and Abnormal Contrast Enhancement Up to this point, discussion has focused on the inherent relaxivity effects of tissue. But the MRI T1 signal can also be altered by certain compounds containing heavy metals, which can artificially accelerate the return to baseline of tissue if such compounds can be retained in voxels throughout the data acquisition period. The basis for abnormal contrast enhancement is similar for both CT (using iodinated contrast medium) and MRI (using gadolinium-diethylenetriamine pentaacetic acid [DTPA]). Abnormal contrast enhancement is related either to an expanded intravascular pool or to an increased endothelial permeability and diffusion of the gadolinium into the brain substance. An expanded vascular pool can occur in areas of brain that have lost autoregulation and increased the capacity of a regional capillary bed. This most commonly is observed in patients with recent head trauma, or a recent seizure. A second cause of an expanded blood pool occurs in vascular malformations where there is an increase in size and number of capillary vessels within the lesion nidus. A third cause for expansion of a vascular pool occurs in neoplastic lesions that generate angioneogenesis. Thus, lesions with an expanded vascular
Neuroradiologic Imaging
49
Figure 2-2. Arachnoid cyst on T2 sequence. An axial, T2weighted, spin-echo, MR image (through the same posterior fossa cyst in Figure 2-1) illustrates how cystic lesions exhibit bright (hyperintense to normal brain) MR signal. The internal cyst intensity is even greater than that of other cerebrospinal fluid–containing spaces, due to reduced normal pulsatile effects during the cardiac cycle; this limitation is created by the wall of the cyst. Figures 2-1 and 2-2 provide standard T1 and T2 information about the cyst nature of this lesion; however, to differentiate between epidermoid cyst and arachnoid cyst requires FLAIR or diffusion imaging, both of which are provided for illustration of their method in Figures 2-6 and 2-7, respectively.
pool, created by either an increased size of and/or increased number of vessels, can also enhance, because the additional vessel capacity physically can accumulate more contrast material than adjacent tissue. A second major cause for abnormal enhancement occurs when the vessels within a region either lack tight endothelial junctions or now leak because of a pathologic process. Enhancement on this basis occurs because of actual extravasation of contrast material out of the vessels and into the interstitium. This form of enhancement occurs normally in the pituitary gland (and stalk), the choroid plexus, and dural structures. However, brain structures whose vessels normally have tight endothelial junctions, and thus an intact bloodbrain barrier, do not normally enhance on this basis. Abnormal contrast enhancement occurs as the result of lost endothelial integrity, increased capillary permeability, and escape of contrast material into the interstitium. It may take the shape of a variety of forms including nodular (or mass like), gyriform (conforming to the cortex), superficial (conforming to the Virchow-Robin spaces), or just be
W9418-02.qxd 10/31/03 3:35 PM Page 50
50
Section I
Introduction
A Figure 2-4. Cerebral infarction with regional hyperemia. An axial, gadolinium-enhanced, T1-weighted MR image through the high convexity demonstrates abnormal enhancement within otherwise normal regional vessels (arrows) in the area surrounding the left frontal stroke illustrated in Figure 2-3A. In this instance, the increased enhancement represents dilated collateral arteries with expanded intravascular space rather than disruption of the blood-brain barrier. This is a second mechanism of abnormal contrast enhancement.
nonspecific. The anatomic configuration of the enhancement pattern can provide significant clues to the underlying pathologic entity (Figs. 2-3 and 2-4).
B Figure 2-3. A, Cerebral infarction. An axial, nonenhanced, T1weighted MR image through the high convexity of brain demonstrates a hypointense area of edema (arrow) consistent with brain infarction in the left frontal cortex. There is no spontaneous hyperintense signal in the area of infarction to suggest a hemorrhagic component. B, Cerebral infarction with focal enhancement. An axial, gadolinium-enhanced, T1-weighted MR image through the high convexity of brain demonstrates a focal area of abnormally increased signal (arrow) following gadolinium injection within the caudal aspect of the ischemic region consistent with altered blood-brain barrier and abnormal contrast enhancement. Disruption of the blood-brain barrier is a common mechanism for abnormal contrast enhancement.
Fast Spin-Echo (or RARE) Techniques The inherently slow relaxation process in standard spin-echo MR makes such imaging particularly susceptible to patient motion artifact. This potential for motion-related image degradation has spawned attempts to shorten the data acquisition process. Unfortunately, this is by necessity at the expense of tissue contrast resolution. In spin-echo sequences, the raw data acquisition must fill a matrix (called K-space) so a mathematical process can be applied that converts electrical signals from a time dimension into an intensity dimension, which provides image contrast resolution. It turns out that some of the data can be predicted rather than measured. This occurs in the commonly used “fast” T2 (or RARE) spin-echo sequences. By acquiring more lines of Kspace within each TR (repetition time) interval, the image acquisition speed of the examination is substantially shortened. The advantage is clearly to improve spatial resolution
W9418-02.qxd 10/31/03 3:35 PM Page 51
Chapter 2
Neuroradiologic Imaging
51
obtained by reducing patient motion–related image degradation. The trade-off is a reduced sensitivity to T2 tissue changes and reduced sensitivity to magnetic susceptibility artifacts. In our experience neither shortcoming appears to significantly lessen the acquisition of essential T2-weighted images as long as the “effective TE (echo time)” exceeds 100 msec.
Gradient-Recalled Echo Sequences All MRI sequences must refocus the inherent signal (of stationary tissue) before the data read-out. This refocusing is obtained with an RF pulse in spin-echo images, but is done with an applied secondary gradient in the GRE examination. For spin-echo image formation, the second 180degree RF pulse reverses most unwanted extrinsic phase shifts, which have occurred up to that point, but does not (by design) reverse the inherent relaxivity that has occurred. The GRE method of refocusing does not reverse extrinsic susceptibility phase effects, and therefore is subject to greater susceptibility artifacts than spin-echo sequences. However, we can use this apparent shortcoming to our advantage to identify one of the prime causes of extrinsic phase artifacts, hemosiderin. The presence of substances such as hemosiderin alters the local magnetic field producing magnetic inhomogeneity and signal loss within the affected voxels. Such intravoxel inhomogeneity creates a signal loss referred to as magnetic susceptibility artifact. Evidence of hemosiderin-related susceptibility artifact provides important diagnostic data because it is easily detected and implies previous hemorrhage. Selection of these GRE parameters must be specifically included in protocols whenever previous hemorrhage is being investigated, as in the diagnosis of cavernous angiomata and other occult vascular malformations. Gradient-recalled imaging has an additional role in volumetric analysis. Three-dimensional (3D) GRE imaging techniques provide contiguous sections without interval section gaps, which are necessary in standard spin-echo imaging. Volumetric data are used commonly in assessment of tumor analysis before radiation surgery, in hippocampal volume assessment in patients with partial complex seizures being considered for seizure surgery. The most effective GRE sequence for postprocessing of brain tumor data is generally a magnetization-prepared GRE. These in conjunction with gadolinium infusion provide excellent T1-weighted background to display the enhancement of the tumor (Fig. 2-5).
Fluid Attenuating Inversion Recovery Imaging with Strong T2 Weighting An innovative technique has been developed that uses a preliminary 180-degree RF pulse to invert the overall signal.5–9 By empiric methods it was possible to suppress the signal of CSF at the same time allowing tissue edema to appear hyper-
Figure 2-5. Example of gradient-recalled echo (GRE) type of T1 sequence. An axial, nonenhanced, T1-weighted GRE image through the mid-convexity of brain demonstrates anatomic detail similar to that of a standard spin-echo T1 sequence, but with some differences. Because this image has been obtained more rapidly than spin-echo image, there is less time to saturate the intravascular blood flow, and, as a result, larger vessels often exhibit spontaneously increased signal, as seen within the middle cerebral arteries and the internal cerebral veins in this example.
intense (bright by comparison to normal brain) on T2weighted images. The advantage of this technique is in the ability to detect tissue edema (T2 brightness) when the pathologic tissue is adjacent to CSF within cisterns or sulci or ventricles. In standard T2 sequences tissue edema is hyperintense but so is CSF, which makes distinction between the two very difficult (“lost conspicuity”). In fluid attenuating inversion recovery imaging with strong T2 weighting (FLAIR) imaging, which is a strongly T2-weighted sequence, CSF remains hypointense (or dark) while the pathology (containing edema) will be hyperintense (or bright), substantially improving conspicuity and lesion. FLAIR imaging is also very sensitive to blood in the subarachnoid and subdural spaces. Again the blood products provide a hyperintense signal relative to brain. FLAIR imaging has the disadvantage of taking longer to perform and therefore is more susceptible to patient motion. This disadvantage has been obviated in part by using the fast (RARE) scanning techniques in conjunction with the FLAIR parameters to achieve a scan time of less than 4 minutes (Fig. 2-6).
W9418-02.qxd 10/31/03 3:35 PM Page 52
52
Section I
Introduction
Figure 2-6. Arachnoid cyst on T2-weighted FLAIR sequence. An axial, T2-weighted FLAIR image through the mid-posterior fossa demonstrates the appearance of the same cyst shown in Figures 2-1 and 2-2. FLAIR images are heavily T2 weighted, but additionally, are performed is such a way that cerebrospinal fluid (CSF) generates a dark signal, while tissue edema generates a brighter signal. In this case the cyst fluid is homogeneous and matches CSF indicating that this mass is an arachnoid cyst.
Special Magnetic Resonance Imaging Techniques Newer technologic developments are constantly unfolding. Some provide new information and some improve contrast or spatial resolution. In this section we will provide an overview of these techniques and the clinical setting in which they are most likely to be used. Magnetic Resonance Diffusion Imaging This form of physiologic imaging relies on the presence of axoplasmic motion of bulk water during the interval between two applied, but opposing, field gradients.10–18 Normal motion of water (protons) within axons limits refocusing of the signal (the protons do not remain stationary during the intrascan interval). As a consequence, less signal output is observed and the background intensity diminishes. The term which describes the rate of diffusion during the time between the gradient pulses is called the apparent diffusion coefficient (ADC). A sequence that is generated from these coefficients is called an ADC map. A map of the consequence of the variability of such motion can also be generated. This is referred to as the diffusion map and it will have a value applied to it that indicates mainly the interval between the first and second diffusion gradient (the “b
value”). Motion of molecules in the brain is not uniform but is affected by the orientation of the white matter bundles. The diffusion gradients are first applied in all three directions (phase, frequency, and major field). The intensity data from the three individual sequences are summed and presented as a single trace image. Diffusion images suffer in one respect from being a T2-weighted sequence. When edema is present on the initial b = 0 sequence (performed without applying the diffusion gradients), the underlying T2 hyperintensity will “shine through” into the diffusion images and simulate abnormal restriction of water motion, which is also detected by a brighter than normal signal. The main, but certainly not the only, use for diffusion imaging is in acute stroke, especially in the hyperacute time frame when standard MRI and CT demonstrate no abnormality. Acute brain infarction axoplasmic motion is reduced, leaving the water protons in virtually the same position during the period between the applied diffusion gradients. Because no motion occurs, less image degradation occurs and the signal within the infarcted area is relatively stronger, which translates into a brighter shade of gray on diffusion images and reduced signal on the ADC (see Fig. 2-32). If pathology has greater motion, as in a cystic lesion, the reverse effect will be observed, as presented in Figure 2-7. Relative Brain Perfusion Imaging Relative brain perfusion can be performed with CT or MRI using a measure of time-concentration graphs following injection of a contrast agent and then monitoring the intensity change in the region of interest during the first pass of the contrast material.19,20 This is especially useful in delineating areas of reduced flow as in acute ischemia, but also in focally increased flow in hypervascular masses as hemangioblastoma and hypervascular metastases, as melanoma or choriocarcinoma. Evidence of increased intratumoral blood volume compared to uninvolved gray matter supports the diagnosis of a higher grade tumor (anaplastic and glioblastoma categories). These methods are qualitative and not quantitative at present but this will be resolved as new applications evolve. We typically standardized our cerebral perfusion to the perfusion of the rostral cerebellum and vermis. Vermic perfusion is seldom restricted with the exception of intercurrent basilar or superior cerebellar artery thrombosis, and hence provides a benchmark to compare areas of altered cerebral perfusion. Magnetic Resonance Angiography and Magnetic Resonance Venography20–22 Arterial and venous MR angiography can be obtained by using the natural magnetic properties of flowing blood, referred to as time-of-flight (TOF) MR angiography (MRA) or magnetic resonance venography (MRV). In standard imaging, the inherent signal of flowing blood in the carotid arteries and superior sagittal sinus is suppressed by a presaturation RF pulse, applied outside the area of interest but in
W9418-02.qxd 10/31/03 3:35 PM Page 53
Chapter 2
Neuroradiologic Imaging
53
A B Figure 2-7. A, Arachnoid cyst on diffusion sequence. An axial, trace (b = 1000) diffusion map through the low convexity of brain demonstrates the appearance of this same arachnoid cyst, as seen in Figures 2-1, 2-2, and 2-6. The cyst is hypodense (implying completely unrestricted water motion) and relatively homogeneous indicating no septation within the cyst. This response is mirrored on the ADC map of B. B, Arachnoid cyst on apparent diffusion coefficient (ADC) sequence. An axial, ADC image through the posterior fossa demonstrates bright signal (*). This implies that the coefficient of motion is high; the water motion is unrestricted. We would expect just the opposite effect in a brain stroke where the apparent diffusion coefficient would be low because tissue water motion is lost. An example of stroke is presented for comparative purposes in Figure 2-32.
the range of antegrade blood flow. Most commonly the presaturation pulse is applied over the neck and vertex of the head in standard head imaging to reduce flow-related signals in the brachiocephalic arteries flowing superiorly and the sagittal sinus blood flowing inferiorly. These presaturation pulses are routinely used to obviate confusing vascularrelated phase shift signal artifacts. In MRA and MRV, a reciprocal process is used. Instead of suppressing flowing blood, suppression of stationary tissue in the brain is allowed, making the inflowing blood appear relatively brighter. The procedure can be altered to emphasize inflowing arterial blood, MRA, or out-flowing venous sinus blood, MRV. Background suppression is further improved using magnetization transfer techniques. Water molecules can exist freely (“unbound water”), as in a fluid compartment, or they may be trapped within the interstices of larger, complex molecules even though they are not chemically attached. This state is described as “bound water.” However, there is a constant exchange of water molecules between the bound and unbound compartments. As it turns out, bound water precesses at a slightly different rate than unbound water. By applying an appropriate off-resonant RF pulse immediately before the imaging sequence, it is possi-
ble to saturate only the bound water molecules thereby obliterating their MR signal. When an exchange of water molecules occurs between the saturated, bound, and unbound molecules, a net reduction in overall background signal is observed for either T1 or T2 sequences. This has been reported to be approximately a 37% reduction in background signal. The percentage drop in signal intensity is less than that in pathologic tissue presumably related to increased regional free water either within the neuropil or within cells. MRA can also be performed using two-dimensional GRE images obtained sequentially while injecting gadolinium intravenously. This works for imaging of the aortic arch and the cervicobrachial arteries, because the examination can be performed with most sections obtained during the first pass arterial opacification phase. After a few more seconds, vein opacification occurs, which obscures arterial detail. This form of MRA is more anatomically correct and suffers fewer artifacts than time-of-flight MRA. Its value is providing high quality noninvasive (other than the intravenous contrast injection) evaluation of the carotid and vertebral arteries in the context of transient ischemic attacks and stroke (Fig. 2-8).
W9418-02.qxd 10/31/03 3:35 PM Page 54
54
Section I
Introduction
Figure 2-8. Time-of-flight (TOF) MRA. This TOF angiographic image demonstrates the normal appearance of both the proximal carotid and vertebrobasilar arteries. TOF imaging depends on the fast inflow of arterial blood to generate adequate signal. Slower flowing blood is saturated during the MR acquisition and will not generate sufficient signal to be visualized.
Magnetic Resonance Venography Examination of intracranial veins and dural sinuses is possible using the same techniques as discussed previously under time-of-flight MRA. The difference is only where the presaturation pulse is applied. In MRV, arterial flow is, by necessity, suppressed by applying a presaturation pulse over the neck. This saturates (essentially eliminates) the flow signal from the carotid and vertebral arteries. The only flowing blood is from the dural sinuses and internal cerebral veins and some major cortical veins. This forms the basis of the MR venogram. It, like MRA, is entirely noninvasive. The only limitation is that where dural sinuses parallel the plane of imaging acquisition some saturation can occur; this most often occurs in the superior sagittal sinus just before the torcula. The other limitation is related to interpretation of the widely disparate variations of normal cerebral venous drainage (Fig. 2-9). Functional Magnetic Resonance Imaging Functional MRI (FMRI) requires rapid acquisition of whole brain MR data in the same time frame that the physiologic alterations are occurring, generally in the range of subsecond brain sections.23–28 The physical basis for current FMRI methods relies on regional changes in blood flow occurring in response to a particular external stimulus. By applying specific neurophysiologic testing while concurrently imaging with rapid MR acquisitions, alterations in regional vascularity can be detected based on the concentration of deoxyhemoglobin in the blood. The stimulus alters neuronal activity affecting regional metabolism, which in turn increases
Figure 2-9. Time-of-flight (TOF) MRV. This TOF venograph demonstrates normal dural sinuses and normal deep venous drainage pathways. Venography requires intentionally saturating the fast inflowing arterial circulation, which by default emphasizes the slower flowing venous efferent circulation.
regional blood flow. Increased local perfusion rates expand local blood volume, while maintaining the same, or even less, oxygen extraction fraction. The net effect is reduced concentration of deoxyhemoglobin. Because deoxyhemoglobin is paramagnetic and normally reduces tissue signal, then obviously less of it in the activated area would contribute to relatively more signal. This is in fact what happens. The clinical challenge is detected by a localized reduction in deoxyhemoglobin compared to the proportion of oxyhemoglobin. This process is called blood oxygen level–dependent imaging. Current use in oncology remains limited but the technique is finding early utilization for motor strip localization before tumor resection when the mass is located near this portion of eloquent cortex. Magnetic Resonance Spectroscopy MR proton spectroscopy is obtained in much the same way routine MR but does not emphasize the proton water peak and is not used to define anatomic data.29–41 However, two differences exist. The applied gradients, which normally establish anatomic location in space, are limited to only a single area or multiple areas in the same axial region. Then the large water peak within the spectra is suppressed, because water is not of interest. Once the large water peak is suppressed, the smaller peaks of organic compounds can be visualized. These include peaks for N-acetyl aspartate, lactate, creatine, and choline to name a few. The heights of these peaks provide insight into the biologic state of the
W9418-02.qxd 10/31/03 3:35 PM Page 55
Chapter 2
tissue being studied. In this way differences in spectra related to pathology emerge. Tumor appears different from radiation necrosis, for instance. Currently, spectroscopy can be used to differentiate the location of tumor versus radiation
55
effects when there is imaging discordance between the tumor site and other areas of abnormal signal in the treatment field. Many other nontumor assessment uses are being explored (Fig. 2-10).
Figure 2-10. A, Brain tumor spectroscopy. An axial, gadolinium-enhanced, T1-weighted MR section through the mid-convexity of brain demonstrates a focal tumor mass (arrow) in the major forceps of the right spleniumcorpus callosum. The central portion of the mass exhibits typical abnormal contrast enhancement produced by the retained proton-relaxing effects of the gadolinium. B, Single voxel proton spectroscopy. Proton spectroscopy provided through the same lesion (illustrated in A) demonstrates the typical spectroscopic features of an anaplastic, nonnecrotic brain tumor. In this instance, there is less-than-normal amount of N-acetyl-aspartate (the “N” peak) and elevation of the choline concentration (the “C” peak). A
B
Neuroradiologic Imaging
W9418-02.qxd 10/31/03 3:35 PM Page 56
56
Section I
Introduction
A
B Figure 2-11. A, Pre–fat suppression in optic neuritis. This coronal, enhanced, T1-weighted MR image through the retroglobal orbital area demonstrates a posterior orbital structure, including the rectus muscle sheath and the optic nerve complex on both sides. The intensity of the right optic nerve (arrow) appears symmetric with the left nerve; no apparent abnormal enhancement. No fat suppression has been applied to this image. B, Fat suppression in optic neuritis. This coronal, enhanced, T1-weighted MR image with fat suppression through the same retroglobal orbital area as illustrated in (A), but now including a fat-suppressive prepulse. The study now demonstrates evidence of optic nerve enhancement in the right optic nerve (arrow). The fat suppression allows this subtle abnormal contrast enhancement to become diagnostic of optic neuritis in this case.
Fat Suppression Techniques Fat suppression is possible on MRI by using a RF presaturation pulse before the data acquisition that is directed at the spin frequency of lipids.42 In doing so, the normally bright fat signal on T1 sequences is substantially reduced. This is a necessary process when trying to detect gadolinium contrast enhancement in a structure (like the optic nerve) that is surrounded by retroglobal fat. By using fat suppression, the conspicuity of the enhancement is substantially improved. Several other fat suppression methods are also available (Fig. 2-11). Cerebrospinal Fluid Flow Studies The pulsatile movement of CSF can be detected using phasesensitive imaging where the direction of the CSF movement affects whether the CSF is portrayed as a positive signal (hyperintense to background) or negative (hypointense to background).43 MR CSF flow studies have value in selective instances. These include determination of blockage or restriction of CSF passage in the ventricles, or in cisterns, especially at the foramen magnum. A common use in our institution is assessment of CSF block at the foramen magnum in the context of Chiari 1 malformations, assessing whether there is a significant foramen magnum impaction
syndrome. Other intracranial uses include determining ventricular blockage at other levels and whether a third ventriculostomy (performed endoscopically) is patent (Fig. 2-12).
Restrictions to the Use of Magnetic Resonance Imaging Pregnancy The current policy for MR examinations during pregnancy states “patients suspected of being pregnant will not be scanned unless the risk/benefit has been evaluated by an attending Radiologist” “If the MR is indicated, an informed consent will be obtained by a Radiologist.”44–51 This policy is based on a recent review of the literature, in which some animal studies report teratogenic effects following MRI exposure performed during the first and second trimester. It should be emphasized to the patient that the MR effects on the human fetus are unknown. Our policy includes the following: “If a patient is known to be pregnant, the MR examination should only be performed if there is clear medical benefit to the patient or fetus, and the examination is deemed medically necessary despite
W9418-02.qxd 10/31/03 3:35 PM Page 57
Chapter 2
A
Neuroradiologic Imaging
57
B
Figure 2-12. A, Cerebrospinal fluid (CSF) MR flow study: positive phase. Phase-contrast CSF flow studies are dependent on the direction of the CSF pulsation during imaging acquisition, which in this circumstance is in a positive direction (bright signal) indicating a cephalad direction. B, CSF MR flow study: negative phase. The flow out of the fourth ventricle is highlighted in this and the following image (arrows) to demonstrate the cyclical change of pulsatile flow, but CSF flow in the prepontine cistern is also apparent. This phase-contrast CSF flow study illustrates the CSF pulsing in a negative direction (dark signal). Observation of the CSF pulsations reflected in the change between positive and negative signal on a video loop allows for a reasonable assessment of the movement of CSF. This type of study is usually used in ventricular or cisternal CSF flow restrictions.
the pregnancy.” For example, the MRI examination would provide information that potentially would be used to change the treatment of the mother or fetus during the pregnancy. This risk-benefit assessment should be discussed with the referring clinician and a radiologist from the service that will be interpreting the examination. It should be emphasized to the referring clinician that there is insufficient data to support, or refute, the safety of MRI during pregnancy, particularly in the first and second trimester. Once the radiologist from the service that will be interpreting the examination agrees that MRI is medically necessary, then that person will: 1. Inform the MR technologist/center that the study has been approved. 2. Give the MR technologist/center the reason for the medical necessity of the case and the name of the referring clinician with which the case was discussed. This information will be recorded by the MR technologist. 3. When the patient undergoes MRI, this information will be written on the consult sheet by the MR technologist so that it may be included in the official dictation by the radiologist interpreting the study. This information will allow the radiologist interpreting the study to defend the medical necessity of the examination at a future date if
an unexpected fetal or postnatal event occurs after the MR examination. Gadolinium contrast administration can be administered only under certain circumstances at the discretion of the radiologist supervising the examination. A conservative approach to the use of contrast administration during pregnancy is recommended. Because gadolinium readily crosses the placenta and is excreted by the fetal kidneys, it ultimately passes through the bladder and into the amnion where it circulates and is swallowed and reabsorbed repeating the process over and over. Thus, the clearance rate of MR contrast agent from the amniotic fluid and fetal circulation is prolonged during which the chelation of the gadolinium can disassociate. The effects of elemental gadolinium on an embryo or fetus are unknown. The manufacturers of contrast material do not support the use of MR contrast agents in pregnant women. Obesity Obesity or a large body habitus presents a problem in MRI. The patients’ body cannot touch the inner surface of the MR scanning gantry or substantial image distortion will result from magnetic interference. To combat this problem, manufacturers have developed large, open bore solid magnets
W9418-02.qxd 10/31/03 3:35 PM Page 58
58
Section I
Introduction
that are capable of handling larger patients. The trade-off is that these are typically permanent magnets running at lower Tesla strength, usually 0.3 T. This limits both the spatial and contrast resolution of the images. Newer, short bore cryomagnets are now being introduced with a shorter area of confined space allowing a larger patient to be imaged.
An Uncooperative Patient MRI suffers significantly from patient motion, because data are acquired sequentially over time and accumulated in a matrix, called K-space. Once the K-space per single phase is filled, the machine then acquires data from the second and so on for 128 or 256 or more phase shifts per slice. Only after the full K-space data set is filled out does mathematical manipulation occur to generate anatomic and intensity information for the image. Any intervening motion of the patient during the image acquisition process actually changes the phase data, which ultimately misrepresents the data when the final mathematical transform is applied. This can create both anatomic and intensity misregistration. Data misregistration means anatomic structures will be placed where they do not exist and intensity misinformation will appear abnormally (spurious pathology). MRI is not like CT scanning where the consequence of motion is merely an unsharp or fuzzy image. Because patient motion artifact has such adverse consequences, any patient likely to be uncooperative (pediatric patients, patients with reduced level of conscience, etc.) should be controlled, monitored and sedated, either with intravenous (IV) medication or with intubation and general anesthesia. The latter presents additional challenges and is not a trivial matter, because the magnetic field precludes the use of all routine anesthesia and intravenous injectors, and monitoring equipment. MRI-compatible anesthesia and monitoring equipment must be purchased to deal with this issue. In our experience general anesthesia should be considered early and not delayed to see whether routine sedation fails.
A Claustrophobic Patient The scanning gantry of most MR scanners used currently are fairly long and tubular in shape. Virtually every patient experiences some degree of claustrophobic response when they are first placed within the gantry. Most patients are able to tolerate such an experience, but not all. Sedatives and antianxiety drugs, such as the benzodiazepines, can be used to limit such claustrophobic effects. However, a certain proportion of patients will be intolerant to MRI because of claustrophobia. The possible solutions for this include scanning within an open bore magnet (widely available), scanning within a newer short bore magnet (currently less available), or use of anesthesiology.
Restriction of Gadolinium Usage Gadolinium is a relatively strong paramagnetic rare earth element, with seven unpaired electrons, which contribute to its effect of proton relaxation enhancement on T1 sequences. Gadolinium is a heavy metal that is toxic. However, the manufacturers of the contrast medium have attached it to a variety of chelating agents, which bind the gadolinium molecule, making it quite safe in most patients. There are reported idiosyncratic reactions to all drugs, including gadolinium compounds, but these are very rare with an incidence in the neighborhood of one per million patients injected. Minor reactions are reported, but for the most part are far less than those of IV iodinated contrast medium. The major current restriction to gadolinium use is in pregnant women, where the gadolinium crosses the placenta, is ingested by the fetus, absorbed then excreted back into the amniotic fluid, where the process repeats itself. Ultimately, disconjugation of the heavy metal can occur from the chelating agent. At this point, the gadolinium becomes a potential hazard. Effect of Corticosteroids on Contrast Enhancement One of the mechanisms for abnormal contrast enhancement is disruption of the blood-brain barrier and extravasation of contrast medium into the neuropil. This mechanism occurs similarly for both intravenous iodinated contrast material in CT and for paramagnetic contrast material in MRI. Corticosteroids, often given to stabilize inflammatory processes, have the ability to stabilize the blood-brain barrier and close the altered endothelial tight junctions. This has an impact on imaging where the presence of abnormal contrast enhancement is often a critical feature in detection of abnormal pathologic conditions. Hence, it is necessary to alert the imager who intends to interpret MRI or CT scans that therapeutic levels of steroids have been given, and that this might prevent or at least alter the appearance of the contrasted portion of the imaging study. This effect of corticosteroids has been reported with a wide variety of lesions but especially lymphoma and inflammatory processes (Fig. 2-13). Patients with Implanted Metal Any metal that has magnetic susceptibility (especially iron) will create a problem for MRI. Several factors require discussion. The high magnetic susceptibility of any ironcontaining metal object means that it is affected by the main magnetic field and can torque. If the metal is securely fastened, such motion effects would be insignificant. If, on the other hand, the object is moveable, as it might be with an aneursym clip, then the torque effect might be disastrous. The second consideration is the effect of heating of the metal. The RF pulses used to activate the spin system are, in many respects, similar to those of a microwave oven. Just as no metal should be put into a microwave because of severe heating effects, the same is true for metal in a MR scanner.
W9418-02.qxd 10/31/03 3:35 PM Page 59
Chapter 2
Neuroradiologic Imaging
59
A B Figure 2-13. A, Effect of steroid therapy in radiation therapy—related contrast enhancement (before treatment). An axial, gadolinium-enhanced, T1-weighted MR image through the posterior fossa demonstrates multicentric areas of abnormal contrast enhancement (arrows). This represents areas of altered blood-brain barrier inferior secondary to cranial radiation therapy. B, Effect of steroid therapy in radiation therapy—related contrast enhancement (following corticosteroid treatment). An axial, gadolinium-enhanced, T1-weighted MR image through the posterior fossa demonstrates no residual abnormal enhancement in the areas seen previously to enhance (see A). The difference in the enhancement pattern is related to stabilization of the blood-brain barrier following corticosteroid therapy.
Heating of the object can be rapid and cause burning of the surrounding tissue. The third effect of the metal would be to alter the local magnetic field. This causes the scanner to misregister anatomic data because the overall magnetic field within the scanning zone is not uniform and hence not predictable. Metal, therefore, typically causes a severe susceptibility artifact, usually rendering the image unreadable. Not all metal has susceptibility properties. Newer metals, like titanium, have little magnetic susceptibility and, hence, create little MR artifact. However, even with these metals, there is a question of whether repeated sterilization, as occurs when aneurysm trays are resterilized between cases, may change the magnetic properties of the clips. The final verdict in this issue remains to be determined.
lists many devices and their MR safety qualifications. In addition to problems related to the metallic content of internal devices, there are issues related to function of the devices. For instance, cardiac pacemakers can malfunction in a strong magnetic field. Therefore, the presence of a cardiac pacemaker is a contraindication to MRI. Morphine pumps for chronic pain therapy can be emptied (to be safe) and then the patient can undergo MRI. The third safety issue is whether the torque effects on the implanted device would cause injury. For instance, ossicular stents can be dislodged, so patients with these devices cannot be studied with MRI.
Pacemakers and Implanted Devices Do not assume any implanted metallic device can be safely scanned. The number of potential devices is endless. Manufacturers of a device for the most part, provide MR compatibility information and usually this information can be accessed from the company directly by telephone or website. There is a noncommercial website, www.mrisafety.com/, that
Introduction
Understanding Computed Tomography
Computed tomography (CT) is a commonly used means of intracranial evaluation.52 But as with MRI, the imaging protocol must match the indication. Imaging protocols are designed to focus on a specific clinical issue and to achieve the most diagnostic information possible during a
W9418-02.qxd 10/31/03 3:35 PM Page 60
60
Section I
Introduction
“one patient” sitting. CT has the advantage of rapid imaging time. The newer scanners have substantially improved heat loading capacities to allow continuous (spiral or helical format) imaging. The principles of CT are similar to those of standard radiography, except that the former uses stationary detectors rather than radiographic film to capture the image. CT uses an x-ray beam and counter-balanced x-ray detector bank situated within the outer ring of the scan gantry. As with traditional radiographs, the degree to which the incident beam is absorbed or scattered determines the radiographic density of the structure. When an x-ray beam enters two separate but contiguous structures, the structure that is composed of the densest material (i.e., has the highest electron density) will absorb more of the beam than its neighbor allowing fewer x-ray photons to reach the detectors. A diminished detector signal is translated into a lighter shade of gray on the image gray scale than that of its less dense neighbor. Such differential x-ray absorptive capacity provides a means of tissue differentiation based on contrast resolution. Once tissue can be resolved on the basis of inherent tissue density then anatomic detail, or anatomic resolution, can be added to the diagnostic capability. The opposite is also true. Insignificant density differences between contiguous structures preclude anatomic delineation. When two larger structures are separated by a smaller intervening structure of different density, the larger structures can be distinguished only if the resolving capacity of the scanner is adequate to detect the smaller of them. Resolution on this basis is referred to as spatial discrimination, and is based mainly on slice thickness. If the resolving ability of the scanner is inadequate, tissues are blended together in a process called partial volume averaging. Spatial resolution, or the ability to distinguish small adjacent structures, is affected by the quality of the x-ray beam collimation, the number and size of the detectors, and the thinness of the beam itself. The thinner the slice, the greater the detail. However, the negative trade-off for thin sectioning is an increased radiation dose (approximately 3 rad per slice for routine imaging). The increased imaging time and x-ray tube heating requirements for more slices has substantially improved with the latest generation of CT scanners. However, there still is a finite limit determined, in part, by x-ray tube limits and, in part, by the time required for the patient to remain still within the scanner.
Filters, Image Segmentation, and Postprocessing CT can be performed with a variety of post-acquisition electronic filters. Filters selectively add or remove various frequencies from the raw data and change the limits of the gray scale, producing either a smoothing or an accentuation of the edges (or borders) of structures, as needed. Filters
are applied to the raw data during post-acquisition manipulation. Postprocessing allows the same set of image data to emphasize either bone or soft tissue contrast. The raw data can be adjusted through image segmentation to deal with the effects of patient motion. Not all the data are needed to produce an adequate image. When motion occurs in only a portion (i.e., a segment) of the tube swing, the data from that portion can be deleted. Segmentation significantly improves the final images in patients who cannot, or will not, lie still.
Computed Tomography Contrast Medium Both ionic and nonionic forms of contrast media are available. The differences between them in terms of contrast enhancement of lesions are minimal. However, other differences between them are distinct and important. Ionic contrast agents are much less costly than are the nonionic forms. While the incidence of severe idiosyncratic complications (i.e., shock or death) is minimal with either type of contrast agent, the nonionic contrast agents have substantially fewer secondary effects. Structurally, the monomeric types of ionic contrast media link a sodium or meglumine cation with an iodine-containing (benzene-ring) anion. A dimeric form of ionic contrast agent links two (iodine-containing) anion molecules together with only one cation. The dimeric contrast medium, thereby, achieves a higher concentration of iodine for the same degree of osmolality, making it less irritating on IV injection. Nonionic forms of contrast agents have the benefit of lower osmolality and yet deliver the same number of grams of iodine per unit volume as the monomeric ionic contrast agents. These attributes result in fewer contrast material– related symptoms. The drawback of nonionic contrast agents is in the area of viscosity. At body temperature, they are nearly twice as viscous as ionic contrast media. Idiosyncratic reactions, although reduced in frequency, still occur with nonionic contrast medium, nonionic contrast agents have significantly fewer or, possibly, no significant anticoagulation effects on blood compared with the ionic contrast agents. This difference has relevance during arterial injections during angiography but has little relevance to intravenous administration during CT. The major advantage of using a nonionic contrast medium is that it causes less tissue injury if it is inadvertently extravasated during an intravenous injection. Most enhanced studies are performed with power injectors to achieve the correct timing and volume of contrast. Thus, the protection of tissues surrounding the injection site is an important consideration in the selection of a contrast medium. Several techniques can be used to achieve contrast enhancement. The standard technique administers the contrast medium slowly, as an infusion for roughly 15 minutes before initiation of the actual CT scanning. This allows time
W9418-02.qxd 10/31/03 3:35 PM Page 61
Chapter 2
Figure 2-14. Abnormal CT enhancement in metastatic tumor. An axial, nonionic, contrast-enhanced, CT scan through the high convexity of brain illustrates the features of abnormal CT contrast enhancement within a metastatic lesion (arrow). The basis of abnormal contrast enhancement on CT is similar to abnormal gadolinium enhancement on MRI. Both are the result of either blood-brain barrier disruption or expanded intravascular spaces.
Neuroradiologic Imaging
61
intravenous contrast whether they have previously received a contrast medium and whether they experienced any untoward reaction to it. If there has been a severe reaction, as described previously, then contrast material should not be given unless absolutely necessary. If critical, a pretreatment regimen (see following discussion) plus anesthesiology stand-by is suggested. Despite severe reactions to intravenous contrast material, the same contrast medium injected intraarterially may not produce the same reaction. There are many less dramatic contrast reactions and effects. These usually include urticaria (especially around the mouth) and mild bronchospasm with wheezing. These types of reaction can be the harbinger of a more severe reaction. Hence, they are important and intravenous contrast material should not be given without pretreatment. The current accepted pretreatment for contrast material reactions, excluding patients with severe reactions, is prednisone 50 mg given at 13, 7, and 1 hour before the procedure plus diphenhydramine hydrochloride (Benadryl) 50 mg given 1 hour before contrast material injection. If a patient is currently taking metformin hydrochloride (Glucophage) for control of diabetes, then IV iodinated contrast material should not be given unless the drug has been stopped at least 24 hours before the study. If metformin has not been stopped on the day of examination, it must be discontinued for 48 hours after the contrast material has been given. A renal laboratory screen should be obtained before reinstituting the drug. Non-immune–Mediated Contrast Effects. While idiosyncratic
for the contrast agent to recirculate and diffuse into the interstices of the pathologic region. This routine injectionto-scan interval is optimal for most tumors, subacute strokes, and inflammatory disease. It is not optimal for lesions that leak contrast material very slowly or those with rapid arteriovenous shunting. These lesions require a different timing sequence that could include either delayed imaging or first pass dynamic imaging (Fig. 2-14). Restrictions for Use of Iodinated Contrast Medium There are restrictions on the use of all IV contrast agents, but these are particularly important for CT contrast medium. Several issues become relevant based on the types of potential contrast effects. Idiosyncratic Reactions: Major and Minor. The most severe reactions that occur with intravenous contrast include an idiosyncratic anaphylactic event, major upper airway edema, and direct cardiac toxicity effects. All can result in major cardiopulmonary complications. Hence, it is always important to ask patients undergoing CT procedures performed with
responses are uncommon, physiologic contrast effects are not. All contrast effects are diminished with the use of nonionic contrast media. Contrast effects are, in part, related to the osmolality or concentration of the agent and, in part, to the contrast or to its carrier directly. As a result, some reactions that are related to the direct effects of the contrast agent can be severe. These mainly affect cardiac rhythm and function. Hence, cardiac arrest can occur with IV contrast medium injections that are not idiosyncratic (immunemediated) but are related to direct drug toxicity. These are very unusual and occur with a very low incidence. Renal Toxicity. Because CT contrast agents are excreted by the kidneys and, when concentrated, have renal toxicity, the status of renal function is important. Relative contraindication to IV contrast agents include borderline renal insufficiency, diabetic patients, or patients with one kidney plus borderline renal insufficiency, and patients younger than 6 years of age. Absolute contraindication for IV contrast agent is evidence of overt renal failure (creatinine serum level > 2) and no plans for dialysis within the next day or two. If dialysis is available within 24 to 48 hours, contrast material may be given. Contrast Material Extravasation. One of the difficulties in
the administration of contrast material is the necessity
W9418-02.qxd 10/31/03 3:35 PM Page 62
62
Section I
Introduction
of obtaining and maintaining venous access. Most contrast agent injections are made at rates of approximately 1 cc/sec. Dynamic scanning requires 3 to 4 mL/sec. The bolus of contrast material is delivered most effectively by mechanical contrast injectors. The problem with this is obvious: if the contrast agent extravasates during the injection, the process cannot be stopped immediately. The effects of extravasated contrast agent depend on the type of contrast medium being used. Ionic contrast is capable of causing severe local pain and possibly local tissue necrosis. Nonionic contrast may cause some regional pain, but has the major advantage of not causing tissue necrosis when extravasated. Use of Computed Tomography and Intravenous Contrast in Pregnant Women. CT is used judiciously in pregnant women
or potentially pregnant women. There should be a clear and pressing need for such imaging. Scanning with CT is not contraindicated in pregnant women, including the pelvis, provided there are no other means of addressing the clinical problem. Head CT is performed routinely with use of pelvic shielding. Intravenous contrast administration, likewise, is
not contraindicated by pregnancy. Again, there ought to be a compelling clinical need for such data, and the total contrast load should be minimized. We do not routinely obtain an “informed consent” for pre- and postcontrast CT examinations of head.
Specialty Computed Tomography Sequences Helical Computed Tomography Scanning The most recent advance in CT technology uses continuous spiral or helical image acquisition.53,54 Helical CT uses a constantly revolving x-ray tube during which the table is indexed, or incremented, in the desired direction. The raw data are later subdivided to produce sections as thin as submillimeters in size, if necessary. This is possible because advances in x-ray tube and detector design allow for substantially improved heat-loading capacity. The advantage of helical scanning is markedly improved image acquisition speed, now producing anatomic submillimeter slices in 1 second or less. Furthermore, there is an isotrophic dataset that allows for excellent multiplanar (MPR), maximum
A B Figure 2-15. A, Multiplanar reformation (MPR) for anterior communicating artery aneurysm. An axial, gradient-recalled echo (high-speed gradient acquisition) single section representative of one of many contiguous images demonstrates the presence of an aneurysm. By using the whole block of images, the MPR software can reformat the data into multiple thin section in other planes (as in B) without actually having to repeat the scan process. These MPR reformations have the advantage of being capable of being performed in oblique projections, which is useful in “straightening” curving objects, especially arteries. B, MPR in sagittal plane for anterior communicating artery aneurysm. This sagittally reformatted MPR is able to delineate the features of the aneurysm (arrow) in sagittal plane. By combining several MPR projections, all aspects of the aneurysm and its relationship to its parent artery can be appreciated.
W9418-02.qxd 10/31/03 3:35 PM Page 63
Chapter 2
intensity projection (MIP), and 3D surface–rendered image reconstruction. The drawback of this procedure is less image data acquired per slice. In standard 2D image acquisition, the CT tube acquires data throughout its entire arc, which we place at 360 degrees for discussion purposes. All the 360-degree tube arc data goes into producing that one section. With helical scanning the table moves during this acquisition process. Thus, less than 360 degrees of data contributes to each slice, and some image detail is necessarily lost. The trade-off becomes one of lessened image detail versus improved image acquisition speed, which translates into less patient motion artifact per slice. As it turns out, this advantage of speed combined with thinner slice thickness, in most instances, outweighs the effect of lost detail. At comparable slice thickness, standard 2D images have more information per slice than helical sections of the same thickness. When submillimeter pathology is sought, especially for instance in the temporal bones, standard imaging should be requested. In most other circumstances, helical scanning will be advantageous. Image Reformations and Computed Tomography Angiography Thin CT sections obtained with helical scanning and performed during a rapid infusion of contrast material produces a dataset that contains vascular information which can be reformatted using several methods, MPR (multiplanar single slice reformation), MIP (maximum intensity projection into a 2D combined image), and 3D surface rendering. To the less sophisticated angiographer, the vascular detail on the 3D and 2D MIP images are the most easily appreciated. However, the most anatomic detail is evident on the MPR images. MPR reformations do not sum the data, but rather allow for reformatting of single sections. The true advantage to MPR is its ability to reformat into any oblique plane necessary to accurately displace the long axis of a vessel, even when it is tortuous. MIP, or maximum intensity projection, image reformatting sums the axial section data and reformats the data into a 2D look at the vascular tree. It also allows for subtraction of background. Having been summed in this manner, the entire 2D image can then be rotated into other planes for interpretation. This is a reasonably good way to portray the data providing accurate background subtraction can be performed. This process requires that anatomic judgements be made and is very labor-intensive. The background subtraction process is subject to error created by attempting to remove bone and calcified structures when they are contiguous to opacified vessels of the same CT density. The same error is created in subtracting the background for 3D surface–rendered images. This method exploits surface-rendering software to create a 3D version of the vascularity in question. It loses internal detail and also loses information from the subtraction of background.
Neuroradiologic Imaging
63
These limitations and the need for labor intensive postprocessing to create the image make both MIP and 3D rendering less practical, especially in acute care situations (Fig. 2-15). Computed Tomography Brain Perfusion CT perfusion can be a useful means of evaluating the presence of regional brain oligemia. Helical scanning combined with dynamic contrast infusion can be exploited to produce a time-concentration data set. By comparing the CT density curves over time in similar regions of brain, a graph of the contrast medium concentration can be created, which in turn, reflects intravascular perfusion. By comparing similar areas in one hemisphere with comparable regions in the opposite hemisphere, a relative brain perfusion analysis can be obtained. Reduction of greater than 20% is considered oligemic. It is uncommon to simultaneously have oligemia in both cerebral hemispheres and in the cerebellum. Thus, the reliability of the CT perfusion is increased by normalizing the cerebral blood flow to the cerebellar blood flow and vice versa. This is much the same method used to define relative brain perfusion using IV technetium-99m and gamma camera analysis. CT perfusion in the next software update is expected ultimately to provide actual brain perfusion analysis in milliliters per 100 g of tissue per minute as can be achieved with Xenon-CT. The clear advantage of CT perfusion and CT angiography is the speed of data acquisition. The primary area of use currently is in the evaluation of hyperacute stroke. Thrombolytic therapy has a window of opportunity of roughly 6 hours after onset of symptoms. Intra-arterial thrombolysis is currently valuable in treating acute proximal middle cerebral artery and acute basilar artery occlusions. The presence of either of these can be easily determined from CT angiography, and detection of significant arterial zone oligemia can be determined by the CT perfusion. Thus, the critical branch points in the thrombolytic treatment scheme can now be obtained on a helical CT scanner within 10 minutes, including all postprocessing. CT perfusion evaluates relative brain flow. It does not provide actual flow values and does not appreciate the full extent of actual brain infarction because of its focused nature. If intra-arterial clot lysis is not being considered and proof of the extent of injury for IV clot lysis is sought, then the drug can be started and diffusion MRI can be performed. This whole treatment scheme is in flux and this method should be viewed in the same manner (Fig. 2-16).
Which Modality to Use? This discussion will be mostly subject to change as old techniques evolve and newer techniques appear. However, for the present, the strengths and weakness of a variety of widely available techniques for MRI and then CT are presented.
W9418-02.qxd 10/31/03 3:35 PM Page 64
64
Section I
Introduction
Figure 2-16. CT perfusion. This technique is performed by repeating CT sections at subsecond speeds through a predetermined slab of brain and then comparing the first pass timeconcentration curves of intravenously injected contrast in symmetric portions of brain. This study can be obtained in less than 2 minutes and provides qualitative information about cerebral blood flow. New software will soon be available for quantitative CT perfusion.
Figure 2-17. STIR imaging in cortical dysplasia. High detail inversion recovery (STIR) image with strong T1 weighting demonstrates abnormal gray matter signal and abnormal gray matter thickness in the left occipital lobe (arrows). These are features of the cortical dysplasia. Such features are often hard to detect on standard spinecho imaging. This inversion recovery technique allows for thinner sections and accentuation of the gray and white matter T1-weighted contrast differences.
W9418-02.qxd 10/31/03 3:35 PM Page 65
Chapter 2
Available Magnetic Resonance Imaging Sequences Standard T1-Weighted and T2-Weighted Sequences T1- and T2-weighted information is necessary for evaluation of virtually every abnormality. Spin-echo techniques provide the most spatially resolved information. However, other imaging sequences provide special information. While spin-echo proton density sequences have been largely supplanted by FLAIR imaging, proton density gradient–recalled imaging is especially sensitive to the presence of hemosiderin and is used routinely in our vascular malformation workup to detect cavernous angiomata. Standard Magnetic Resonance Imaging Plus Intravenous Gadolinium The addition of intravenous gadolinium is warranted in many circumstances. This is true especially for the detection and staging of tumors, both primary and secondary. Enhanced scanning is important for both inflammatory and infectious etiologies. Generally, however, enhanced scanning is unnecessary for screening of developmental lesions or for follow-up of patients with partial complex epilepsy or ventriculomegaly. These workups should have included contrast material to exclude tumors on their initial MR study. Screening patients for first-onset seizures usually necessitates the use of IV contrast material as part of their initial evaluation. Dynamic MR studies with power-injected intravenous gadolinium have occasional use. This is most commonly used in our protocols for evaluation of pituitary microadenomas. The differential enhancement rates between the normal pituitary gland versus the microadenoma, can in many cases, provide the only clue to the presence of the microadenoma on imaging. Dynamic imaging is the basis for MR perfusion studies, as well. Fat-Suppressed T1-Weighted Magnetic Resonance Imaging with or Without Gadolinium There are circumstances and locations when it is necessary to rid the image of fat-related hyperintensity of T1-weighted images, for example, when it is necessary to differentiate between subacute hematoma (hyperintense because of methemoglobin) and lipoma (hyperintense because of lipid content). Another instance occurs when trying to visualize abnormal gadolinium enhancement in the orbit, as might occur in radiation-induced optic neuritis, optic nerve ischemia, and multiple sclerosis (acute phase). The periorbital fat makes such enhancement difficult to detect without fat-suppression techniques (see Fig. 2-11). High Detail Inversion Recovery (STIR) T1-Weighted Inversion recovery imaging uses a preparatory RF pulse that emphasizes T1-weighting. This creates hyperintense white matter and a relatively hypointense gray matter. The result is excellent gray matter to white matter differentiation. This,
Neuroradiologic Imaging
65
combined with thin sections, is very helpful in detecting subtle migrational abnormalities affecting the cortex. This process is time intensive. Therefore, it is not a screening exam and necessarily must be used in conjunction with electroencephalograph data to pick the site of interest (Fig. 2-17). Gradient-Recalled Imaging with T1 Weighting and/or with and Without Flow Compensation The GRE allow nearly isotropic evaluation of tissue and thin section capability without gaps between sections. These suffer from less signal quality, but provide better spatial resolution. The most common use of such images is for radiation surgery planning and for volumetric analyses, for instance, for hippocampal volumes in patients with partial complex epilepsy. A variation of GRE imaging, which performs the examination with and without flow compensation gradients, is useful in detecting low flow states, especially in radiation therapy–treated arteriovenous malformations. Low-flow vessels can exhibit signal characteristics similar to thrombosed vessels. Scanning the site without flow compensation followed by rescanning with flow compensation gradients allows differentiation of thrombosis from slow flow. This sequence is a routine part of our post-radiation surgery assessment of the arteriovenous malformation nidus (Fig. 2-18).
Figure 2-18. Gradient-recalled echo (GRE) imaging of hemosiderin in multifocal cavernous angiomata. GRE image through the low convexity of brain demonstrates focally reduced signal in the left lateral mesencephalic and left occipital areas (arrows) representing hemosiderin found within cavernous angiomata. Susceptibility artifacts produced by the presence of hemosiderin are more conspicuous with a gradient type image sequence, as illustrated in this examination.
W9418-02.qxd 10/31/03 3:35 PM Page 66
66
Section I
Introduction
Magnetic Resonance Angiography (Time of Flight) MRA is acquired using the intrinsic signal of fast-flowing arterial blood. The process uses a presaturation pulse over the top of the head to obviate dural sinus venous blood flow signal. The process of imaging the brain plus the use of magnetic transfer prepulses causes dampening of stationary brain signal. The unsaturated inflowing arterial blood in the carotid and vertebral arteries provide the tissue contrast nec-
A
B
essary to generate the MRA signal. Postprocessing, back-projection, computer manipulation creates the arterial images. The limitation of this method is that saturation occurs in distal intracranial vessels, precluding their opacification. The other significant limitation is patient motion during data acquisition resulting in misregistration of vessel segments. Time-of-flight MRA remains the mainstay of intracranial MR arterial evaluation (Fig. 2-19).
Figure 2-19. A, Time-of-flight (TOF) MRA in carotid occlusion. Threedimensional TOF MRA in dark-contrast mode demonstrates abnormally reduced flow in the left intracranial carotid distribution (arrow). Because some left cerebral circulation is present, this does not represent a complete occlusion of the middle cerebral circulation, but rather, a slow flow state with collateral vessels derived through the circle of Willis. The collateralized circulation is less apparent because its slower flow is subject to additional saturation of the inherent signal (less available signal than faster flowing blood). B, TOF MRA in segmental basilar artery occlusion. Three-dimensional TOF MRA with contrast in white-contrast mode demonstrates an absent signal (thrombosis) in the mid-portion of the basilar artery (arrow). Both the remaining antegrade flow from the vertebral arteries and the retrograde flow through the posterior communicating arteries can be appreciated on MRA.
W9418-02.qxd 10/31/03 3:35 PM Page 67
Chapter 2
Neuroradiologic Imaging
67
Magnetic Resonance Angiography (Two-Dimensional with Gadolinium) MRA of the brachiocephalic vessels can be performed using a simultaneous intravenous infusion of gadolinium and acquiring 2D coronal sections. Postprocessing provides a high-quality MRA that rivals digital angiography. The limitation is timing the contrast material bolus. If filming occurs too soon there is insufficient arterial detail. If image acquisition is too late then the cervical veins flood the image and obscure arteries. Nevertheless, when performed correctly in a cooperative patient, infusion MRA can accurately image the aortic arch and cervicobrachial vessels. This method is less suitable for intracranial arterial studies (Fig. 2-20). Magnetic Resonance Venography MRV can be performed to evaluate patency of the major dural sinuses and the internal cerebral vein/vein of Galen/straight sinus complex. The difference between MRA and MRV relates to where the presaturation pulse is placed. In cranial MRV, the saturation bands are placed over the cervical arterial vessels obviating any arterial inflow signal. No presaturation RF pulse is used over the top of the head,
Figure 2-21. Dural sinus thrombosis on MRV. Three-dimensional time-of-flight MRV demonstrates very little signal within the superior sagittal sinus. Some cortical veins are visualized. These findings are indicative of sagittal sinus thrombosis. For comparison, review the normal cranial MRV in Figure 2-9.
allowing signal from efferent flow in dural sinuses to generate the vascular signal. Hence, this study, for the most part, images veins and not arteries. The cavernous sinus is imaged only fairly reliably. MRV is used, in large measure, for evaluation of superior and transverse dural sinus thrombosis. It is occasionally used for deep vein thromboses. It is seldom used for cavernous sinus thrombosis. The features of the latter are usually apparent on standard imaging sequences (Fig. 2-21).
Figure 2-20. Gadolinium-enhanced MRA of the aorta and brachiocerebral arteries. Two-dimensional gradient-recalled echo images with gadolinium being actively injected (early in the infusion to emphasize arteries) demonstrate opacification of the thoracic aorta and the brachiocephalic arteries to the level of the skull base. The good visualization of vessels in this instance depends not on the inflow effects but rather on the physical presence of gadolinium within the vessels to create the image. This makes artifacts related to turbulence substantially less of a problem, which translates into greater lesion conspicuity.
Magnetic Resonance Diffusion The most common current use of diffusion MR is for the diagnosis of hyperacute cerebral infarct. It has the advantage of delineating the area of central ischemia before being detectable by routine CT or MR sequences. It has other uses, for instance, distinguishing between epidermoid cysts and arachnoid cysts where both exhibit CSF features on standard imaging. Examples of both of these issues are presented in Figures 2-7 (arachnoid cyst) and 2-33 (stroke). Magnetic Resonance Cerebrospinal Fluid Flow Study CSF flow studies can detect abnormal CSF pulsation dynamics. This becomes relevant in cases of possible hydrocephalus when minimal ventriculomegaly is present. Differentiation of arrested or low-grade hydrocephalus and mild atrophy is often a problem. In most cases, the intracranial CSF flow studies evaluate CSF dynamics within the cerebral aqueduct,
W9418-02.qxd 10/31/03 3:35 PM Page 68
68
Section I
Introduction
the outlet of the fourth ventricles and within the foramen magnum. The latter is especially relevant when there is downward tonsillar ectopia (Chiari 1 malformations; see Fig. 2-12 for example). Magnetic Resonance Spectroscopy MR proton spectroscopy is used currently in brain tumor staging as mentioned previously. In this capacity, it is used to help distinguish between radiation necrosis and persistent tumor. It is helpful in identifying gliomatosis cerebri and in distinguishing sequestered stroke or organized abscess from tumor. It has other nontumor uses as well, including assessment of leukodystrophy. Spectroscopy is not limited to proton spectroscopy alone. Phosphorus imaging is also possible. This can be used to map adenosine triphosphate and brain metabolism. While not in common use today, phosphorus imaging likely will be in time. Phosphorus spectroscopy is likely to be valuable in assessment of metabolic abnormalities and dementia (Fig. 2-22).
Available Computed Tomography Imaging Sequences Computed Tomography with or Without Intravenous Contrast Medium CT performed without contrast material is used as the standard screening examination for the brain in patients being evaluated for both spontaneous intracranial hemorrhage and posttraumatic brain imaging. Combined with sections through the maxillofacial regions, this provides a screen for uncomplicated sinusitis and headache. Bone windows allow diagnosis of skull and maxillofacial fractures. CT performed with and without contrast material is commonly used for screening of most infectious entities, inflammatory lesions (including vasculitis), and tumors. It is occasionally helpful in stroke. CT perfusion is likely to replace precontrast and postcontrast CT for stroke evaluation. Ventricular assessment seldom needs the postcontrast examination unless the cause of the ventriculomegaly is
B A Figure 2-22. A, Brain tumor spectroscopy. An axial, FLAIR image (strongly T2 weighted) demonstrates signal abnormalities evident in the right cerebral hemisphere in two locations. Review of Figure 2-10A, which is the same case, demonstrates that the more medial lesion is contrast enhancing and represents the primary tumor (single arrow). The more lateral lesion (double arrows) is a problem and could represent either radiation therapy effect or a satellite tumor implant. Mulitvoxel spectroscopy was applied in this case to differentiate these two possibilities. B, Multivoxel spectroscopy scout film. Mulitvoxel MR proton spectroscopy can be applied to assess the presence of tumor versus radiation change in both of the suspected areas. This scout image demonstrates how the spectroscopy is positioned to include both lesions and intervening brain. Spectroscopy revealed that the more lateral lesion exhibited features of radiation change and not tumor. Differentiation of tumor from radiation effects when multicentric abnormalities are evident represents one current role for proton spectroscopy.
W9418-02.qxd 10/31/03 3:35 PM Page 69
Chapter 2
unknown or if infectious-based shunt malfunction is suspected. Routine sinusitis evaluation seldom requires contrast material, whereas the complication of acute sinusitis may require a postcontrast examination. Computed Tomography Brain Perfusion CT perfusion is currently being used in the evaluation of hyperacute stroke. A preceding noncontrast examination should be acquired to evaluate the presence of blood products before the contrasted study. CT angiography is added if evidence of stroke or vascular thrombosis is suspected (see Fig. 2-16). Computed Tomography Angiography CT angiography is now available as a screen for extracranial or intracranial vascular thromboses. The detail is excellent. It also provides assessment of both mural thickness and residual lumen size. It can be used for diagnosis of arterial
Neuroradiologic Imaging
69
dissections as well as stenosis or thrombosis. Because of the ease and speed of providing appropriate data, the combination of CT perfusion and CT angiography will likely become the standard for pretreatment assessment in the context of acute brain attack (Fig. 2-23).
Relevant Radiologic-Clinical Issues The following discussion focuses on radiologic-clinical issues often affecting treatment decisions for the neurointensivist. In order of presentation, these issues include understanding the changes of brain herniations, imaging of hyperacute stroke/hypoxia/metabolic abnormalities, MR and CT manifestations of hemorrhage, early changes in hydrocephalus and intracranial hypotension, and finally, exclusion of tumefactive lesions from intrinsic brain tumor. Each of these subjects requires common understanding between the neuroradiologist and the neurointensivist because each has important clinical consequences and may affect outcome. Brain Herniations
Figure 2-23. Maximum intensity projection reconstruction (MIP). Individual, contiguous, thin, CT sections can be post processed in a number of ways. They can, as in Figure 2-15, be reformatted as individual sections in another plane (multiplanar reconstruction), or as in this instance, the entire data set can be combined to make a two-dimensional MIP image that can be rotated in any projection. Or, the data can be put into a three-dimensional mode using surface rendering. Both the MIP image and the three-dimensional surface rendering subtract out the bone. The MIP form of reconstruction with subtraction is presented in this case, and demonstrates the same anterior communicating aneurysm (arrow) as presented in Figure 2-15. The process of subtracting bone is subject to interpretive error, which limits the usefulness of both two-dimensional MIP and three-dimensional surface modeling techniques.
One of the common indications for CT imaging is “change in mental status” and, as a corollary, answering whether it is “safe” to perform lumbar puncture on a patient for CSF analysis.55–59 What is being asked, in essence, is whether there is a space-occupying lesion of any type and whether it is producing a brain herniation syndrome. This discussion will review the topographic arrangement of structures adjacent to the tentorial incisura and at the foramen magnum. It will present both simulated and actual imaging data to illustrate both pertinent normal relationships and abnormal consequences of the major herniation syndromes. Certain elements need to be present for herniation to occur. For supratentorial masses to cause herniation they must exist anatomically in reasonable proximity to the falx (for subfalcine herniation) or to the tentorial incisura (for downward transtentorial herniation). The size of the mass must be large enough to displace pertinent anatomic structures into specific cisternal spaces. In other words, herniations occur in the context of a mass exerting pressure in a direction, or “a vector.” It is the direction of the vector that defines the type of herniation. Herniation syndromes, as discussed in this chapter, include subfalcine herniation (produced by a mesial mid-convexity vector), downward transtentorial uncal herniation (produced by a mesial, lowconvexity, mid-temporal vector), downward transtentorial parahippocampal herniation (produced by a mesial, lowconvexity, posterior-temporal vector), upward herniation (produced by a superior cerebellar-region vector), and downward cerebellar tonsillar herniation (produced by a mid to inferior, cerebellar-region vector). Larger, nonfocal brain swellings can produce herniations, but these seldom
W9418-02.qxd 10/31/03 3:35 PM Page 70
70
Section I
Introduction
Figure 2-24. Subfalcine herniation. An axial, nonenhanced, CT section through the mid-convexity of brain demonstrates displacement of brain structures from left-to-right related to a large left cerebral infarction affecting the ipsilateral middle and posterior cerebral artery territories. In addition to the midline shift, there is also dilatation of the contralateral right-sided ventricle related to distortion of the foramina of Monro. These findings are typical of subfalcine type of brain herniation.
exert a single vector, but rather combine multiple types of herniation. For instance, global unilateral brain swelling might produce uncal herniation, parahippocampal herniation, and subfalcine herniation simultaneously. Subfalcine Herniation Subfalcine herniation implies a shift of midconvexity brain structures into the aperture formed beneath the falx cerebri. Generally, it is produced by an asymmetrical hemispheric mass, which displaces brain structures (cingulate gyrus and corpus callosum) across the midline and compresses them against the free margin of the falx cerebri. Injury to midline brain structures occurs more readily from acute shifts than from chronic or more slowly evolving masses. In the orbitofrontal region, the falx is small and midline shift does not result in focal tissue compression, but merely a midline shift. Masses in the midconvexity of the brain, usually in the frontoparietal area, the basal ganglia and the sylvian region most commonly account for subfalcine shifts. Lesions in the cerebral high convexity and those in the occipital polar region usually produce little shift of midline because they exist in proximity to the solid portions of the falx where no aperture is available through which to herniate brain. Similarly, anterior temporal lesions that reside in the temporal
fossa produce no shift until they expand into the midtemporal or lateral orbitofrontal regions (Fig. 2-24). Downward Transtentorial Herniation of the Uncus Downward transtentorial herniation of the uncus results from an asymmetric supratentorial mass of the midtemporal region that shifts the uncal portion of the parahippocampal gyrus medially into the suprasellar space (a medial vector) and then downward into the crural cistern (an inferior vector). The medial uncal shift produces an asymmetry of the suprasellar cistern initially, but obliterates the cistern as herniation becomes more severe. The downward vector results in compression of the crural portion of the circummesencephalic cistern initially, but, as it becomes more severe, results in displacement of the mesencephalon, ultimately compressing it upon the free margin of the opposite tentorial edge. Ipsilateral cisterns below the tentorium, as the cerebellopontine angle cistern, progressively enlarge as the brainstem is shifted to the contralateral side (Fig. 2-25). Downward Transtentorial Herniation of the Parahippocampal Gyrus Downward transtentorial herniation of the parahippocampal gyrus results from an asymmetric supratentorial mass of
W9418-02.qxd 11/21/03 2:39 PM Page 71
A
B
C
D
Figure 2-25. A, Downward transtentorial herniation: axial plane. Simulated MR image through the low convexity of brain demonstrating normal incisural anatomy and relationships. Cisterns surrounding the mesencephalon include the interpeduncular fossa and suprasellar cisterns anteriorly, the crural cisterns outlining the cerebral peduncles anterolaterally, the ambient cisterns outline the lateral aspect of the mesencephalon, and the quadrigeminal plate cistern outlines the tectal surface of the mesencephalon. The uncus (the uncal portion [u] of the parahippocampal gyrus) is also delineated by the crural cistern and contiguous to the cerebral peduncle. Brain adjacent to the ambient cistern is the parahippocampal gyrus (p). Brain adjacent to the quadrigeminal plate cistern is the superior vermis. All these cisterns combine to comprise the circummesencephalic cisterns which, along with the mesencephalon, fill the tentorial hiatus (or incisura). B, Downward transtentorial herniation (uncal herniation). The first stage of downward tentorial herniation is related to both medial and downward displacement of the uncal portion of the parahippocampal gyrus creating deformity of the suprasellar space, effacement of the crural cistern and early deformity of the ipsilateral cerebral peduncle. This simulated MR image with an apparent temporal lesion illustrates the findings of early downward transtentorial herniation with mesial shift of uncus and secondary compression of the crural cistern and ipsilateral cerebral peduncle (arrow). C, Downward transtentorial herniation (parahippocampal gyrus and uncal herniation). As downward incisural herniation worsens, there is further mesial displacement of the remaining parahippocampal gyrus into the incisura with further compression and contralateral midbrain displacement and loss of both the crural cistern and the ambient cistern. This simulated MR image through the tentorial hiatus illustrates these changes with substantial mesial shift of the parahippocampal gyrus including both the uncus (arrow) and the more posterior portions (arrow) with further contralateral displacement of the mesencephalon and effacement of the ipsilateral crural and ambient cisterns. D, Downward transtentorial herniation (prior to herniation). An axial, nonenhanced, CT section through the tentorial hiatus (incisura) demonstrates the normal anatomic relationships of the hippocampal gyrus in the mesencephalon surrounded by the circummesencephalic cisterns.
Figure continues
W9418-02.qxd 10/31/03 3:35 PM Page 72
E
F
G
H
Figure 2-25 continued. E, Downward transtentorial herniation (early stage). An axial, nonenhanced, CT section through the tentorial hiatus (incisura) demonstrates features of early uncal herniation. The crural cistern is compressed, while the ambient cistern is preserved. There is also minimal deformity of the left cerebral peduncle. F, Downward transtentorial herniation (advancing stage). An axial, nonenhanced, CT section through the tentorial hiatus (incisura) demonstrates features of progressive downward transtentorial herniation. There is now obliteration of the suprasellar, crural, and ambient cisterns. There is early obstructive hydrocephalus with dilatation of the left temporal horn. The herniated brain is edematous (low CT density; arrows). Notice that the quadrigeminal plate cistern is preserved and even accentuated in the case of downward herniation. This effect should be compared and contrasted to the effects of upward transtentorial herniation illustrated in Figure 2-26. G, Downward transtentorial herniation (advanced stage). Three-dimensional TOF MRA image demonstrates preservation of the basilar (b) arterial flow, but virtual elimination of the intracranial carotid circulation. These are MRA features of severe brain swelling with incisural compression and obstructed carotid circulation, but with preservation of the posterior fossa circulation. H, Downward transtentorial herniation (advanced stage). An axial, proton-density MR image through the occipital lobes demonstrates evidence of focally reduced signal (edema) in the left posterior cerebral artery territory (arrow). Downward herniation has occluded the parietal-occipital branch of the left posterior cerebral artery at the point where it crossed over the free edges of the left tentorium as it passes between the infra and supratentorial compartments.
W9418-02.qxd 10/31/03 3:35 PM Page 73
Chapter 2
Neuroradiologic Imaging
73
Figure 2-26. Upward transtentorial herniation. An axial, nonenhanced, CT scan through incisura demonstrates features of upward transtentorial herniation. Upward displacement of the superior vermis secondary to cerebellar swelling results in first compression and then complete effacement of the quadrigeminal plate cistern (arrow).
the posterior temporal region that shifts the mid and caudal portions of the parahippocampal gyrus medially into the ambient cistern. The first evidence of herniation is obliteration of the ambient cistern followed by displacement of the mesencephalon, ultimately compressing it upon the free margin of the contralateral tentorial edge. Parahippocampal gyrus herniation has another important consequence. The parietal occipital branch of the posterior cerebral artery (PCA) crosses from an infratentorial position to a supratentorial position in its post tectal portion. Antegrade blood flow in this portion of the PCA may be compromised when the parahippocampal herniation becomes significant. Characteristically, this results in brain infarction in the mesial portion of the ipsilateral occipital lobe. Upward Transtentorial Herniation Upward transtentorial herniation occurs as the result of a mass within the upper half of the posterior fossa, either intra-axial or extra-axial types. The upward vector displaces the superior vermis and the contiguous portion of the rostral cerebellum into the tentorial notch, while displacing the rostral brainstem against the clivus. The presenting feature of upward transtentorial shift is obliteration of the quadrigeminal plate cistern and deformity of the inferior colliculi. As mass effects worsen, there will be obliteration of the prepontine cistern and deformity of the interpeduncular fossa. It is important to recognize the differences between
upward and downward herniation. Upward herniation seldom affects the crural and ambient portions of the circummesencephalic cisterns, and downward herniation seldom affects the quadrigeminal plate cistern. When both vectors are present, it is likely that a global brain insult has occurred affecting both supratentorial and infratentorial brain (Fig. 2-26). Downward Cerebellar Tonsillar Herniation Caudal-half posterior fossa masses produce a downward vector which shifts the cerebellar tonsils inferiorly into the cistern of the cisterna magna. If the mass effect is lateralized, an asymmetric tonsillar herniation is produced. A generalized or midline oriented mass will produce a symmetric tonsillar herniation. A mass arising within the fourth ventricle will separate the tonsils as they are downwardly displaced. The two cisterns, which must be identified to exclude tonsillar herniation, include the vallecula, a midline extension of the cisterna magna that separates the mesial surfaces of the cerebellar tonsils, and the circummedullary cisterns located in the plane of the foramen magnum. If either of these cisterns is obliterated, tonsillar herniation is imminent. If it is visualized but compressed and displaced away from its normal midline position, then there is an asymmetric mass in the posterior fossa. Low tonsillar position by itself is not abnormal. Three conditions exist where low tonsillar position is not related to
W9418-02.qxd 11/21/03 12:39 PM Page 74
A
C
B
D
Figure 2-27. A, Downward tonsillar herniation. Simulated MR image through the low posterior fossa illustrating the normal relationships between the medulla, the cerebellar tonsils (T), and cerebrospinal fluid within the cisterna magna, the vallecular extension of the cisterna magna (called the vallecula), and the remaining circummedullary cisterns. The vallecula is the midline postmedullary cistern that separates the mesial surfaces of the cerebellar tonsils. This cistern is used to delineate the midline structures in the lower posterior fossa, and therefore, must be visualized in all normal head scans to exclude tonsillar shifts. B, Downward tonsillar herniation. Simulated MR image through the low posterior fossa illustrating left-to-right shift of the cerebellar tonsils compressing and displacing the vallecula to the opposite side (arrow). Lower sections would show displacement of the inferior pole of the left cerebellar tonsils below the plane of the foramen magnum. The plane of the foramen magnum is defined by the imaginary line interconnecting the anterior and posterior margins of the foramen magnum. The anterior margin is delineated by the anatomic points of reference which include the basion (or caudal tip of the clivus) and the opisthion (or posterior margin of the occipital bone as it forms the foramen magnum). The caudal poles of the tonsils lie above this line, as shown in Figure 2-28. C, Downward tonsillar herniation. An axial, T2-weighted, spin-echo MR image through the plane of the foramen magnum demonstrates early bilateral downward tonsillar herniation. The caudal tonsillar poles (arrows) are positioned at or slightly below the plane of the foramen magnum. D, Downward tonsillar herniation (same patient as in C). An axial, T2-weighted, spin-echo MR image through the plane of the cisterna magna demonstrates substantial downward tonsillar herniation with obliteration of the circummedullary cisterns, obliteration of the vallecula, and anterior displacement of the medulla against the clivus. The caudal tonsillar poles are displaced well below the plane of the foramen magnum.
W9418-02.qxd 10/31/03 3:35 PM Page 75
Chapter 2
Neuroradiologic Imaging
75
Figure 2-28. Downward tonsillar herniation; sagittal plane. This sagittal, nonenhanced, T1weighted MR image through the tonsillar region demonstrates features of downward tonsillar herniation. The caudal poles of the tonsils project below the plane of the foramen magnum (dotted line). Normally, the caudal tonsillar poles lie above this line. In this case, the tonsillar herniation was related to hyponatremia and generalized brain swelling. Once corrected, the tonsillar herniation improved over a very short time without clinical sequelae.
mass effect. Two are developmental and the last is the consequence of a suboccipital craniotomy. It is not uncommon for the caudal poles of the cerebellar tonsils to be positioned normally a little below the plane of the foramen magnum. In this instance, the foramen magnum tends to be large and the tonsils are more laterally placed. Significantly, the vallecula is unusually large and remains in a midline position. The second developmental variation is that of a Chiari malformation. In this instance, the cervicomedullary portion of the brain stem is also displaced inferiorly and there is no mass evident. The tonsils are frequently both ectopic and dysmorphic, appearing elongated and ending in a pointed shape. It may, however, be difficult to differentiate between the effects of intracranial hypotension and a mild Chiari-1 malformation without clinical history (Figs. 2-27, 2-28). Evaluation of Oligemia, Stroke, and Dural Sinus Thrombosis Brain perfusion is normally autoregulated maintaining perfusion in the range of 40 to 80 mL per 100 gm of tissue per minute.60–67 Oligemia results when perfusion falls beneath 20 mL/100 gm of tissue, placing the brain at risk of infarc-
tion. Perfusion exceeding 80 mL/100 gm of tissue per minute is considered hyperemic and usually implies a loss of autoregulation; it takes the appearance of luxury perfusion on angiographic images. Clinical function is typically affected by ischemia earlier than most of our radiologic methods can detect, but the gap is rapidly closing with diffusion MRI and both perfusion CT and MRI. Most brain perfusion modalities (other than Xenon-CT studies) provide relative measures of perfusion comparing abnormal to more normal areas of circulation, and not actual values, although this is changing as well. When oligemia becomes symptomatic, it implies that some portion of brain is at imminent risk of infarction. This area at risk has two components: the central ischemic core, which is likely to infarct, and the surrounding, oligemic zone, or penumbra, which is potentially at risk for infarction. The extent of the penumbra depends on collateralization by adjacent brain circulation. Diffusion MRI can facilitate assessment of the extent of the ischemic core, and either CT or MRI perfusion estimates the overall area of ischemia penumbra. Superimposing these studies determines the likely ischemic core and the likely oligemic penumbra. Arterial patency can be determined using either CT or MRA. Routine CT is used to evaluate for acute clot in
W9418-02.qxd 10/31/03 3:35 PM Page 76
76
Section I
Introduction
A B Figure 2-29. A, Early brain stroke on CT. An axial, nonenhanced CT scan through the mid-convexity of brain demonstrates features of early brain swelling (*). In this case, the findings are subtle, but there is effacement of the gray matter-white matter junction along the right insula. The follow-up CT study, performed 24 hours later, is illustrated in B. The earliest stage (hyperacute) of cytogenic edema is often difficult to detect on CT without reviewing the images in a high-contrast mode. B, Early brain stroke diagnosis on CT. An axial, nonenhanced CT scan through the mid-convexity of brain demonstrates features of acute stroke with focal cytogenic changes in the subinsular centrum semiovale regions on the right side. Once focal edema is present, the site of the ischemic area is relatively easy to detect with CT. However, this usually requires more than 12 hours postictus, which is most likely beyond the time frame for intra-arterial clot lysis therapy.
a proximal cerebral artery and/or parenchymal cerebral hemorrhage (Figs. 2-29 to 2-31). The first component of a stroke workup requires documentation that brain injury has actually occurred. Time is of the essence if either intravenous or intra-arterial treatments are to be effective. Currently, MR diffusion has become the current means of early stroke diagnosis, although with time it may be supplanted by perfusion CT. MR diffusion methods are very sensitive and can detect an ischemic event, usually within minutes to hours. The limitation of this method is spatial discrimination. Several cubic centimeters of tissue need to be affected to confidently make a radiologic diagnosis of stroke in progress. The features that characterize an acute infarct on MR diffusion imaging include a reduced diffusion coefficient (low ADC values implying
restricted tissue water movement) and increased signal on the trace diffusion images (high values implying failure of the molecular motion to reduce the overall signal as it occurs in normal brain). The limitation of the trace images is “T2 shine through.” This means that increased tissue water (cytogenic edema) also creates increased signal on the diffusion images, because diffusion is a T2-sensitive sequence. In practical terms, cerebral infarction diagnosis is most secure when the ADC map has a negative defect (dark area) and the trace images have a bright area. These changes are detectable before cytogenic edema produces an abnormality on the standard MR sequences. The perfusion portion of the study, which is an estimate of the signal change following a first pass method after an intravenous gadolinium injection, will estimate the total area of regional oligemia. Comparing the
W9418-02.qxd 10/31/03 3:35 PM Page 77
Chapter 2
A
Figure 2-30. A, Hyperdense artery sign in stroke. An axial, nonenhanced, CT scan through the low convexity of brain near the base of the sylvian fissure demonstrates a hyperdense segment of the middle cerebral artery (M2 segment). The hyperdensity of the vessel is related intravascular clot. B, Basilar artery thrombosis. Sagittal, nonenhanced, T1weighted MR scan through the posterior fossa demonstrates acute thrombus (arrows) within the basilar artery in its mid-portion. The flowing portion of blood is hypointense, while the mid-portion of the basilar artery contains isodensity material consistent with intraluminal clot.
B
Neuroradiologic Imaging
77
W9418-02.qxd 10/31/03 3:35 PM Page 78
78
Section I
Introduction
Figure 2-31. Laminar necrosis. An axial T1-weighted MR scan through the mid-convexity of brain demonstrates features of a large zone infarct but also gyriform blood products collecting in the cortex (arrows). This gyriform type of blood distribution is indicative of thrombosis at a capillary level and corresponds pathologically to laminar necrosis. The spontaneous hyperintensity indicates the presence of methemoglobin.
perfusion study with the diffusion study allows differentiation of the ischemic core versus the oligemic penumbra versus the remaining normal brain volume. The second component of a stroke includes evaluation of actual cervical and intracranial arterial patency. Both MRA and CT angiography can provide these data. However, TOF MRA (which is used intracranially) suffers in two respects. First, it is motion sensitive and requires a cooperative patient. Second, acute thrombus containing methemoglobin can generate signal similar to blood flow on TOF sequences. Thus, MR can provide spurious information if care is not used to correlate the standard MRI sequences with the MRA. Gadolinium-enhanced MRA is more accurate than TOF sequences, but is only suitable in the neck. Our standard stroke MRA sequences include a gadolinium-enhanced MRA for the neck vessels, a TOF MRA for the intracranial vessels, and thin section T1 sequences through the skull base and upper cervical region to assess for intraluminal thrombus and to detect concomitant dissection, when it is present.
The latter study is, in our experience, the most sensitive method of delineating mural dissection. Both direct angiography and MRA only image the status of the lumen and not the vessel wall. CT angiography obtained with newer generation helical scanners produces vascular patency data that are very accurate. CT will likely replace MRI in our institution for acute stroke evaluation when the patient is imaged in a period that does not preclude intraarterial clot lysis (less than 6 hours post ictus) (Fig. 2-32). The diagnosis of arterial dissection is sometimes a problem. Although the features of dissections are well known on angiography, this procedure only evaluates the lumen of the affected vessel. What is needed is evaluation of the lumen as well as the vessel wall (i.e., mural and extramural thickening). Hence, direct transfemoral arteriography is used only in indeterminate cases and thin section T1-weighted MRI or thin section spiral CT with contrast material are used as initial procedures. In most cases, these studies are diagnostic without being invasive (Fig. 2-33).
W9418-02.qxd 10/31/03 3:35 PM Page 79
Chapter 2
A
B
Figure 2-32. A, Early stroke diagnosis on MRI. An axial, nonenhanced, T1-weighted MR image through the low convexity of brain in a patient with acute neurologic deficit demonstrates no abnormality. An asterisk (*) indicates the region of actual ischemia presented in the following images. B, Early stroke diagnosis on MRI. An axial, trace (b = 1000), diffusion image through the low convexity of brain demonstrates features of a hyperacute infarct with a focal (hyperintense) diffusion abnormality in the mesial right occipital lobe (*). Diffusion MR imaging can detect hyperacute stroke, when routine MRI and CT typically fail in the absence of hemorrhage. C, Early stroke diagnosis on MRI. An axial, apparent diffusion coefficient (ADC) image through the same low convexity brain demonstrates reduced apparent diffusion (hypointensity) on the ADC map. This indicates less inherent motion in this portion of brain compared to the remainder of brain.
C
Neuroradiologic Imaging
79
W9418-02.qxd 10/31/03 3:35 PM Page 80
80
Section I
Introduction
Venous obstruction is a less common form of vascular disease producing global or focal symptoms. Imaging is helpful in delineating this problem, especially in the major dural sinuses and in the cavernous sinus. Individual cortical vein thrombosis can also be diagnosed occasionally. The imaging features of dural sinus thrombosis include evidence of clot within the lumen of the dural sinus
on nonenhanced images and abnormal enhancement of the wall of the dural sinus. Brain edema is often bilateral, straddling the dural sinus. Blood products within cortical veins or extravasated into the brain parenchyma are common. MRI and MRV are often the noninvasive diagnostic modalities of choice in venous occlusive disease of the head (Fig. 2-34).
A
B
Figure 2-33. A, Thrombosis of the cervical carotid in acute stroke. An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates the presence of a focal zone of edema (arrow) in the anterior sylvian region on the right side consistent with an acute ischemic event (stroke in progress). B, Thrombosis of the cervical carotid in acute stroke. An axial, T2-weighted section through the high cervical region in the same patient demonstrates asymmetric intensity within the high cervical carotid arteries. There is a normal flow void on the left side (arrow), but not on the right (double arrows). The usual hypointense lumen of the right carotid is replaced with nearly isodense clot. These findings are indicative of lumenal thrombosis in the high cervical portion of the right carotid artery.
W9418-02.qxd 10/31/03 3:35 PM Page 81
Chapter 2
A
B
Figure 2-34. A, Dural sinus thrombosis (superior sagittal sinus). An axial, nonenhanced, CT section through the low convexity of brain demonstrates spontaneous hyperdensity within the central lumen of the posterior portion of the superior sagittal sinus (arrow), representing acute clot within its lumen (“delta sign”). B, Dural sinus thrombosis. An axial, gadolinium-enhanced, T1-weighted MR image through the low convexity of brain demonstrates abnormal enhancement of the wall of the superior sagittal sinus (arrow) with little or no enhancement within the central lumen (“empty delta sign”). Spontaneous, hyperdensity within the lumen of the vein of Galen and proximal straight sinus (double arrows) represents acute intravascular clot, as well. C, Acute dural sinus thrombosis. Sagittal, nonenhanced, T1-weighted MR image through the superior sagittal sinus (*) demonstrates spontaneous hyperintensity consistent with clot already in a methemoglobin phase. Similar changes are also evident in the deep venous system (arrow). These are features of acute dural sinus thrombosis on MRI.
C
Evaluation of Hypoxia and Metabolic Insufficiency This is a diverse group of abnormalities.68–72 The basic physiologic defect common to most hypoxic/metabolic (excluding storage diseases and dysmyelination) diseases is a failure to produce adequate energy to maintain cell function and/or integrity. Global hypoxia, carbon monoxide poisoning, and genetic defects in aerobic metabolism typically develop brain
Neuroradiologic Imaging
81
changes affecting regions of highest adenosine triphosphate (energy) demands. These include, in order of susceptibility, mesial globus pallidus, periaqueductal gray matter, dorsal pontine nuclei, dentate cerebellar nuclei, hippocampi, cerebellar cortex, and cerebral cortex. We refer to these regions as areas of high adenosine triphosphate demand and when abnormalities are evident in these zones, effects of global hypoxia, diabetes (global glucose loads), or mitochondrial
W9418-02.qxd 10/31/03 3:35 PM Page 82
82
Section I
Introduction
A
C
dysfunction become prominent in the differential diagnosis (Figs. 2-35 to 2-37). Evaluation of Intracranial Hemorrhage Intracranial hemorrhage is a common cause of mental status change.73–76 The source of the hemorrhage is always a problem. Extra-axial blood products typically arise from
B
Figure 2-35. A, Diffuse cytogenic brain edema-global insult. An axial, nonenhanced, CT section through the mid-convexity of brain demonstrates abnormal CT density characterized by virtual homogenization of the gray and white matter tissue density. Such change is indicative of a global insult with either a hypoxic or a hypoperfusion basis. Generalized cytogenic (or cytoxic) edema in its early stage often produces little mass effect other than effacement of superficial sulci. The global injury in this patient is seen to progress in B and C. B, Diffuse cytogenic brain edema-global insult. An axial, nonenhanced, CT scan through the low convexity of brain demonstrates evidence of severe cytogenic edema affecting both hemispheres. The degree of cytogenic edema is often underappreciated unless the cerebellar density (*) is used for comparison. In this instance, the cytogenic temporal edema stands out because it can be contrasted with the more normal cerebellar density. C, Diffuse cytogenic brain edema-global insult. An axial, nonenhanced, CT scan through the low convexity of brain demonstrates features of diffuse brain swelling with compression of the ventricular spaces, loss of gray-white junction differentiation, effacement of the superficial sulci, and compression of the incisural cisterns.
trauma or vasculopathy, and especially chronic anticoagulation. Aneurysm and arteriovenous malformations figure prominently when subarachnoid blood is present. These lesions figure prominently in discussions by other authors and, hence, will not be further discussed. Intra-axial hemorrhage, on the other hand, occurs in a wide assortment of lesions and often requires specialized imaging to resolve the basic lesion. Discussion in this section will deal with the fea-
W9418-02.qxd 10/31/03 3:35 PM Page 83
Chapter 2
Figure 2-36. Central nuclear injury in diabetic ketoacidosis. An axial, T2-weighted FLAIR image through the basal ganglia region demonstrates features of metabolic brain injury. In this instance, the patient has been treated for diabetic ketoacidosis. But, despite usually appropriate therapy, the patient had a negative clinical outcome. The changes of a metabolic insult to brain tend to affect areas of higher adenosine triphosphate utilization. These regions include the mesial basal ganglia, the periaqueductal gray matter, the dentate nuclei, thalamic, and hippocampi. Injury to these regions (as in this case) results in abnormal T2 hyperintensity (arrows) distributed in the structures mentioned previously. Abnormalities of this type are based on a metabolic insult and are usually related either to hypoxia, carbon monoxide poisoning, or underlying abnormalities of aerobic metabolism as occurs in patients with cytochrome oxidase enzymatic and/or mitochondrial deficiencies.
tures of resolving blood on MRI and CT in general and then examine some of the underlying causes. CT and FLAIR MRI sequences are both sensitive to subarachnoid and other acute extra-axial blood products. CT, because of its easy accessibility, is typically used in the acute state. The location and likely cause of the hemorrhage depends on its location. Subarachnoid blood collects in basilar (suprasellar space, cerebellopontine angle, and interpeduncular fossa) cisterns and in communicating (sylvian and interhemispheric) cisterns. Subarachnoid or subpial blood, over the convexities of brain, takes on a “gyriform” appearance. If the gyriform appearance is well delineated, the blood is likely within the sulci adjacent to the cortex; if less well delineated, the blood is likely intraparenchymal. Blood collections delimited by a margin in brain are likely within lesions of the brain and
Neuroradiologic Imaging
83
Figure 2-37. Progressive necrotizing leukoencephalopathy (PNL). Coronal, enhanced, CT scan through the posterior frontal region demonstrates features of acquired small vessel vascular disease associated with combination chemotherapy and radiation resulting in PNL. The imaging features typically include bilateral (although often asymmetric) distribution, abnormal marginal contrast-enhancement (arrows), and leukomalacic changes.
not the brain per se. These include thrombus within lesions as cavernous angioma, metastatic deposits, or within the fundus of a giant aneurysm. The following sections will review the appearance of blood as it evolves toward ultimate resorption. This evolution of blood, where pertinent, can help with the dating of the time of onset of the hemorrhage. Hemorrhage and the Evolution of Blood Products on Computed Tomography In the acute phase of clot evolution, which generally lasts 1 to 3 days, acute blood on CT is hyperdense relative to brain and CSF alike. Over the next few days, the blood passes into an isodense phase and then over 2 weeks into a hypodense phase. This entire process of clot evolution is accelerated for subarachnoid blood and slowed for extra-axial (epidural and subdural) hematomas. The intermediate, isodense stage is sometimes a problem because differentiation between isodense blood and intra-axial swelling may be difficult. This is less of a problem on current generation CT scanners than in the past. It may at times, however, be necessary to administer IV contrast material to increase the background brain density to exclude an extra-axial subacute hematoma. Intra-
W9418-02.qxd 10/31/03 3:35 PM Page 84
84
Section I
Introduction
ventricular blood usually clears readily while subependymal hematomas do not (Fig. 2-38).
Hemorrhage and the Evolution of Blood Products on Magnetic Resonance The appearance of hemorrhage on MRI is more complex than that on CT. CT density varies in proportion to the density of the hematoma. In MRI, the signal is affected by both the physical and chemical status of the clot and of
hemoglobin. The hemoglobin effects are related mainly to changes in the magnetic properties of the iron contained in the hemoglobin molecule and are directly influenced by the iron’s oxidation state and biochemical form. These changes in the products of blood-iron breakdown are influenced by local oxygen saturation, age of the hematoma, size of the hematoma, degree of clot retraction, and status of the local circulation, as well as by other factors. Despite knowledge of the evolution of blood clot, dating the time of the hemorrhage exactly is somewhat of a problem after the hyperacute phase.
A B Figure 2-38. A, Evolution of blood on CT; subdural hematoma (acute phase on CT). An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates acute extra-axial hematoma on CT. There is hyperdense blood (arrow) layered along the lateral aspect of the right brain margin separating brain from the inner table of the skull consistent with an acute subdural hematoma. There is a concomitant right-to-left subfalcine shift. B, Evolution of blood on CT; subdural hematoma (subacute phase on CT). An axial, nonenhanced, CT scan through the mid-convexity of brain (as in A) but approximately 1 week later. The subdural hematoma density has changed from acute clot to a subacute hematoma (arrow). The hyperdense clot evident on A has now become nearly isodense to brain with only a minimal hyperdense component. Intracranial hemorrhages pass through an intermediate phase when they become virtually isodense to brain. If the hematomas happen to be bilateral, their mass effect can be missed. This error has been partially eliminated with recent vintage CT scanners. However, if an isodense subdural is suspected, but remains unconfirmed, contrastenhanced CT should be performed to increase brain density, making the subdural hematoma become apparent.
W9418-02.qxd 10/31/03 3:35 PM Page 85
Chapter 2
The first stage in hematoma evolution is termed the hyperacute stage. This stage lasts only a few hours and is therefore encountered less frequently in routine imaging. During this hyperacute period, red blood cell integrity is preserved, as is the oxygenation of hemoglobin. The blood does not have paramagnetic properties during this time, and the appearance of the blood is related mainly to its proton density. The clot at this time is isointense or slightly hypointense on T1-weighted images and slightly hyperintense on T2-weighted sequences. The clot, during this stage, may not be distinguishable from other pathologic processes on the basis of its intrinsic signal intensity. The hematoma then passes into the acute stage, which lasts for approximately 1 week. The first dramatic change that occurs is that the oxyhemoglobin is converted to deoxyhemoglobin. This deoxyhemoglobin within the red blood cells has little effect on the signal intensity of the blood on T1-weighted images but produces a susceptibility artifact on T2-weighted images with reduced signal in the clot. This effect is more clearly seen in magnets of higher field strength. In the next stage of hematoma evolution, the subacute stage, the hematoma undergoes two changes: a conversion of hemoglobin to methemoglobin through oxidation of the iron, and a break down of red cell membranes with influx of tissue water. The methemoglobin transformation may be seen as early as 2 days after the hemorrhage, but depending on the location of the hematoma, may not be very noticeable until about a week has passed. Methemoglobin results in T1 shortening (proton relaxation enhancement) resulting in a high signal intensity on T1-weighted scans. This high signal intensity first appears at the outer margin of the hematoma, then progresses inward. Initially, the area containing methemoglobin remains hypointense on T2-weighted images. When red blood cell membranes burst and the methemoglobin is released, the extracellular methemoglobin combined with increased water content produces high signal on T2-weighted scans (hyperintensity on T2 sequences). The hyperintense T1 signal of methemoglobin persists well into the chronic phase of the hematoma and may remain visible for months or perhaps years. The last phase is the chronic stage of hematoma evolution. The principal feature of the chronic hematoma is the appearance of hypointensivity evident on T2-weighted sequences in or around the hematoma and the adjacent tissue. This signal change is related to susceptibility artifacts produced by accumulation of iron in the form of ferritin and hemosiderin within macrophages. The zone of hypointensity on T2-weighted images appears to remain indefinitely at the border of a hemorrhage, even after the high signal intensity of methemoglobin may have disappeared and the hematoma cavity has been converted to a thin, slit-like
Neuroradiologic Imaging
85
space. The preceding description of the changes within hematomas applies mainly to bleeds into neural tissue. Hemorrhages into other anatomic compartments may have different characteristics. For example, subarachnoid blood is notoriously difficult to detect by magnetic resonance imaging, except on FLAIR imaging. This difficulty is probably related to the relatively high oxygen content of cerebrospinal fluid, which does not promote the formation of deoxyhemoglobin. The signal intensity of a chronic subdural hematoma is dictated primarily by methemoglobin, which remains for a very long time. Hemosiderin does not appear to form at the margins of subdural hematomas. Hemorrhagic malignant tumors in the brain may demonstrate a slower evolution of blood product changes. Deoxyhemoglobin may persist longer in tumors than in parenchymal hematomas, a delay in methemoglobin may also be apparent, and a clear hemosiderin-ferritin ring may never develop (Fig. 2-39).
Evaluation of Ventricular Size Ventriculomegaly is readily defined with either MRI or CT. The grading scheme we use for estimation of ventricular enlargement is based on a scale of four steps beyond normal. Determination of normal ventricular size requires experience because it varies by chronologic age. Once the ventricular size is estimated to exceed normal limits for age, then a grading scheme is useful. We use a method that examines the mid-convexity images where the bodies of the lateral ventricles are best seen. Grade 2 (of 4) is defined as having bodies of the lateral ventricles taking up roughly half the hemibrain volume and the cerebral mantle the remaining half. Grade 1/4 ventriculomegaly exceeds normal-for-age, but has not reached the level of grade 2/4. Grade 3/4 is moderately severe ventriculomegaly but leaving at least 1.5 cm of cerebral mantle, and grade 4/4 reduces the cerebral mantle to under 1 cm. Signs of elevated CSF pressure include acute expansion of ventricle size and developing transependymal fluid migration. Acute changes in ventricular size in either direction become important from the perspective of critical care. Symptoms can be produced by ventricular size, which is too small from overshunting or from intracranial hypotension, as well as acutely increased ventricular volume. The ventriculomegaly can be generalized or occur in an asymmetric fashion in cases of sequestered ventricular components. Acute change in ventricular size is best assessed by comparing temporal horn size. It is difficult to appreciate a 10% change in the lateral ventricular size. However, the same change is usually easily seen in temporal horn volume. Other signs of increasing ventricular size are effacement of high convexity sulci and evidence of transependymal fluid migration. Any acute change in ventricular size can be symptom
W9418-02.qxd 10/31/03 3:35 PM Page 86
86
A
C
Section I
Introduction
B
Figure 2-39. A, Evolution of blood on MR; subdural hematoma in an acute phase on MRI. An axial, T2-weighted FLAIR, MR image through the high convexity of brain demonstrates features of mixed age subdural hematomas. The blood products on the right are very hyperintense indicating completed transition to a fluid state. On the left side the hematoma (*) is relatively less bright indicating that the clot is relatively more solid and, hence, younger. B, Evolution of blood on MR: subdural hematoma (subacute phase on MR). An axial, T1-weighted, spin echo, MR scan through the high convexity of brain demonstrates that both the right and left (*) subdural hematomas contain methemoglobin, but there is more on the right indicating a slightly older age. The methemoglobin does not require being within cells to be hyperintense on a T1weighted sequence. C, Evolution of blood on MR; methemoglobin within cavernous angioma of the mesencephalon. Sagittal, T1-weighted, spin-echo MR image through the mesencephalon demonstrates spontaneous hyperintensity within a well-delineated mass (*) within the mesencephalon. This hyperintensity is related to methemoglobin within the clot, which remains confined within the borders of the lesion. The outer margin of the mass is hypointense, indicating the presence of hemosiderin. These are the features of the intra-axial mesencephalic cavernous angioma.
W9418-02.qxd 10/31/03 3:35 PM Page 87
Chapter 2
producing. Thus, these early signs of hydrocephalus become important (Figs. 2-40 to 2-42).
Neuroradiologic Imaging
87
tures that discriminate between aggressive and nonaggressive intracranial tumors.
Screening patients with a depressed level of consciousness includes evaluation for intracranial mass lesions.77–93 Focused imaging protocols tailored to the clinical setting are the most suitable means of establishing an initial diagnosis, excluding nontumoral conditions, and answering questions related to staging. The first step in this diagnostic algorithm is to determine whether the scan is normal or not. Assuming a significant lesion has been detected, then an attempt must be made to distinguish between tumor and tumor mimics (“tumefactive lesions”). In adults, stereotactic biopsies of mass lesions (all thought to be tumors before biopsy) reveal an 8% to 12% incidence of nontumoral, tumefactive lesions. Thus, a 10% incidence of tumefactive lesions becomes a useful number to remember when evaluating mass lesions on imaging procedures. A corollary to the tumefactive issue (nontumors masquerading as tumors) is the reverse scenario where tumors masquerade as nontumoral conditions, most notably strokes. Both of these issues are considered in the following section, along with the fea-
Inflammatory or Infectious Tumefactive Lesions The more common inflammatory tumefactive lesions include “tumefactive MS” (multiple sclerosis or Schilder’s disease in children), myelin basic protein hypersensitivity disorders (i.e., acute disseminated leukoencephalopathy), adrenoleukodystrophy, sequestered cerebral infarction, herpes encephalitis, and resolving sequelae of occult trauma or hemorrhage. Each of these entities has features that aid in their diagnosis, but full analysis is beyond the scope of this discussion; references are, however, included in the bibliography. Infectious lesions can occasionally simulate tumors as well. Notable among this group are chronic empyema or brain abscess, neurocystercercosis, tuberculoma, and cryptococcoma. Chronic empyema is usually associated with adjacent paranasal sinusitis or otomastoid inflammatory disease. Patients undergoing immune therapy and bone marrow transplantation as part of their tumor treatment plan are subject to the opportunistic superinfections. Infections that might simulate persistent tumor include progressive multifocal leukoencephalopathy associated with papova viral encephalitis and cryptococcus (forming cryptococcomas).
A
B
Evaluation of Brain Masses and Tumor Mimics
Figure 2-40. A, An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates small ventricular size in a patient with a right frontal ventriculo-peritoneal shunt in place. Although slightly small for age, the ventricles are still well delineated. B, An axial, nonenhanced CT scan through the mid-convexity of brain demonstrates features of overshunting. This patient has become symptomatic in the interim from the first scan (A). Ventricle size has become slitlike, indicating overdrainage.
W9418-02.qxd 10/31/03 3:36 PM Page 88
88
A
C
Section I
Introduction
B
Figure 2-41. A, Ventriculomegaly. An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates lateral ventricular size that is at the upper limits of normal but not clearly enlarged for chronologic age. Subtle change in lateral ventricular size is often difficult to appreciate when examining the bodies of the lateral ventricles. But examination of the temporal horn size is often more revealing for early ventricular enlargement (B). B, Ventriculomegaly. An axial, nonenhanced, CT scan through the temporal horns in the low convexity of brain demonstrates unequivocal enlargement of the temporal horn size compared to anticipated normal size for age. Temporal horn size increases are more easily appreciated when detecting early hydrocephalus in middle age and younger patients. It is less effective in elderly patients where temporal atrophy may coexist. The ventriculomegaly in this case is related to a previous subarachnoid hemorrhage and secondary external hydrocephalus. C, Ventriculomegaly. An axial, nonenhanced, CT scan through the high convexity of brain demonstrates the additional feature of early hydrocephalus of sulcal.
W9418-02.qxd 10/31/03 3:36 PM Page 89
Chapter 2
Figure 2-42. Intracranial hypotension. This midline, sagittal, nonenhanced, T1-weighted MR image demonstrates tonsillar herniation (arrow) related to intracranial hypotension produced by over drainage through an implanted lumboperitoneal subarachnoid shunt. Low tonsillar position combined with dural enhancement are the main features of intracranial hypotension.
Abnormalities related to surgical intervention also create tumor mimics. Surgical resection commonly produces a thin rim of enhancement in the tumor bed and regional dural structures; occasionally this may appear somewhat nodular. The distinguishing feature of the surgically altered tumor bed is the presence of a thin rim of hemosiderin evident on the T2-weighted images. We call these findings “bovey tracks” (from the name of the commonly used electrocautery unit). Occasionally, patients exhibit reactive ventricular marginal changes to the catheters inserted for chemotherapy (i.e., an idiopathic reactive ventriculitis). These changes are quite dramatic. Ventricular shunting produces a generalized dural thickening with abnormal enhancement that can mimic dural implantation by tumor. This is probably related to chronic intracranial hypotension created by the ventricular drainage. It is contrasted to the more nodular dural thickening found in dural tumor implants. Delayed effects of radiation surgery or stereotactic radiation therapy add additional considerations to the tumefactive list including subacute radiation effects, chronic radiation-induced necrosis (see Fig. 2-13A), and rarely telangiectasis or radiation-induced tumor formation. Radiation necrosis can be distinguished by the presence of focal enhancement confined to the zone of radiation and hemosiderin deposition. Radiation necrosis typically reveals reduced uptake on both thallium-enhanced single-photon
Neuroradiologic Imaging
89
emission CT (SPECT) and positron emission tomography imaging, while persistent tumor generally exhibits increased uptake. Radiation-induced telangiectasia results in multicentric deposition of hemosiderin (evident on MRI) and simulates cryptic vascular lesions. In patients who have received both intrathecal chemotherapy and radiotherapy, the possibility of progressive necrotizing leukoencephalopathy also exists. This necrotizing vasculopathy causes destructive, multicentric necrosis of the brain, mainly in the white matter. It is generally associated with whole brain irradiation. Other less common events can occur associated with some chemotherapeutic agents (i.e., tacrolimus and cyclosporin). These patients exhibit capillary zone infarctions, usually in the parietal-occipital artery territories, which are probably related to elaboration of thromboxane associated with a regional vasculopathy (“capillary leak”). Additional imaging observations essential for distinguishing a nontumoral mass from a recurrent tumor include lack of progression or improvement following in the absence of radiation therapy or chemotherapy on serial studies. MRI has the advantage of distinguishing blood products from other substances thus obviating the confusion with occult trauma, and hemorrhagic infarction. This is important because hemorrhagic cerebral metastases, or hemorrhage within primary tumors, are uncommon in children and the presence of a hemorrhage would therefore suggest other histologically benign etiologies such as arteriovenous malformation. CT plays an important role in defining and distinguishing extra-axial disease, especially those entities involving the skull base, orbits, and cranial nerves. Thin section CT adds imaging value whenever evidence of bone (skull or skull base) destruction is being considered. Direct extension of a primary tumor to bone is rare. The presence of a mass with both brain and bone involvement widens the differential diagnosis to include lesions metastatic to bone and dura, as neuroblastoma, histiocytosis, or hemangiopericytoma. CT is also helpful in confirming the presence of paranasal sinus or otomastoid inflammatory disease, and thereby supports a diagnosis of chronic empyema. Neither CT nor MRI has been particularly reliable in distinguishing intra-axial inflammatory lesions from tumors. This has prompted the adjunctive use of thallium-SPECT imaging. Providing that the intravascular thallium can find access to the brain (usually through sites of blood-brain barrier disruption or areas of gadolinium enhancement), it can then be metabolized by tumors but not by inflammatory processes. Thallium positivity is consistent with the presence of residual tumor, or the onset of immune-suppression– related tumors, such as primary cerebral lymphoma. Thallium negativity is generally indicative of radiation necrosis or inflammatory lesions of the brain (i.e., progressive multifocal leukodystrophy) (Fig. 2-43).
W9418-02.qxd 10/31/03 3:36 PM Page 90
A
B
C D Figure 2-43. A, Tumefactive lesion (focal multiple sclerosis). An axial, gadolinium-enhanced, T1weighted MR section through the posterior fossa demonstrates a focal nodular enhancing mass (arrow) in the low part of the right cerebellar hemisphere. Whenever masses are identified both tumor and tumefactive (nonneoplastic) masses must be considered, which, in this case, proved to be an active plaque of multiple sclerosis with marginal enhancement. B, Tumefactive lesion (pontine myelinolysis). Sagittal, nonenhanced, T1-weighted MR image through the pons demonstrates a hypodense lesion (*) simulating a pontine glioma. Etiology, however, is that of acute pontine myelinolysis. C, Tumefactive lesion (cortical heterotopia). An axial, nonenhanced, CT scan through the high convexity of brain demonstrates an additional tumefactive lesion. There is a focal gray matter hypodensity (arrow) on the left side, which, on biopsy, was found to be an area of cortical heterotopia. D, Tumefactive lesion (resolving intra axial hematoma). An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates a mass lesion in the right frontal region (arrow) that has the appearance of a metastatic focus, because of its ring enhancement. Ring enhancement, however, is nonspecific and in this instance is related to reaction surrounding a resolving intra-axial hematoma.
W9418-02.qxd 10/31/03 3:36 PM Page 91
Chapter 2
Dyshistiogenesis Certain lesions, especially the neurocutaneous syndromes, are associated with what can be described as, for lack of a better term, dyshistiogenesis. These changes take two forms. The first is an increased T2 signal occurring mainly within the basal ganglia, pons, and deep cerebellar nuclei, and are most evident in neurofibromatosis type 1 (NF-1). They do not exhibit evidence of mass effect. There are other lesions in NF-1 commonly seen in the medulla, which have mass effect but little signal alteration. Their cause remains unknown. The T2-hyperintense lesions are most evident in childhood and diminish or resolve later in childhood or adolescence. The other form of dyshistiogenesis is associated with neurocutaneous syndromes involving melanin concentrations (or neurocutaneous melanosis). These syndromes cause signal changes in the brain (mainly T1 hyperintensity), but are significantly more rare. The presence of dyshistiogenetic changes is useful in confirming the presence of a genetic based disease as NF-1. Focal masses in patients with neurocutaneous syndromes are likely to represent specific cell types that exhibit reasonably predictable biologic behavior (Fig. 2-44).
Figure 2-44. Dyshistiogenesis in neurofibromatosis type 1 (NF1). An axial, T2-weighted, spin-echo, MR image through the mid convexity of brain demonstrates multicentric areas of abnormality consisting of bright signal (arrows) seen on T2weighted images in the basal ganglia in the context of NF1. This abnormality is consistent with the intracellular cystic changes seen in NF1. This process is self-limited and usually resolves by the age of 12. It is a common MR finding in NF1 and does not evolve or represent infiltrative tumor.
Neuroradiologic Imaging
91
Tumefactive Lesions of Vascular Nature The most common vascular masses simulating tumors include venous and cavernous angiomata and giant aneuryms, especially when they are partially thrombosed. Cavernous and venous angiomas are separate and distinct entities but often occur in conjunction with each other. Venous angiomas are misnamed and in fact represent anomalies of venous drainage rather than actual masses. They seldom cause significant clinical deficit. Cavernous angiomata, however, are vascular malformations of capillary formation that lack vessels with muscular walls as well as arterial or venous components. They may be multicentric or unicentric. When they present clinically, symptoms are usually the result of internal thrombosis and expansion. They have a hemosiderin-containing pseudocapsule that is evident on MRI and represents a major diagnostic feature. Occasionally, these lesions actually hemorrhage outside the pseudocapsule into the adjacent neuropil. Although occurring infrequently, giant aneurysms occur spontaneously in both adults and children. Large saccular aneurysms are most commonly seen arising from the ophthalmic, cavernous, or internal carotid artery bifurcation region of the carotid arteries. Most giant aneurysms in the posterior fossa arise from the basilar tip. Large fusiform aneurysms are likely to be the result of dissection of the parent artery, occurring spontaneously or following a known traumatic event. Not all aneurysms are arterial; some are associated with arteriovenous malformations (AVMs), most often of dural type with aneurysms forming in the venous system. Brain AVMs that drain into the deep venous system, most notably the vein of Galen, can also form large venous aneurysms. The preoperative exclusion of a giant aneurysm is important to avoid biopsy of such a lesion. Aneurysms cause variable degrees of turbulence and distortions of flow. This can create a variety of appearances on imaging, especially on MRI, where flow-related enhancement and turbulence-related signal dropout can simulate masses with heterogeneous signals. In these circumstances, the actual diagnosis can usually be confirmed with either MRA, or CT angiography. The advent of spiral image techniques has substantially improved the reliability and clarity of CT angiography. Both are essentially noninvasive procedures capable of excluding vascular causes of mass lesions (Fig. 2-45).
Tumors Masquerading as Other Lesions Most brain tumors are centered in white matter and grow along white matter tracts. In adults, some gliomas, mainly oligodendrogliomas, have a gray matter epicenter with secondary white matter involvement. When such lesions lack nodular enhancement, they can easily mimic strokes on standard imaging. We expect that spectroscopy will make
W9418-02.qxd 10/31/03 3:36 PM Page 92
92
A
C
Section I
Introduction
B
Figure 2-45. A–C, Tumefactive lesion (giant intrasellar carotid aneurysm). A, An axial, nonenhanced, CT scan through the low convexity of brain and sellar area demonstrates abnormal enhancement within the sella (*). Pituitary adenoma is included in the differential diagnosis but intrasellar giant aneurysm must always be excluded. B, An axial, gadolinium-enhanced, T1weighted MR scan through the sella demonstrates MR features of a hypodense mass within the sella, plus pulsation artifact creating the bright linear densities (arrows) overlying the mesencephalon and vermis. This pulsation artifact is a clue that the mass is indeed vascular in nature. C, Digital, cerebral angiography in this case demonstrates a large aneurysm (arrow) arising from the left carotid artery projecting medially into the sella. Exclusion of aneurysm is essential in evaluation of any enhancing sellar or parasellar masses.
W9418-02.qxd 10/31/03 3:36 PM Page 93
Chapter 2
Neuroradiologic Imaging
93
A
B Figure 2-46. A, Tumor simulating a stroke. An axial, nonenhanced, T1-weighted MR scan through the mid-convexity of brain demonstrates an area of subtle hypodensity (arrow) in the caudal aspect of the insula, which corresponds to the posterior division of the right middle cerebral artery. These findings are consistent with stroke. However, with every stroke there always remains a possibility of an underlying brain tumor simulating a stroke, as in this case. Those tumors most likely to be confused with strokes include primary cerebral lymphoma, angiocentric lymphoma, and oligodendroglioma. All of these tumors can primarily involve the gray matter rather than an underlying white matter, which accounts for the confusion. Patients with apparent stroke often present with seizures rather than typical stroke symptoms and, if the diagnosis is not clear, the patient needs to be rescanned after 6 to 8 weeks. Persistence of the abnormality at that time may indicate biopsy. B, Tumor simulating a stroke. An axial, T2-weighted, spin-echo MR scan through the mid convexity of brain on a follow-up examination after 6 weeks demonstrates that the lesion has both persisted and minimally increased in size rather than evolve to a post-infarct status. This course is indicative of a brain tumor (arrow).
inroads in this delineation, but experience is limited (Figs. 2-46, 2-47). Predicting the Biologic Behavior of Tumors Once the biopsy provides evidence of an underlying tumor, imaging features can then be combined with pathologic features to predict biologic activity. Pathology may be limited if selection-sampling bias of the biopsy-sampled tissue results in it not being reflective of the most aggressive cell line. Additionally, many tumors demonstrate poor to limited correlation between their respective cytologic appearance and their biologic or clinical behavior. The following section addresses those imaging features that help predict likely clinical aggressiveness of a tumor. Features Suggesting a Less Aggressive Neoplasm. Low-grade,
true neoplasms tend to occur in children and adolescents rather than in infants. These are true neoplasms and can
exhibit malignant behavior. Lesions that are somewhat unique to children include mixed ganglion-glial cell gangliogliomas, pilocytic astrocytomas, desmoplastic astrocytomas, and pleomorphic xanthoastrocytomas. Geneticbased neoplasia also begins to occur in children and adolescents. These are lesions associated with neurocutaneous syndromes, especially tuberous sclerosis, von HippelLindau, and neurofibromatosis types 1 and 2. The features typical of lower grade neoplasms include slow growth potential on serial imaging; masses containing nonenhancing cysts often with a mural nodule; and no tumoral necrosis or tumor growth away from the epicenter. Importantly, there is concordance between the tumor size on both T1 and T2 images (Fig. 2-48). Features Suggesting a More Aggressive Neoplasm. Higher grade primary neoplasms tend to occur primarily in infants, younger children, and then later in life. In older children they
W9418-02.qxd 11/21/03 2:47 PM Page 94
A A
B B Figure 2-47. A, Tumor presenting as hypertensive hemorrhage (melanoma metastasis). An axial, nonenhanced, CT scan through the mid-convexity of brain demonstrates a focal parenchymal hematoma within the left thalamus and a small amount of intraventricular blood. In older patients, as indicated by the agerelated brain atrophy, most intra-axial hemorrhages away from the circle of Willis are seldom related to aneurysm or vascular malformation but more often small vessel vasculopathy of amyloid angiopathy or hypertension. Again, as with brain stroke, tumor must also be considered in cases of focal hematoma. B, Tumor presenting as hypertensive hemorrhage (melanoma metastasis). Follow-up axial, contrast-enhanced CT scan through the mid-convexity of brain demonstrates how the solid hematoma has evolved to mass with a peripheral enhancing ring and a mural nodule (arrow) consistent with a metastatic tumor focus. These findings were related to metastatic melanoma, a common tumor presenting with brain hemorrhage.
Figure 2-48. A, Low-grade brain tumor. An axial, gadoliniumenhanced, MR image through the high convexity of brain demonstrates features of a relatively noninvasive low-grade brain neoplasm. There is no abnormal contrast enhancement within the microcystic (low intensity) mass (arrow) seen in this right, high convexity, lateral cortex. Most importantly, there is concordance between the margins of the mass on the T1 images compared with the T2 images. B, Low-grade brain tumor. An axial, T2-weighted, spin-echo MR image through the high convexity of brain demonstrates features of a low-grade, low-aggressive type of tumor. Findings illustrated include hyperintense T2 changes that closely match the margins on T1 imaging, with no evidence of subependymal or subpial extension of tumor and with no intratumoral necrosis. Lack of enhancement is also evident and supportive of a low-grade lesion, but intratumoral enhancement is not a reliable indicator of tumor aggressiveness.
W9418-02.qxd 10/31/03 3:36 PM Page 95
Chapter 2
usually occur as anaplastic transformation of lower grade tumors. Metastatic tumor occurs in all age groups, but tends to be seen in older patients. The features that suggest a higher grade tumor include growth on serial imaging, tumor necrosis, T1 and T2 spatial discordance, evidence of subpial or subependymal extension, and increased efferent venous drainage away from the epicenter of the tumor (related to angioneogenesis). There are proposed anatomic reasons why aggressive tumors extend into the subependymal and subpial spaces. Tumor extension in most cases is oriented along fiber pathways; nodular tumor growth occurs in areas of high dendritic concentration. It is suggested that the dendritic concentration is lower in the subpial zone beneath the basement membrane of the pia. Similar reasons for growth are suggested for the subependymal region. Subependymal growth along the third ventricle is common
Neuroradiologic Imaging
95
in germinoma and lymphoma. Other features of aggressive tumors include invasion of dura, satellitosis, and the most obvious are “drop metastases” with seeding of distant portions of intracranial or intrathecal CSF-filled spaces. Additional risk factors for brain tumor formation include immunosuppression (with resultant lymphoma) and prior cranial radiation. Patients with a history of cranial radiation for any reason have a substantial increased likelihood of development of another intracranial tumor, often brain sarcoma (Figs. 2-49, 2-50). Post-treatment Imaging. Post-treatment imaging is critical to correctly determine follow-up therapy and to correctly assess patients enrolled in treatment protocols.94,95 The success, or relative success, of the therapeutic regimen is monitored in part by MRI or occasionally by CT. The greatest problem
B A Figure 2-49. High-grade (primary) brain tumor. A, Coronal, gadolinium-enhanced, T1-weighted MR image through the parietal occipital junction region of brain demonstrates features of a high-grade, highly infiltrative brain tumor. There is dense contrast enhancement in the epicenter (*) of the mass but also enhancement in brain adjacent to the apparent mass (arrow) and in the subpial region (arrows) well away from the primary lesion. Subependymal spread was seen on other images. There is obvious discordance between the T1 image (this image) and the T2 image (B), indicating tumor infiltration. B, An axial, T2-weighted, FLAIR scan through the mid-convexity of brain demonstrates discordance between the enhancing mass (seen in A) and the overall zone of peritumoral edema.
W9418-02.qxd 10/31/03 3:36 PM Page 96
96
Section I
Introduction
B A Figure 2-50. A, Metastatic brain tumor. Coronal, gadolinium-enhanced, T1-weighted MR image through the occipital polar region of brain demonstrates an intra-axial neoplasm with marginal enhancement, central necrosis and dural involvement. These are always features of a high-grade neoplasm, either of primary or secondary type. B, Highly aggressive (metastatic) brain tumor. An axial, T2-weighted, spin-echo MR image through the low convexity of brain demonstrates unusual hypodensity within the mass itself. There is hyperintense peritumoral edema. Hypodensity within the mass is suggestive of a metastatic lesion of colon (as occurred in this patient) because of intratumoral mucin production, which reduces the T2 signal.
arises because of insufficient baseline information. Thus, as a routine, we obtain a postoperative MRI with contrast enhancement as soon as the patient is able to tolerate the procedure, generally within the first postoperative week. In doing so, we establish a functional baseline with which we can compare later developments. The goal of subsequent imaging is to define the presence or absence of residual disease; establish which changes are related to surgery, including those associated with ventricular shunting; and to identify recurrent tumor when and if it occurs. The following discussion focuses on the usual and the untoward consequences of therapy. Postoperative Changes. Following brain resection, a variety
of changes are commonly observed.96,97 The most obvious of these is an encephalomalacic defect. The second is evidence of the surgical excision and bipolar coagulation of bleeding sites within the tumor bed. Postoperative findings include linear, non-nodular contrast enhancement conforming to the margins of the encephalomalacic cavity in the same areas as hypodense susceptibility artifact related to hemosiderin
deposition on T2-weighted images. Persistent tumor generally appears more nodular and appears hyperintense on T2-weighted sequences. Occasionally, persistent operative material, such as Gelfoam, may be intentionally left in the operative site. In some cases, this can be misconstrued on repeat imaging as a persistent mass or herniation. The next surgically related abnormality is dural thickening and enhancement associated with the presence of an indwelling ventricular-peritoneal (VP) shunt catheter. The dural changes may be regional, near the entrance of the shunt, but more often are widespread. Benign dural thickening is not nodular and generally measures in the 1- to 3mm range of thickness. VP shunts also alter ventricular size. The shunt typically will reduce the volume of the ipsilateral lateral ventricle more so than the contralateral side. This ventricular asymmetry can cause diagnostic concern at times. VP shunts can infrequently induce an untoward reaction within the ventricular cavity, presumably thought to be a hypersensitivity response to the catheter material. This hypersensitivity response produces an inflammatory ventriculitis that can easily be confused with tumor seeding.
W9418-02.qxd 10/31/03 3:36 PM Page 97
Chapter 2
Loculation (or sequestration) produced by ventricular inflammation and cicatrix formation can create cystic appearing spaces also simulating tumor. A third postoperative change that can occur is unexpected postoperative hemorrhage occurring away from the site of surgical intervention. This, in our experience, often occurs in the cerebellum. In all likelihood, it is related to transient hypertension sometime in the perioperative period. This is an uncommon occurrence, but must be considered when any neurosurgical patient fails to wake in the anticipated period (Fig. 2-51). Postradiation Effects. Radiation therapy is commonly used
to control brain tumors. The imaging manifestations vary according to time and dose. The most common anticipated manifestations include mucositis within the paranasal sinuses and mastoid air cells, and fat-replaced marrow space. There are subacute changes in the neuroaxis, which often occur following stereotactic radiation therapy or radiation
A
Neuroradiologic Imaging
97
surgery. In the vicinity near these areas of concentrated radiation doses, T2 hyperintensity commonly occurs. These changes generally resolve over 1 to 2 years. Full neuroaxis radiation, on the other hand, elicits little acute or subacute change. Mild diffuse brain atrophy and chronic ischemic demyelination similar to that of arteriosclerosis may develop in patients receiving whole brain or whole neuroaxis irradiation (usually longer than 6 months). Radiation dosage that exceeds neural tolerance can produce other untoward effects including radiation necrosis, progressive necrotizing leukoencephalopathy (usually seen in conjunction with chemotherapy), and radiationinduced second primary tumor formation (usually brain/dural sarcoma). The appearance of radiation necrosis on MRI and CT studies can closely mimic a residual or recurrent tumor. As a consequence, we have resorted to thallium-SPECT studies to help differentiate the two. Radiation necrosis is thallium-negative, while recurrent tumor is generally thallium-positive. As the technology evolves, it is
B Figure 2-51. A, Postoperative changes simulating pathology. Sagittal, nonenhanced, T1-weighted MR image through the foramen magnum region demonstrates pseudotonsillar herniation. During a suboccipital craniotomy, Gelfoam was applied over the dural closure. This material can, as in this instance, take on the appearance of cerebellar tonsils, which have been downwardly displaced (arrow). Closer inspection reveals a normal tonsillar position (*). The craniotomy in this instance was performed to remove a mesencephalic cavernous angioma. B, Postoperative unexpected cerebellar hemorrhage. An axial, nonenhanced CT scan through the posterior fossa demonstrates evidence of a spontaneous hemorrhage into the left cerebellar hemisphere (*) following a recent anterior temporal lobectomy. Exact etiology for such a cerebellar hemorrhage is not clear, but presumably is related to transient acute hypertension.
W9418-02.qxd 11/21/03 12:41 PM Page 98
Figure 2-52. Radiation necrosis. Coronal, gadolinium-enhanced, T1-weighted MR image through the posterior temporal lobe region demonstrates nodular-appearing enhancement mesial aspect of the temporal lobe (*). Despite its mass-like appearance, which simulates residual tumor, this mass corresponds to a previous radiation field and represents the MR changes of radiation necrosis. In cases where treatment was originally performed for anaplastic tumor, it is likely that residual tumor cells will remain on repeat biopsy, even when most of the changes are related to radiation necrosis. MR spectroscopy and thallium-SPECT help differentiate whether such findings are dominantly related to recurrent tumor or predominantly to radiation necrosis.
likely that MR perfusion imaging will play a role in this diagnosis as well (Fig. 2-52). Postchemotherapeutic Complications. The onset of progres-
sive necrotizing leukoencephalopathy is characterized on MRI and CT as multicentric areas of contrast enhancement that are scattered throughout cerebral white matter in a somewhat random manner. They do not necessarily correspond anatomically with the site of the primary tumor or the region of maximal radiation therapy. The findings may be more intense along the tract of the Omaya reservoir catheter, presumably because of a higher concentration of the chemotherapeutic agent.
Recently, there have been brain changes associated with the use of particular chemotherapeutic agents that have been delivered not by an intrathecal route but intravenously. These agents can infrequently produce a zone of infarction in deep capillary beds. For the most part, these have occurred in the parieto-occipital regions and produced visual loss. The mechanism is presumed to be related to a regional vasculopathy causing elaboration of intrinsic thromboxane. The thromboxane results in localized vasospasm and accelerated clotting, hence the capillary zone infarctions (Fig. 2-53).
Figure 2-53. Reversible neurotoxicity (immune-suppression drug effect). An axial, T2-weighted FLAIR scan through the low convexity of brain demonstrates bilateral occipital region cerebral edema (arrows). The posterior distribution, the mainly white matter involvement, and the bilaterality are typical of reversible vasogenic edema associated with acute hypertension or (as in this case) exposure to immunosuppressive drugs.
W9418-02.qxd 10/31/03 3:36 PM Page 99
Chapter 2
Neuroradiologic Imaging
99
P earls 1. Fundamental to providing adequate sectional imaging is access to sufficient clinical information to appropriately protocol the examination. No single test fits all, and not all patients can undergo all varieties of imaging studies. 2. T1 hyperintensity (compared to isointense normal brain) will appear “brighter” (approaching a shade toward white) and implies the presence of substances that naturally exhibit rapid proton relaxation (a more rapid return to baseline), such as structures with high lipid content. 3. T1 hypointensity (compared to normal brain signal), is displayed as a darker shade of gray, and indicates that tissue in these voxels are not like fat but are either rigidly bound (an anisotrophic effect), as seen in fibrosis or matrix calcification, or are virtually unbound (an isotropic effect), as seen in edema, necrosis, or cyst formation. 4. Tissues that possess more unbound water molecules (like CSF or tissue edema) will preserve phase coherence longer, which translates into slower T2 rate, and a brighter signal or T2 hyperintensity. 5. Any intervening motion of the patient during the image acquisition process actually changes the phase data, which ultimately misrepresents the data when the final mathematical transform is applied. This can create both anatomic and intensity misregistration. Data misregistration means anatomic structures will be placed where they do not exist and intensity misinformation will appear abnormally (spurious pathology). MRI is not like CT scanning where the consequence of motion is merely an unsharp or fuzzy image.
References 1. Atlas SW: Magnetic Resonance Imaging of the Brain and Spine. New York, Raven Press, 1991. 2. Fujita N, Hirabuki N, Kashiwagi N, Watanabe Y, Nakamura H: Epidermoid tumor of the cerebellopontine angle: signal loss in the contralateral cistern. Am J Neuroradiol 1998;19:333–335. 3. Miyati T, Kasai H, Shundo H, et al: Fast RARE MR imaging with variable flip angle excitation. Radiographics 1996;595. 4. Sheppard S, Davis PC, Kater G, Peterson JE: Inversion-recovery echoplanar MR in adult brain neoplasia. Am J Neuroradiol 1998;19: 267–273. 5. Kates R, Atkinson D, Brant-Zawadzki M: Fluid-attenuated inversion recovery (FLAIR): Clinical prospectus of current and future applications. Top Magn Reson Imaging 1996;8:389–396. 6. Rydberg JN, Riederer SJ, Rydberg CH, Jack CR: Contrast optimization of fluid-attenuated inversion recovery (FLAIR) imaging. Magn Reson Med 1995;34:868–877.
6. The high magnetic susceptibility of any iron containing metal object means that it is affected by the main magnetic field and can torque. If the metal were securely fastened, such motion effects would be insignificant. If, on the other hand, the object is moveable, as it might be with an aneurysm clip, then the torque effect might be disastrous. 7. The major advantage of using a nonionic contrast medium is that it causes less tissue injury if it is inadvertently extravasated during IV injection. 8. Helical scanning combined with dynamic contrast infusion can be exploited to produce a timeconcentration data set. By comparing the CT density curves over time in similar regions of brain, a graph of the contrast medium concentration can be created, which in turn reflects intravascular perfusion. By comparing like-areas in the one hemisphere with comparable regions in the opposite hemisphere, a relative brain perfusion analysis can be obtained. 9. The features typical of lower grade neoplasms include slow growth potential on serial imaging; masses containing nonenhancing cysts often with a mural nodule; and no tumoral necrosis or tumor growth away from the epicenter. Importantly, there is concordance between the tumor size on both T1 and T2 images. 10. The features that suggest a higher grade tumor include growth on serial imaging, tumor necrosis, T1 and T2 spatial discordance, evidence of subpial or subependymal extension, and increase efferent venous drainage away from the epicenter of the tumor (related to angioneogenesis).
7. Rydberg JN, Hammond CA, Grimm RC, et al: Initial clinical experience in MR imaging of the brain with a fast fluid-attenuated inversion recovery pulse sequence. Radiology 1994;193:173–180. 8. Tsuchiya K, Mizutani Y, Hachiya J: Preliminary evaluation of fluidattenuated inversion recovery MR in the diagnosis of intracranial tumors. AJNR 1996;17:1081–1086. 9. Tsuchiya K, Inaoka S, Mizutani Y, Hachiya J: Fast fluid-attenuated inversion-recovery MR of intracranial infections. Am J Neuroradiol 1997;18:909–913. 10. Beaulieu C, Allen PS: Water diffusion in the giant axon of the squid— Implication for diffusion-weighted MRI of the nervous system. Magn Res Med 1995;32:579–583. 11. Castillo M, Smith JK, Kwock L, Wilber K: Apparent diffusion coefficients in the evaluation of high-grade cerebral gliomas. Am J Neuroradiol 2001;22:60–64. 12. Ebisu T, Naruse S, Horikawa Y, et al: Discrimination between different types of white matter edema with diffusion-weighted MR imaging. J Magn Reson Imaging 1993;3:863–868.
W9418-02.qxd 10/31/03 3:36 PM Page 100
100
Section I
Introduction
13. Filippi CG, Edgar MA, Ulu AM, Prowda JC, Heier LA, Zimmerman RD: Appearance of meningiomas on diffusion-weighted images—Correlating diffusion constants with histopathologic findings. Am J Neuroradiol 2001;22:65–72. 14. Meyer JR, Gutierrez A, Mock B, et al: High-b-value diffusion-weighted mr imaging of suspected brain infarction. Am J Neuroradiol 2000;21:1821–1829. 15. Nusbaum O, Tang CY, Buchsbaum MS, Wei TC, Atlas SW: Regional and global changes in cerebral diffusion with normal aging. Am J Neuroradiol 2001;22:136–142. 16. Pierpaoli C, Righini A, Linfante I, Tao-Cheng, Alger J, Di Chiro G: Histopathologic correlates of abnormal water diffusion in cerebral ischemia: Diffusion in cerebral ischemia: Diffusion-weighted MR imaging and light and electron microscopic study. Radiology 1993;189:439–444. 17. Schwartz RB, Mulkern RV, Gudbjartsson H, Jolesz F: Diffusionweighted MR imaging in hypertensive encephalopathy—Clues to pathogenesis. Am J Neuroradiol 1998;19:859–862. 18. Yonemura K, Hasegawa Y, Kimura K, et al: Diffusion tensor mr imaging of the brain—Effect of diffusion weighting on trace and anisotropy measurements. Am J Neuroradiol 2000;21:1813–1820. 19. Aronen HJ, Gazit IE, Louis DN, et al: Cerebral blood volume maps of gliomas: Comparison with tumor grade and histologic findings. Radiology 1994;191:41–51. 20. Le Bihan D, Douek M, Argyropoulou M, Turner R, Patrona N, Fulham M: Diffusion and perfusion magnetic resonance imaging in brain tumors. Top Magn Reson Imaging 1993;5:25–31. 21. Ducreux C, Oppenheim X, Vandamme D, et al: Diffusion-weighted imaging patterns of brain damage associated with cerebral venous thrombosis. Am J Neuroradiol 2001;22:261–268. 22. Patrizia T, Danilo CF, Falcou A, Fiorelli M, Argentino C, Fantozzi LM, Bozzao L: Relationship between vascular enhancement, cerebral hemodynamics, and MR angiography in cases of acute stroke. Am J Neuroradiol 2001;22:255–260. 23. Belliveau JW, Kennedy DN, McKinstry RC, et al: Functional mapping of the human visual cortex using magnetic resonance imaging. Science 1991;254:716–719. 24. Connelly A, Jackson GD, Frackowiak RS, Belliveau JQ, VarghaKhadem F, Gadian DG: Functional mapping of activated human primary cortex with a clinical MR imaging system. Radiology 1993;188:125–130. 25. Holodny AI, Ollenschleger MD, Liu WC, Schulder M, Kalnin AJ: Identification of the corticospinal tracts achieved using bloodoxygen-level-dependent and diffusion functional mr imaging in patients with brain tumors. Am J Neuroradiol 2001;22:83–88. 26. Moritz CH, Haughton VM, Cordes D, Quigley M, Meyerand ME: Whole-brain functional MR imaging activation from a finger-tapping task examined with independent component analysis. Am J Neuroradiol 2000;21:1629–1635. 27. Rombouts S, Barkhof F, Veltman DJ, et al: Functional MR imaging in Alzheimer’s disease during memory encoding. Am J Neuroradiol 2000;21:1869–1875. 28. Yousry TA, Schmid UD, Jassoy AG, et al: Topography of the cortical motor hand area: Prospective study with function MR imaging and direct motor mapping at surgery. Radiology 1995;195: 23–29. 29. Angelie E, Bonmartin A, Boudraa, Gonnaud PM, Mallet JJ, SappeyMarinier D: Regional differences and metabolic changes in normal aging of the human brain—Proton MR spectroscopic imaging study. Am J Neuroradiol 2001;22:119–127. 30. Castillo M, Smith JK, Kwock L: Proton MR spectroscopy in patients with acute temporal lobe seizures. Am J Neuroradiol 2001;22:152–157. 31. Chang KH, Song IC, Kim SH, et al: In vivo single-voxel proton MR spectroscopy in intracranial cystic masses. Am J Neuroradiol 1998; 19:401–405.
32. Coo D, Van De Kerckhove T, De Reuck J, Caemaert J, Kunnen M: Singlevoxel proton MR spectroscopy and positron emission tomography for lateralization of refractory temporal lobe epilepsy. Am J Neuroradiol 1998;19:1–8. 33. DeLano MC, Cooper TG, Siebert JT, Potchen MJ, Kuppusamy K: High-b-value diffusion-weighted MR imaging of adult brain—Image contrast and apparent diffusion coefficient map features. Am J Neuroradiol 2000;21:1830–1836. 34. Filippi CG, Ulu AM, Ryan E, Ferrando SJ, van Gorp W: Diffusion tensor imaging of patients with HIV and normal-appearing white matter on MR images of the brain. Am J Neuroradiol 2001;22:277– 283. 35. Haseler LJ, Sibbitt WL, Mojtahedzadeh HN Jr, Reddy S, Agarwal VP, McCarthy DM: Proton MR spectroscopic measurement of neurometabolites in hepatic encephalopathy during oral lactulose therapy. Am J Neuroradiol 1998;19:1681–1686. 36. Itoh R, Melhem ER, Folkers PJM: Diffusion-tensor MR imaging of the human brain with gradient- and spin-echo readout—Technical note. Am J Neuroradiol 2000;21:1591–1595. 37. Kadota T, Horinouchi T, Kuroda C: Development and aging of the cerebrum—Assessment with proton MR spectroscopy, Am J Neuroradiol 2001;22:128–135. 38. Krouwer HG, Kim TA, Rand SD, et al: Single-voxel proton MR spectroscopy of nonneoplastic brain lesions suggestive of a neoplasm. Am J Neuroradiol 1998;19:1695–1703. 39. Melhem ER, Itoh R, Jones L, Barker PB: Diffusion tensor mr imaging of the brain—Effect of diffusion weighting on trace and anisotropy measurements. Am J Neuroradiol 2000;21:1813–1820. 40. Pavlakis SG, Lu D, Frank Y, Wiznia A, Eidelberg D, Barnett T, Hyman RA: Brain lactate and N-acetylaspartate in pediatric AIDS encephalopathy. Am J Neuroradiol 1998;19:383–385. 41. Sinson G, Bagley LJ, Cecil KM, et al: Magnetization transfer imaging and proton MR spectroscopy in the evaluation of axonal injury—Correlation with clinical outcome after traumatic brain injury. Am J Neuroradiol 2001;22:143–151. 42. Amano Y, Amano M, Kumazaki T: Normal contrast enhancement of the extraocular muscles—Fat-suppressed MR findings. Am J Neuroradiol 1997;18:161–164. 43. Bhadelia RA, Bogdan AR, Wolpert SM: Analysis of cerebrospinal fluid flow waveforms with gated phase–contrast MR velocity measurements. Am J Neuroradiol 1995;16:389–400. 44. Beer GJ: Biological effects of weak electromagnetic fields from 0 hz to 200 mhz—A survey of the literature with special emphasis on possible magnetic resonance effects. Magn Reson Imaging 1989;7: 309–331. 45. Colletti PM, Sylvestre PB: Magnetic resonance imaging in pregnancy. Magn Reson Imaging Clin North Am 1994;2:291–307. 46. Heinrichs WL, Fong F, Flannery M, et al: Midgestational exposure of pregnant BALBIC mice to magnetic resonance imaging conditions. Magn Reson Imaging 1988;6:305–313. 47. Kanal E: An overview of electromagnetic safety considerations associated with magnetic resonance imaging. Ann NY Acad Sci 1992; 649:20424. 48. Kanal E: Pregnancy and the safety of magnetic resonance imaging. Magn Reson Imaging Clin North Am 1994;2:309–317. 49. Shellock FG, Kanal F: Magnetic Resonance Bioeffects, Safety, and Patient Management. Philadelphia, Lippincott Williams & Wilkins, 1996. 50. Salvolini U, Provinciali L, Signorino M: Functional effects of contrast media on the brain. Am J Neuroradiol 2001;22:229. 51. Shehock PG: Biological effects and safety aspects of magnetic resonance imaging. Magn Reson Imaging 1989;5:243–261. 52. Quisling RG, Peters K: Computed tomography. In Youmans, ed: Neurological Surgery, 2nd ed, part 2. Philadelphia, Saunders, 1996, pp. 93–147.
W9418-02.qxd 10/31/03 3:36 PM Page 101
Chapter 2 53. Alberaico RA, Patel M, Casey S, Jacobs B, Maguire W, Decker R: Evaluation of the circle of Willis with three-dimensional CT angiography in patients with suspected intracranial aneurysms. Am J Neuroradiol 1995;16:216–217. 54. Brink JA, Heiken JP, Wang G, McEnery KW, Schlueter FJ, Vannier MW: Helical CT: Principles and technical considerations. Radiographics 1994:887–893. 55. Dillon WP, Fishman RA: Some lessons learned about the diagnosis and treatment of spontaneous intracranial hypotension. Am J Neuroradiol 1998;19:1001b–1002b. 56. Laine FJ, Shedden AI, Dunn MM, Ghataki NR: Acquired intracranial herniations—MR imaging findings. AJR 1995;165:967–973. 57. Messori A, Polonara G, Salvolini U: Dilation of cervical epidural veins in intracranial hypotension. Am J Neuroradiol 2001;22:224–225. 58. Rabin BM, Roychowdhury S, Meyer JR, Cohen BA, LaPat KD, Russell EJ: Spontaneous intracranial hypotension—Spinal MR findings. Am J Neuroradiol 1998;19:1034–1039. 59. Quisling RG, Lotz P: Correlative Neuroradiology. Baltimore, Williams and Wilkins, 1985. 60. Augustin M, Bammer R, Simbrunner J, Stollberger R, Hartung HP, Fazekas F: Diffusion-weighted imaging of patients with subacute cerebral ischemia—Comparison with conventional and contrast-enhanced MR imaging. Am J Neuroradiol 2000;21:1596–1602. 61. Bryan RN: Diffusion-weighted imaging of stroke—A brief follow-up. Am J Neuroradiol 1998;19:1003–1004. 62. Chong J, Lu D, Aragao F, et al: Diffusion-weighted MR of acute cerebral infarction—Comparison of data processing methods. Am J Neuroradiol 1998;19:1733–1739. 63. Ducreux D, Oppenheim C, Vandamme X, et al: Diffusion-weighted imaging patterns of brain damage associated with cerebral venous thrombosis. Am J Neuroradiol 2001;22:261–268. 64. Karonen JO, Partanen PLK, Vanninen R, Vainio PA, Aronen HJ: Evolution of MR contrast enhancement patterns during the first week after acute ischemic stroke. Am J Neuroradiol 2001;22:103–111. 65. Lovblad KO, Jakob PM, Chen Q, et al: Turbo spin-echo diffusionweighted MR of ischemic stroke. Am J Neuroradiol 1998;19:201–208. 66. Meyer JR, Gutierrez A, Mock B, et al: High-b-value diffusion-weighted MR imaging of suspected brain infarction. Am J Neuroradiol 2000;21:1821–1829. 67. Rumpel H, Ferrini B, Martin E: Lasting cytotoxic edema as an indicator of irreversible brain damage—A case of neonatal stroke. Am J Neuroradiol 1998;19:1636–1638. 68. Coskun A, Lequin M, Segal M, Vigneron DB, Ferriero DM, Barkovich AJ: Quantitative analysis of MR images in asphyxiated neonates—Correlation with neurodevelopmental outcome. Am J Neuroradiol 2001;22:400–405. 69. Barkovich AJ, Ali FA, Rowley HA, Bass N: Imaging patterns of neonatal hypoglycemia. Am J Neuroradiol 1998;19:523–528. 70. Bryan RN: Diffusion-weighted imaging—To treat or not to treat? That is the question. Am J Neuroradiol 1998;19:396–397. 71. Dubowitz DJ, Blum S, Arcinue E, Dietrich RB: MR of hypoxic encephalopathy in children after near drowning—Correlation with quantitative proton MR spectroscopy and clinical outcome. Am J Neuroradiol 1998;19:1617–1627. 72. Falini A, Barkovich AJ, Calabrese G, Origgi D, Triulzi F, Scotti G: Progressive brain failure after diffuse hypoxic ischemic brain injury—A serial MR and proton MR spectroscopic study. Am J Neuroradiol 1998;19:648–652. 73. Atlas SW, Thulborn KR: MR detection of hyperacute parenchymal hemorrhage of the brain. Am J Neuroradiol 1998;19:1471–1477. 74. Horowitz M, Kondziolka D: Multiple familial cavernous malformations evaluated over three generations with MR. Am J Neuroradiol 1995;16:1353–1355. 75. Levy RA, Allen R, McKeever P: Pleomorphic xanthoastrocytoma presenting with massive intracranial hemorrhage. Am J Neuroradiol 1996;17:154 –156.
Neuroradiologic Imaging
101
76. Offenbacher H, Fazekas F, Schmidt R, Koch M, Fazekas G, Kapeller P: MR of cerebral abnormalities concomitant with primary intracerebral hematomas. Am J Neuroradiol 1996;17:573–578. 77. Berry I, Osaki L, Brasch R, et al: Gd-DTPA in clinical MR of the brain— Intra-axial lesions. Am J Radiol 1986;147:1223–1230. 78. Bowen BC: Proton MR spectroscopy and the ring-enhancing lesion. Am J Neuroradiol 1998;19:589–590. 79. Brant-Zawadzki M, Badami P, Mills CM, Norman D, Newton TH: Primary intracranial tumor imaging: A comparison of magnetic resonance and CT. Radiology 1984;150:435–440. 80. Brown MS, Stemmer SM, Simon JH, et al: White matter disease induced by high-dose chemotherapy—Longitudinal study with MR imaging and proton spectroscopy. Am J Neuroradiol 1998;19:217–221. 81. Castillo M, Smith JK, Kwock L, Wilber K: Apparent diffusion coefficients in the evaluation of high-grade cerebral gliomas. Am J Neuroradiol 2001;22:60–64. 82. Castillo M, Smith JK, Kwock L: Correlation of myo-inositol levels and grading of cerebral astrocytomas. Am J Neuroradiol 2000;21:1645– 1649. 83. Dean B, Drayer B, Bird C, et al: Gliomas—Classification with MR imaging. Radiology 1990;174:411– 415. 84. Gideon P, Sorensen PS, Thomsen C, Stahlberg F, Gjerris F, Henriksen O: Increased brain water self-diffusion in patients with idiopathic intracranial hypertension. Am J Neuroradiol 1995;16:381–387. 85. Hwang JH, Egnaczyk GF, Ballard E, Dunn RS, Holland SK, Ball WS Jr: Proton MR spectroscopic characteristics of pediatric pilocytic astrocytomas. Am J Neuroradiol 1998;19:535–540. 86. Iwana T, Yamada H, Era S, et al: Proton nuclear magnetic resonance studies on water structure in peritumoral edematous brain tissue. Magn Reson Med 1992;24:53–63. 87. Krouwer HG, Kim TA, Rand SD, et al: Single-voxel proton MR spectroscopy of nonneoplastic brain lesions suggestive of a neoplasm. Am J Neuroradiol 1998;19:1695–1703. 88. Maheshwari SR, Mukherji SK, Neelon B, et al: The choline/creatine ratio in five benign neoplasms—Comparison with squamous cell carcinoma by use of in vitro MR spectroscopy. Am J Neuroradiol 2000;21:1930–1935. 89. Medina LS, Zurakowski D, Strife KR, Robertson RL, Poussaint TY, Barnes PD: Efficacy of fast screening MR in children and adolescents with suspected intracranial tumors. Am J Neuroradiol 1998;19: 529–534. 90. Poptani RK, Gupta, RR, R Pandey, Jain VK, Chhabra DK: Characterization of intracranial mass lesions with in vivo proton MR spectroscopy. Am J Neuroradiol 1995;16:1593–1603. 91. Roberts HC, Dillon WP: MR imaging of brain tumors—Toward physiologic imaging. Am J Neuroradiol 2000; 21:1570–1571. 92. Tovi M, Lilja A, Bergstrom M, et al: Delineation of gliomas with magnetic resonance imaging using Gd-DTPA in comparison with computed tomography and positron emission tomography. Acta Radiol 1990;31:417– 429. 93. Waldrop SM, Davis PC, Padgett CA, Shapiro MB, Morris R: Treatment of brain tumors in children is associated with abnormal MR spectroscopic ratios in brain tissue remote from the tumor site. Am J Neuroradiol 1998;19:963–970. 94. Cloft HJ, Matsumoto JA, Lanzino G, Cail WS: Posterior fossa hemorrhage after supratentorial surgery. Am J Neuroradiol 1997;18: 1573–1580. 95. Poussaint TY, Siffert J, Barnes PD, et al: Hemorrhagic vasculopathy after treatment of central nervous system neoplasia in childhood—Diagnosis and follow-up. Am J Neuroradiol 1995;16:693–699. 96. Gaensler EHL, Dillon WP, Edwards MSB, Larson DA, Rosenau W, Wilson CB: Radiation-induced telangiectasia in the brain simulates cryptic vscular malformation at MR imaging. Radiology 1994;193: 629–636. 97. Valk PE, Dillon WP: Radiation injury of the brain. Am J Neuroradiol 1991;12:45–62.
W9418-03.qxd 11/3/03 2:48 PM Page 103
Chapter 3 Introduction to Neurophysiology Dean Lin, MD, PhD, Ehud Mendel, MD, and Bernard H. Guiot, MD, FRCSC
Introduction The neuron (Fig. 3-1), embryologically derived from epithelial precursor cells, is charged with the highly specialized function of signal transmission. It contains the basic cellular organelles required to carry out cellular metabolism and maintenance. These include a cytoskeleton, a nucleus, an endoplasmic reticulum, a Golgi apparatus, mitochondria, and liposomes. The neuronal membrane is however, unique in that it contains a very high concentration of proteins embedded into its phospholipid bilayer. The proteins form specialized pores that function as either ion-specific pumps or channels. These allow for the passage of ions through the impermeable plasma membrane (Fig. 3-2). All ionic species, if allowed to pass freely through open channels, would flow across the neuronal membrane until the electrical and the chemical driving forces equilibrate. An electrical potential, specific for every ion, exists at this state of equilibrium and can be calculated using the Nernst equation.1 The net equilibrium potential for a given neuron depends on the state of the various ionic channels. Each ion channel, when open, will shift the membrane potential toward the equilibrium potential of that particular ion (ENa = +55 mV, ECl = -60 mV, and EK = -75 mV). At rest, the neuron is relatively impermeable to the passage of all ions except potassium. The charge imbalance that arises therefore produces a resting membrane potential (RMP) that approximates the equilibrium potential for potassium (RMP = -65 mV).2 When the membrane depolarizes, the sodium channels open, thereby shifting the equilibrium potential closer to that of sodium.
Most ion channels are composed of multiple protein subunits, which form a central pore. Ion channels permit the nearly instantaneous translocation of thousands of charged ions from one side of the plasma membrane to the other. A single channel can gate up to one billion ions in a second.2 Channels are generally very specific with regard to the ionic species they gate. Thus sodium channels are 10 to 20 times more specific for sodium than for potassium, despite the fact that both are monovalent cations.3 Only two broad classes of channels will be considered in this chapter, namely voltage-gated channels and ligand-gated channels. Voltagegated channels (Fig. 3-3, p. 106) are activated by a change in the transmembrane potential, which triggers a conformational change in the channel’s tertiary structure. This conformational change opens the channel’s central pore, thereby permitting a flux of ions. Common examples include voltage-gated sodium- and calcium-channels (Fig. 3-4, p. 106), which allow ions into the cell, and voltage-gated potassium channels, which cause a potassium efflux. Ligandgated channels undergo a conformational change when activated by agonist binding to the channel’s extracellular domain. These specialized channels are receptors. Their activation may be excitatory or inhibitory onto the postsynaptic neuron. The receptors play a vital role in maintaining homeostasis between the excitatory and inhibitory tone in the central nervous system (CNS); therefore, derangements in their function can lead to a number of pathologic states. Receptors can be grouped into two main categories based on the mechanism by which they implement their effects. 103
W9418-03.qxd 11/3/03 2:48 PM Page 104
104
Section I
Introduction
Figure 3-1. Anatomy of a neuron. Although there are numerous neuronal morphologies, all neurons have several basic features in common that allow signal transmission. Incoming signals are received at the postsynaptic terminals of dendrites. These signals subsequently travel along dendritic trees until they arrive at the cell body. Action potentials may be formed at the axon hillock and subsequently propagate along the axons with or without the assistance of a myelin sheath, until the signal reaches the presynaptic terminal. Arrival of an action potential at the presynaptic terminal induces release of neurotransmitter into the synaptic cleft, thereby propagating the signal to another neuron.
Ionotropic Receptors
Glutamate Receptors
Activation of ligand-gated ionotropic receptors produces a conformation change, which causes the ion channel to open. Immediately, a large ionic flux occurs resulting in a small, transient alteration in the membrane potential. The direction of this change determines whether an excitatory or an inhibitory signal is conveyed. As ligand unbinds from the receptor, thereby reversing the initial opening conformational change and closing the central pore, the flow of ions ceases just as rapidly as it began. Some examples of ligandgated channels include the ionotropic glutamate receptors (N-methyl-D-aspartate [NMDA], a-amino-3-hydroxyl-5methyl-4-isoxazolepropionic acid [AMPA], and kainate), the gamma-aminobutyric acid (GABAA) receptor, glycine receptors, and nicotinic acetylcholine receptors.
AMPA and kainate receptors (Fig. 3-5) are rapidly activated by ligand binding, and thus are involved in the rapid upstroke of an excitatory postsynaptic potential. These channels primarily gate sodium and potassium, but a small percentage of AMPA receptors also gate calcium, depending on the presence or absence of the glu-R2 subunit. If this subunit is present, as it is in the vast majority of AMPA receptors, the channel is impermeable to calcium ions.4 In the receptor subset lacking the glu-R2 subunit, however, calcium is able to flow through the pore, although to a much lesser degree than is gated by NMDA receptors. NMDA receptor channels are much slower to open and close than the AMPA and kainate receptors. They are therefore responsible for the latter phases of the excitatory poten-
W9418-03.qxd 11/3/03 2:48 PM Page 105
Chapter 3
Introduction to Neurophysiology
105
Figure 3-2. Electrochemical gradients determine the flow of ions across open ion channels. Channel opening results in an ionic flux either into or out of the neuron, depending on the driving forces. The intracellular concentration of potassium ions, for example, is significantly higher than the extracellular concentration. The neuronal interior, however, is negatively charged, and subsequently the electrical potential tends to drive cations into a neuron. Opening of a potassium channel results in an efflux of potassium ions because the concentration gradient generates an electromotive force much greater than the force generated by the opposing electrical gradient. As long as the channel remains open, ions will flow until electrochemical equilibrium is achieved.
tial. Their major role is in determining intracellular calcium levels. Indeed, under physiologic conditions, NMDA receptor activation and the subsequent increase in intracellular calcium may be vital to the formation of long-term memory.2 Hyperstimulation of NMDA receptors and pathologically elevated levels of intracellular calcium, however, are believed to be key in various CNS pathologies, such as stroke and traumatic brain injury.5,6 There are several features unique to the NMDA receptor. First, binding of glutamate alone is insufficient to open the NMDA channel; the cofactor glycine must also be present and bind to the receptor concurrently to affect channel opening. Typically, the ambient glycine in the extracellular space is sufficient to activate the receptor. Another unique property of the NMDA receptor is the magnesium blockade. Even after both glutamate and glycine have bound, the NMDA receptor will remain shut at resting membrane potential because extracellular magnesium ions obstruct the pore in a voltagedependent manner. Only after the membrane has been depolarized to between -40 mV and -20 mV by AMPA and kainate receptors is the electrostatic repulsion sufficient to
expel magnesium, relieving the blockade and permitting sodium and calcium influx.4 Gamma-aminobutyric Acid and Glycine Receptors Activation of inhibitory ionotropic receptors such as GABAA and glycine receptors generates an inhibitory postsynaptic potential. GABA mainly localizes to the brain, where it serves as the primary inhibitory neurotransmitter, whereas glycine is the primary inhibitory neurotransmitter in the spinal cord. These channels gate chloride into neurons. The open chloride channels can be inhibitory in two different ways. First, a depolarized membrane is repolarized toward the chloride equilibrium potential (ECl = -60 mV) by the anion flux. Second, when a membrane is at resting potential, opening chloride channels would not hyperpolarize the membrane any further because the ECl and the resting potential are nearly equal. Instead, open chloride channels essentially short-circuit incoming excitatory postsynaptic potentials (EPSPs) by gating anions and clamping the membrane at ECl as an excitatory influx of cations attempts to
W9418-03.qxd 11/3/03 2:48 PM Page 106
106
Section I
Introduction depolarize the membrane. This phenomenon is known as shunting.2 Because the GABAA and glycine receptors essentially regulate inhibitory tone in the CNS, they play a crucial role in shaping neuronal firing patterns. Thus, derangements in their function have been shown to result in hyperexcitability states. For example, the muscular hypertonicity and rigidity resulting from strychnine poisoning are mediated by strychnine’s antagonism of the glycine receptor.2 As another example, research has shown that GABAA receptor dysfunction in the hippocampal region is implicated in temporal lobe epilepsy.7
Figure 3-3. Selectivity of voltage-gated channels. Many of the channels lining the polarized neuronal membrane are voltagesensitive, ion-specific channels. This figure shows a channel specific for potassium ions. The basis for the selectivity filter is not size alone; although sodium is a smaller monovalent cation than potassium, potassium-selective channels are many times more selective for potassium ions. A change in the membrane voltage potential induces a conformational change that causes voltage-sensitive potassium channels to open, thereby selectively gating potassium down its electrochemical gradients.
B
Metabotropic Receptors These receptors act quite slowly through intracellular second messengers. As such, they do not convey fast, discrete signals. Rather, they serve a modulatory role in neurons, regulating basal levels of excitation that last much longer than the brief inhibitory postsynaptic potentials (IPSPs) or EPSPs. Activation of metabotropic receptors rarely generates enough depolarization to result in an action potential. Although they do facilitate opening and closing of
Figure 3-4. Selectivity filter of voltage-gated channels. Cations are surrounded by a sphere of water molecules. Theoretically, the basis of the selectivity filter is active sites within the channel pore. Sodium ions traversing bind weakly to active sites within the channel pore. As a sodium ion passes through the sodium-selective channel, the sodium ion fleetingly binds to a negatively charged amino acid residue lining the channel pore while an adjacent water molecule is stabilized by a negatively charged residue on the opposite wall. Within a sodium-selective channel, active site binding is specific for the sodium’s unique cationic charge and hydration sphere. Subsequently, other cations are excluded.
W9418-03.qxd 11/3/03 2:49 PM Page 107
Chapter 3
A
Introduction to Neurophysiology
107
B
Figure 3-5. Glutamate receptors. There are three basic ionotropic glutamate receptors, each with numerous binding sites for a number of ligands. A, NMDA receptors require both binding of glutamate and glycine, and membrane depolarization to initiate the conformation change that results in the gating of sodium, potassium, and calcium. Membrane depolarization must occur in order to relieve the magnesium pore blockade. The NMDA receptor also has binding sites to which modulators such as zinc and PCP bind. B, Non-NMDA receptors such as kainate and AMPA gate sodium and potassium ions. These receptors also have numerous binding sites for modulators.
ion channels, binding of ligand to metabotropic receptors may have more profound effects on resting membrane potential, action potential duration, passive membrane properties, and presynaptic neurotransmitter release.1 Furthermore, metabotropic receptors may be excitatory or inhibitory. Onset of action may be delayed several seconds after ligand binding because metabotropic receptor activation typically initiates a cascade of events with multiple intracellular players. In turn, their modulatory effects can last for minutes or longer. Metabotropic receptors initiate reaction cascades, which may activate transcription factors. They may also generate numerous second messengers, which results in signal amplification. For example, norepinephrine binding to a catecholamine receptor may activate a G-protein.2 Each stimulatory G-protein then induces activation of multiple adenyl cyclase enzymes, which subsequently generate copious amounts of cyclic adenosine monophosphate (cAMP) from adenosine triphosphate (ATP). Each cAMP molecule can then proceed to activate numerous protein kinase A molecules. In this manner, a relatively small number of metabotropic receptors can cause widespread and longlasting effects in postsynaptic neurons. Some examples of metabotropic receptors include catecholamine receptors, neuropeptide receptors, metabotropic glutamate receptors, the GABAB receptor, and muscarinic receptors.
Signal Transmission Neurons communicate using both electrical and chemical forms of signaling. The fundamental unit by which a neuron conducts information is the action potential, a localized, ephemeral, self-regenerating membrane depolarization that typically forms at a trigger zone, which is anatomically correlated to the initial segment of the axon. The process begins with the localized influx of sodium into the neuron, producing a zone of focal membrane depolarization. If the resulting membrane potential reaches a threshold, then an action potential is generated. This threshold, known as the action potential threshold, is -55 to -35 mV.3 At the onset of the action potential, a large number of sodium channels are suddenly opened, resulting in an almost instantaneous and massive regional depolarization of the plasma membrane. The rapid influx of sodium, known as the upstroke or rising phase of the action potential, produces membrane potential changes of 70 to 90 mV. These potential changes can, in turn, depolarize the membrane in adjacent segments. If the action potential threshold is reached in these neighboring regions, then the action potential is spread to contiguous membrane segments. The process may repeat itself along the entire length of the neuron. Membrane depolarization does not continue unchecked. Two voltage-dependent mechanisms work to halt the action potential upstroke.
W9418-03.qxd 11/3/03 2:49 PM Page 108
108
Section I
Introduction
1. Voltage-sensitive sodium channels open and gate sodium when the membrane becomes slightly depolarized. Further depolarization, however, causes these channels to close.3 Less than a millisecond after they open, the sodium channels rapidly transition into an inactive, closed state. The influx of sodium ceases and the depolarizing drive stops. 2. Depolarization of the membrane also causes voltagesensitive potassium channels to open. Potassium flows down its electrochemical gradient out into the extracellular space, causing the membrane to repolarize, and counteracting the sodium channel-mediated depolarization. Potassium channels cycle from closed to open to closed states more slowly than their sodium-channel counterparts and, subsequently, the potassium efflux does not peak until the majority of sodium channels transition to their closed and inactive states.1,3 The membrane subsequently repolarizes as the membrane is driven back toward potassium’s equilibrium potential (EK = -75 mV). Because potassium’s equilibrium potential is more polar than the cell’s resting membrane potential (-65 mV), the membrane is transiently hyperpolarized before returning to its resting potential. These depolarization-mediated mechanisms repolarize the cell membrane relatively quickly. As a result, action potentials last only a few milliseconds. If action potential propagation were to occur by simply depolarizing adjacent axonal segments, signal transmission would be quite slow. Indeed, a signal originating in the brain and transmitted to one’s foot, for example, could take up to 2 seconds to reach its destination. This delay is incompatible with survival. Rather, signal transduction occurs by saltatory conduction. The basis for this rapid form of conduction is the myelin sheath.3 Most axons are invested by myelin, a phospholipid extension of a glial cell membrane. In the CNS, the glial cells responsible for myelinating axons are oligodendrocytes. In the peripheral nervous system, Schwann cells provide the myelination. Each oligodendrocyte projects myelin appendages to the axons of several neurons. These appendages form concentric lipid layers around the axon, with each extension covering 1 to 2 mm of axonal membrane. The insulated regions of the axon, or internodes, are interrupted by 2-mm segments of bare membrane known as nodes of Ranvier. These nodes are specialized portions of membrane that contain a high density (1000 to 2000 channels/mm2) of sodium channels.2 These nodal regions are therefore extremely sensitive to membrane depolarization. The myelin sheath functionally increases the membrane thickness 100-fold, thereby decreasing the membrane’s electrical capacitance and increasing membrane resistance. The increased resistance precludes sodium influx in an internodal segment.3 Rather, current flow is forced down the longitudinal axis of the axon to the next node of Ranvier, where
there are exposed sodium channels. In this way, the action potential effectively skips from node to node and is regenerated one internode away from its original location. This skipping action, known as saltatory conduction, increases action potential velocity up to 100-fold, so that signals can propagate as briskly as 120 m/sec.2 This velocity is clearly more suitable for transmitting signals. At the axon terminal, the action potential initiates a cascade of events, which results in the transmission of a signal from one neuron to another. The interneuronal site of communication is called a synapse. The presynaptic terminal is separated from the postsynaptic terminal by a synaptic cleft. Two basic kinds of synapses exist, namely electrical and chemical. Electrical Synapse The neurons in this type of synapse are linked by gap junctions. These form narrow (3.5 nm) bridges through which cytoplasm and small intracellular metabolites can pass.2 The cytoplasm is essentially continuous between neurons, so that current is conducted from one neuronal membrane to the next with negligible delay. Chemical Synapse These account for the vast majority of synapses in the central nervous system. In contrast to electrical synapses, neurons in a chemical synapse are not in physical contact. The synaptic clefts are large, typically measuring between 20 and 40 nm.2,8 The action potential is not able to span this distance. Rather, depolarization at the presynaptic terminal results in the release of neurotransmitters into the synaptic cleft.
Neurotransmitters and Synaptic Transmission Neurotransmitters are broadly divided into two classes: small-molecule transmitters and neuroactive peptides (Table 3-1). Both are stored within vesicles at the axon terminal and are released into the synapse through exocytosis.8 Neurotransmitters are packaged into membrane-bound vesicles, concentrated at the axon terminal. At rest, the vesicles are anchored to presynaptic cytoskeletal elements by a class of proteins known as synapsins. The vesicles are attached just proximal to the active zone. There are four known members of the synapsin family: Ia, Ib, IIa, and IIb. In their basal, unphosphorylated states, the synapsins form a bridge between the cytoskeleton and the vesicles. When an action potential arrives at the presynaptic terminal, these protein bridges break down, permitting vesicular fusion with the presynaptic plasma membrane and subsequent release of neurotransmitter into the synaptic cleft. They are released at a synapse, resulting in changes in a postsynaptic cell.
W9418-03.qxd 11/3/03 2:49 PM Page 109
Chapter 3
Table 3-1 Two Major Classes of Neurotransmitters Small-Molecule Transmitters
Neuroactive Peptides
Acetylcholine Dopamine Norepinephrine Epinephrine Serotonin Histamine Gamma-aminobutyric acid Glycine Glutamate
Hypothalamic-releasing hormones Neurohypophyseal hormones Pituitary peptides Gastrointestinal peptides Others
Transmitter release begins with presynaptic depolarization (Fig. 3-6). The depolarization, caused by a shift of sodium ions, results in the activation of voltage-gated calcium channels. Although these channels are sparsely dis-
Introduction to Neurophysiology
109
tributed through most of the axon, they are concentrated at the active zone, which is the site of neurotransmitter release. Activation of the calcium channels results in an influx of calcium into the cell, in keeping with the large gradient at the axon terminal. The abrupt increase in intracellular calcium levels leads to the rapid activation of calcium/calmodulin protein kinases, which phosphorylate the synapsins.2,8 This phosphorylation triggers a conformational change that extricates the synaptic vesicles from cytoskeletal incarceration. The vesicles then proceed to the active zone with the assistance of the guanosine triphosphate (GTP)—bound proteins, Rab3A and Rab3C. The Rab3vesicle complex is diverted toward the active zone, and on arrival, GTP is hydrolyzed to guanosine diphosphate (GDP) and an inorganic phosphate as the vesicle dissociates from the Rab3 molecule near the active zone. The Rab3 protein then exchanges the GDP for a GTP molecule in priming itself to ferry another vesicle to the active zone. The vesicle then
Figure 3-6. Release of neurotransmitter vesicles at the presynaptic terminal. Numerous steps are involved in the fusion of a vesicle with the presynaptic membrane. (1) Mobilization: As local calcium concentrations skyrocket, synapsins that normally tether vesicles to the neuronal cytoskeleton release the vesicles. (2) Targeting: Rab3 proteins subsequently hydrolyze bound GTP molecules and guide the vesicle toward its release site. (3) Docking-priming: v-SNARE proteins such as synaptobrevin then dock the vesicles with t-SNARES (syntaxin and SNAP-25) and prime the vesicles for membrane fusion. (4) Fusion pore: Through mechanisms not entirely elucidated, a slender pore subsequently forms between the vesicle and the plasma membrane, initiating the release of neurotransmitter molecules into the synaptic cleft.
W9418-03.qxd 11/3/03 2:49 PM Page 110
110
Section I
Introduction
must dock in the active zone.2 In docking a vesicle, proteins bound to the vesicular membrane (vesicular-SNAREs or vSNAREs) interact with proteins bound to the plasma membrane (target-SNAREs or t-SNAREs). In neurons, the only v-SNARE identified thus far is VAMP (otherwise known as synaptobrevin). Linkage of VAMP with the t-SNAREs, syntaxin and SNAP-25, forms a stable protein complex, and this complex serves to secure the vesicle in position for impending fusion with the plasma membrane. This protein complex is later disassembled by N-ethylmaleimide-sensitive fusion (NSF) protein and soluble NSF attachment protein (SNAP) as the vesicle is recycled. Synaptotagmin is another protein that plays an integral role in neurotransmitter release. Although its exact functions have yet to be elucidated, possible actions, among others, include regulation of neurotransmitter release, a role as a v-SNARE, and facilitation of vesicular recycling.2,8 Soon after the vesicle is properly positioned, a narrow fusion pore (1 to 2 nm in diameter) temporarily forms, joining the vesicular membrane with the plasma membrane. Formation of the pore is the first step in fusion of the vesicle with membrane. After the pore has been established, the pore promptly dilates, dumping the vesicle’s contents into the cleft, thereby completing fusion of the vesicle with the presynaptic membrane.2 The neurotransmitter molecules then rapidly diffuse into and across the synaptic cleft to interact with postsynaptic receptors. Neurotransmitter is cleared from the synaptic cleft within milliseconds. The mechanisms vary from neurotransmitter to neurotransmitter, but three basic mechanisms exist. Reuptake The most common means of terminating a signal is reuptake. High-affinity transporter proteins embedded in neuronal and glial membranes rapidly convey neurotransmitter molecules from within the cleft back to the intracellular compartment shortly after neurotransmitter is released. The neurotransmitter may be repackaged into vesicles and then re-released with subsequent depolarizations. Numerous medications antagonize neurotransmitter reuptake. These agents effectively permit continued and protracted postsynaptic neurotransmitter-receptor interaction, thereby amplifying the postsynaptic signal. The selective serotonin reuptake inhibitors, as well as cocaine and other sympathomimetics, are examples of agents that use this mechanism of action.1 Enzymatic Degradation Neurotransmitters may be enzymatically degraded within the synaptic cleft. The classic example of degradation is the breakdown of acetylcholine by acetylcholinesterases near the postsynaptic acetylcholine receptors at the neuromuscular junction. The choline generated by acetylcholine hydrolysis
then is rapidly taken up to be recycled. Other neurotransmitters that are broken down enzymatically include GABA and some of the neuropeptide neurotransmitters.1,2 Diffusion Diffusion plays a role in clearance of every neurotransmitter. After achieving high synaptic cleft concentrations, neurotransmitter rapidly diffuses away from its site of action into the extracellular space.
Postsynaptic Signal Transduction When synaptic vesicles fuse with presynaptic membranes, neurotransmitter concentrations in the synaptic cleft increase exponentially in less than a millisecond. The neurotransmitter molecules rapidly diffuse to the postsynaptic membrane and saturate, or fully occupy, the postsynaptic receptors. Binding of neurotransmitter to a receptor induces conformational changes that initiate a signal cascade originating at the postsynaptic membrane.8 In this manner, the neurotransmitters carry a message from a presynaptic membrane, across the void of the synaptic cleft, to a postsynaptic neuron. Activation of postsynaptic receptors generates a postsynaptic potential. A postsynaptic potential is a sudden, brief alteration in the local membrane potential followed by a gradual return to baseline potential over a few hundred milliseconds. There are two kinds of postsynaptic potentials; an EPSP is generated by activation of excitatory receptors, whereas an IPSP results from inhibitory receptor activation. Similar to an action potential, a postsynaptic potential propagates along neurites; however, a postsynaptic potential is neither an all-or-none nor a self-regenerating phenomenon. Rather, it is a graded response the amplitude of which is dependent on the number of activated receptors and which propagates electrotonically, or passively, independent of sodium channel activation. Therefore, a postsynaptic potential degenerates rapidly, unlike a self-regenerating action potential that travels by saltatory conduction. All three ionotropic glutamate receptors are involved in an EPSP. AMPA and kainate receptors are rapidly activated by ligand binding, and thus are involved in the rapid upstroke of an excitatory postsynaptic potential. As the NMDA receptor is slow to open and close, its activation produces the tail of an excitatory potential. IPSPs, although usually mediated by a single kind of receptor, may also be mediated by a mixed population of receptors.9 CNS neurons are constantly bombarded by excitatory and inhibitory inputs from other neurons. Tens of thousands of synapses, in fact, may grace a single motor neuron, and part of the neuron’s function is to integrate these signals into the all-or-none response of an action potential. How, then, does a neuron know when to fire an action potential? To begin with, it is important to realize that a single EPSP is typically
W9418-03.qxd 11/3/03 2:49 PM Page 111
Chapter 3
insufficient to generate an action potential. Even without input from inhibitory synapses, tens or even hundreds of EPSPs in proximity both temporally and spatially may be necessary to produce enough depolarization to reach action potential threshold.2 At the same time, IPSPs are also arriving and counteracting the EPSPs. A neuron integrates these opposing inputs into a coherent signal to pass along the information that it is receiving. Charles Sherrington termed the summation of inputs the “integrative action of the nervous system.”10 With the highest density of voltage-sensitive sodium channels, the axon hillock is the usual site of synaptic integration. At the hillock, depolarization has the highest probability of opening the most sodium channels, thereby bringing the membrane closer to threshold potential. Only a 10-mV depolarization from resting potential is necessary to fire an action potential at the hillock. This relatively small depolarization is in sharp contrast to the 30-mV depolarization typically required elsewhere along the neuron. Thus, if an action potential is to fire from anywhere on a neuron, it is most likely to fire from the hillock.2 How does the hillock summate the synaptic inputs? First, it is important to understand two passive properties of neuronal membranes, the length constant and the time constant. The length constant determines the degree to which a potential degrades over a fixed distance. The amount of degradation is crucial because most excitatory synapses are axo-dendritic and thus lie a long distance from the axon hillock. The result is that the depolarization generated by a single EPSP is typically too weak to fire an action potential because it is but a fraction of its initial amplitude by the time it reaches the hillock. How much signal remains is dictated by the membrane’s length constant; a large length constant indicates that a potential will degrade only slightly over a long distance whereas a small length constant indicates that the potential will decay completely over a short distance.2 In contrast, the time constant is another passive membrane property that influences summation. As the name implies, the time constant determines how rapidly a postsynaptic potential decays over time. A large time constant implies that a postsynaptic potential will decay very little over a given period whereas a small time constant indicates that a potential will decay rapidly.11 Although the length and time constants determine the degree to which postsynaptic potentials have decayed before they reach the hillock, it is rare that a single potential will arrive without having encountered another potential during its journey. Postsynaptic potentials, when they interact with one another, have a tendency to combine to form a single potential. This property is especially important for EPSPs because individual EPSPs are usually too small to generate an action potential. When two EPSPs merge, however, they summate and form a larger-amplitude EPSP,11 one more capable of reaching action potential threshold. Thus, summation is critical to synaptic communication because
Introduction to Neurophysiology
111
without summation, very few action potentials would be fired. There are two kinds of summation, spatial summation and temporal summation. The first is heavily influenced by the length constant whereas the time constant largely determines the latter. With spatial summation, two inputs, traveling along different dendrites from distal synapses, encounter each other at a common branch point and merge to form a single potential. Because the length constants of individual dendrites dictate the extent to which these disparate signals degrade, the length constant influences the fraction of signal remaining from each distal dendrite. With dendrites that have small length constants, spatial summation will occur more infrequently or be rather inconsequential, because most potentials will degenerate completely before they summate. In this manner, the magnitude of the length constant impacts the integration of multiple synaptic inputs; a neuron with a larger length constant is more likely to summate potentials. Thus, this neuron will fire more action potentials when it receives a constant number of EPSPs than another neuron with smaller length constants.2 On the other hand, temporal summation occurs when multiple inputs arrive at a single point in rapid succession. In this situation, the second potential may ride the coattails of the first potential and summate with the initial potential if it has not yet decayed completely. Therefore, temporal summation is governed by the time constant because this passive membrane property determines the fraction of signal remaining after a given period.2 Thus, the magnitude of the time constant can have a significant effect on the integrative properties of a neuronal membrane. Similar to spatial summation, a neuron with larger time constants will tend to integrate more potentials than those with smaller time constants and may thus fire more action potentials. As a neuron receives input from other neurons, the multitudes of excitatory signals arriving on distal dendritic branches propagate toward the soma, converging on the axon hillock. As previously discussed, action potentials typically do not originate along these dendrites because of the low concentration of sodium channels. By the time these EPSPs reach the soma, however, they may have undergone enough summation that they are of sufficient amplitude to generate an action potential at the hillock. Before an EPSP can reach the axon hillock, however, it must propagate through the soma. Most inhibitory neurons synapse at the soma and hillock, and the soma functions as a gatekeeper and filter.2 In traveling along dendrites, most interpotential interactions have been excitatory up to this point and therefore additive in nature, resulting a large EPSP. Interactions between EPSPs and IPSPs at the soma, however, have a negating effect, as one would expect, and the opposite potentials short-circuit one another. Because most GABA-ergic synapses occur at the soma, the incipient EPSPs reaching the soma on their way to the hillock are heavily curtailed. In this manner, IPSPs increase the size of the EPSPs necessary to
W9418-03.qxd 11/3/03 2:49 PM Page 112
112
Section I
Introduction
reach and depolarize the axon hillock. Should the EPSPs surviving the expedition still be of sufficient amplitude, the axon hillock, loaded with its high density of sodium channels, will fire an action potential. The cycle is then complete; this action potential, like the presynaptic action potentials that resulted in the EPSPs that ultimately depolarized this neuron’s hillock, will travel down this neuron’s axon, acting as the combined expression of numerous signals arriving at this central neuron. As with other action potentials, this one will eventually result in the release of neurotransmitter onto other postsynaptic receptors, thereby continuing to propagate information.
Pathophysiology Under normal efficiently and however, the described may function. Two
circumstances, neurotransmission proceeds effectively. In some situations (Fig. 3-7), well-tuned neurophysiologic mechanisms fail, leading to significant neurologic dystypical injuries that disrupt physiologic
neurotransmission are traumatic CNS injury and ischemic cerebrovascular accidents, or stroke. CNS trauma is a result of mechanical injury to the brain or spinal cord, whether it be by motor vehicle accident, gunshot wound, or any mechanism that physically disrupts CNS tissue. Strokes, on the other hand, occur as a result of ischemia. The CNS receives the greatest blood flow given its volume of any organ in the body, and it is exquisitely sensitive to decreases in its supply of oxygen and glucose. Classically, both trauma and ischemia cause two stages of CNS damage. The initial insult is the instantaneous damage that occurs at the time of the trauma or ischemia. The primary insult results in irreversible neuronal dysfunction and is typically characterized pathologically by necrosis at the primary site of injury, in which membrane failure and local inflammation lead to swelling of the neuron and its organelles.12 The lack of oxygen and glucose results in the depletion of cellular energy stores. The deficiency of ATP causes the energy-dependent ion pumps to fail, so that ionic homeostasis fails and neurons depolarize. This phenomenon, known as anoxic depolarization, has serious repercus-
Figure 3-7. Pathophysiologic states. Synaptic transmission may be hindered or arrested at multiple sites along a neuron. For example, numerous medications augment postsynaptic receptor function by inhibiting neurotransmitter re-uptake (selective serotonin release inhibitors, tricyclic antidepressant drugs) or augmenting neurotransmitter binding (benzodiazepines). Far too often, however, a pathophysiologic state is present that is pernicious to normal signal transduction (tetrodotoxin). For example, demyelinating diseases such as Guillain-Barré syndrome or multiple sclerosis hinders action potential propagation, thereby precluding effective signal transduction.
W9418-03.qxd 11/3/03 2:49 PM Page 113
Chapter 3
sions at axonal terminals.6,13 Depolarization results in a large influx of sodium and calcium into neurons and glia, both at and around the primary injury site; chloride ions and water soon follow. A large efflux of potassium ions also occurs. The overall cellular energy failure prevents ion pumps from correcting the derangements in ionic homeostasis, and cellular edema rapidly ensues.13 At the axon terminals, depolarization opens voltage-gated calcium channels, which then results in the massive release of vesicular glutamate, thereby increasing the extracellular glutamate concentration. Extracellular concentrations of other neurotransmitters such as GABA and adenosine also increase at this time.14,15 The increased glutamate levels, however, are believed to result in excitotoxicity, a pathologic state characterized by the hyperstimulation of glutamate receptors, which may inevitably lead to neuronal degeneration and death. The glutamate level peaks at 5 minutes after injury, before returning to basal concentrations. Studies have demonstrated that the magnitude and duration of the increased glutamate levels are directly correlated with the severity of injury.14,15 Although the initial step in a neuron’s downward spiral is the dramatic increase in glutamate levels, it is the resultant increase in intracellular calcium that directly causes cellular damage. One must keep in mind that cytoplasmic calcium concentrations are approximately 100 nM, and the extracellular concentrations are typically greater than 1 mM. Thus, there is a powerful driving force for calcium ions to enter neurons. There are many means by which calcium can enter neurons. Entry occurs through the various glutamate receptors, as well as via voltage-gated calcium channels, opened as a result of membrane depolarization.5,16,17 In addition, calcium enters through the sodium-calcium antiporter, which, under normal circumstances, transports calcium out of cells. After injury, however, the antiporter operates in reverse, moving calcium into the neuron while pumping sodium into the extracellular space. Finally, metabotropic glutamate receptors are activated and induce secondmessenger cascades that increase cytoplasmic inositol 1,4,5triphosphate concentrations, thereby releasing calcium from intracellular stores. The calcium buffering mechanisms are rapidly overloaded, and free calcium ion concentrations increase rapidly, setting into motion a number of potentially pernicious chain reactions.14,15 How does calcium wreak havoc within cells? Under physiologic circumstances, calcium serves a number of roles. Many cytosolic proteins are calcium-dependent and become activated when calcium is bound. These include kinases, phospholipases, and proteases. When intracellular calcium levels go unchecked, these proteins are activated with devastating consequences. Activated phospholipases such as phospholipase A2 and phospholipase C begin to digest cellular phospholipids such as the plasma membrane and membranes lining organelles. Calcium-activated intracellular proteases such as the calpains chew up structural and regulatory proteins such as actin and spectrin, while nucleases
Introduction to Neurophysiology
113
induce deoxyribonucleic acid (DNA) strand breaks.13,18,19 Some of these DNA breaks may set off a chain of events that irrevocably leads to cellular apoptosis. High intracellular calcium concentrations also stimulate the arachidonic acid cascade, activating 5-lipoxygenase, prostaglandin synthase, and neuronal nitric oxide synthase, which in turn stimulate free radical generation. These free radicals are mostly reactive oxygen species that are potentiated by iron liberated from hemoglobin, transferrin, and ferritin secondary to increased environmental acidity.6 Iron serves as a catalyst for free radical reactions, such as lipid peroxidation and protein destruction. The resultant damage takes a toll as membranes begin to leak and transmembrane ion pumps and other membrane proteins begin to fail. Consequently, derangements in ionic homeostasis worsen. Similar damage is dealt to the mitochondrial membrane, resulting in mitochondrial swelling and loss of cellular energy stores as the electron transport system is disrupted. Furthermore, the mitochondrial permeability transition pore is affected, liberating cytochrome C from its intramembranous home in the mitochondria into the cytoplasm. Cytochrome C sets off a cascade of events that ultimately activates caspases, which are aspartate-specific cysteine proteases. The activation of caspases, especially caspase-3, has been shown in a number of studies to be a late-stage event in apoptosis.13,20 The connection between free radical generation and neuronal damage is corroborated by in vitro studies demonstrating that free radical scavengers and antioxidants substantially curtail the damage incurred by neurons after trauma or ischemia.21,22 In addition, cyclosporin A, an immunosuppressant that has been shown to preclude the translocation of cytochrome C into the cytoplasm, appears to decrease tissue damage following both stroke and trauma in animal models.23,24 The secondary insult is the damage that occurs several hours to weeks after the primary insult and affects regions neighboring the initial injury site—known as the penumbra—where damage is initially insufficient to kill. In this region, cell death may occur as a result of apoptosis, or programmed cell death, a process in which a cell actively undergoes an orderly characteristic pattern of degenerative changes, such as chromatin condensation, cell shrinkage, internucleosomal DNA fragmentation, and the appearance of membrane-bound apoptotic bodies.15 Because disparate processes may result in cell death at different times following CNS damage, distinct interventional strategies may be more effective in halting apoptosis than in saving cells dying by necrosis. Apoptosis versus Necrosis As discussed previously, several of the reactions initiated by excitotoxicity lead to cellular necrosis while others tend to induce cellular apoptosis. What, then, are the factors that lead a dying neuron toward one and not the other? Certainly, activation of some genes have clearly been shown to be pro-
W9418-03.qxd 11/3/03 2:49 PM Page 114
114
Section I
Introduction
apoptotic (bax, bad, bcl-xS) while others are anti-apoptotic (bcl-2, bcl-xL),5 but what events activate or inhibit these gene products? First of all, the boundaries dividing necrosis and apoptosis are not as clear-cut as they were once thought to be. Recently, it has been shown that some of the classic pathologic features of apoptosis, namely DNA laddering and a positive TUNEL reactivity, may also be observed during necrosis. In addition, it appears that the two processes are not mutually exclusive of one another; a dying neuron may not follow a specific pathway but rather exhibit a mixture of morphologic features from both processes. Nevertheless, some researchers have proposed that several factors may be crucial to determining which path a cell takes: (1) the nature and severity of the insult, (2) the kind of cell damaged, (3) the cell’s stage of maturation, and (4) the complexity of the neuron’s dendritic tree. In addition, some investigators have proposed that intracellular calcium concentration may play a large role in determining the neuron’s fate, with lower calcium levels, perhaps, resulting in a propensity toward apoptosis.5 Such a premise, if it were shown to be true, would go a long way toward explaining why glutamate receptor antagonists have been such a disappointment after stroke or traumatic CNS injury. It is possible that the well-intentioned efforts to block excitotoxicity cause calcium concentrations to dip too low, thereby encouraging cellular apoptosis. This hypothesis still lacks a preponderance of evidence, however, and a great deal of debate remains. Other Mechanisms—Zinc, Inflammation, and Microvascular Changes There are other mechanisms, some of which are only peripherally related to excitotoxicity, that may be actively contributing to CNS injury. These mechanisms include zinc translocation, regional inflammation, and vascular changes in the CNS. The heavy metal zinc is a divalent cation found in all cells and is typically bound to metalloproteins. In the CNS, zinc’s physiologic role is unclear, although recent studies have shown that it may play a significant role as a modulator of neurotransmission.25–27 This heavy metal has been shown to be co-localized with glutamate in a number of excitatory pathways.28,29 Typically, zinc is only released with glutamate under conditions of high neuronal stimulation, and excess zinc liberation has been shown to be associated with a number of pathologic states such as epilepsy, stroke, and traumatic brain injury. In models of stroke, zinc translocates from nerve terminals into postsynaptic neurons. Zinc ions are believed to gain entry into neurons by the same routes by which calcium ions enter during states of membrane depolarization, namely voltage-gated calcium channels, NMDA channels, calcium-permeable AMPA channels, and sodium/calcium antiporters.30–33 In support of its possible role in neuronal damage, zinc has been shown to be present in neurons in concentrations up to 0.5 mM just before neu-
ronal degeneration.33 Additionally, the administration of a zinc chelator, calcium-ethylenediamine tetraacetic acid, administrated intracerebroventricularly blocks the translocation of zinc into postsynaptic cells and greatly attenuates cellular damage after an acute event.18 Again, a great deal of research remains to be performed with regard to this mysterious ion before a definitive causal relationship between zinc and neuronal damage is uncovered. Inflammatory processes may also play a role in exacerbating CNS damage. In some models of stroke, messenger ribonucleic acid transcripts for tumor necrosis factor-a and interleukin-1b have been increased as early as 1 hour after ischemia. These cytokines have been shown to increase brain water content, as well as exacerbate blood-brain barrier breakdown and increase production of free radicals. In addition, cellular adhesion molecules such as intracellular adhesion molecule–1, P-selectins, and E-selectins are upregulated on the surface of vascular endothelial cells in and around ischemic regions. These factors increase the inflammatory cells’ adhesion to the vessel walls, and augment vessel permeability to these cells. Neutrophils are the first cells to arrive at the scene after ischemia, followed shortly thereafter by macrophages and monocytes. The presence of these inflammatory cells not only leads to microvascular obstruction, further contributing to ischemia, but also to the release of toxic mediators onto wounded neurons. Corroborating the relationship between the inflammation and brain damage, numerous investigators have shown that blockade of the inflammatory response has been shown to decrease infarct volume substantially.5,34,35 Thus, antiinflammatory measures, such as corticosteroids, have been used with some success in traumatic injury. Microvasculature changes have also been demonstrated following traumatic brain injury. Under physiologic conditions, CNS arteries have a remarkable ability to autoregulate cerebral blood flow, maintaining an adequate supply of oxygen and glucose despite drastic changes in perfusion pressure. Several investigators have shown that, after traumatic brain injury, the microvasculature no longer responds appropriately to physiologic challenges. For example, acetylcholine application to microvasculature normally causes vasodilation; after trauma, however, acetylcholine causes vasoconstriction. Changes such as this may reduce blood flow to damaged brain tissue, exacerbating injury. Similarly, under hypocapnic conditions, vessels vasoconstrict under physiologic conditions; after traumatic injury, however, vessels vasodilate.12,36,37 In general, there is loss of normal vascular autoregulation, which may decrease brain tissue perfusion, resulting in more severe injury.
Conclusions The process of signal transmission within the central nervous system is complex. Membrane characteristics
W9418-03.qxd 11/3/03 2:49 PM Page 115
Chapter 3
Introduction to Neurophysiology
115
P earls 1. Most ion channels are composed of multiple protein subunits, which form a central pore. 2. Activation of ligand-gated ionotropic receptors produces a conformation change, which causes the ion channel to open. 3. Metabotropic receptors act quite slowly through intracellular second messengers. 4. This skipping action, known as saltatory conduction, increases action potential velocity up to 100-fold, so that signals can propagate as briskly as 120 m/sec. 5. . . . a large length constant indicates that a potential will degrade only slightly over a long distance whereas a small length constant indicates that the potential will decay completely over a short distance. 6. A large time constant implies that a postsynaptic potential will decay very little over a given period, whereas a small time constant indicates that a potential will decay rapidly.
underlie the separation of ionic species, required for the maintenance of a membrane potential. The action potential is propagated along the axon by saltatory conduction and reaches the axon terminal where neurotransmitters, bundled in vesicles, are released into the synapse. The postsynaptic receptor sites in turn give rise to excitatory or inhibitory potentials, which are integrated with the inputs from thousands of other neurons at the soma. These processes are disrupted in trauma and ischemia, leading to cell death through necrosis and apoptosis.
References 1. Nicholls JG, Martin AR, Wallace BG, Fuchs PA: From Neuron to Brain. Sunderland, England, Sinauer Associates, 2001. 2. Kandel ER, Schwartz JH, Jessell TM: Principles of Neural Science. New York, McGraw-Hill, 2000. 3. Waxmann SG, Kocsys JD, Stys PK: The Axon. Oxford, Oxford University Press, 1995. 4. Hollmann M, Heinemann S: Cloned glutamate receptors. Annu Rev Neurosci 1994;17:31–108. 5. Lee JM, Zipfel GJ, Choi DW: The changing landscape of ischaemic brain injury mechanisms. Nature 1999;399:A7–14. 6. Narayan RK, Wilberger JE, Povlishock JT: Neurotrauma. New York, McGraw-Hill, 1996. 7. Gibbs JW III, Shumate MD, Coulter DA: Differential epilepsyassociated alterations in postsynaptic GABA(A) receptor function in dentate granule and CA1 neurons. J Neurophysiol 1997;77:1924–1938. 8. Cowan WM, Sudhof TC, Stevens CF: Synapses. Baltimore, Johns Hopkins University Press, 2001. 9. Jonas P, Bischofberger J, Sandkuhler J: Corelease of two fast neurotransmitters at a central synapse. Science 1998;281:419–424.
7. The lack of oxygen and glucose results in the depletion of cellular energy stores. The deficiency of ATP causes the energy-dependent ion pumps to fail, so that ionic homeostasis fails and neurons depolarize. This phenomenon, known as anoxic depolarization, has serious repercussions at axonal terminals. 8. Although the initial step in a neuron’s downward spiral is the dramatic increase in glutamate levels, it is the resultant increase in intracellular calcium that directly causes cellular damage. 9. High intracellular calcium concentrations also stimulate the arachidonic acid cascade, activating 5-lipoxygenase, prostaglandin synthase, and neuronal nitric oxide synthase, which in turn stimulate free radical generation. 10. The activation of caspases, especially caspase-3, is shown in a number of studies to be a late-stage event in apoptosis.
10. Sherrington C: Integrative Action of the Nervous System. New Haven, Yale University Press, 1947. 11. Stuart G, Spruston N, Hausser M: Dendrites. Oxford, Oxford University Press, 1999. 12. Povlishock JT: Pathophysiology of neural injury: Therapeutic opportunities and challenges. Clin Neurosurg 2000;46:113–126. 13. Dirnagl U, Iadecola C, Moskowitz MA: Pathobiology of ischaemic stroke: An integrated view. Trends Neurosci 1999;22:391–397. 14. Obrenovitch TP, Urenjak J: Is high extracellular glutamate the key to excitotoxicity in traumatic brain injury? J Neurotrauma 1997;14: 677–698. 15. Zipfel GJ, Babcock DJ, Lee JM, Choi DW: Neuronal apoptosis after CNS injury: The roles of glutamate and calcium. J Neurotrauma 2000;17:857–869. 16. Choi DW: Calcium-mediated neurotoxicity: Relationship to specific channel types and role in ischemic damage. Trends Neurosci 1988;11:465–469. 17. Choi DW: Calcium: Still center-stage in hypoxic-ischemic neuronal death. Trends Neurosci 1995;18:58–60. 18. Lee JM, Grabb MC, Zipfel GJ, Choi DW: Brain tissue responses to ischemia. J Clin Invest 2000;106:723–731. 19. Zipfel GJ, Lee JM, Choi DW: Reducing calcium overload in the ischemic brain. N Engl J Med 1999;341:1543–1544. 20. Green DR, Reed JC: Mitochondria and apoptosis. Science 1998;281:1309–1312. 21. Hall ED, Braughler JM: Free radicals in CNS injury. Res Publ Assoc Res Nerv Ment Dis 1993;71:81–105. 22. Kontos HA, Povlishock JT: Oxygen radicals in brain injury. Cent Nerv Syst Trauma 1986;3:257–263. 23. Okonkwo DO, Buki A, Siman R, Povlishock JT: Cyclosporin A limits calcium-induced axonal damage following traumatic brain injury. Neuroreport 1999;10:353–358. 24. Uchino H, Elmer E, Uchino K, Lindvall O, Siesjo BK: Cyclosporin A dramatically ameliorates CA1 hippocampal damage following transient forebrain ischaemia in the rat. Acta Physiol Scand 1995;155:469–471.
W9418-03.qxd 11/3/03 2:49 PM Page 116
116
Section I
Introduction
25. Barberis A, Cherubini E, Mozrzymas JW: Zinc inhibits miniature GABAergic currents by allosteric modulation of GABAA receptor gating. J Neurosci 2000;20:8618–8627. 26. Lin DD, Cohen AS, Coulter DA: Zinc-induced augmentation of excitatory synaptic currents and glutamate receptor responses in hippocampal CA3 neurons. J Neurophysiol 2001;85:1185–1196. 27. Mozrzymas JW, Barberis A, Michalak K, Cherubini E: Chlorpromazine inhibits miniature GABAergic currents by reducing the binding and by increasing the unbinding rate of GABAA receptors. J Neurosci 1999;19:2474–2488. 28. Assaf SY, Chung SH: Release of endogenous Zn2+ from brain tissue during activity. Nature 1984;308:734–736. 29. Howell GA, Welch MG, Frederickson CJ: Stimulation-induced uptake and release of zinc in hippocampal slices. Nature 1984;308: 736–738. 30. Choi DW, Yokoyama M, Koh J: Zinc neurotoxicity in cortical cell culture. Neuroscience 1988;24:67–79. 31. Choi DW, Weiss JH, Koh JY, Christine CW, Kurth MC: Glutamate neurotoxicity, calcium, and zinc. Ann NY Acad Sci 1989;568:219–224.
32. Kerchner GA, Canzoniero LM, Yu SP, Ling C, Choi DW: Zn2+ current is mediated by voltage-gated Ca2+ channels and enhanced by extracellular acidity in mouse cortical neurones. J Physiol 2000; 528:39–52. 33. Sensi SL, Canzoniero LM, Yu SP, et al: Measurement of intracellular free zinc in living cortical neurons: Routes of entry. J Neurosci 1997;17: 9554–9564. 34. Becker KJ: Inflammation and acute stroke. Curr Opin Neurol 1998;11:45–49. 35. Zhang RL, Chopp M, Li Y, et al: Anti-ICAM-1 antibody reduces ischemic cell damage after transient middle cerebral artery occlusion in the rat. Neurology 1994;44:1747–1751. 36. Ellison MD, Erb DE, Kontos HA, Povlishock JT: Recovery of impaired endothelium-dependent relaxation after fluid-percussion brain injury in cats. Stroke 1989;20:911–917. 37. Wei EP, Dietrich WD, Povlishock JT, Navari RM, Kontos HA: Functional, morphological, and metabolic abnormalities of the cerebral microcirculation after concussive brain injury in cats. Circ Res 1980;46:37–47.
W9418-04.qxd 10/31/03 3:37 PM Page 117
Chapter 4 Neurologic Disease: An Overview William A. Friedman, MD, Kelly D. Foote, MD, and David Peace
Introduction Central and peripheral nervous diseases are different from most other medical problems in one fascinating and important way: The precise site of the neurologic disorder can almost always be determined from a careful history and neurologic examination. Once the site of disease is elucidated, a differential diagnosis, drawing from the basic categories of neurologic illness can be constructed. The history is particularly valuable in ordering the differential diagnosis from most to least probable. When this intellectual process is finished, appropriate diagnostic tests can be ordered. Hopefully, the precise diagnosis will then be established.
The Neurologic Examination The neurologic examination is generally divided into the following parts: mental status, cranial nerves, motor examination (including cerebellar function), reflexes, and sensory examination. As in other parts of the physical examination, adhering to a strict order helps the physician avoid errors of omission. Mental Status Examination of mental status focuses on level of consciousness, orientation, memory, emotional state, and higher cortical functions (including language). Orientation is typically tested to person, place, and time. Patients are frequently described as “oriented times three.” Typically recent memory
is tested by asking the patient to remember three objects and then testing them 2 and 5 minutes later for recall. Long-term memory may be tested by asking for their address, telephone number, or the names of presidents, capitals, and so forth. Altered states of consciousness are described by a variety of terms including, in order of severity, clouded, lethargic, obtunded, stuporous, and comatose. Clouding of consciousness refers to a mildly depressed level of awareness and slowing of mentation. A “lethargic” patient will lie quietly or sleep in the absence of stimulation but can interact fairly well when prompted. An “obtunded” patient will sleep in the absence of stimulation, can be aroused with some difficulty, and has generally depressed intellectual function. A “stuporous” patient requires vigorous stimulation to provoke any arousal and is incapable of meaningful verbal exchange. A “comatose” patient fits the following precise definition: “incapable of following commands, does not speak, and does not open eyes to pain.” Dementia, unlike the previous descriptors, refers only to a loss of intellectual function and does not imply any alteration of consciousness. Unfortunately, these terms are rather imprecise. To eliminate interobserver variability in describing decreased levels of consciousness, the Glasgow coma scale (GCS) was developed. It relies on three simple tests: best eye opening response, best verbal response, and best motor response (Fig. 4-1). The worst score is 3 and the best score is 15. The definition of coma cited previously corresponds to a GCS score of 8 or less. The differential diagnosis of coma is broad and includes toxic/metabolic disorders such as electrolyte imbalance, endocrine dysfunction, toxin ingestion, infection, nutri117
W9418-04.qxd 10/31/03 3:37 PM Page 118
118
Section I
Introduction
Eye Opening Spontaneous To Speech To Pain None
4 3 2 1
Best Verbal Response Oriented Confused speech Inappropriate Incomprehensible None
5 4 3 2 1
Best Motor Response Obeys commands Localizes pain Withdraws Abnormal flexion Extension None
6 5 4 3 2 1
Figure 4-1. The Glasgow coma scale.
tional deficiency, organ failure, and epilepsy. A variety of structural disorders can also cause coma, including hemorrhage, ischemic stroke, brain abscess, brain tumor, and trauma. A variety of disorders of higher cortical function can occur without alteration of consciousness. The most important are those which disturb language function. Disorders affecting the ability to use speech are called aphasias. Broca’s aphasia refers to a disorder affecting the posterior inferior left frontal lobe (Brodmann’s area 45). It produces difficulty with speech output, but has little effect on speech understanding. Wernicke’s aphasia involves the posterior superior left temporal lobe (Brodmann’s area 22). It causes loss of ability to understand spoken or written language but speech output remains fluent (although nonsensical). A rarer disorder, called conduction aphasia, refers to a condition in which understanding spontaneous speech is relatively preserved, with loss of the ability to repeat. This condition results from lesions in the pathways that connect Broca’s and Wernicke’s areas (Figs. 4-2, 4-3). In general, peri-Sylvian
Figure 4-2. Brodmann’s cytoarchitectonic map. Most cerebral cortex consists of six layers. The anatomy of each layer may vary depending on whether that area is more sensory, motor, or associative in function. Brodmann sliced the brain horizontally, starting from the top, and assigned numbers to each new area of cortex. These “Brodmann’s numbers” are still used to refer to different anatomic areas of the brain.
W9418-04.qxd 10/31/03 3:37 PM Page 119
Chapter 4
Neurologic Disease: An Overview
119
Figure 4-3. A functional map of the brain shows the expected deficit from focal lesions. Lesions in Wernicke’s area (Brodmann 22) tend to produce an aphasia where comprehension is lost but speech is fluent. Lesions in Broca’s area (Brodmann 45) produce aphasia with relatively preserved comprehension, but lack of speech output.
lesions are associated with loss of the ability to repeat, while deeper lesions produce the transcortical aphasias, which are associated with preserved repetition. Gerstmann’s syndrome refers to a constellation of four neurologic findings: finger agnosia, alexia, acalculia, and agraphia. It is usually found with lesions of the left inferior parietal lobe. Alexia without agraphia is seen with lesions of the dominant occipital lobe, when they extend into the splenium of the corpus callosum. A detailed discussion of the fascinating varieties of higher cortical dysfunction and their localization is beyond the scope of this chapter.
nerve will produce funduscopically visible optic pallor (optic atrophy). The most valuable localizing test of visual pathway function involves visual fields. Visual fields are tested in the clinic by having the patient occlude one eye. The examiner then directs the patient to look straight ahead and tests his ability to count fingers in the four quadrants of that eye’s visual field. Formal visual fields can be performed by an ophthalmologist if this test is abnormal. A variety of visual field abnormalities are described, unilateral visual loss, junctional scotoma, bitemporal hemianopia, homonymous hemianopia, which very accurately localize the site of the lesion within the nervous system (Fig. 4-6). CN III. The oculomotor nerve innervates the levator palpe-
Cranial Nerves Cranial Nerve (CN) I. The olfactory nerve mediates the sense
of smell (Fig. 4-4). This can be tested by asking the patient to identify vials containing common substances (coffee, orange extract, etc.) by smell alone, while alternately occluding each nostril. The olfactory nerve is the most frequently injured by head trauma. It can also be affected by neoplasms growing near the olfactory groove area intracranially, including meningiomas and esthesioneuroblastomas. CN II. The optic nerve connects the retina to the optic chiasm and, hence, to the posterior visual pathways (Fig. 4-5). Unlike other cranial nerves, it can be directly viewed via the funduscopic examination. Increased intracranial pressure will often be manifest as papilledema of the optic nerve head. Longstanding pressure or inflammation of the optic
brae, the medial rectus, the superior rectus, and inferior oblique, the inferior rectus, the pupilloconstrictor muscle, and the muscle that controls accommodation of the lens within the eye (Fig. 4-7). Lesions of CN III result in movement of the globe into a “down-and-out” position, ptosis, and pupillary dilatation. Temporal lobe herniation can produce a unilateral injury of CN III with, usually, contralateral hemiplegia. Direct compression of CN III by an aneurysm or tumor can produce these findings, as can diabetes or stroke. CN IV. The trochlear nerve innervates the superior oblique muscle (Fig. 4-8). Lesions of this nerve result in vertical diplopia, with the affected eye elevated and externally rotated. The patient will attempt to correct the condition by tilting the head away from the affected side (to internally rotate the eye). The trochlear nerve is rarely affected in
W9418-04.qxd 10/31/03 3:37 PM Page 120
120
Section I
Introduction
Figure 4-4. Anterior inferior view of the brain shows the 12 paired cranial nerves.
Figure 4-5. The optic nerves meet at the optic chiasm. The optic tracts run from the chiasm to the lateral geniculate nuclei of the thalamus. The optic radiations run from these nuclei to the primary visual cortex in the occipital lobe. The nasal retinal fibers cross in the chiasm, so a lesion there tends to produce the characteristic bitemporal visual field cut. Lesions of the tract, radiations, or occipital lobe tend to produce complete loss of vision on the opposite side (a homonymous hemianopia).
W9418-04.qxd 10/31/03 3:37 PM Page 121
Chapter 4
Neurologic Disease: An Overview
121
isolation, but frequently affected in combination with cranial nerves III and VI by lesions of the cavernous sinus area. CN V. The trigeminal nerve provides sensation to the face
(Fig. 4-9). It has three divisions: the ophthalmic division innervates the forehead and eye; the maxillary division innervates the upper jaw and side of nose; and the mandibular division innervates the lower jaw, teeth, and tongue. The trigeminal nerve also innervates the muscles of mastication. Injuries to this nerve will result in ipsilateral loss of facial sensation and atrophy and weakness of the masseter muscle. This most frequently results from tumors (acoustic schwannoma, trigeminal schwannoma, nasopharyngeal carcinoma, etc.). The trigeminal nerve is also involved in a severe, lancinating pain disorder, called trigeminal neuralgia, which is frequently treated with surgery. CN VI. The abducens nerve innervates the lateral rectus Figure 4-6. Lesions in various parts of the visual pathways lead to very characteristic and localizing visual field defects: A, Optic nerve; B, Optic chiasm; C, Optic tract; D, Temporal optic radiations; E, Parietal optic radiations; F, Primary visual cortex.
muscle, which moves the globe laterally (Fig. 4-10). Lesions of this nerve will produce inward deviation of the eye. Because of its long subarachnoid course, CN VI is commonly affected by increased intracranial pressure, which can produce unilateral or bilateral palsy (a false localizing sign).
Figure 4-7. The oculomotor nerve exits the brainstem between the mesencephalon and pons and travels to the cavernous sinus and on to the orbit. The motor nucleus innervates multiple extraocular muscles. The Edinger-Westphal nucleus provides parasympathetic innervation to the pupil and the lens of the eye.
W9418-04.qxd 10/31/03 3:37 PM Page 122
122
Section I
Introduction
Figure 4-8. Cranial nerve IV is the only one that decussates (crosses midline). It exits just below the collicular plate and travels along the tentorium to the cavernous sinus. It enters the orbit and innervates one extraocular muscle— the superior oblique.
CN VII. The facial nerve innervates the muscles of facial expression (Fig. 4-11). It is also responsible for salivary gland activity, lacrimation, and taste over the anterior two thirds of the tongue. Peripheral lesions of the nerve (such as trauma or Bell’s palsy) cause weakness of the upper and lower face. Lesions of the upper motor neuron pathways from the motor cortex innervating the facial nerve nucleus in the brainstem cause weakness of the contralateral lower face only. The upper face receives bilateral cortical innervation and remains normal. This type of facial paralysis is called a central seventh. CN VIII. The vestibulocochlear nerve innervates the cochlea
(the organ of hearing) and the vestibular complex (the organ of balance). Lesions of this nerve are seen after basilar skull fractures and with tumors of the cerebellopontine angle (usually acoustic schwannomas). Hearing is usually tested in the clinic by determining whether a patient can hear fingers rubbing together outside the external ear canal. Formal audiometry is performed if this test result is abnormal.
Lesions of CN IX rarely occur in isolation and sectioning of the nerve usually does not result in any significant deficit. CN X. The vagus nerve innervates most of the muscles responsible for swallowing and supplies sensory input to the pharynx as well (Fig. 4-13). Lesions of the nerve result in asymmetry of palatal movement. Of course, the vagus nerve also supplies parasympathetic input to the heart and gastrointestinal tract. Cranial nerves IX, X, and XI exit through the jugular foramen, where they can be jointly affected by tumors (especially jugular foramen schwannomas and meningiomas). CN XI. The spinal accessory nerve innervates the trapezius and sternocleidomastoid muscles (Fig. 4-14). These muscles are tested by shoulder shrug and head turning. Trauma is the most frequent cause of an isolated injury of CN XI. CN XII. The hypoglossal nerve innervates the muscles of the
CN IX. The glossopharyngeal nerve innervates the stylopha-
ryngeus muscle, which is involved in swallowing (Fig. 4-12). This nerve supplies taste sensation to the posterior tongue.
tongue. Dysfunction leads to protrusion of the tongue toward the affected side. This nerve may be affected by tumor, trauma, or stroke.
W9418-04.qxd 10/31/03 3:37 PM Page 123
Chapter 4
Neurologic Disease: An Overview
Figure 4-9. The trigeminal nerve exits the brainstem at the pontine level. The trigeminal (Gasserian) ganglion is just lateral to the cavernous sinus. Three branches enter the cavernous sinus and innervate the face: the ophthalmic branch (V1), the maxillary branch (V2), and the mandibular branch (V3). The trigeminal nerve also mediates motor function (the muscles of mastication and the tensor tympani).
Figure 4-10. The CN VI nucleus is in the floor of the fourth ventricle. The nerve exits the midline at the junction of the pons and medullar. It travels along the clivus and through Dorello’s canal to reach the cavernous sinus. This nerve innervates one extraocular muscle—the lateral rectus.
123
W9418-04.qxd 10/31/03 3:37 PM Page 124
Figure 4-11. Cranial nerve VII exits laterally to cranial nerve VI, in the pontomedullary groove. It innervates the muscles of facial expression through multiple branches after exiting the stylomastoid foramen. This nerve also provides parasympathetic innervation to the geniculate and sphenopalatine ganglia, and hence, to the lacrimal and salivary glands. The chorda tympani branch provides taste sensation to the anterior two thirds of the tongue.
Figure 4-12. The glossopharyngeal nerve provides motor innervation to the stylopharyngeus muscle, and parasympathetic innervation to the parotid gland and carotid body.
W9418-04.qxd 10/31/03 3:37 PM Page 125
Chapter 4
Figure 4-13. The vagus nerve provides motor innervation to the pharyngeal muscles and parasympathetic innervation to the heart and gut.
Neurologic Disease: An Overview
125
W9418-04.qxd 10/31/03 3:37 PM Page 126
126
Section I
Introduction
Figure 4-14. The spinal accessory nerve has branches from the brainstem (nucleus ambiguus) and the upper cervical spinal cord. It provides motor innervation to the trapezius and sternocleidomastoid muscles.
Motor Examination Motor function involves motor cortex of the brain, the corticospinal (and other descending) tracts, the motor pathways within the spinal cord, the motor neurons within the brainstem and spinal cord, the nerve roots, peripheral nerves, neuromuscular junction, and muscles. In addition, normal motor function is heavily dependent on complex circuitry within the basal ganglia and the cerebellum. Disease at any of these sites will lead to alterations during the motor examination and, in many cases, will clearly localize to a specific part of the nervous system (Fig. 4-15). Diseases of the motor cortex and descending motor pathways are called upper motor neuron disorders and are characterized by weakness, lack of atrophy, spasticity, and increased reflexes (Fig. 4-16). Diseases of the motor neurons in the brainstem and spinal cord, as well as the peripheral and cranial nerves, are called lower motor neuron disorders (Fig. 4-17). They are characterized by weakness, atrophy, decreased tone, and decreased reflexes. Diseases of the neuromuscular junction (such as myasthenia gravis) result in
fluctuating weakness affecting cranial and limb muscles, normal tone and reflexes, and no atrophy. Diseases affecting the muscle (like polymyositis) result in atrophy, weakness, decreased reflexes and tone. Diseases affecting the basal ganglia and its related pathways (like Parkinson’s disease) are called extrapyramidal disorders. They are frequently associated with normal strength, increased tone, unchanged reflexes, and tremor. Diseases of the cerebellar pathways can result in decreased tone and reflexes, normal strength, limb incoordination, and gait ataxia. Examination of the motor systems should include evaluation for atrophy and changes in tone, strength, and coordination. Strength is usually tabulated on a 0 to 5 scale: 0 is total paralysis; 1 is a visible flicker of movement only; 2 is movement weaker than antigravity; 3 is full movement against gravity; 4 is full movement overcome by resistance; and 5 is normal strength. Coordination is tested by having the patient touch fingertip to nose and by observing gait. Lateral cerebellar lesions tend to affect extremity coordination; vermian lesions tend to produce ataxic gait. In Parkinson’s disease and
W9418-04.qxd 10/31/03 3:37 PM Page 127
Chapter 4
Neurologic Disease: An Overview
127
Figure 4-15. Coronal section through the primary motor strip of the brain, illustrating the motor homunculus. The parts of the body that are innervated by different motor strip areas are shown.
other extrapyramidal disorders, gait tends to be stooped and shuffling. A hemiparesis will typically cause a spastic gait. Waddling gait is associated with muscular dystrophy affecting the hips. Reflexes The reflex examination centers on determination of the deep tendon reflex responses at the elbows, knees, and ankles. This reflex arc requires integrity of sensory neurons, motor neurons, and muscle. The rating is 0 to 4, with 0 signifying absence, 2+ being normal, and 4+ being very hyperactive. Hyperactive reflexes may be seen with upper motor neuron disease. Hypoactive reflexes may be seen with diseases of the peripheral nerve (e.g., polyneuritis), sensory root (e.g., tabes dorsalis), anterior horn cell (e.g., polio), proximal nerve root (e.g., lumbar disk herniation), peripheral motor nerve (e.g., trauma), and muscle (myopathy). Various superficial (cutaneous) reflexes are released in upper motor neuron disease. The most commonly tested are the Babinski (Fig. 4-18), a dorsiflexion of the toes to plantar stimulation, and the Hoffman, twitching of the distal thumb in response to flicking the distal fingers (Fig. 4-19).
Sensation Examination Pain and temperature sensation are mediated via the spinothalamic tracts, which cross in the spinal cord and ascend to the opposite side of the brainstem and cortex (Figs. 4-20, 4-21). Touch and proprioception are mediated by the dorsal columns of the spinal cord. These pathways remain ipsilateral until they reach the sensory decussation in the lower brainstem. Consequently, lesions of the hemispinal cord will result in loss of pain and temperature on the opposite side of the body and loss of touch and proprioception on the ipsilateral side of the body. The sensory examination should include tests of pain fibers (pin), light touch (fingers), proprioception (checking that the patient can tell whether toes or fingers are being moved up or down), and vibration (tuning fork). Various patterns of sensory loss can localize to specific areas of the nervous system and/or to certain well-known disease processes (Figs. 4-22, 4-23). For example, loss of all sensory modalities in the distribution of one peripheral or cranial nerve clearly localizes to that nerve. Loss of all modalities below a given spinal level localizes to the spinal cord. Loss of sensation on one side of the face and the opposite side of the Text continued on page 134
W9418-04.qxd 10/31/03 3:37 PM Page 128
Figure 4-18. The Babinski sign.
Figure 4-16. The upper motor neuron fibers descend through the internal capsule into the ventral mesencephalon to the pyramids of the medulla. There the pathways decussate and form the corticospinal tracts in the posterolateral spinal cord.
Figure 4-17. The corticospinal (and other) axons synapse with the motor neurons of the ventral spinal gray matter. The lower motor neuron axons exit through the ventral root and travel through the spinal roots and peripheral nerves to reach the neuromuscular junction. The basic spinal reflex arc, whereby stretch of the muscle spindle leads to synaptic connections with the motor neurons leading to the muscle involved in the reflex, is also shown.
W9418-04.qxd 10/31/03 3:37 PM Page 129
Chapter 4
Neurologic Disease: An Overview
Figure 4-19. The Hoffman (A) and Tromner (B) reflexes produce flexion of the distal thumb if upper motor neuron disease is present.
Figure 4-20. All sensory neurons reside in the dorsal root or brainstem ganglia. Sensory fibers enter the spinal cord through the dorsal root. Those mediating proprioception, vibration, and fine touch enter the dorsal columns and ascend ipsilaterally. Those mediating pain and temperature sensation cross the midline and ascend in the spinothalamic tracts.
129
W9418-04.qxd 10/31/03 3:37 PM Page 130
130
Section I
Introduction
Figure 4-21. The spinothalamic tracts ascend to the sensory thalamic nuclei and on to the sensory cortex. The dorsal column fibers synapse in the dorsal column nuclei of the medulla, decussate, ascend to the thalamus, and then to the cortex.
W9418-04.qxd 10/31/03 3:37 PM Page 131
Chapter 4
Neurologic Disease: An Overview
Figure 4-22. Each spinal sensory root, from C2 to coccygeal 5, innervates a specific skin area called a dermatome.
131
W9418-04.qxd 10/31/03 3:37 PM Page 132
132
Section I
Introduction
Figure 4-23. Spinal nerves combine to form a multiple of peripheral nerves. Each peripheral nerve also has a known sensory skin representation. Careful attention to the sensory examination can localize the nervous system lesion to a peripheral nerve, a spinal nerve, the spinal cord, the brainstem, or the cerebrum.
W9418-04.qxd 10/31/03 3:37 PM Page 133
Chapter 4
Figure 4-23. continued.
Neurologic Disease: An Overview
133
W9418-04.qxd 10/31/03 3:37 PM Page 134
134
Section I
Introduction
body localizes to the brainstem. Loss of sensation of one side of the body and face localizes to the opposite cerebrum. Loss of dorsal column modalities only usually indicates a metabolic disease such as vitamin B12 deficiency.
Clinical-Anatomic Correlation in Neurologic Disease In this section of the chapter we will briefly review the neurologic functions that commonly localize to the major anatomic divisions of the nervous system. Brain Frontal Lobe The frontal lobes of the brain subserve many important neurologic functions. First, the frontal association areas are responsible for personality and level of energy. Frontal lobe lesions can result in apathy, inactivity, depression, changes in personality, inappropriate actions, and so forth. Severe bilateral frontal lobe injury leads to a condition called akinetic mutism. Alternatively, some frontal lesions lead to a sense of euphoria accompanied by inappropriate jocularity (witzelsucht). The frontal eye fields (Brodmann’s area 8) direct both eyes to the opposite side of the body. Damage to this area of the brain will frequently result in head and eye turning toward the side of injury. This is seen after trauma and stroke. The posterior portion of the frontal lobes contains the primary motor cortex (Brodmann’s area 4). This strip is organized in a somatotopic pattern, with face most inferior, then hand, then arm. Hip and shoulder are near the top of the strip. Leg function is localized to the mesial hemisphere. Injury to this area will result in paralysis of an upper motor neuron variety. The inferior posterior surface of the dominant (usually left) frontal lobe is called Broca’s area (Brodmann’s area 45). This area of the brain mediates speech output. Injury to this area results in Broca’s aphasia, which is characterized by loss of spontaneous speech, inability to write, and inability to repeat, but with relatively preserved speech comprehension and reading ability. Temporal Lobe The temporal lobes, especially the mesial limbic structures (including the hippocampus) are an important part of the neurologic circuits involved in memory. Disease of the temporal lobes, especially when bilateral, can lead to profound recent memory loss. The dominant temporal lobe is involved in language function. The posterior superior temporal gyrus is called Wernicke’s area (Brodmann’s area 22). Lesions of this area cause Wernicke’s aphasia, which is characterized by inability
to understand spoken or written speech. Speech output is fluent, but nonsensical. Lesions of both temporal tips can produce a condition called Klûver-Bucy syndrome, which is characterized by oral automatisms and hypersexuality. This condition is rare and is most commonly seen with head trauma. Parietal Lobe The anterior parietal lobes contain the primary somatosensory cortex (Brodmann’s areas 3, 1, 2). Like the motor strip, this area is organized somatotopically. Lesions of the primary sensory cortex cause loss of all sensory modalities in the corresponding opposite body parts. Posterior to the primary sensory cortex, in the superior parietal lobules, are the sensory association areas. Lesions of this area produce agnosias wherein objects cannot be recognized by sensory input even though the primary sensory modalities are intact. Parietal lesions can produce neglect of the opposite hemibody, wherein the patient tends to ignore stimuli on that side. Severe neglect can lead to autotopagnosia, the inability to recognize one’s own body. Anosognosia refers to ignorance of the existence of disease and has been specifically applied to denial of hemiplegia. A more subtle test for parietal lobe dysfunction involves simultaneous stimulation of bilateral body parts. With parietal lesions, the sensation over the opposite body frequently will not be perceived. The left inferior parietal lobe integrates many sensory modalities. Lesions in this area can produce Gerstmann’s syndrome, which is characterized by finger agnosia, acalculia, left-right confusion (allochiria), and agraphia. Lesions in the angular gyrus area of the inferior parietal lobe are particularly likely to cause anomia, the inability to remember the names of objects. The right parietal lobe is implicated in higher sensory processing as well. Lesions in this area can produce dressing apraxia, wherein a patient cannot figure out how to properly put on clothing because of inability to integrate sensory information.
Occipital Lobe The occipital lobes contain primary visual cortex. Lesions of the occipital lobe tend to produce loss of vision in the opposite visual fields of both eyes (homonymous hemianopia). The macular (central) fields may be spared. More anterior occipital areas contain visual association cortex. Lesions here will produce a variety of visual agnosias. The inability to recognize faces is called prosopagnosia. Brainstem The brainstem is usually divided into three parts: mesencephalon, pons, and medulla. The brainstem connects the cerebral hemispheres and deep structures to the spinal cord, so it contains the long motor and sensory tracts that connect
W9418-04.qxd 10/31/03 3:37 PM Page 135
Chapter 4
these structures, as well as all of the cranial nerve nuclei. Because so many important structures are crowded into such small areas, even minor injuries to the brainstem produce highly localizable deficits. Large brainstem insults frequently produce coma or death. Ventral mesencephalic lesions tend to produce an ipsilateral third nerve palsy and contralateral hemiplegia. This symptom complex, called Weber’s syndrome, may be seen with stroke and tumors. Dorsal mesencephalic lesions often are associated with Parinaud’s syndrome. Parinaud’s syndrome is characterized by decreased pupillary light reflex, preserved pupillary constriction to accommodation, paralysis of upgaze, retraction-convergence nystagmus, and other anomalies. Parinaud’s is most often seen with pineal region tumors but can also be seen with severe hydrocephalus. Ventral pontine lesions produce an ipsilateral sixth nerve palsy and contralateral weakness (Millard-Gubler syndrome) that is most often seen with stroke. More dorsolateral lesions are associated with hearing loss, vertigo, facial weakness, and ataxia. Ventral medullary lesions produced an ipsilateral CN XII palsy and contralateral hemiplegia (Hughlings-Jackson syndrome). Dorsolateral lesions produce Wallenberg’s syndrome, which is usually associated with vertebral artery occlusion and characterized by dysphonia, Horner’s syndrome, dysphagia, ataxia, ipsilateral facial numbness, and contralateral body numbness. Many other brainstem syndromes have been described— a comprehensive review is beyond the scope of this chapter. Spinal Cord The spinal cord may be injured by trauma, tumor, degenerative disease, infection, and metabolic disorders. A complete spinal cord injury results in loss of all motor and sensory function below the level of the lesion. If the thoracic cord is involved, this results in paraplegia. If the cervical cord is involved, quadriplegia is seen. In the acute phase of an injury—spinal shock—all reflexes are lost. Because sympathetic outflow to the body is interrupted, hypotension associated with bradycardia is often seen. Gradually, reflexes return and, because of the upper motor neuron nature of the injury, become hyperactive. A variety of partial spinal cord injury syndromes are commonly described. Brown-Sequard syndrome refers to a hemispinal cord injury, which can be produced by trauma or by neoplasm. This is characterized by loss of ipsilateral motor function, as well as proprioception and vibratory sensation below the lesion. Contralateral pain and temperature sensation are lost. The anterior spinal artery syndrome is usually associated with trauma, but can be caused by vascular disease of the aorta. It is characterized by bilateral paralysis and loss of pain and temperature sensation, with preservation of dorsal column function (touch and proprioception).
Neurologic Disease: An Overview
135
The dorsal column syndrome refers to isolated injury of the dorsal columns, most commonly associated with metabolic disease. This results in loss of touch and proprioception below the lesion, with preservation of other function. The motor neurons of the spinal cord can be selectively injured in certain diseases, including poliomyelitis, and some forms of amyotrophic lateral sclerosis. This results in weakness and atrophy of the associated muscles, without sensory loss.
Peripheral Nerve The peripheral nerves contain motor and sensory axons. The cell bodies of origin of the sensory axons are found in the dorsal root ganglia. The cell bodies of origin of the motor axons are found in the ventral horn of the spinal cord. The sensory axons subserve the many peripheral sensory organs, which convey pain, temperature, touch, special sensory function, and so forth. The motor axons conduct impulses from the lower motor neurons to the muscles. Hundreds of disorders can affect the peripheral nerves. Peripheral nerve disease typically leads to weakness and atrophy of the involved muscles. Disease of a specific nerve will lead to sensory loss in the known distribution of that nerve. Generalized peripheral neuropathy tends to cause loss of sensation in a “stocking-glove” distribution, although there are many exceptions. The reader is referred to several excellent texts for a detailed discussion of peripheral neuropathy.1–6
Neuromuscular Junction The synapse between the motor axon and the muscle is called the neuromuscular junction. It is probably the best studied and understood synapse in the body. Diseases of the neuromuscular junction (like myasthenia gravis) tend to produce fluctuating weakness, no signs of atrophy, and involvement of cranial and peripheral muscles. These disorders respond to anticholinesterase inhibitors.
Muscle Diseases of muscle are called myopathies. Myopathy can be congenital, metabolic, inflammatory, infectious, or can present as a remote effect of carcinoma. Myopathies are characterized by weakness, most frequently affecting the proximal more than the distal musculature. Often atrophy is present and reflexes are reduced. Muscle pain may be present and muscle enzyme concentrations may be elevated.
W9418-04.qxd 10/31/03 3:37 PM Page 136
136
Section I
Introduction
Major Categories of Neurologic Disease Once the history and neurologic examination are completed, the neurologic specialist should, in most cases, be able to localize the disease to a specific nervous system area, based on the principles discussed previously. Then, one draws from the extensive lists of neurologic disorders that are most likely to cause a lesion in that location based on historical details.
Subsequent laboratory investigations, including radiographs, cerebrospinal fluid analysis, electroencephalography, electromyography, biopsy, will help to narrow the differential to a specific diagnosis for appropriate treatment. The major categories of neurologic disease, and the most common diseases in each category, are outlined in Table 4-1. A detailed discussion of all of these disease categories is the topic of many excellent textbooks of neurology.
Table 4-1 Major Categories of Neurologic Disease Disturbances of cerebrospinal fluid circulation Hydrocephalus Congenital Acquired Normal pressure hydrocephalus Pseudotumor cerebri Intracranial neoplasms Gliomas Glioblastoma Astrocytoma Oligodendroglioma Ependymoma Medulloblastoma/peripheral neuroectodermal tumor Meningioma Pituitary adenoma Schwannoma Metastatic tumor Craniopharyngioma, dermoid, epidermoid, etc. Infections of the nervous system Bacterial meningitis Subdural empyema Epidural abscess Brain abscess Tuberculous meningitis Neurosyphilis Fungal infections of the brain Acquired immunodeficiency syndrome (AIDS) and AIDS-related infections Viral syndromes Poliomyelitis Herpes zoster Viral meningitis/encephalitis Cerebrovascular diseases Transient ischemic attacks Thrombotic stroke Embolic stroke Aneurysm or arteriovenous malformation Arteritis Intracerebral hemorrhage Trauma Basilar skull fracture Concussion Cerebral contusion Epidural hematoma
Subdural hematoma Intracerebral hematoma Diseases of the spinal cord Spinal cord trauma Myelitis Infectious Inflammatory/demyelinating Spinal vascular malformations Spinal neoplasms Arnold-Chiari malformation/hydromyelia Multiple sclerosis and other demyelinating diseases Inherited metabolic disorders (Too numerous to list here) Diseases due to nutritional deficiency Wernicke’s disease and Korsakoff psychosis Beriberi Pellagra Subacute combined system disease (vitamin B12) deficiency Acquired metabolic disorders Hypoxia Hypo/hyperglycemia Hepatic failure Uremia Disturbances in sodium metabolism Alcohol intoxication or withdrawal Degenerative diseases of the nervous system Dementia Huntington’s disease Parkinson’s disease Progressive ataxias Amyotrophic lateral sclerosis Hereditary sensory neuropathies Developmental disorders of the nervous system Hydrocephalus Craniosynostosis Myelodysplasia Arnold-Chiari malformation Phakomatoses (e.g., neurofibromatosis) Cerebral palsy Idiopathic epilepsy Mental retardation
W9418-04.qxd 10/31/03 3:37 PM Page 137
Chapter 4
Neurologic Disease: An Overview
137
P earls 1. The precise site of the neurologic disorder can almost always be determined from a careful history and neurologic examination. 2. The neurologic examination is generally divided into the following parts: mental status, cranial nerves, motor examination (including cerebellar function), reflexes, and sensory examination. As in other parts of the physical examination, adhering to a strict order helps the physician avoid errors of omission. 3. In general, peri-Sylvian lesions are associated with loss of the ability to repeat, while deeper lesions produce transcortical aphasias, which are associated with preserved repetition. 4. Diseases of the motor cortex and descending motor pathways are called upper motor neuron disorders, and are characterized by weakness, lack of atrophy, spasticity, and increased reflexes. 5. Lower motor neuron disorders are characterized by weakness, atrophy, decreased tone, and decreased reflexes. 6. The right parietal lobe is implicated in higher sensory processing as well. Lesions in this area can produce
References 1. Aronson AE, et al: Clinical Examination in Neurology, 4th ed. Philadelphia, WB Saunders, 1977, pp 1–235. 2. Heilman KM, Watson RT, Greer M: Differential Diagnosis of Neurologic Signs and Symptoms. London, Appleton-Century-Crofts, 1977, pp 1– 231. 3. Newman NJ: Practical neuro-ophthalmology. In Tindall GT, Cooper PR, Barrow DL (eds): The Practice of Neurosurgery. Baltimore, Williams and Wilkins, 1996, pp 159–185.
7.
8.
9.
10.
dressing apraxia, wherein a patient cannot figure out how to properly put on clothing because of inability to integrate sensory information. Parinaud’s syndrome is characterized by decreased pupillary light reflex, preserved pupillary constriction to accommodation, paralysis of upgaze, retractionconvergence nystagmus, and other anomalies. Parinaud’s is most often seen with pineal region tumors but can also be seen with severe hydrocephalus. The anterior spinal artery syndrome is usually associated with trauma, but can be caused by vascular disease of the aorta. It is characterized by bilateral paralysis and loss of pain and temperature sensation, with preservation of dorsal column function (touch and proprioception). Diseases of the neuromuscular junction (such as myasthenia gravis) tend to produce fluctuating weakness, no signs of atrophy, involvement of cranial and peripheral muscles, and respond to anticholinesterase inhibitors. Myopathies are characterized by weakness, most frequently affecting the proximal more than the distal musculature.
4. Rengachary SS: Cranial nerve examination. In Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, pp 50–70. 5. Victor M, Ropper AH: Adams and Victor’s Principles of Neurology. New York, McGraw-Hill, 2001, pp 1–1644. 6. Youmans JR: Neurological Surgery, 4th ed., vol. I. Philadelphia, WB Saunders, 1996.
W9418-05.qxd 10/31/03 3:38 PM Page 139
Chapter 5 Central Nervous System Neoplasia William A. Friedman, MD
Epidemiology of Central Nervous System Tumors According to Rubinstein, central nervous system tumors account for less than 2% of all autopsied deaths and for approximately 9% of all primary tumors.1 However, the central nervous system (CNS) is the second most common site of primary tumor formation in children. Notably, 70% of CNS tumors in children are infratentorial and 70% of CNS tumors in adults are supratentorial. The American Cancer Society estimates that 16,800 new intracranial tumors were diagnosed in 1999, more than double the number of diagnosed cases of Hodgkin’s disease.2 The incidence of symptomatic intracranial tumors is approximately 12 per 100,000 persons per year. Additional information about brain tumor epidemiology is shown in Tables 5-1 to 5-3. Gliomas tend to be more common in males, whereas schwannomas are slightly more common in females. Intracranial meningiomas are twice as likely to develop and intraspinal meningiomas are four times more likely to develop in females. Certain diseases, called phakomatoses, present as a neurocutaneous syndrome wherein characteristic skin changes are coupled with CNS tumors. The most common is Von Recklinghausen’s disease, associated with CNS meningiomas, bilateral acoustic schwannomas, as well as gliomas. Von Hippel-Lindau disease presents with retinal lesions, as well as multiple hemangioblastomas of the brain and spinal cord. Tuberous sclerosis classically presents with the triad of adenoma sebaceum, mental retardation, and
seizures. Cortical tubers and subependymal giant cell tumors tend to develop in these patients.
Symptoms and Signs of Central Nervous System Tumors The symptoms and signs of CNS tumors can generally be divided into three groups: those due to increased intracranial pressure, those due to focal irritative effects on the brain, and those due to focal destructive effects on the brain.3 Increased Intracranial Pressure Tumors produce increased intracranial pressure (ICP) by local mass effect, surrounding edema and, sometimes, by accompanying hydrocephalus, either obstructive or communicating. Increased ICP usually causes progressively increasing headache. This type of headache is usually pancephalic and is frequently worse in the morning, after prolonged supine posture. With time, this headache is frequently accompanied by nausea and vomiting. Further increases in pressure produce a decreased level of consciousness. Patients with increased ICP often have papilledema on funduscopic examination. Occasionally, increased ICP will produce unilateral or bilateral palsy of cranial nerve VI as well. The symptoms and signs of increased ICP are generally regarded as urgent indications for therapy. Frequently, steroids will provide dramatic, although transient, relief. 139
W9418-05.qxd 10/31/03 3:38 PM Page 140
140
Section I
Introduction
Table 5-1 The Percentage Incidence of Primary Brain Tumors, All Ages
Table 5-3 The Percentage Incidence of Intracranial Gliomas in Children
Glioma Meningoma Pituitary adenoma Schwannoma Other
Astrocytomas Medulloblastomas Ependymomas
50% 20% 15% 10% 15%
Tumors presenting in this fashion will frequently require surgical debulking and/or shunting. Focal Irritative Symptoms The incidence of epilepsy associated with brain tumors is approximately 35% overall. Tumor is an important cause of a first seizure in adults (Table 5-4). Seizures may be partial, indicating no alteration of consciousness. Partial seizures can involve focal motor activity, focal sensory effects (paresthesiae), or focal visual phenomena. Complex partial seizures involve a short alteration of consciousness, with inattention to the surrounding environment. They are typically associated with unusual smells, or unusual emotions, such as fear, anxiety, or a déjà vu sensation. Complex partial seizures usually originate in the temporal lobe. Grand mal seizures involve a loss of consciousness, with tonic-clonic activity of the arms and legs. Urinary and bowel control are usually lost as well. Grand mal seizures result when focal seizures propagate throughout the brain. All types of seizures can usually be controlled well with medication. The most commonly used medications are Dilantin (phenytoin, fosphenytoin) and Tegretol (carbamazepine). Focal Destructive Symptoms Tumors can usually be distinguished from strokes, because tumor symptoms develop relatively slowly and are clearly progressive. The exceptions occur when tumors present with a hemorrhage or a seizure (sudden symptoms). These slowly progressive symptoms are dependent almost entirely on the location of the tumor and provide valuable clues as to where
Table 5-2 The Percentage Incidence of Intracranial Gliomas, Regardless of Age Glioblastomas Astrocytomas Ependymomas Medulloblastomas Oligodendrogliomas Choroid plexus papillomas Colloid cysts
55% 20.5% 6% 6% 5% 2% 2%
48% 44% 8%
the tumor might be, even before radiographic imaging is performed. Tumors with weakness generally involve the frontal motor cortex. Those with sensory changes frequently involve the parietal, sensory cortex. Speech problems usually involve the left inferior frontal area (Broca’s area) or the left temporal lobe (Wernicke’s area). Visual changes are particularly valuable in localization. Optic nerve compression affects the vision in one eye only. Optic chiasm compression typically produces a bitemporal field cut. Optic tract, radiation, or occipital lobe lesions usually cause a homonymous hemianopia (loss of the visual field on the same side in each eye). Tumors in certain locations may produce a characteristic constellation of symptoms, called a syndrome. Pineal region tumors, regardless of histologic findings, often produce Parinaud’s syndrome, which is characterized by loss of upgaze, loss of light reflex, and preservation of papillary constriction with near gaze (accommodation). Suprasellar neoplasms frequently produce a bitemporal field cut and endocrine disturbances. Intra-axial brainstem lesions often produce a hemiplegia alternans syndrome, characterized by cranial nerve findings on one side, and motor or sensory findings involving the opposite side of the body.
Radiographic Evaluation of Central Nervous System Tumors Tumors can produce radiographic changes on plain skull or spine radiographs, computed tomography (CT) scan, bone scan, ventriculography, or pneumoencephalography. Magnetic resonance imaging (MRI), however, produces superior diagnosis of almost all CNS tumors, so this section will be confined to that modality alone. The interpretation of MR images is a complex undertaking, requiring many years of training to refine. Nonetheless, a small number of characTable 5-4 Incidence of Cause of First Seizure in Adults Idiopathic Stroke Alcohol withdrawal Tumor Central nervous system infection Vascular malformation Trauma Drug toxicity
28% 24% 11% 8% 8% 6% 4% 2%
W9418-05.qxd 10/31/03 3:38 PM Page 141
Chapter 5
teristic changes are helpful in narrowing down the differential diagnosis of CNS neoplasms. These characteristics include multiplicity, enhancement, surrounding edema, dural attachment, and location. Multiplicity Metastatic brain tumors are frequently multiple. However, approximately 40% of these lesions will present with a solitary lesion. Conversely, malignant gliomas are usually solitary, but at least 10% are radiographically multicentric, even involving different lobes or hemispheres of the brain. Infectious and inflammatory lesions of the brain are also frequently multiple. Enhancement Most CNS tumors enhance with the administration of gadolinium. Benign lesions, such as meningiomas and acoustic schwannomas, tend to be homogeneously enhancing. Malignant lesions, such as glioblastoma and metastasis, tend to have less regular enhancement because of their necrotic centers (ring enhancement is common). Low-grade gliomas, especially those that are not pilocytic, rarely enhance. Surrounding Edema Malignant tumors are usually associated with surrounding edema. Metastatic tumors tend to have more associated edema than gliomas. Benign lesions usually do not have surrounding edema, although there are certainly exceptions. Infectious lesions, such as abscesses, are frequently associated with edema.
Central Nervous System Neoplasia
141
Treatment of Central Nervous System Tumors The most commonly used methods of treating CNS tumors are surgery, radiation therapy, radiosurgery, and chemotherapy. The evolving field of molecular biology promises to offer sophisticated immunotherapy and/or gene therapy alternatives in the near future. Surgery Stereotactic Biopsy Surgical procedures on the brain fall into two general categories: stereotactic biopsy and craniotomy. A typical stereotactic biopsy procedure is as follows: The patient presents in the morning to preoperative holding. After the injection of local anesthetic, a stereotactic head ring is applied (Fig. 5-1). The patient is then transported to the CT scanner. There, a series of 1-mm-thick CT scan slices are taken from the top to the bottom of the head. These images are transferred, via Internet, to the stereotactic computer system, where each slice is quickly converted to a set of pixels, each of which has a defined anteroposterior, lateral, and vertical coordinate relative to the fixed head ring. Usually, a nonstereotactic MRI, performed the day before the procedure, is then fused, with special software, to the stereotactic CT scan. The patient is then transported to the operating room. In the operating room, the stereotactic MRI is viewed on computer. The desired target points and a precise trajectory, designed to avoid blood vessels and other danger spots, are computed. The target point is set up on a device called a phantom. The stereotactic frame is set to the desired coor-
Dural Attachment Meningiomas, one of the more common primary CNS neoplasms, almost always have a clear dural attachment, as well as enhancement running away from the attached area (“a dural tail”). Occasionally, metastatic tumors or gliomas will arise from or secondarily attach to the dura. Location Certain neoplasms have very characteristic locations within the brain. Lesions commonly encountered in the suprasellar space include pituitary adenoma, craniopharyngiomas, meningioma, and germinomas. Lesions that occur in the cerebellopontine angle are usually acoustic schwannomas or meningiomas. Common pineal region tumors are germinomas, pineocytoma, and meningioma. Meningiomas tend to occur in the following locations: parasagittal, falcine, sphenoid wing, tentorial, foramen magnum, and cavernous sinus.
Figure 5-1. The stereotactic head ring is applied under local anesthesia. No shaving or prepping is required. This head ring becomes the platform upon which subsequent imaging and instrument guidance is performed. It allows the entire imaged brain to be redrawn in a Cartesian coordinate system. Each pixel on a CT scan or MR image has an anteroposterior, lateral, and vertical coordinate in relation to the fixed head ring.
W9418-05.qxd 10/31/03 3:38 PM Page 142
142
Section I
Introduction This complication may necessitate rapid conversion of the procedure to a general anesthetic and open craniotomy. Other reported complications are seizure (rare), focal deficit without hemorrhage, infection, and lack of diagnostic tissue. Craniotomy Craniotomy is the most commonly performed procedure for all types of brain tumors. Craniotomy not only provides tissue diagnosis (usually via frozen section neuropathologic examination during surgery), but it provides the means for relieving symptoms of increased ICP and, sometimes, focal irritation or destruction of brain tissue. The myriad details of the craniotomy procedure are beyond the scope of this text. In general, the following steps are used.
Figure 5-2. A variety of computer systems are used to determine the stereotactic coordinate target coordinates. A stereotactic frame is then attached to the top of the ring. A burr hole or twist drill hole is placed at the desired entry point. The stereotactic frame is then used to direct a biopsy needle directly to the target point.
dinates and connected to the phantom to verify that no errors of setup have occurred. The skin is shaved and prepped over a small scalp area where the entry point is anticipated. The stereotactic frame is attached to the head ring. At the point where the biopsy probe touches the scalp, local anesthetic is injected and an incision made. A single burr hole is then placed (Fig. 5-2). The dura is coagulated and opened. A biopsy needle is advanced through the burr hole to the target point and several biopsy specimens are taken. A neuropathologist examines the tissue. Once a pathologic diagnosis is confirmed, the needle is withdrawn and the scalp wound closed in layers. The stereotactic frame is removed and the patient returned to the recovery room and, later, the floor. The following morning the patient is discharged. Anesthetic Considerations. These procedures are almost
always “local-standby” in adults. A small amount of sedation and pain control is frequently helpful. However, if the patient’s mental status is already altered, one should be careful not to render them uncooperative with too much medication. Complications. The most feared complication of stereotactic biopsy is arterial hemorrhage. Fortunately this occurs in less than 2% of cases.4,5 Hemorrhage will usually be manifest in surgery, with arterial blood in the biopsy needle and, sometimes, with the onset of new neurologic deficit.
1. Induction of anesthesia: The type of anesthesia may be influenced by the perceived intracranial pressure or coexistent medical problems. Rarely, brain tumor excision may be performed under local anesthesia, to allow electrical stimulation of the brain and mapping of function (typically speech, memory, and motor function), during brain tumor excision. 2. Positioning: The majority of craniotomies for frontal lesions are performed with the patient in the supine position, perhaps slightly rolled to one side (Fig. 5-3). Many parietal tumors are removed with the patient in the lateral position. Most occipital and suboccipital (posterior fossa) tumors are removed with the patient in the three-quarter prone position (Fig. 5-4). The most important surgical consideration in this position is to get the shoulder out of the way of surgery. This is accomplished by rolling the shoulder away from the head, flexing the head away from the shoulder, and turning the head away from the shoulder. Rarely, neurosurgical tumor procedures are performed with the patient in the semi-sitting position. This facilitates drainage of blood and CSF out of the field, but increases the risk of air embolism. 3. Skin incision: The skin flap is designed to encompass the involved area of the brain while preserving blood supply to the flap. The incision is made sharply and clips are applied to the skin edges to control bleeding. Often, the underlying muscle is separated from the skull and elevated with the skin flap. Occasionally, a separate muscle flap (such as the Yasargil flap) will be elevated to facilitate skull exposure (Fig. 5-5). 4. Skull flap: The skull overlying the tumor is removed by placing one or more burr holes around its periphery with a high-speed drill. The burr holes are connected with a power saw and the skull is lifted away from the dura. When skull flaps cross venous sinuses in the dura (like the sagittal sinus), there is a risk of significant hemorrhage (Fig. 5-6).
W9418-05.qxd 10/31/03 3:38 PM Page 143
Chapter 5
Central Nervous System Neoplasia
Figure 5-3. Artist’s depiction of a typical supine position craniotomy. The standard positions of patient, surgeon, microscope, anesthesiologist are shown.
Figure 5-4. A, “Bird’s-eye” view of patient in “park-bench” or three-quarter prone position, as would typically be used for a posterior fossa tumor. B, Side view of three-quarter prone position.
143
W9418-05.qxd 10/31/03 3:38 PM Page 144
144
Section I
Introduction 6. Tumor excision: A variety of instruments, including microdissectors, ultrasonic dissectors, lasers, and so forth, are used to remove brain tumors. Probably, the most important technical development over the past several decades has been the routine use of the operating microscope, which allows the surgeon to identify and preserve normal neural and vascular structures. 7. Closure: Once the tumor is removed and hemostasis is obtained, the wound is closed in layers, including the dura, skull, muscle, galea, and skin. Anesthetic Considerations. Craniotomy for tumor is most
Figure 5-5. The skin is incised. Clips are applied to the skin edges to control bleeding. The temporalis muscle is elevated separately (this is a called a Yasargil flap) to provide better exposure of the anterior skull base. Burr holes are placed as indicated. A saw is used to incise the skull as indicated by the dotted line.
5. Dural opening: The dura, which covers the entire brain, is incised and opened. Usually, the dura is left attached to a pedicle and remains viable. Occasionally, the dura is involved with the tumor (i.e., meningioma) and must be removed. In this case, a dural substitute is often sewn in place at closure.
Figure 5-6. Once the skull is removed, the dura is incised and retracted. The arachnoid overlying the Sylvian fissure is incised with a knife and dissection proceeds under the operating microscope.
frequently performed under general anesthesia. If ICP is elevated, induction may need to be modified to avoid transient exacerbations of this problem. If facial nerve or motorevoked potential monitoring is used, muscular paralytic agents must be minimized. If somatosensory-evoked potentials are used, halogenated agents should be minimized. If the surgical procedure has the potential to traumatize lower cranial nerves or the brainstem, consideration should be given toward leaving the endotracheal tube in place until neurologic function (such as gag reflex) can be fully assessed. Finally, most neurosurgeons will be interested in a rapid emergence from anesthesia, so that they can assess the patient’s neurologic function and detect and correct any surgical complications (see following discussion). Complications. A variety of complications can occur during and after brain tumor surgery. The most common are listed following.
1. Intracranial hemorrhage: Any tumor case may be complicated by immediate or delayed intracerebral, subdural, or epidural hematoma. If the neurologic examination is abnormal at any time postoperatively, a noncontrast CT scan is urgently needed to rule out intracranial hemorrhage. 2. Hydrocephalus: Residual mass effect from tumor may cause obstructive hydrocephalus. Subarachnoid blood or infection may cause communicating hydrocephalus. Neurologic abnormalities should prompt an urgent CT scan, which may lead to the detection and treatment of hydrocephalus. 3. Seizures: Surgery of the posterior fossa is rarely complicated by seizure activity. However, all supratentorial surgery may be followed by postoperative seizures. Although focal motor or grand mal seizure activity is obvious, subclinical seizures may cloud consciousness and are sometimes only detected by bedside electroencephalography (EEG). Any patient with an unexplained decreased level of consciousness should be considered for EEG. 4. Meningitis: Acute bacterial meningitis may complicate any surgical procedures. Altered mental status, meningismus, and fever should prompt an early lumbar
W9418-05.qxd 10/31/03 3:38 PM Page 145
Chapter 5
puncture (if not otherwise contraindicated) and antibiotic coverage while culture results are pending. 5. Hyponatremia: The syndrome of inappropriate antidiuretic hormone secretion, as well as cerebral salt wasting, may lead to hyponatremia. When the sodium concentration is less than 130 mmol/L, alterations in consciousness and/or seizure activity may occur. See Chapter 18 for more information. Radiation Therapy Radiation therapy is of value in the treatment of many types of tumors, once tissue diagnosis is established. Radiation therapy involved the administration of high-energy photon beams to a well-defined target area including the tumor and a variable degree of surrounding normal tissue. The most commonly used device for manufacturing and focusing these radiation beams is the linear accelerator (LINAC) (Fig. 5-7). LINACs accelerate electrons very close to the speed of light. These highly energetic electrons then collide with a heavy metal. The collision produces mainly heat, but a small percentage of the energy is converted to braking radiation, whereby high-energy photons are electronically produced. These photon beams are focused by a series of shaping devices called collimators onto the tumor. To take advantage of the increased sensitivity of tumors versus normal brain to the effects of radiation, the treat-
Central Nervous System Neoplasia
145
ments are typically delivered in small daily doses (measured in centiGray), Monday through Friday, for approximately 6 weeks. There are many variations in dosing schedules, dose per daily fraction, and total dose delivered, depending on tumor type and treatment philosophy. Radiation ultimately produces its effects by damaging the deoxyribonucleic acid (DNA) strands in tumor cells. Double strand breaks eliminate the ability of the tumor cell to further reproduce. Cells that are not normally dividing (like normal neurons) are much more resistant to the effects of radiation. However, the blood vessels and glial cells may be damaged by radiation, leading to either short-term or longterm side effects. The following brain tumors are routinely irradiated in most institutions: • • • • • • •
Anaplastic astrocytoma Glioblastoma Medulloblastoma Pineal germinomas Metastatic brain tumor Pituitary adenoma (when not curable with surgery alone) Craniopharyngioma (when not curable with surgery alone) • Meningiomas (when not curable with surgery alone and/or radiosurgery) Expected Outcomes Conventional radiation therapy produces long-term tumor control rates of at least 75% for medulloblastoma, germinoma, pituitary adenoma, craniopharyngioma, and meningioma. Fewer than 50% of metastatic tumors will be cured by radiation therapy alone. Median survival after radiation therapy for anaplastic astrocytoma is approximately 18 months, and for glioblastoma less than 1 year. Complications Radiation therapy may be associated with acute complications, including nausea, vomiting, malaise, hair loss and, rarely, neurologic deficit. Radiation necrosis of the brain may develop years after treatment but is quite uncommon. The delayed onset of cognitive deficit, including severe memory loss, is commonly seen after large fields of radiation are used in the treatment of brain tumors. Radiosurgery
Figure 5-7. A linear accelerator accelerates electrons, which collide with a heavy metal. A small percentage of the energy from this collision generates highly energetic photons, called x-rays. The x-rays are focused on the patient’s tumor.
Stereotactic radiosurgery (SRS) is a minimally invasive treatment modality that delivers a large single dose of radiation to a specific intracranial target while sparing surrounding tissue. SRS treatments are administered using special devices. The gamma knife is one such device that uses 201 fixed cobalt sources, all focused on one spot within the head. LINAC radiosurgical systems use the linear accelerator as the source of radiation. Particle beam devices use cyclotrons.
W9418-05.qxd 10/31/03 3:38 PM Page 146
146
Section I
Introduction
Following is a description of the SRS protocol at the University of Florida that, with minor changes, could apply to most SRS centers.6,7 Almost all SRS procedures in adults are performed on an outpatient basis. The patient reports to the neurosurgical clinic the day before treatment for a detailed history and physical examination, as well as an in-depth review of the treatment options. If SRS is deemed appropriate, the patient is sent to the radiology department for a volumetric MRI scan. The next morning, the patient arrives at 7:00 am. A stereotactic head ring is applied under local anesthesia. No skin shaving or preparation is required. Subsequently, stereotactic CT scanning is performed. One-millimeter slices are obtained throughout the entire head. The patient is then transported to an outpatient holding area where he and his family have breakfast and relax until the treatment planning process is complete. The stereotactic CT scan and the nonstereotactic volumetric MR image are transferred via Internet to the treatment-planning computer. The CT scans are quickly processed so that each pixel has an anteroposterior, lateral, and vertical stereotactic coordinate, all related to the head
ring previously applied to the patient’s head. Using image fusion software, the nonstereotactic MRI is fused, pixel for pixel, with the stereotactic CT. Dosimetry then begins and continues until the neurosurgeon, radiation therapist, and radiation physicist are satisfied that an optimal dose plan has been developed. A variety of options are available for optimizing the dosimetry. The fundamental goal is to deliver a radiation field that is precisely conformal to the tumor shape (Fig. 5-8), while delivering a minimal dose of radiation to all surrounding neural structures. Dosimetric options include arc elimination, arc weighting, arc tilting, and the use of multiple isocenters. A detailed review of dosimetry is beyond the scope of this chapter. When dose planning is complete, the radiosurgery device is attached to the LINAC. The patient then is attached to the device and treated (Fig. 5-9). The head ring is removed and, after a short observation period, the patient is discharged. The radiosurgery device is disconnected from the LINAC, which is then ready for conventional usage. Close clinical and radiologic follow-up is arranged at appropriate intervals depending on the pathology treated and the condition of the patient.
Figure 5-8. In SRS, the goal is to create a very high radiation application that conforms to the shape of the tumor. Many hundreds of small beams of radiation are aimed at multiple target points. The computer snapshot shows the treatment isodose, one half the treatment isodose, and 20% of the treatment isodose displayed around a cavernous sinus area meningioma in the axial, sagittal, and coronal MRI views.
W9418-05.qxd 10/31/03 3:38 PM Page 147
Chapter 5
Central Nervous System Neoplasia
147
apy is complicated by the blood-brain barrier, which limits the penetration of many chemotherapeutics to the central nervous system and, hence, reduces their potential efficacy. Nonetheless, chemotherapy has a well-established role in the treatment of a variety of brain tumors, especially when they are recurrent after surgery and radiation treatment. The following sections classify the most commonly used brain tumor drugs. Alkylating Agents Most alkylating agents form positively charged carbonium ions that attack nucleophilic sites on nucleic acids, proteins, and amino acids. The result of DNA base alkylations can be cross-linking of DNA, single-strand and double-strand breaks, with subsequent misreading of the DNA code and cell death.
Figure 5-9. Depiction of procedure once the computer plan is completed. The patient’s head ring is clamped to the modified linear accelerator. Multiple arcs of radiation are delivered, always focused on the target. The end result is a very high dose of radiation delivered to the target.
SRS is commonly used to the treat the following brain tumors (surgery and conventional radiation therapy are often used as part of the treatment of these patients, as well): • • • • •
Acoustic schwannoma Meningioma Pituitary microadenoma, endocrine active Metastatic tumor Small malignant gliomas
Expected Outcomes SRS can cure as many as 90% of treated acoustic schwannomas and meningiomas. Approximately two thirds of endocrine active pituitary microadenomas will normalize within 18 months of treatment. Approximately 70% of metastatic brain tumors can be cured with SRS. It is only palliative in the treatment of malignant gliomas. Complications SRS may lead to delayed (3 to 18 months after treatment) radiation necrosis in a small percentage of cases. Neurologic symptoms will depend on the precise area of the brain involved. A course of oral steroids will often lead to resolution. In severe cases, surgical resection of the necrotic area may be required. Chemotherapy As is the case with most cancers, chemotherapy for brain tumors is palliative, not curative. Brain tumor chemother-
1. Nitrosoureas: These are generally considered to be the most active CNS drugs, especially for malignant gliomas. 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) is given intravenously and N-(2-chloroethyl)-N’cyclo-hexyl-N-nitrosourea (CCNU) is given orally. Dose-limiting adverse effects are myelosuppression and pulmonary fibrosis. 2. Procarbazine: This oral agent has been used as a single agent or in combination with others in the treatment of malignant gliomas. Limiting adverse effects are myelosuppression and gastrointestinal complications. 3. Platinum compounds: Cis-platinum is a water-soluble intravenous agent that primarily acts on the guanine bases. It is active against medulloblastoma, primitive neuroectodermal tumors (PNET), lymphoma, and germ cell tumors. Limiting toxicities are myelosuppression, nephrotoxicity, ototoxicity, and peripheral neuropathy. Plant Alkaloids 1. Vinca alkaloids: Vincristine is a water-soluble, intravenously administered alkaloid that acts as a spindle toxin. In combination with CCNU and procarbazine, it is used for the treatment of anaplastic oligodendrogliomas and medulloblastomas. Its main toxicity is peripheral neuropathy. 2. Podophyllotoxins: VM-26 (tenoposide) and VP-16 (etoposide) are lipophilic, intravenously administered drugs that arrest cells in the G2 phase by binding to topoisomerase II. VP-16 is used in combination chemotherapy of pediatric brain tumors, including medulloblastoma and germ cell tumors. 3. Methotrexate: This drug is an antifolate antimetabolite. It is active only on cells in the S phase. It is used in the treatment of lymphoma, medulloblastoma, and PNET. It is administered intravenously or intrathecally.
W9418-05.qxd 10/31/03 3:38 PM Page 148
148
Section I
Introduction
Future directions for brain tumor chemotherapy include blood-brain barrier disruption, and intratumoral chemotherapy (e.g., Gliadel wafers). More importantly, new molecular biologic methods may enable the profiling of individual brain tumors to determine which drugs would be most effective, either singly or in combination.
Common Brain Tumors Gliomas Low-Grade Gliomas Approximately 25% of all gliomas are low-grade astrocytomas or oligodendrogliomas. Low-grade gliomas tend to occur in children and younger adults. The pediatric tumors frequently have pilocytic morphology on pathologic examination and are considered truly benign in that curative surgical excision is often possible. Pilocytic tumors tend to occur in the cerebellum, optic nerve/hypothalamic region, and thalamus. Adult low-grade tumors have a propensity for the convexity of the brain, in relative proportion to the size of the various lobes. Thus frontal locations are most common. Pilocytic tumors often have homogenous enhancement on MRI, with no surrounding edema. Adult low-grade gliomas
A
typically have no enhancement on MRI and are best seen on T2-weighted or FLAIR (fluid attenuated inversion recovery) sequences (Fig. 5-10). As with all brain tumors, low-grade gliomas may present with symptoms referable to increased ICP, focal brain irritation (seizures), or focal destruction (neurologic deficit). Those presenting with mass effect and ICP increase are typically treated with craniotomy. Most, however, present with focal symptoms and the best treatment is controversial. Recent guidelines for the management of presumed lowgrade tumors call for attempted resection when locations are lobar and complete resection can be accomplished without high risk of neurologic deficit. In other tumors, stereotactic biopsy is recommended to confirm diagnosis. Conservative follow-up, with serial MRI scanning, is then pursued because there is no evidence that radiation therapy or chemotherapy alter the natural history of the disease. In practice, many presumed low-grade tumors presenting with seizure are followed without biopsy until evidence of growth occurs. Adult low-grade gliomas should not be regarded as benign tumors. If unresectable, most will eventually undergo malignant degeneration. Average life expectancy with lowgrade astrocytoma is between 5 and 10 years from the time of diagnosis. Oligodendrogliomas have a better prognosis, with longer than 10 year survival commonly seen.
B Figure 5-10. A, T1-weighted, enhanced, axial MRI shows a typical low-grade glioma. The lesion does not enhance and it has no surrounding edema. B, T2-weighted axial MR image shows hyperintensity characteristic of a low-grade glioma.
W9418-05.qxd 10/31/03 3:38 PM Page 149
Chapter 5
A
Central Nervous System Neoplasia
149
B
Figure 5-11. A, T1-weighted, enhanced axial MR image shows ring enhancement and surrounding edema characteristic of a malignant glioma. B, T2-weighted axial MR image reveals a pattern of hyperintensity characteristically caused by the vasogenic edema around a malignant glioma.
Malignant Gliomas Malignant gliomas account for about 40% of the approximately 17,000 new cases of CNS malignancy in the United States every year, and an exceptionally high fatality rate makes their clinical impact dramatic. Glioblastoma multiforme (GBM) accounts for roughly 80% of malignant gliomas and has an annual incidence of over 5000 cases. The natural history of untreated GBM results in a median survival of 3 months. With current standard therapy (resection plus conventional fractionated radiation therapy), median survival is typically 9 to 10 months with a 5-year survival rate of approximately 5%. Five-year survival rate for anaplastic astrocytoma (AA) is typically less than 20%. AAs are pathologically defined by the presence of mitoses, endothelial proliferation, and nuclear pleomorphism. Glioblastomas have the added pathologic characteristic of necrosis. The anaplastic lesions have a peak incidence in the sixth decade of life, with glioblastomas tending to occur later than the sixth decade. Because of their rapid growth, ICP symptoms are more frequent than those in low-grade gliomas. Malignant gliomas characteristically have irregularly enhancing areas and surrounding edema on MRI (Fig. 5-11). Most studies support the concept that cytoreductive surgery (craniotomy) is beneficial in terms of patient survival and quality of life.8 For those lesions that cannot be resected safely, stereotactic biopsy is performed for diagno-
sis. Radiation therapy, typically to a dose of 6000 cGy, is routinely administered because it clearly increases survival. A variety of chemotherapy protocols are used, both as upfront and salvage treatments, with generally low response rates. The observation that local control and median survival can be extended through dose escalation is the basis for the application of brachytherapy and radiosurgery to malignant gliomas.9 While achievable radiation doses with conventional external beam irradiation are limited by induced toxicity to around 70 Gy, the addition of a stereotactically focused boost of radiation allows total cumulative doses in excess of 100 Gy to be delivered to residual focal tumor. Ependymoma Intracranial ependymomas are relatively uncommon tumors; at least one half occur in the first two decades of life. Tumors of the posterior fossa predominate, although supratentorial ependymomas are also seen, especially in older patients. In the posterior fossa, the most common presenting symptoms are headache, nausea, vomiting, and imbalance, related to obstructive hydrocephalus. Small ependymomas, because of their typical origin from the floor of the fourth ventricle, near the area postrema, may present with insidiously progressive nausea in the absence of clear neurologic symptoms. Typically, these tumors are noncystic,
W9418-05.qxd 10/31/03 3:38 PM Page 150
150
Section I
Introduction of at least 3000 cGy will produce a 50% to 70% 5-year eventfree survival. Chemotherapy has a treatment role in two situations: the high risk patient, with large invasive tumors and/or metastatic disease; and the very young patient in whom chemotherapy has been used in an attempt to avoid the harmful effects of radiation therapy on the immature nervous system. Cisplatin, CCNU, vincristine, and other agents have demonstrated efficacy in these situations. Vestibular Schwannomas
Figure 5-12. Noncontrast MR image shows mass filling fourth ventricle. This has caused severe obstructive hydrocephalus.
relatively homogenously enhancing masses, within the lower fourth ventricle area on MRI (Fig. 5-12). The treatment of choice for ependymoma is surgery. Surgery has three goals: tissue diagnostic confirmation, tumor resection, and relief of obstructive hydrocephalus. The most important prognostic factor is the extent of surgical resection—gross total resection confirmed by postoperative scan correlates with a high cure rate. Resection is so important that residual disease on postoperative MRI frequently leads to “second-look” surgery. When complete resection is deemed impossible secondary to brainstem invasion, adjuvant radiation therapy and chemotherapy protocols have been followed, with low response rates.
These common tumors (representing approximately 10% of all primary brain tumors) are a benign proliferation of Schwann cells arising from the myelin sheath of the vestibular branch of cranial nerve VIII. These tumors are slightly more common in women, present at an average age of 50 years, and occur bilaterally in patients with neurofibromatosis type II. The most common presenting symptoms are hearing loss, tinnitus, and ataxia. MRI typically shows an enhancing, well-defined mass in the cerebellopontine angle, extending into the internal auditory meatus (Fig. 5-13). The mainstay of treatment for vestibular schwannoma has long been surgical resection, which has been significantly refined during the past 20 years.10,11 Results have been improved by use of the operating microscope, improved understanding of the involved microsurgical anatomy, modifications of the various approaches to access the tumor site, and intraoperative neurophysiologic monitoring of the facial and cochlear nerves. Recent reports from centers that have extensive experience with surgical management of vestibu-
Medulloblastoma Medulloblastomas account for 7% to 8% of all primary CNS tumors and nearly one third of all childhood brain tumors. They are also called infratentorial PNETs. In children, the tumors are almost always within the fourth ventricle whereas in adults a more lateral, cerebellar hemispheric location may be found. The overwhelming majority of children present with symptoms of increased ICP due to obstructive hydrocephalus. MRI usually reveals an enhancing, sharply delineated fourth ventricle mass. Much like ependymoma, the role of surgery is to relieve the hydrocephalus and to attempt a radical tumor resection. Extent of tumor resection is thought to correlate with survival. Unlike ependymomas, these are clearly malignant tumors, with a propensity to metastasize throughout the cerebrospinal fluid. Adjuvant radiation therapy, with posterior fossa doses of at least 5000 cGy, and craniospinal doses
Figure 5-13. T1-weighted enhanced axial MR image shows enhancing lesion in the right cerebellopontine angle. Note that the lesion extends into the internal auditory canal, in a manner characteristic of vestibular schwannoma.
W9418-05.qxd 10/31/03 3:38 PM Page 151
Chapter 5
Central Nervous System Neoplasia
151
lar schwannomas confirm the efficacy and relatively low morbidity of this mode of treatment.12 Many institutions have also documented their successful experience with radiation surgery for small acoustic tumors. For tumors less than 3 cm in diameter, radiosurgery offers a one-time outpatient alternative to surgery. Success rates of greater than 90% are routinely reported. Modern dosimetry methods have reduced the complication rate to less than 5% facial or trigeminal neuropathy, which compares favorably with most surgical series.
Meningiomas Meningiomas are common tumors that result from proliferation of meningothelial cells. They account for approximately 20% of primary brain tumors, affect predominantly middle-aged patients, and have a 2 : 1 predilection for females. Like vestibular schwannomas, they are generally noninvasive, pathologically benign, and tend to behave indolently, but the natural history of any particular case is unpredictable. Clinical presentation is variable, includes seizures, hemiparesis, visual field loss, aphasia, and other focal findings, and is determined in large part by the location of the tumor. MRI typically shows a homogeneously enhancing, well-defined mass, with a dural attachment (Fig. 5-14). Surgical resection is the long-standing treatment of choice for all operative candidates.13 In Black’s 1993 review of the contemporary management of meningiomas, overall surgical mortality rates ranged from 7% to 14% in modern published series.14 Common surgical complications include hemorrhage, significant blood loss, and cortical deficits. Cranial nerve deficits may occur in skull base meningiomas, and these deficits, especially in the posterior fossa, may be very disabling. Like clinical presentation, surgical outcomes are largely dependent on the location of the tumor. Convexity, parasagittal and lateral sphenoid wing meningiomas are the most accessible and resectable. Cavernous sinus, petroclival, and foramen magnum locations present considerably greater technical challenges. The rate of postoperative recurrence of meningiomas is clearly dependent on extent of resection. In a classic paper from 1957, Simpson defined five classes of resection:15 grade I, complete removal, including resection of dural attachment and any abnormal bone; grade II, complete tumor removal with coagulation of dural attachment; grade III, complete tumor removal without resection or coagulation of dural attachments; grade IV, subtotal removal; and grade V, decompression only. Simpson reported 10-year recurrence rates of 9%, 19%, 29%, and 40% for grade I through grade IV resections, respectively. This trend has been confirmed in more recent series, and the higher recurrence rates associated with meningiomas in certain locations correlate with a
Figure 5-14. T1-weighted, contrast-enhanced axial MR image shows a homogenously enhancing lesion arising from the falx—a characteristic location and MRI pattern for a meningioma.
high frequency of subtotal resection due to poor accessibility or involvement of critical structures (e.g., sagittal sinus, cranial nerves, carotid artery). Radiation for meningiomas has been somewhat controversial given its implication as an etiologic factor in some meningiomas. However, most reports have demonstrated a beneficial effect.16 Several reports have demonstrated the efficacy of radiation therapy in preventing recurrence after subtotal resection. For example, Condra and associates reported 15-year local control rates of 76% after total excision, 87% after subtotal excision plus radiation therapy, and 30% after subtotal excision alone.17 Survival rates at 15 years were 51% for subtotal excision alone versus 88% and 86% for total excision and subtotal excision plus radiation, respectively. Because local recurrence was associated with lower survival rates, prompt postoperative radiation therapy was recommended after subtotal resection, rather than waiting until regrowth. Goldsmith and colleagues16 reported similar findings after a retrospective analysis of 140 patients at the University of California at San Francisco who received postoperative external beam radiation therapy after subtotal resection from 1967 to 1990. For patients with benign lesions, the
W9418-05.qxd 10/31/03 3:38 PM Page 152
152
Section I
Introduction
overall 10-year progression-free survival rate was 77%. Results improved significantly with the advent of CT and MRI. After 1980, the 5-year progression-free survival rate for benign meningiomas thus treated improved to 98% from a pre-1980 rate of 77%. No secondary neoplasms occurred. These findings strongly suggest that postoperative radiation is beneficial if complete resection cannot be achieved. Successful outcomes from radiosurgery treatment of vestibular schwannomas and the encouraging results of conventional radiation therapy for meningiomas have led to enthusiasm about radiosurgery as a possible alternative treatment for meningiomas. A large number of reports now document the efficacy and safety of radiosurgery in the treatment of small meningiomas. Cure rates of greater than 90% are commonly reported.13,18 Radiosurgery is regarded by some as the treatment of choice for meningiomas that are high risk for surgery (e.g., cavernous sinus). Pituitary Adenomas Pituitary adenomas account for as many as 15% of all primary intracranial neoplasms. Symptoms are caused by pituitary hypersecretion, pituitary hyposecretion, or mass effects in the sellar areas.19 The most common hypersecretion syndromes are amenorrhea-galactorrhea (from prolactin hypersecretion), acromegaly (from growth hormone hypersecretion), and Cushing’s disease (from adrenocorticotropin hormone hypersecretion). Nonsecreting tumors are much more common and are often accompanied by pituitary hypofunction. Low levels of luteinizing hormone and follicle-stimulating hormone lead to amenorrhea in women and impotence in men. Additional hyposecretion symptoms include fatigue (hypocorticalism) and hypothyroidism (low thyroid-stimulating hormone secretion). An enlarging mass in the sella most commonly impinges on the optic chiasm, producing a bitemporal field cut. A sudden hemorrhage into a large pituitary tumor, called pituitary apoplexy, can produce sudden visual loss, endocrine failure (Addisonian crisis), and extraocular muscle dysfunction secondary to cavernous sinus compression. Pituitary tumors present at all ages. MRI typically shows an enhancing mass within the pituitary fossa (sella turcica) (Fig. 5-15). The treatment of choice for pituitary adenoma is complete surgical excision. This is most commonly accomplished through a transnasal, transphenoidal approach. For the majority of hypersecreting microadenomas, and many macroadenomas, surgery is curative. Complications of surgical treatment include cerebrospinal fluid leakage, visual dysfunction, hypopituitarism, and, rarely, carotid injury. If complete resection is not possible, multiple studies have demonstrated a greater than 90% 10-year tumor control rate with conventional radiation therapy.20–22 Microadenomas are also treatable with radiosurgery.
Figure 5-15. Coronal contrast-enhanced MR image shows pituitary tumor with impingement on the optic chiasm.
Figure 5-16. Contrast-enhanced axial MR image shows multiple enhancing lesions, most at gray-white junction locations, characteristic of metastatic brain tumors.
W9418-05.qxd 10/31/03 3:38 PM Page 153
Chapter 5
Cerebral Metastases Metastatic brain tumors outnumber all other brain tumors combined. Each year, 75,000 to 140,000 new cases of brain metastases are diagnosed. Half are solitary. The incidence of metastatic tumors is increasing as cancer therapy leads to longer control of systemic disease. The most common primary tumors to metastasize to the brain are lung, breast, and renal carcinoma. Although less common tumors, melanomas have a very high propensity for spreading to the brain. Metastases can present with seizures (20%), focal neurologic deficit, or symptoms of increased ICP. Although gradually progressive symptoms are most common, sudden neurologic deficit may result from hemorrhage into a metastatic tumor. This is especially common with melanoma and choriocarcinoma. Contrast-enhanced MRI typically shows multiple lesions that enhance and are surrounded by significant edema (Fig. 5-16). Metastases tend to occur at the gray-white junction areas of the brain, where the blood vessels narrow to the point that metastatic deposits lodge and grow.
Central Nervous System Neoplasia
153
The medical therapy of metastatic brain tumors includes the administration of anticonvulsant medications and the use of high-dose steroids (usually dexamethasone, 4 mg four times daily). Steroids usually produce rapid and dramatic relief of mass-effect symptoms by reducing the edema surrounding these tumors. Unfortunately the effect of steroids is short-lived (weeks at most) and steroids do have significant adverse effects when used for longer periods. Surgical treatment of a metastatic brain tumor is clearly indicated when it presents with significant mass effect.23,24 In addition, when there is no known primary cancer, surgical resection or stereotactic biopsy may be needed to establish a diagnosis. Conventional radiation therapy usually involves whole brain treatment to a dose of 3000 cGy. There is serious debate over the value of such therapy, because it is rarely curative and may cause significant neurocognitive side effects in long-term survivors. Radiosurgery has become an attractive alternative to whole-brain radiation therapy because it produces a higher local control rate and does not expose the whole brain to potential injury.25 The relative roles of surgery, radiation therapy, and radiosurgery are currently under investigation and are controversial.
P earls 1. The incidence of symptomatic intracranial tumors is approximately 12 per 100,000 persons per year. 2. The symptoms and signs of CNS tumors can generally be divided into three groups: those due to increased intracranial pressure, those due to focal irritative effects on the brain, and those due to focal destructive effects on the brain.3 3. Malignant tumors are usually associated with surrounding edema. Metastatic tumors tend to have more associated edema than gliomas.
References 1. Rubinstein LJ: Tumors of the Central Nervous System. Washington DC, Armed Forces Institute of Pathology, 1972. 2. DeAngelis LM: Brain tumors. N Engl J Med 2001;344:114–123. 3. Kaye AH, Laws ER: Brain Tumors: An Encyclopedic Approach. Edinburgh, Churchill Livingstone, 1995. 4. Apuzzo MLJ, Chandrasoma PT, Cohen D, Zee CS, Zelman V: Computed imaging stereotaxy: Experience and perspective related to 500 procedures applied to brain masses. Neurosurgery 1987;20:930–937. 5. Ostertag CB, Mennel HD, Kiessling M: Stereotactic biopsy of brain tumors. Surg Neurol 1980;14:275–283. 6. Friedman WA, Bova FJ: The University of Florida radiosurgery system. Surg Neurol 1989;32:334–342.
4. Conventional radiation therapy produces long-term tumor control rates of at least 75% for medulloblastoma, germinoma, pituitary adenoma, craniopharyngioma, and meningioma. Fewer than 50% of metastatic tumors will be cured by radiation therapy alone. 5. Radiosurgery can cure as many as 90% of treated acoustic schwannomas and meningiomas.
7. Friedman WA, Buatti JM, Bova FJ, et al: LINAC Radiosurgery—A Practical Guide. Berlin, Springer-Verlag, 1998. 8. Salcman M: Survival in glioblastoma: Historical perspective. Neurosurgery 1980;7:435–439. 9. Kondziolka D, Flickinger JC, Bissonette DJ, et al: Survival benefit of stereotactic radiosurgery for patients with malignant glial neoplasms. Neurosurgery 1997;41:776–785. 10. Ebersold MJ, Harner SG, Beatty CW, et al: Current results of the retrosigmoid approach to acoustic neurinoma. J Neurosurg 1992;76:901– 909. 11. Gormley WB, Sekhar LN, Wright DC, et al: Acoustic neuromas: Results of current surgical management. Neurosurgery 1997;41:50–60. 12. Samii M, Matthies C: Management of 1000 vestibular schwannomas (acoustic neuromas): Surgical management and results with an empha-
W9418-05.qxd 10/31/03 3:38 PM Page 154
154
13.
14. 15. 16.
17.
18.
19.
Section I
Introduction
sis on complications and how to avoid them. Neurosurgery 1997;40: 11–23. Hakim R, Alexander III E, Loeffler JS, et al: Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998;42:446–454. Black PM: Meningiomas. Neurosurgery 1993;32:643–657. Simpson D: The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957;20:22–39. Goldsmith BJ, Wara WM, Wilson CB, et al: Postoperative irradiation for subtotally resected meningiomas: A retrospective analysis of 140 patients treated from 1967 to 1990. J Neurosurg 1994;80:195–201. Condra KS, Buatti JM, Mendenhall WM, et al: Benign meningiomas: Primary treatment selection affects survival. Int J Radiat Oncol Biol Phys 1997;39:427–436. Shafron DH, Friedman WA, Buatti JM, et al: LINAC radiosurgery for benign meningiomas. Int J Radiation Oncology Biol Phys 1999; 43:321–327. Kramer S: Diagnosis and Treatment of Pituitary Tumors. Amsterdam, Exerpta Medica, 1973, pp 217–229.
20. Grigsby PW, Stokes S, Marks JE, et al: Prognostic factors and results of radiotherapy alone in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988;15:1103–1110. 21. McCollough WM, Marcus RB, Rhoton AL, et al: Long-term follow-up of radiotherapy for pituitary adenoma: The absence of late recurrence after greater than or equal to 4500 centigray. Int J Radiat Oncol Biol Phys 1991;21:607–614. 22. Sheline GE, Tyrrell B: Pituitary adenomas. New York, Raven Press, 1984, pp 1–35. 23. Noordijk EM, Vecht CJ, Haaxma-Reiche H, et al: The choice of treatment of single brain metastasis should be made based on extracranial tumor activity and age. Int J Radiat Oncol Biol Phys 1994;29:711– 717. 24. Patchell RA, Tibbs PA, Walsh JW, et al: A randomized trial of surgery in the treatment of single metastases to the brain. N Engl J Med 1990;322:494–500. 25. Flickinger JC, Kondziolka D, Lunsford LD, et al: A multi-institutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994;28:797–802.
W9418-06.qxd
11/3/03 2:50 PM Page 155
Chapter 6 Hemorrhagic Cerebrovascular Disease Pascal M. Jabbour, MD, Issam A. Awad, MD, MSc, FACS, and Daniel Huddle, DO
General Principles of Critical Care of Hemorrhagic Stroke
stroke. In this chapter we review the general principles of critical care for hemorrhagic stroke and specific interventions in the setting of various etiologies.
Epidemiology and Significance Hemorrhagic stroke, including intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH), constitutes 23% to 25% of all stroke cases.1,2 It is a catastrophic event, occurring when a cerebral vessel ruptures spontaneously and causes brain damage. Hemorrhagic stroke kills or disables most of its victims, and accounts for more than half of stroke-related deaths, disabilities, and costs.3 Hemorrhagic stroke typically refers to a spontaneous hemorrhage, as opposed to bleeding caused by trauma. The age-adjusted incidence of ICH is 10 to 15 per 100,000 per year, with a mean age of 65, and its incidence doubles with each decade of life above age 45.1,4,5 Most ICHs in younger patients are caused by a vascular malformation, coagulopathy, or drug use, while the vast majority of intracerebral bleeds affecting the elderly are caused by hypertensive or amyloid angiopathy. The incidence of SAH has been estimated at 6 to 30 per 100,000 per year, with a mean age of 50 years.6,7 Most nontraumatic SAHs are caused by a ruptured intracranial aneurysm affecting the circle of Willis or its branches. It has been estimated that up to 30% of SAH cases die from initial effects of hemorrhage, with half of the survivors dying or being disabled from rebleeding or other sequelae. The prompt recognition, acute resuscitation, and early diagnostic evaluation and therapeutic intervention have vastly improved the outcome of patients with hemorrhagic
Recognition, Transport, and Acute Resuscitation Acute Evaluation The hallmark characteristics of the clinical presentation of hemorrhagic stroke are the sudden onset of headache, loss of consciousness, and focal neurologic symptoms. A careful history from prehospital witnesses or family can identify some features of onset of the symptoms (e.g., headache preceding loss of consciousness), or other risk factors (e.g., use of anticoagulation, other drugs, untreated hypertension, or known preexisting vascular anomaly) raising suspicion about a hemorrhagic stroke. Regardless, the possibility of hemorrhagic stroke must be raised in any case of unexplained loss of consciousness or severe headache, with or without other neurologic symptoms.8 Whenever hemorrhagic stroke is suspected, and after stabilization of vital signs (airway, respiratory support, and blood pressure), urgent referral to a center where diagnostic evaluation can be carried out and subsequent treatment can be instituted must be considered. Coagulation parameters are tested and a toxic screen performed for possible drug use. During transport, patients must be kept comfortable, with good control of pain, anxiety, and agitation. The airway must be secured in all but fully conscious patients; blood pressure is closely controlled; anticonvulsant prophylaxis is administered; and reversal of coagulopathy is initiated. 155
W9418-06.qxd
11/3/03 2:50 PM Page 156
156
Section I
Introduction
Diagnosis and Etiology of Hemorrhagic Stroke A computed tomography (CT) scan of the brain is the imaging of choice when a hemorrhagic stroke is suspected, with near absolute sensitivity and specificity. It will differentiate between an ischemic stroke and hemorrhagic one, and between an SAH and an ICH (Figs. 6-1 and 6-2).8 The localization of the hemorrhage combined with the characteristics of the onset of symptoms, and the patient’s risk factors will help determine a possible etiology and the need for further diagnostic studies. The mere diagnosis of hemorrhagic stroke, its broad type (ICH versus SAH), and any associated hydrocephalus can allow the institution of early management interventions to minimize neurologic and systemic sequelae. In ICH, a search for etiology is often undertaken more electively, when the patient is otherwise stabilized, or if urgent surgical intervention is being entertained for hematoma evacuation. Administration of contrast material can often highlight suspected vascular abnormalities on CT scan (Fig. 6-3), and a CT angiogram (Fig. 6-4) can be useful when the patient is not stable enough for formal catheter angiography. For cases in which the clinical presentation is typical of SAH urgent definition of the etiology is indicated, because early rebleeding from ruptured aneurysm is common and potentially catastrophic. If the CT scan is negative or questionable in a case where there remains clinical suspicion of SAH, a lumbar puncture should be performed and CSF should be examined for xanthochromia and for red blood cell count. Lumbar puncture is not indicated, and may be harmful, if the CT scan already confirms the diagnosis of hemorrhagic stroke.
Figure 6-1. CT scan showing polycisternal subarachnoid hemorrhage. The cause is most likely aneurysmal.
Figure 6-2. CT scan showing lobar intracerebral hemorrhage. In older patients, the most likely etiology is amyloid angiopathy. In younger patients, an underlying tumor or vascular malformation is likely. Coagulopathy may cause such a bleed at any age.
Four-vessel cerebral angiography remains the gold standard for detecting aneurysms and for planning their most optimal therapy (Fig. 6-5). It should be performed as soon as possible in patients with SAH. Angiographic techniques have recently been enhanced by three-dimensional rotational angiography, allowing a dramatic spatial rendition of the vascular anatomy (Fig. 6-6). Magnetic resonance angiography (MRA) and computed tomographic angiography (CTA) may detect intracranial aneurysms but with lesser sensitivity and spacial resolution than angiography (Figs. 6-4 and 6-7). These tests may be performed urgently if a catheter angiogram is not promptly available. They are also quite useful in highlighting the anatomy of giant aneurysms, especially thrombosed portions not filling on angiography (Fig. 6-8). Magnetic resonance imaging (MRI) is more sensitive than CT scan for detecting structural abnormalities such as tumors and vascular malformations9 and is used when these etiologies are suspected, as in ICH in younger patients (Fig. 6-9). Catheter angiography remains the most sensitive study for detecting arteriovenous malformations and should be performed in younger patients with ICH to exclude this
W9418-06.qxd
11/3/03 2:50 PM Page 157
Chapter 6
Hemorrhagic Cerebrovascular Disease
157
Figure 6-5. Angiogram revealing basilar summit region aneurysm.
Management Strategy
Figure 6-3. Contrast-enhanced CT scan showing middle cerebral artery aneurysm.
etiology (Fig. 6-10). It also should be performed, urgently, as in cases of SAH, whenever an ICH communicates with the circle of Willis or its branches (Fig. 6-11), or if contrast CT or MRI/MRA suggest a possible cerebral aneurysm (see Figs. 6-3 and 6-7).
Figure 6-4. CT angiogram performed by computer reconstruction of thin-cut high-resolution CT scan with contrast material reveals an aneurysmal dilation at the middle cerebral artery.
ABCs of Multisystem Support The acute resuscitation of hemorrhagic stroke follows the general guidelines of the “ABC” rules of airway, blood pressure, and cerebral perfusion. The airway should be cleared, and any patient with a Glascow coma scale (GCS) score of 8 or less, or unable to protect the airway, should be intubated. A good oxygen saturation is not sufficient and does not reflect the arterial partial pressure of carbon dioxide (PaCO2), so even if the saturation is normal, one should make sure that the patient is not hypercapnic because this may worsen
Figure 6-6. Three-dimensional (3D) rotational angiogram in the same case depicted in Figure 6-5, revealing much more spacial resolution than in the conventional angiogram images. The information in 3D angiography can help guide therapeutic decisions regarding endovascular versus surgical intervention.
W9418-06.qxd
11/3/03 2:50 PM Page 158
158
Section I
Introduction
Figure 6-7. MRA revealing a carotid summit berry aneurysm. This is an excellent modality for screening patients for aneurysms. A negative MRA is insufficient to exclude aneurysm in a patient with subarachnoid hemorrhage.
Figure 6-9. Magnetic resonance image of intracranial arteriovenous malformation showing nidus of abnormal vessels as flow voids in the brain.
intracranial hypertension. A central line and an arterial line should be inserted. Blood pressure should be controlled aggressively (Table 6-1); hypertension and hypotension should be avoided.
A
Intracranial Pressure and Cerebral Perfusion The cerebral perfusion pressure (CPP) is the pressure gradient responsible for cerebral blood flow and its compromise results in cerebral ischemia. The CPP is defined as mean arterial pressure minus intracranial pressure (MAP - ICP). Monitoring of ICP can be used to guide CPP management1,10 whenever intracranial hypertension is suspected or is potentially compromising cerebral perfusion. Elevated ICP is defined as intracranial pressure exceeding 20 mm Hg for longer than 5 minutes. The goal of treatment for elevated ICP is less than 20 mm Hg and CPP greater than 60 to
B Figure 6-8. Magnetic resonance imaging of partially thrombosed giant aneurysm. A, T1-weighted imaging. B, T2weighted imaging.
Figure 6-10. Angiogram revealing arteriovenous malformation nidus.
W9418-06.qxd
11/3/03 2:50 PM Page 159
Chapter 6
Hemorrhagic Cerebrovascular Disease
159
Table 6-1 Blood Pressure Management in Hemorrhagic Stroke
Figure 6-11. CT scan in a comatose patient with massive subarachnoid hemorrhage and intracerebral hemorrhage (plus subdural hemorrhage). Such a patient is taken emergently for surgical evacuation of hematoma with or without contrastenhanced CT scan, but without taking additional time for conventional angiography.
70 mm Hg.11 An ICP monitor should be placed in all patients with a GCS score of 8 or less or who could be suffering elevated ICP and cannot be followed by a neurologic examination. Intraparenchymal fiberoptic ICP monitors and intraventricular monitors are commonly used, with the former being more accurate and less vulnerable to obstruction, while the latter allows simultaneous drainage of cerebrospinal fluid to treat elevated ICP. Intracranial hypertension can be treated by draining cerebrospinal fluid, by decreasing brain tissue bulk or cerebral blood volume, or by sedation and decreasing brain metabolic demands. External ventricular drainage allows diversion of cerebrospinal fluid in cases of hydrocephalus (Fig. 6-12), or whenever ICP exceeds a certain level. This may be performed continuously (by titrating the level of the drip chamber) or intermittently depending on intracranial pressure. External ventricular drainage is ineffective if the ventricles are “slit-like” from brain edema, overdrained, or if the catheter is obstructed by clotted blood. Brain bulk may be treated to lower ICP with osmotherapy using mannitol (0.25 to 0.5 g/kg every 4 hours) and furosemide (10 mg every 2 to 8 hours), and these are administered alternately or simultaneously as needed for ICP waves. Serum osmolarity and sodium concentrations should
From Broderick JP, Adams HP, Barsan W, et al: Guidelines for the management of spontaneous intracerebral hemorrhage. A statement of healthcare professionals from a special writing group of the stroke council, American Heart Association. Stroke 1999;30:905, with permission.
be measured at least twice a day to target an osmolarity less than 310 mOsm/L,8 and fluid administration should aim to maintain euvolemia and normonatremia. Osmotherapy
W9418-06.qxd
11/3/03 2:50 PM Page 160
160
Section I
Introduction
cannot be used to treat ICP if extremes of hypovolemia and hypernatremia are allowed to develop. Hypocarbia (PaCO2 25 to 35 mm Hg) decreases the ICP by causing a cerebral vasoconstriction, and this can be very effective in acute crises with waves of elevated ICP. Extreme hyperventilation (PaCO2 < 20 mm Hg) can exacerbate brain ischemia by decreasing cerebral blood flow. Hyperventilation should also not be used for a long time because it can become ineffective with metabolic adjustment to respiratory alkalosis. Further response to life-threatening ICP waves becomes ineffective after chronic hyperventilation, and the patient becomes vulnerable to rebound increased ICP when restoring normocapnea.8 Sedation (propofol, benzodiazepine, or morphine) with neuromuscular paralysis can reduce elevated ICP, and it does so by preventing agitation and straining, and by decreasing brain metabolic demands. If the ICP is still high despite maximizing all the previous medical treatments, induced barbiturate coma may be instituted with continuous electroencephalographic (EEG) monitoring. Central and arterial lines are used, and even a pulmonary artery catheter and pressors, if needed, to maintain hemodynamic support during barbiturate-induced coma.8 Barbiturates can decrease ICP in proportion to the level of sedation, down to EEG burst suppression. Further administration of barbiturates beyond effective EEG burst suppression offers no additional benefits of ICP control, while increasing toxic complications. Decompressive craniotomy allows control of ICP by opening the cranial vault and dura, allowing therapeutic herniation of cerebral tissue. It is used in cases of diffuse hemispheric cerebral edema from ischemic or hemorrhagic stroke, and is discussed elsewhere in this book (see Chap. 25).
Anticipation and Treatment of Delayed Sequelae In patients with SAH, the major cause of early death during the first 24 hours is rebleeding from an unsecured aneurysm.6,12 Controlling blood pressure helps to prevent rebleeding after aneurysmal rupture and may minimize the risk of hematoma expansion in ICH.8 The coagulation parameters should also be checked and corrected if needed to avoid rebleeding. In the case of SAH and intraventricular hemorrhage, or any hemorrhage compromising the CSF circulation pathways, acute hydrocephalus should be suspected and treated with a ventriculostomy, performed as a quick bedside procedure (see Fig. 6-12). Subsequent lumbar punctures or permanent ventriculoperitoneal shunting may be performed for subacute and chronic hydrocephalus. The breakdown products of subarachnoid blood are probably responsible for spasm in cerebral arteries following SAH.6,13 Vasospasm begins 3 to 5 days after the SAH and may last for 2 to 3 weeks, during which time the brain is vulner-
able to further ischemic insults. Nimodipine at 60 mg orally every 4 hours for 21 days should be started on the day of admission as it has been shown to decrease neurologic sequelae from vasospasm and/or reperfusion injuries.6,14,15 Noninvasive monitoring for vasospasm should be instituted during this vulnerable period, with other treatments instituted (hypertensive, hyperdynamic, hemodilution [HHH] therapy and endovascular interventions) if there is evidence of progression or neurologic symptoms. All patients with hemorrhagic stroke should be loaded with anticonvulsant to prevent early seizures,1 which can increase rebleeding and elevate ICP. Patients who experience acute seizures and those with cortical parenchymal hemorrhages may benefit from longer-term anticonvulsant prophylaxis. The breakdown of the blood-brain barrier in ICH will cause edema in the brain around the hematoma. This edema can be managed by fluid intake restriction, diuretics, and osmotherapy.
Systemic Complications of Hemorrhagic Stroke Hemorrhagic stroke has been shown to induce subendocardial ischemia, proportional to the severity of neurologic insult. Electrocardiographic changes, elevation of myocardiac enzymes, ventricular wall motion abnormalities, cardiogenic pulmonary edema, and life-threatening arrhythmias can occur in patients with hemorrhagic stroke especially in the acute phase.16,17 These do not typically alter the course of the illness, and should not generally prevent timely interventions for diagnosis or therapy. Cardiac complications can be life threatening in the setting of preexisting cardiac disease, or after multisystem complications of illness or therapies (i.e., myocardiac depression by barbiturates). These complications should be monitored and treated prophylactically. Pulmonary complications may develop for a variety of reasons. The patients with poor neurologic grade are at increased risk of aspiration, atelectasis, pneumonia, and pulmonary embolism.18 Neurogenic pulmonary edema is a complication that occurs after significant neurologic insult, and consists of leakage of protein-rich fluid into the pulmonary alveoli. It is believed to be due to the disruption of the endothelial barrier in response to massive sympathetic discharge.19 Volume overload during HHH therapy for vasospasm can cause or exacerbate pulmonary edema; for this reason pulmonary artery catheter-optimized hemodynamics are often used with this therapeutic modality. Hyponatremia is common in patients with hemorrhagic stroke.7 It may result from two mechanisms, the syndrome of inappropriate antidiuretic hormone (SIADH) with free water retention, or inappropriate natriuresis (also known as cerebral salt wasting) mediated by the atrial natriuretic
W9418-06.qxd
11/3/03 2:50 PM Page 161
Chapter 6
Hemorrhagic Cerebrovascular Disease
161
A B Figure 6-12. Patient with subarachnoid hemorrhage and severe hydrocephalus. Before (A) and after (B) placement of ventriculostomy catheter.
factor (ANF) and brain natriuretic peptide (BNP).20,21 Determining the likely cause is important because the two syndromes are managed differently. Hyponatremia is treated with fluid restriction in the setting of SIADH, as is common after ICH. Hyponatremia is treated with hypertonic fluid and salt replacement in natriuresis syndrome, common after SAH. The fluid balance (intake versus output), urine sodium concentrations, urine and sodium osmolality, and the clinical setting assist in differentiating the two syndromes. Patients who suffer SAH have been shown to be predominantly fluid contracted from inappropriate natriuresis. Especially during periods of vulnerability to vasospasm, these patients can suffer irreversible brain infarctions if they are further fluid restricted. Volume and salt maintenance are the mainstay of therapy during vasospasm after SAH. Hypernatremia may be caused by diabetes insipidus (DI) and should be treated with free water replacement and vasopressin.6 Gastroduodenal erosions may develop in patients with sustained cerebral injuries (Cushing ulcers). Prophylaxis against this complication is instituted using gastric ulcer
pharmacotherapy, stomach pH management, and early enteric feeding. Patients’ hematocrit and stool guaiac status should be determined daily, and any decrease in the hematocrit should be investigated. Gastrointestinal motility is often compromised in acute neurologic illness, or as a result of pharmacologic sedation, especially with barbiturates and opiates. Monitoring of gastric emptying should be ongoing via nasogastric tube in unconscious patients, and early enteric feeding encouraged whenever possible. Early percutaneous gastrojejunostomy may prevent aspiration complications in patients with decreased level of consciousness or impaired swallowing, and allow removal of nasal tubes with enhanced comfort, ease of care and rehabilitation, and prevention of nasal septomucosal erosion and sinusitis. Patients with diarrhea should be investigated for treatable causes, including alteration of feeds and possible intoxication from Clostridium difficile. Fever can be caused by noninfectious sources such as medications, deep venous thrombosis, or neurogenic causes. Nevertheless close monitoring of patients for any source of infection is important, keeping in mind the vascular catheter
W9418-06.qxd
11/3/03 2:50 PM Page 162
162
Section I
Introduction
insertion sites, ventriculostomies, pneumonitis, and urinary tract infections. Early tracheotomy in patients with impaired level of consciousness can minimize pulmonary complications and allow earlier mobilization and rehabilitation. Pneumonia should be treated aggressively with antibiotics, mobilization of secretions, and bronchoscopy, if necessary. Timing and Management of Underlying Cause The timing and management of the underlying cause of hemorrhagic stroke depend on the patient’s neurologic status, the type of hemorrhagic stroke, and the underlying etiology. This subject will be addressed in detail later in this chapter. Any accessible hematoma with serious mass effect and impending herniation should be evacuated emergently regardless of the cause (see Fig. 6-11). In SAH, securing the aneurysm should be done early to prevent rebleeding. In arteriovenous malformations (AVMs), treating the etiology is generally deferred until the patient is stable and until the brain edema decreases because risk of early rebleeding from most AVMs is not high. Psychosocial Support, Mobilization, and Rehabilitation Hemorrhagic stroke is a catastrophic illness, with lifealtering consequences to patients and families. Aggressive care and invasive interventions should be tempered by the patient’s advanced directives, his or her expressed wishes communicated through the family, and evidence-based assessment of realistic prognosis. Survival in severely disabled or vegetative states has been described by families as “worse than death.” This is particularly true if there is evidence of catastrophic brain damage or if the patient’s quality of life or life expectancy is already significantly compromised by advanced age or previous illness. Otherwise, outcome predictions remain less than certain, and patients with apparently dire illnesses can still achieve dramatic recoveries. This is especially true in younger patients.3 An ongoing evaluation of response to treatments, reassessment of prognosis, and open lines of communication with family are essential throughout the illness, with frequent revision of therapeutic stance. Counseling and spiritual support services can be very helpful during every phase of the illness. Decisions on withdrawal of care and determination of brain death are handled with the utmost sensitivity and a sense of patient advocacy (see Chap. 30). Organ donation options are best presented by a separate team uninvolved in the patient’s acute critical care, so there is never the remotest perception of conflict of interest or compromise of advocacy. On recovery from the acute and most critical and dangerous phases of the illness, there must be ongoing education about expectations during recovery and rehabilitation, and about expected quality of life and the risks and benefits of various interventions, including treatment of residual lesions which might cause recurrent hemorrhagic stroke. All
management considerations are guided by considerations of quality of life3 integrated over the patient’s remaining life expectancy.22 Options of rehabilitation are reviewed in detail, considering the availability of ongoing neurologic and medical care, and also the proximity of services to patients’ home and family (see Chap. 29). Special rehabilitation needs are considered, such as higher executive functions in relation to prior occupation. Advocacy is maintained in negotiations with insurance carriers and managed care plans. Early evaluation for rehabilitation and the coordination and overlap between acute care interventions and rehabilitation services are encouraged, to hasten recovery. Communication is maintained with the patient’s primary care providers and local physicians to ensure optimization of long-term prophylaxis, therapies, and follow-up.
Subarachnoid Hemorrhage Etiology Intracranial Aneurysm The prevalence of cerebral aneurysms in the population is estimated to range between 0.2% and 7.9%, with greater prevalence in older patients.23 This etiology is considered to be responsible for 70% to 80% of spontaneous SAHs. Aneurysms are known to develop at vessel bifurcations, points of maximum hemodynamic stress. The ones associated with infection or trauma tend to occur more distally in the circulation. Eighty percent to 90% of aneurysms affect the anterior (or carotid) circulation, at the anterior communicating artery, posterior communicating artery, middle cerebral artery, and other locations. Ten percent to 20% of aneurysms affect the posterior (or vertebrobasilar) circulation, most likely at the basilar summit and at the posterior inferior cerebellar arteries, and other locations. Aneurysms can be classified by shape, with the great majority of aneurysms saccular or berry shaped, involving an eccentric pathology of the arterial wall, usually at a branching point. A small fraction of aneurysms are fusiform, with or without saccular protrusions, reflecting more diffuse vessel wall pathology, including arteriopathy, dissection, or infection. Saccular aneurysms are classified by size; small, if less than 10 mm in diameter (78%); large, from 10 to 24 mm in diameter (20%); and giant, if more than 24 mm in diameter (2%). The pathogenesis of saccular aneurysms is not fully understood, although their risk factors appear to be both congenital and acquired. Some systemic conditions are associated with the presence of cerebral aneurysms. These include connective tissue disorders (including EhlersDanlos syndrome, Marfan syndrome), autosomal dominant polycystic kidney disease, fibromuscular dysplasia, and atherosclerosis, but these account for only a small fraction of all aneurysms. Approximately 20% of patients with
W9418-06.qxd
11/3/03 2:50 PM Page 163
Chapter 6
Hemorrhagic Cerebrovascular Disease
163
aneurysms have a family history of aneurysm affecting a first-degree blood relative.24 Hypertension and smoking appear to contribute to the risk of aneurysm formation, and also the risk of hemorrhage.6 Risk of hemorrhage increases with larger aneurysm size.7 The annual risk of hemorrhage for unruptured aneurysms varies between 0.1% and 5% to 10% per year, with highest risks in giant aneurysms. Higher bleed risk also occurs in patients who have bled from another aneurysm, and in aneurysms at certain locations (basilar summit and anterior communicating arteries).25,26 Twenty percent of patients harbor multiple aneurysms.
Vascular Malformations Vascular malformations are a heterogeneous group of lesions categorized as proposed by McCormick in 1966,27 including the AVMs, cerebral cavernous malformations (CCMs), capillary malformations or telangiectasia, and venous malformations or angiomas (also known as venous developmental anomalies). Less than 10% of patients with SAH are found to have an AVM as etiology for the bleeding,28 and this usually occurs in association with a parenchymal ICH or intraventricular hemorrhage. The prevalence of AVMs is 14 per 10,000 population.29 Patients diagnosed with AVMs have an average age 10 years younger than those diagnosed with aneurysms.30 The average risk of hemorrhage from an AVM varies between 2% and 4% per year.31 Seven percent to 10% of the AVMs have an aneurysm enclosed in the malformation, located on the feeding arteries or in the nidus, and when present, this is more often the source of hemorrhage than the nidus proper.32 The CCMs are uncommonly the cause of SAH, when adjacent to a pial or ependymal surface, and these are often associated with an adjacent venous angioma. The venous angioma is rarely the sole source of hemorrhage. Another vascular anomaly that can cause SAH or ICH is the dural arteriovenous malformation or fistula (Fig. 6-13), an indirect communication between a meningeal artery and a vein.33 Dural fistulae are usually acquired and frequently associated with thrombosis of a major venous sinus. They are composed of a network of dilated dural meningeal arterial capillaries, usually within the wall of a dural sinus, that shunt in a retrograde fashion into a leptomeningeal vein, or antegrade into the dural sinus itself. They are classified according to the lesion location and pattern of venous drainage.33,34 The cardinal feature affecting risk of aggressive clinical behavior of dural arteriovenous malformations, including hemorrhage, is the presence of leptomeningeal venous drainage. Spinal vascular malformations35 can rarely cause subarachnoid hemorrhage, and they should be sought and excluded in cases with associated myelopathy, or if the symptoms of SAH include severe neck or back pain.
Figure 6-13. Angiogram of external carotid injection showing a dural arteriovenous fistula with cortical venous drainage. Such a lesion commonly presents with intracerebral or subarachnoid hemorrhage.
Arterial Dissection Arterial dissection is the extravasation of blood from the true lumen of a vessel through the arterial wall and it can affect extracranial or intracranial vessels. Yamaura36 suggested three types: (1) dissection between the intima and the media with luminal compromise; (2) dissection between the media and the adventitia with aneurysm formation (Fig. 6-14); and (3) artery rupture and encapsulation of the hematoma causing a pseudoaneurysm (Fig. 6-15). When the dissected artery involves an intracranial segment within the subarachnoid space, there is a risk of subarachnoid hemorrhage due to the second or third mechanisms described by Yamaura. Arterial dissection can be spontaneous, associated with collagen vascular disease or fibromuscular dysplasia, or other arteriopathies. It can be post-traumatic following blunt or penetrating injury, or iatrogenic due to angiography catheters. Arterial dissection occurs in young adults.37 A large series of 260 cases36 found that the vertebral artery was the most common intracranial site, and the most common source of SAH from dissection. Arterial dissections account for less than 5% of SAH. Idiopathic Subarachnoid Hemorrhage and Benign Perimesencephalic Bleeds Other SAHs are idiopathic. The incidence of angiogramnegative SAH is estimated to be between 10% and 15%.38,39 A small fraction of these include an occult aneurysm, with
W9418-06.qxd
11/3/03 2:50 PM Page 164
164
Section I
Introduction
Figure 6-16. Perimesencephalic SAH. Cerebral angiogram did not reveal an aneurysm. A perimesencephalic SAH may also be attributed to basilar aneurysm or other causes. Figure 6-14. Angiogram of patient with severe subarachnoid hemorrhage, revealing dissection of the vertebral artery, with double lumen and small aneurysmal dilatation.
incomplete or misread angiogram, poor filling of aneurysm due to flow patterns or thrombosis, or parent vessel vasospasm. A second or third angiogram may reveal occult aneurysm. The pretruncal nonaneurysmal SAH is a benign entity.40,41 It includes primary SAH in perimesencephalic
Figure 6-15. Angiogram showing traumatic pseudoaneurysm arising from disruption of the internal carotid artery by basilar skull fracture.
cisterns without involvement of other cisterns (Fig. 6-16). Schwartz and Solomon presented follow-up data on 169 patients with perianeurysmal SAH and negative angiogram for 8 to 51 months, and did not document any instance of rebleeding. The cases also were found to have lower risk of vasospasm and better overall outcome than aneurysmal SAH.42 It is thought that the cause of hemorrhage in such cases may be a small perimesencephalic vein.43 It must be emphasized that perimesencephalic SAH may also be caused by ruptured basilar or superior cerebellar or posterior cerebral artery aneurysms, so it remains a diagnosis of exclusion after a negative, good quality cerebral angiogram. An SAH involving other cisterns reflects a higher risk of missed or occult aneurysm or dissection. This can be revealed on repeat angiograms or with MRI or CTA scans. Some experts have recommended surgical exploration of SAH with negative angiograms, involving the interhemispheric or Sylvian cisterns, likely revealing occult anterior communicating or middle cerebral artery aneurysms, respectively.44 Other Causes Other causes of SAH include unrecognized trauma, coagulation disorder, vasculitis, tumor,45 pituitary apoplexy,46 cocaine use,47 sickle cell disease, and infection (endarteritis or septic embolism). Hemorrhagic stroke including SAH has been reported at higher incidence in women using certain vasoactive drugs including decongestant cold remedies and diet pills containing phenylpropranolamine.48
W9418-06.qxd
11/3/03 2:50 PM Page 165
Chapter 6
Clinical Presentation The typical presentation of SAH is a sudden onset of severe headache, “thunderclap,” or the worst headache the patient has ever experienced.1 The presentation could also be a seizure, loss of consciousness, or altered level of consciousness. Some circumstances may precipitate the rupture of an aneurysm like physical efforts, postural modification, or intense emotions. Other associated symptoms in the wake patient may include photophobia, nausea, double vision, or neck stiffness. Clinical signs may include hypertension, meningismus, cranial neuropathy, altered level of consciousness, focal neurologic deficit, and ocular hemorrhage. Subarachnoid hemorrhage is frequently misdiagnosed, with dire clinical consequences and a poorer outcome among misdiagnosed cases.49 Hence a concerted level of clinical suspicion should always accompany any clinical presentation suggestive of possible SAH. Among chronic headache sufferers, the headache of SAH is typically different and more severe than other previous headaches, and may be associated with a sense of impending doom.50 The patient’s level of consciousness is a cardinal determinant of outcome after subarachnoid hemorrhage, and it can affect treatment decisions as well as prognostication. It can be assessed using the Hunt and Hess grade (Table 6-2), or the World Federation of Neurological Surgeons scoring system (Table 6-3). The former is quite simple and widely used, while the latter grade has been shown to have better positive and negative predictive power in relation to outcome, especially among high-grade patients.6,51–54 Cases with cranial neuropathy may represent elevated ICP (abducens or oculomotor palsies), or aneurysmal compression on the cranial nerve (posterior communicating artery or superior cerebellar artery aneurysm compression on the oculomotor nerve). Other focal neurologic deficits likely imply an ICH in addition to the SAH, as is common with middle cerebral artery aneurysms, associated ischemic event, or mass effect from giant aneurysm. Vascular malformations and fistulae are more likely to cause ICH than SAH, but could result in both. Aneurysmal bleeds from anterior
Hemorrhagic Cerebrovascular Disease
165
Table 6-3 World Federation of Neurologic Surgeons (WFNS) Subarachnoid Hemorrhage Grade WFNS Grade
Glasgow Coma Scale Score
Major Focal Deficit
0 (intact aneurysm) 1 2 3 4 5
— 15 13–14 13–14 7–12 3–6
— Absent Absent Present Present or absent Present or absent
From Drake CG: Report of World Federation of Neurological Surgeons Committee on a Universal subarachnoid hemorrhage grading scale. J Neurosurg 1988;68:985, with permission.
communicating artery and basilar summit, or from posterior inferior cerebellar artery aneurysms may cause intraventricular hemorrhage (Fig. 6-17), and this in turn may cause ventricular obstruction and account for decreased level of consciousness.
Table 6-2 Hunt and Hess Classification Scale Grade I II III IV V
Neurologic Status Asymptomatic; or minimal headache and slight nuchal rigidity Moderate to severe headache; nuchal rigidity; no neurologic deficit except cranial nerve palsy Drowsy; minimal neurologic deficit Stuporous; moderate-to-severe hemiparesis; possibly early decerebrate rigidity and vegetative disturbances Deep coma; decerebrate rigidity; moribund appearance
From Hunt WE, Hess RM: Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 1968;28:14, with permission.
Figure 6-17. Severe intraventricular hemorrhage from rupture of anterior communicating artery aneurysm. Note bilateral intraventricular catheters used to drain the ventricles pending endovascular treatment of the aneurysm. Subsequent clotting of the catheters required intraventricular tissue plasminogen activator administration for clot lysis and restoration of ventricular drainage (this treatment is contraindicated if aneurysm is not secured).
W9418-06.qxd
11/3/03 2:50 PM Page 166
166
Section I
Introduction
Diagnostic Evaluation The diagnosis of an SAH must be executed rapidly because of dire consequences of rebleed or misdiagnosis. It is usually done with a simple nonenhanced brain CT scan, which can detect the SAH in as many as 95% of the cases (see Fig. 6-1).55 The amount of SAH is evaluated by the Fischer grading system, which has a prognostic value in predicting the risk of vasospasm and overall patient outcome (Table 6-4).56 In the cases where the presentation is typical and the CT scan is negative or questionable, a lumbar puncture should be done and CSF should be examined for xanthochromia and for red blood cell count. The latter should not drop between the first and the last tube because this would more likely reflect a traumatic tap rather than SAH. Once the diagnosis of SAH has been made, a four-vessel cerebral angiogram will typically reveal the etiology of the bleed. In the cases where the angiogram fails to show the etiology, MRI should be performed to rule out any angiographically occult lesion. The angiogram should be repeated 1 to 2 weeks after the first study if the source of the SAH is still indeterminate. In arterial dissection the cerebral angiogram may reveal one of the following findings: luminal stenosis, complete occlusion, double lumen sign (see Fig. 6-14), fusiform dilation (see Fig. 6-14), frank extravasation of dye, or pseudoaneurysm (see Fig. 6-15). An arterial dissection may be associated with normal luminal filling on angiography, and is not excluded by negative results on angiography. The MRI image with axial T1 sequences and MRA source images are more sensitive than catheter angiography (and the reconstructed MRA) for the diagnosis of an arterial dissection. They reveal a crescent sign, which is the hematoma in the vessel wall, as a bright signal surrounding the signal void of the carotid or vertebrobasilar arteries on axial T1-weighted or source images. If angiography does not reveal the source of SAH, a systematic search of other causes is undertaken, including an MRI of the brain and spine, performed with and without contrast and with dissection detection protocol. This would reveal occult vascular malformations, dissections, or tumors. If none is found, a repeat cerebral angiogram is performed, Table 6-4 Fischer Grading System of Severity of Subarachnoid Hemorrhage
From Fisher CM, Kistler JP, Davis JM: Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by CT scanning. Neurosurgery 1980;6:1, with permission.
1 week or more later, this time with external carotid selective injections in addition to traditional four-vessel views, to exclude dural fistula (see Fig. 6-13). A second angiogram is not performed if another definite etiology of SAH is found or if the bleed was solely limited to perimesencephalic cisterns and if the first angiogram was of excellent quality. Critical Care Management The acute resuscitation of SAH follows the general guidelines discussed previously for hemorrhagic stroke in general. • Any patient with a GCS score of 8 or less or unable to protect the airway should be intubated, especially for planned transport. • Anticonvulsant prophylaxis should be administered as soon as possible because seizures can increase morbidity and brain edema. A central venous line and an arterial line should be inserted to assist with acute management. • Blood pressure should be controlled aggressively, and both hypertension and hypotension should be avoided (see Table 6-1). Hypertension should be avoided because of the presumed presence of an unsecured vascular lesion. Hypotension should be avoided because it could compromise cerebral perfusion, especially in the setting of elevated ICP in an unconscious patient. • Coagulation parameters should be examined and corrected promptly. • The patient should be given stool softeners to avoid any physical effort that could cause rebleeding of the cerebral aneurysm. • Pain management should be optimized, and the patient should be transferred as soon as possible to a critical care environment where these measures are maintained along with other multisystem homeostasis, as further diagnostic and therapeutic interventions are planned. Timing of Intervention The major cause of death in patients who survive an initial aneurysmal SAH is rebleeding. The timing of intervention should consider this risk, and there is a general consensus that good-grade patients should have early intervention to eliminate the aneurysm from the circulation within the first 48 hours.12 In experienced neurovascular centers, early treatment of aneurysm is also performed on poor-grade patients if they are hemodynamically stable.57 Arterial dissection that has caused an SAH also mandates early therapeutic intervention aimed typically at excluding the dissected segment from the circulation. Other causes may require urgent treatment, such as correction of coagulopathy. Rebleeding from tumors and vascular malformations is not common in the early phase, and treatment of these lesions is typically deferred until the patient is stabilized and optimal plans for lesion therapy are made.
W9418-06.qxd
11/3/03 2:50 PM Page 167
Chapter 6
Specific Treatment of Underlying Cause of Subarachnoid Hemorrhage: Endovascular Endosaccular aneurysm occlusion with platinum coils has been shown to be a safe and effective treatment option when performed by an experienced team. The goal of treatment is to obliterate and promote thrombosis of the aneurysm sac (see Chap. 7). This technique appear to be most effective in nongiant aneurysms with narrow necks and in poor surgical candidates.1 The recently published International Subarachnoid Aneurysm Trial (ISAT)58 is the only prospective randomized trial that compared endosaccular coiling to surgical clipping in ruptured aneurysms suitable for both treatments and observed for 1 year; this study concluded that coiling is significantly more likely to result in survival free of disability 1 year after SAH than surgical treatment.58 Yet fewer than 25% of eligible patients (1% to 40% at the various centers) in ISAT were randomized, while the others received treatment thought best by their physicians. Hence the results of this study cannot be generalized to patients with aneurysmal SAH, but rather are applicable to cases where a well-thoughtout discussion between endovascular and surgical experts does not favor one treatment over another. Aneurysms were more likely to rebleed after coiling than surgery, and to require retreatment, especially aneurysms with broader neck and at certain locations.58 Morbidity rate after coiling was close to 4% with 1% of mortality in a systemic review.59 One series of 75 patients treated with coiling found a 5% incidence of rebleeding within 6 months of treatment, complete obliteration was achieved in 40%, 37% had residual necks, and angiographic recurrences occurred in 24%.60 While coiling may be favored in certain cohorts such as in the ISAT, or in aneurysms at certain locations, such as the basilar summit,61,62 and in older or sicker patients,63 there is no evidence that the introduction of coiling has improved overall outcome of aneurysm treatment at large neurovascular centers.64 Other endovascular interventions have an important role in the management of SAH. Parent vessel occlusion, performed with endovascular coils or glue, can be used for occlusion of distal vessels harboring aneurysm, as in mycotic or traumatic aneurysms, and aneurysms on feeding vessels of AVMs. More proximal parent vessel occlusion, using endovascular balloons or coils, is performed in cases of fusiform aneurysm or dissection.65 This is preceded by balloon test occlusion under full anticoagulation and clinical monitoring, and is often deferred until patient is stabilized and awakened to tolerate test occlusion, and also until vasospasm has subsided because occlusion of a major artery poses a significant ischemic risk during vasospasm. Newer endovascular treatments include stents and balloon-assisted coiling of aneurysms with broad neck. Endovascular adjuncts, including proximal control, suction
Hemorrhagic Cerebrovascular Disease
167
decompression, and intraoperative angiography have enhanced surgical treatment of certain aneurysms such as giant lesions and those at paraclinoid locations.66 Endovascular coiling may be used in cases where surgery has failed to completely clip an aneurysm, and the residual neck is narrow. Specific Treatment of Underlying Cause of Subarachnoid Hemorrhage: Surgical Microsurgical clipping remains the definitive method for exclusion of an aneurysm and prevention of rebleeding.1 Surgery is not favored if the patient is unstable, in poor medical condition, or in the setting of intractable elevated ICP. Surgery is specifically indicated and should be performed emergently for cerebral aneurysm whenever there is an associated ICH causing or threatening herniation syndrome (see Fig. 6-11), as is common with middle cerebral aneurysms with temporal clots and anterior communicating or carotid aneurysms with deep frontal clots. Surgery for hematoma and aneurysm clipping should be perfomed emergently, and if necessary without cerebral angiography. An emergent enhanced CT scan should follow a regular CT scan before surgical intervention whenever there is a blood clot with a suspicion of aneurysm or vascular malformation rupture. This will often show the suspected lesion and avoid unexpected findings at surgery, such as more complex lesion than anticipated. An aneurysm should always be sought and clipped at the time of hematoma evacuation, but an AVM does not require excision at the same setting unless the patient is stable and the lesion is simple and well defined. Intraoperative or immediate postoperative angiography (before awakening) may be considered if there is any question about adequacy of treatment of aneurysm, especially if angiography was not performed preoperatively. Complications of Therapy A range of potential complications are associated with endovascular or surgical therapy. Aneurysm rupture may occur during attempted coiling or surgical clipping. This is typically handled with emergent technical maneuvers, but may result in untoward sequelae. Coils or clips may compromise parent vessels, their branches, or perforating vessels, and thromboembolism may occur during endovascular or surgical manipulation of blood vessels—all causing a spectrum of ischemic complications. These are prevented by judicious anticoagulation during endovascular interventions, and by verification of vessel patency by micro-Doppler insonation or intraoperative angiography. They are treated according to the specific clinical scenario, as with interventions for brain ischemia discussed elsewhere in this text.
W9418-06.qxd
11/3/03 2:50 PM Page 168
168
Section I
Introduction
Hydrocephalus Hydrocephalus is documented in 15% to 20% of patients with SAH.67,68 It is explained by blood interfering with CSF circulation in the ventricles, sylvian aqueduct, or the basal cisterns.69 A ventriculostomy should be performed whenever there is ventriculomegaly on the CT scan, especially if associated with altered mental status or with intraventricular hemorrhage (see Figs. 6-12 and 6-17). It is a bedside procedure, using a sterile technique and compact cranial access kits for twist drill or burr hole. In one study there was an improvement in 80% of the patients in whom this was used.67 Overdrainage should be avoided because it can provoke aneurysmal rebleeding67,70,71 by rapid modification of the transmural pressure. Overdrainage may also precipitate slit ventricles and prevent further cerebrospinal fluid drainage for treatment of elevated ICP. Optimal ventricular drainage aims to keep ICP below 10 to 15 mm Hg. Ventriculostomy is also performed in conjunction with aneurysm surgery to enhance brain relaxation for access to the aneurysm, and in cases of decreased level of consciousness, regardless of ventricular size, to monitor and assist in ICP management. Infection occurs in 5% to 10% of cases undergoing ventriculostomy.1,72,73 This is minimized by optimizing sterile technique at catheter insertion, by tunneling and carefully caring for the catheter exit site, by avoiding unsterile breach of the draining system, and by prophylactic intravenous antibiotics. Ventricular infections are best detected before fulminant ventriculitis and meningitis, by frequent (every 1 or 2 days) cerebrospinal fluid sampling for gram stain, cell counts, glucose, protein, and cultures. These infections are treated by optimizing intravenous antibiotic coverage, intrathecal antibiotics, and by changing the infected catheter. More than half of patients who undergo ventriculostomy are weaned from cerebrospinal fluid drainage in the first 2 weeks after SAH. This is assisted by a gradual increase of drainage threshold, intermittent clamping of ventriculostomy, or by intermittent lumbar punctures. Ventriculoperitoneal shunting, or the permanent implantation of a ventricular diversion system, is performed in cases where external ventricular drainage cannot be weaned, and symptomatic ventriculomegaly persists 1 to 2 weeks after SAH.74
infarctions in up to one third of patients with SAH. The risk of vasospasm increases with the severity of SAH as assessed on Fisher grade. Delayed neurologic deterioration after SAH is presumed due to ischemic sequelae of vasospasm unless proven otherwise, or attributed to other causes. The diagnosis of symptomatic vasospasm is supported by clinical evidence of spasm by transcranial Doppler (TCD) or angiography (see later discussion), and/or evidence of ischemia on diagnostic tests of cerebral blood flow (although such tests are not widely available, and are by themselves nonspecific to vasospasm). Symptomatic vasospasm is exacerbated by dehydration (hypovolemia) and by hypotension. The precise pathogenesis of vasospasm is not yet totally understood. The breakdown products of subarachnoid blood are probably responsible for initiating the vasospasm response in arteries of the circle of Willis and its branches.13 Vasospasm may be monitored noninvasively by insonating the circle of Willis vessels and its branches using TCD (Fig. 6-18). This bedside procedure has a high sensitivity
A
Vasospasm Monitoring, Prophylaxis, and Therapy Monitoring and Prophylaxis of Vasospasm Several days after SAH, there is an inflammatory reaction in blood vessels bathed in subarachnoid blood, resulting in luminal narrowing. The phenomenon, known as vasospasm, affects 60% to 70% of patients after SAH, and results in symptomatic ischemia in approximately half those cases. It reaches its maximal severity in the second week after SAH, and typically resolves spontaneously in the third or fourth weeks. Vasospasm causes death or serious disability from
B Figure 6-18. Transcranial Doppler insonation of intracranial artery. A, Normal tracing with mean velocity below 100 cm/sec and peak systolic velocity below 140 cm/sec. B, Severe vasospasm tracing with mean velocity exceeding 150 cm/sec and peak systolic velocity exceeding 240 cm/sec.
W9418-06.qxd
11/3/03 2:50 PM Page 169
Chapter 6
and specificity for vasospasm, but requires technical expertise and experience.75 The course of TCD documented vasospasm correlates closely with the course and clinical sequelae of vasospasm detected on angiography, and its severity closely reflects clinical sequelae of brain ischemia. Angiography may be used to confirm vasospasm in clinical situations in which the cause of delayed neurologic deterioration is questionable, TCD findings are nonconcordant with clinical progress, or endovascular therapy for vasospasm is being contemplated. Otherwise, TCD has largely replaced catheter angiography for the mere diagnosis of vasospasm, as catheter angiography carries definite risks, and the contrast dye load may exacerbate hypovolemia and precipitate or worsen clinical manifestations of spasm. Additionally, the dye load may impair renal function. Vasospasm prophylaxis includes pharmacologic therapy, judicious hydration, and volume and blood pressure support. Nimodipine, 60 mg orally or by nasogastric tube administered every 4 hours, for 21 days after SAH has been shown to significantly decrease the clinical sequelae of symptomatic vasospasm. Curiously, orally administered calcium channel blockers have not been shown to decrease the incidence of angiographic or TCD vasospasm. It has also been suggested that early surgery, with washing of the subarachnoid cistern, may help prevent vasospasm by clearing blood breakdown products. The same has been suggested for external ventricular drainage after SAH. Intravascular volume and blood pressure are judiciously maintained during the period of vulnerability to vasospasm, commencing 2 to 3 days after SAH. By then, risk of aneurysm rebleeding ought to have been eliminated by surgery or endovascular treatment, as the “spasm watch” phase of the illness is entered. During this phase, TCD is monitored every day or every other day. Close attention is given to monitoring of fluid balance, electrolytes, and central venous pressure, as indices of adequate hydration. Hyponatremia is treated with hypertonic saline rather than dehydration. Blood pressure parameters are liberalized, by withholding antihypertensives. Nimodipine is given more frequently in divided doses (30 mg every 2 hours) or held altogether if blood pressure is low. Any hint of neurologic deterioration is closely correlated with TCD findings to diagnose symptomatic vasospasm. Volume maintenance is gradually tapered following the period of vulnerability to vasospasm, and confirmed by resolving TCD velocities. Hypertensive, Hyperdynamic, Hemodilution Therapy Hypervolemia and induced hypertension are instituted in cases of TCD velocities indicating severe vasospasm (mean TCD velocities exceeding 150 cm/sec or peak systolic velocities exceeding 200 cm/sec in the anterior circulation vessels), or if there is any hint of neurologic deterioration attributed to vasospasm.76 Central venous pressure monitoring is mandatory in these cases, with a low threshold for
Hemorrhagic Cerebrovascular Disease
169
introducing a pulmonary artery catheter for optimizing volume resuscitation. Pressors are used to induce hypertension, titrated in proportion to TCD velocities, or to reverse ischemic neurologic deficits. Typically, volume status is aimed at a CVP value between 8 and 10 mm Hg. If the step to a pulmonary artery catheter is made, pulmonary artery wedge pressure (PAWP) or end diastolic volume indices (EDVI) are titrated through the creation of a pressure-volume (Starling) curve. This curve details the lowest PAWP or EDVI with the highest cardiac index. The fluid loads used for the generation of this relationship may be either isotonic crystalloid or 5% albumin solution. Simply pushing fluids to make the PAWP an arbitrary number is not an acceptable manner of handling hemodynamics, as one may either overshoot the optimal cardiac filling volume (or pressure) or keep the patient volume “underloaded.” Systolic blood pressure is maintained at blood pressure greater than 160 to 180 mm Hg. True hemodilution is no longer advocated as the goal of HHH therapy, for fear of inducing anemia and limiting oxygen carrying capacity to the brain. In fact, a hematocrit less than 30% is treated with red blood cell transfusion, while a hematocrit of between 40% and 45% is generally treated with volume expansion and iatrogenic hemodilution in the ICU setting. Alternated crystalloids and colloids are used for volume resuscitation, and dopamine or neosynephrine infusions are used for induced hypertension, after withholding all antihypertensive agents.6 Endovascular Therapy Cases of worsening vasospasm despite hyperdynamic therapy are considered for endovascular treatment.77–80 The precise threshold for endovascular interventions remains controversial, with some centers advocating early and frequent endovascular treatment of spasm, while other centers reserve endovascular intervention for cases where symptomatic vasospasm does not respond to hyperdynamic therapy. It is clear that not all cases of severe TCD spasm will require endovascular intervention, and such therapy introduces an added risk that must be considered and weighed. Conversely, endovascular treatment of spasm should not be delayed until actual infarction has developed. Endovascular treatment of spasm consists of balloon angioplasty, best used for large-vessel spasm (Fig. 6-19), or intra-arterial vasodilator infusions, papaverine or verapamil, best used for more distal branch vasospasm (Fig. 6-20). Angioplasty is associated with greater risk of arterial rupture or dissection, especially if applied to more distal vessels, but its effect is more durable than intra-arterial pharmacologic infusions.81–83 The latter may need to be repeated daily, driven by clinical response and monitoring of TCD velocities. With aggressive monitoring, prophylaxis, HHH therapy, and judicious use of interventional techniques, morbidity from vasospasm can be minimized to less than 5% of SAH patients.84
W9418-06.qxd
11/3/03 2:50 PM Page 170
170
Section I
Introduction
B
A
Figure 6-19. Angiogram of severe basilar artery vasospasm. Before (A) and after (B) balloon angioplasty.
Outcome after Subarachnoid Hemorrhage Subarachnoid hemorrhage is associated with significant morbidity and mortality. Despite affecting patients in the middle years of their lives, often without other preexisting or associated diseases, it is estimated that 25% to 50% of all SAH patients will die as a result of their bleed. A common cause of death is neurologic damage from the initial bleed, with 10% of patients estimated to die before reaching a medical facility, while others reach medical attention in poor condition. Many survivors are left with persistent physical, cognitive, behavioral, and emotional changes that affect their day-to-day lives. The most common predictor of death or major disability after SAH is the patient’s clinical condition at presentation.50,53,54 Age, medical morbidities, severity of hemorrhage on CT, and aneurysm type (giant or posterior circulation) are also correlated with poorer outcome.85–90 Other patients are initially in good condition, and deteriorate in the setting of misdiagnosis, from rebleed, as a result of therapeutic complications, or from vasospasm or other medical or neurologic sequelae of the disease.
Much of what is discussed in this chapter and elsewhere in this book, as well as ongoing advances in surgical and endovascular therapy will further reduce mortality and morbidity from SAH. Those who survive will benefit from early rehabilitation.91 Some studies showed an improvement in the Functional Independence Measure (FIM) after an inpatient rehabilitation stay.91,92 The quality of life of those who apparently recover with minimal disability may still be impaired by cognitive, psychological, and emotional sequelae. Recognition and intervention for these higher functional deficits may further improve the quality of life of patients inflicted by this disease. The clinical outcome after SAH cannot be addressed at a single point in time, without regard to whether the aneurysm has been treated effectively. It is essential to note whether the aneurysm may still pose a risk of future rupture, and what additional follow-up and retreatments may be indicated. These questions of long-term durability of treatment, and the impact on quality of life and future risks, are essential when addressing relative benefits of endovascular versus surgical interventions.
W9418-06.qxd
11/3/03 2:50 PM Page 171
Chapter 6
Hemorrhagic Cerebrovascular Disease
171
B A Figure 6-20. Angiogram of severe middle cerebral artery vasospasm. Before (A) and after (B) intraarterial verapamil infusion.
Intracerebral Hemorrhage Etiology Hypertensive Small Vessel Disease The major risk factors for primary ICH are age, hypertension, and alcohol abuse.1,93,94 The most frequent sites of hypertensive ICH are the putamen (50%) (Fig. 6-21), thalamus (15%), pons (10% to 15%), and cerebellum (10%).2 The relative risk of ICH with hypertension is increased 3.9- to 5.4-fold.93 Hypertensive arteriopathy includes arterial and arteriolar sclerosis, dilated perivascular Virchow-Robin spaces (“état criblé”), and deep brain matter ischemic degeneration (“leukoaraoisis”), all likely predisposing to hemorrhage. Deep ganglionic hypertensive bleeds are often caused by miliary aneurysms,95 also known as microaneurysms of Charcot-Bouchard. They occur at bifurcation of small perforators of lenticulostriate arteries mainly in hypertensive patients.96 Deep ganglionic hemorrhages occur at any age in the setting of uncontrolled hypertension, but their prevalence increases exponentially with each decade above age 60.
Outcome of hypertensive bleeds is mostly determined by the volume of hemorrhage, patient age, neurologic condition at presentation, and clot location (anterior ganglionic bleeds of caudate and putamen have a better outcome than thalamic or pontine bleeds). Amyloid Angiopathy Amyloid angiopathy, also known as congophilic angiopathy, is associated with degeneration of lobar parenchymal vessels in advancing age. It is found at autopsy in the brains of one third of patients older than 60 years of age.97 It consists of the deposition of beta amyloid protein within the media of small meningeal and cortical vessels.98 Fibrinoid necrosis of the vessel wall is also present in some vessels.99,100 This arteriopathy is thought to predispose to lobar ICH (see Fig. 6-2), mostly in subcortical white matter.101 It is thought to account for 10% to 32% of nontraumatic ICH.102 It has a better prognosis than deep basal ganglia bleed.102 Coagulopathy The risk of ICH is increased in patients on warfarin, estimated at 0.3% per year and increasing with advancing age
W9418-06.qxd
11/3/03 2:50 PM Page 172
172
A
C
Section I
Introduction
B
Figure 6-21. Deep (ganglionic ICH). Most common cause is hypertension, although an underlying vascular anomaly must be sought if patient is young or has no history of hypertension. A, At presentation. B, After placement of catheters in the ventricle for treatment of hydrocephalus, and in the clot for thrombolytic drainage of ICH. C, After five doses of intra-clot ICH performed twice daily, with serial aspiration of ICH.
W9418-06.qxd
11/3/03 2:50 PM Page 173
Chapter 6
and with uncontrolled hypertension.103 The risk of hemorrhage is the highest during the first 3 months of anticoagulation104 and with extremes of anticoagulation. In one series, the only fatal bleeding complication related to warfarin therapy was ICH.104 Thrombolytic therapy for ischemic stroke and acute MI is also associated with increased risk of ICH.105–107 Aspirin treatment is associated with a risk of ICH at a rate of 0.2% to 0.8% per year.103,108 Patients with amyloid angiopathy are also at increased risk of ICH with anticoagulant therapy.109 Vascular Malformations The average risk of hemorrhage from an AVM varies between 2% and 4% per year, typically causing an ICH (Fig. 6-22), and less commonly intraventricular hemorrhage and SAH.31 Hemorrhage is frequently associated with arterial feeder aneurysm, nidal aneurysm, venous varices, and venous outflow obstruction of AVM nidus.32 An AVM hemorrhage is associated with 10% mortality and 30% to 50% morbidity.110
Hemorrhagic Cerebrovascular Disease
173
Angiographically occult vascular malformations, including cavernous malformation, can cause ICH and yet be occult on angiography. These also are difficult to detect on MRI when obscured by overlying hematoma. It is important to suspect such lesions as the cause of ICH, especially in younger patients. If surgical evacuation and exploration of hematoma is not undertaken, a follow-up MRI is mandatory, after resolution of ICH, to detect underlying lesions. Intracranial Aneurysm ICH is an uncommon presentation of an aneurysm rupture (see Fig. 6-11). When ICH is present, it is almost always associated with an SAH or IVH component, or the hematoma is contiguous with a major subarachnoid artery. The middle cerebral artery aneurysm is most frequently associated with ICH, although aneurysms at all locations can also do it.111 Venous Occlusive Disease Many conditions have been incriminated in cerebral venous thrombosis including birth control pills,112 otitis media,113,114 pregnancy and puerperium,115 dehydration, closed head injury, hypercoagulable state, and postcraniotomy. Venous occlusion results in hyperemic brain edema and extravasation of blood into the surrounding brain parenchyma, predisposing to SAH or ICH.116 Postoperative Hemorrhage The frequency of clinically significant postoperative intracerebral hemorrhage varies between 0.8%117 and 3.9%.118 Other minor contusional changes are quite common after craniotomy. Usually the hemorrhage occurs at the site of surgery, but cases are also reported remote from the surgical bed.119,120 Blood pressure and coagulation parameters have been implicated in the majority of these bleeds.117 ICH can also occur after carotid endarterectomy or stenting, and is known as cerebral hyperperfusion syndrome.121
Figure 6-22. CT scan with contrast in a young patient revealing a frontal ICH and an adjacent AVM. The location of AVM would not have been expected from CT scan without contrast. Contrast-enhanced CT scan should always be performed before surgical evacuation of ICH to avoid unexpected findings and complications at surgery. The hematoma was evacuated emergently at primary procedure, without disrupting the AVM. Following recovery from acute event, the AVM was embolized and excised surgically.
Drug Use Hemorrhagic stroke and ICH in particular have been associated with the use or abuse of a variety of drugs including cocaine,122,123 amphetamines,124 cannabis,125 and nasal decongestant and diet drugs containing phenylpropranolamine.126,127 Women are more predisposed to the latter etiology.127 The mechanisms of hemorrhage are thought to include drug-induced vasculopathies, hemodynamic effects of drug use (notably hypertension), and bleeding from an associated brain pathology (vascular malformation or aneurysm). Other Causes There are other known etiologies of ICH, including hemorrhagic transformation of an ischemic stroke,128 primary vasculitis of the central nervous system, endarteritis from septic emboli, and primary and metastatic brain tumors. In the latter category, the most important etiology is glioblastoma
W9418-06.qxd
11/3/03 2:50 PM Page 174
174
Section I
Introduction
multiforme and metastases from melanoma, renal cell carcinoma, and choriocarcinoma.129–131 Tumors often exhibit contrast enhancement on MRI, and may be detectable within or adjacent to ICH. Multiple metastases may also be seen on MRI. Otherwise, a late MRI is indicated, as with occult vascular malformations, to exclude underlying lesions after resolution of ICH. Presentation and Diagnosis Clinical Presentation The classic presentation of ICH is sudden onset of headache, decreased consciousness, a focal neurologic deficit slowly progressing over hours, nausea, vomiting, and elevated blood pressure.1,8 The acute onset of thunderclap headache is typical of aneurysm rupture and is not documented classically with other etiologies. The neurologic examination including hemiparesis, ataxia, sensory disturbances, gaze syndromes, and cranial neuropathies can often suggest the ICH location.1 The level of consciousness at clinical presentation often deteriorates in the first hours after ICH, correlating with expansion of hematoma volume. Diagnostic Evaluation It is important to obtain a detailed history about the onset of symptoms, medical risk factors, anticoagulation, drug abuse, and head trauma. The history by itself can orient toward a possible etiology, as with drug use or uncontrolled hypertension. Preliminary laboratory testing should include coagulation parameters, blood cell counts, electrolytes, drug screen, and liver function tests.1 A CT scan is the best initial imaging modality for the diagnosis of acute ICH. An enhanced brain CT scan should typically be performed before any invasive interventions, to highlight suspected vascular abnormalities or underlying structural lesions as etiology of ICH. The CT scan also assists in determining the volume of the ICH in cubic centimeters by a simple equation, based on ellipsoid volume: volume = (A ¥ B ¥ C)/2, where A is the maximal diameter of hematoma in centimeters on its largest axial cut, B is its diameter in centimeters in the dimension perpendicular to A on the same axial cut, and C is the height of the hematoma. C is calculated by multiplying the number of slices on which the hematoma is visible (excluding the first and last cuts just showing its extremes) by the slice thickness of the CT scans in centimeters.132,133 The estimated hematoma volume in cubic centimeters is an important prognostic indicator, and may influence surgical indications. The MRI/MRA can be useful in identifying ICH etiologies including AVMs, cavernomas, tumors, and aneurysms. Although MRI may miss small aneurysms and vascular malformations, it is superior to CT and angiography in detecting cavernous malformations. Lesions may often be
obscured by overlying blood on early imaging after ICH. A delayed MRI will often reveal underlying structural lesions that may have been obscured by overlying ICH in the acute state. Cerebral four-vessel angiography remains the gold standard for the diagnosis of aneurysms and AVMs, and should be performed in younger patients where the etiology of ICH is not known. An angiogram should be performed urgently when ICH extends into a subarachnoid cistern, and could represent intracranial aneurysm. A contrast-enhanced CT or a CT angiogram can be performed in lieu of four-vessel angiography in unstable patients, or in patients taken for urgent surgery because of impending herniation from ICH. For the evaluation of AVM, the angiogram may be delayed until the patient is stable, because early rebleeding is not common from AVMs. An AVM may be compressed by mass effect on early angiography after ICH, and may fill more readily on delayed angiograms. Critical Care Management Acute Resuscitation The principles of airway management are as considered previously in this chapter for hemorrhagic stroke in general and for SAH. Blood pressure should be controlled closely8,134 because hypertension may contribute to rebleeding within the first hours after ICH.135 However, excessive reduction of blood pressure should be avoided so as not to compromise the cerebral perfusion pressure. In patients with a history of hypertension, the mean arterial pressure should be maintained below 130 mm Hg.11 Vasopressors, albumin, and crystalloid boluses can be used to treat hypotension. The American Heart Association evidence-based guidelines for blood pressure control in ICH are summarized in Table 6-1.8 Patients with ICH should be kept euvolemic or relatively dry to avoid exacerbations of brain edema. Hyponatremia should be monitored and treated as appropriate for SIADH because it can exacerbate brain edema and lower seizure threshold. Seizures can contribute to neuronal injury and systemic instability, especially in the acute state after ICH, so anticonvulsant prophylaxis is administered acutely (phenytoin 17 mg/kg as a loading dose, then 100 mg every 8 hours) and tapered after 1 month if the patient is seizure free.8 Intracranial Pressure Intracranial hypertension is the main cause of death in ICH. Elevated ICP is defined as intracranial pressure greater than or equal to 20 mm Hg for greater than 5 minutes. The goal of treatment is ICP less than 20 mm Hg and cerebral perfusion pressure greater than 60 to 70 mm Hg (see Chap. 25).136 Patients with ICH and GCS score less than 9, or who cannot be followed up by neurologic examination are considered for hematoma evacuation or ICP monitoring. The type of device
W9418-06.qxd
11/3/03 2:50 PM Page 175
Chapter 6
depends on availability and experience. Intraventricular ICP monitors and intraparenchymal fiberoptic ICP devices are two commonly used methods of monitoring ICP.8 Ventriculosomy is favored in cases without slit ventricles because it allows cerebrospinal fluid drainage for treatment of elevated ICP, and is essential in cases of hydrocephalus or intraventricular hemorrhage associated with ICH. Cerebral compartment herniation can occur from ICH in the setting of low or moderate ICP, especially with ICH in the frontal or temporal lobes, and in cases of cerebellar or brainstem hematomas. Direct destructive effects of expanding ICH may cause much neurologic damage without elevated ICP. Intracranial pressure monitoring and management are not considered as substitutes for hematoma evacuation if ICH volume and the clinical situation warrant it (see later discussion). Medical Therapy Therapeutic measures for blood pressure control, correction of coagulopathy, and seizure prophylaxis are continued after the acute resuscitation. Measures for ICP control are instituted as discussed previously, but these should never be considered in lieu of hematoma evacuation as appropriate.8 Fluid management is aimed at minimizing brain edema, and hence judicious fluid restriction is instituted with close follow-up of electrolytes. There is no evidence that steroids are generally beneficial in the treatment of ICH, and they may increase systemic complications. Attention to multisystem homeostasis is maintained, including judicious prevention and treatment of pneumonia and other sepsis, and the use of tracheotomy and percutaneous feeding tube if long-term impairment of consciousness is anticipated. Surgical Therapy Ventricular Drainage and Intraventricular Thrombolysis Obstructive hydrocephalus may require ventriculostomy placement and external cerebrospinal fluid drainage. This can be a lifesaving maneuver, and has contributed to many dramatic recoveries after intraventricular hemorrhage. Ventricular catheters often occlude from clot (see Fig. 6-17), which limits their effectiveness. This drawback has recently been effectively and safely managed with intraventricular instillation of thrombolytic agent (1 to 2 mg of tissue plasminogen activator, administered through the catheter every 8 to 12 hours as needed to maintain patency). This intervention is only considered if a structural cause (tumor, AVM, or unsecured aneurysm) or coagulopathy have been excluded as causes of ICH. Ventricular catheters are managed as discussed previously for SAH, with gradual weaning and consideration of ven-
Hemorrhagic Cerebrovascular Disease
175
triculoperitoneal shunting if hydrocephalus persists after the acute stage. Hematoma Evacuation The patient’s clinical condition, co-morbidities and life expectancy, the size and location of the hematoma, and the presence of an underlying etiology are all primary factors in the decision whether to evacuate an ICH. Aggressive interventions should be avoided in patients in poor neurologic state, those with premorbid conditions limiting life quality or expectancy, or with specific advance directives. Hematoma evacuation is also avoided when ICH affects the brainstem or diencephalon and the patient is elderly or deeply comatose. Patients with favorable life expectancy, especially younger patients, who are deteriorating rapidly from large lobar ICH should undergo emergent hematoma evacuation regardless of the underlying etiology and without waiting for further elaborated imaging. Cerebellar or temporal lobe ICH causing any mass effect or neurologic impairment should be evacuated urgently because of high risk of dire complications from subsequent herniation.8,137 A contrast-enhanced CT scan may be helpful at revealing unexpected etiology, and may avoid unexpected findings and complications at surgery. Associated aneurysm is clipped at the time of hematoma evacuation if at all possible. In younger patients, clot evacuation is performed with a thorough microsurgical exploration of the cavity because of high likelihood of underlying structural lesion as etiology of ICH. If a simple vascular malformation or tumor is associated with ICH, it is dealt with during the same procedure. Otherwise, the blood clot is removed and thorough hemostasis and brain relaxation are ensured. The patient is allowed to recover and any underlying lesion is further investigated and tackled at a subsequent stage when brain edema has subsided and under the most optimal conditions and preparations for addressing a complex lesion.137 This may include subsequent preparatory embolization, or a modified surgical approach (see Fig. 6-22). If the brain is swollen or anticipated to likely swell, the dural layer is not closed, and the bone flap is left out (frozen or implanted in the patient’s preperitoneal fat) for subsequent autogenous cranioplasty after brain swelling has subsided. In certain situations, imaging cannot exclude an underlying occult lesion as the cause of ICH in a young patient with favorable life expectancy. Even if the patient is neurologically stable or the ICH is not particularly large, ICH evacuation may be performed if it can be accomplished at very low risk, as in young patients with superficial bleeds, along with microsurgical exploration for underlying etiology (see Fig. 6-2). Often, such patients recover faster because of evacuation of clot, and earlier surgery provides a more generous cavity for exploring and resecting underlying lesion that may have required surgery anyway at a later time. Early surgery also avoids any risk, however small, of rebleeding while
W9418-06.qxd
11/3/03 2:50 PM Page 176
176
Section I
Introduction
waiting for hematoma to resolve. If early surgery is not performed in a young patient with ICH, delayed follow-up is always planned, with subsequent imaging for underlying lesion after ICH has resolved. When underlying structural lesion is not suspected, as in elderly or severely hypertensive patients with a negative contrast-enhanced CT scan, the role of hematoma evacuation is controversial.1 Lobar ICH in a patient neurologically stable and with a volume less then 20 cc should be treated conservatively, with close clinical and radiographic follow-up,1,8 while surgery is recommended in deteriorating young patients with large lobar hematomas.138,139 Cerebellar hemorrhage is usually a surgical emergency because of the risk of brainstem compression and sudden death, and some nonrandomized studies have favored the evacuation of every cerebellar hematoma greater than 3 cm in diameter.136,138–140 The management of deep thalamic and ganglionic hematomas is complex. A large Japanese prospective registry concluded that medical therapy was superior to surgery in patients with minimal neurologic impairment or clots smaller than 15 cc. In comatose patients there was an apparent reduction in mortality in the surgical group but very poor functional recovery.141 Randomized studies reported to date on surgical removal of ICH have not shown a clear benefit of open surgical evacuation (most of these studies did not randomize younger patients with active neurologic deterioration who were most likely to benefit from surgery). Several studies have indicated a potential benefit of ICH evacuation when performed using minimally invasive techniques (endoscopic or stereotactic aspiration rather than open surgery).142–145 The American Heart Association has published recommendations for surgical evacuation of ICH and these are summarized in Table 6-5.8 It may be that the benefits of ICH volume reduction are obviated by the morbidity of open surgery in older patients, while hematoma reduction may be beneficial if it can be performed noninvasively. One such technique of noninvasive evacuation is the thrombolysis of ICH using tissue plasminogen activator (tPA) or urokinase, and aspiration of hematoma through a catheter (see Fig. 6-21).146–150 One commonly proposed protocol advocates tPA infusion in the catheter in doses of 1 to 2 mg at 12-hour intervals, with open catheter drainage between instillation for deep or lobar ICH with volume greater than 15 to 20 cc.146 Early results are very encouraging regarding the safety and effectiveness of this ICH volume reduction technique in cases without underlying structural lesion and/or coagulopathy.146–148 The optimization of this technique, dose escalation studies, thorough assessment of clinical efficacy, and quality of life among survivors awaits more rigorous trials.
Table 6-5 Recommendations for Surgical Treatment of Intracerebral Hemorrhage
From Broderick JP, Adams HP, Barsan W, et al: Guidelines for the management of spontaneous intracerebral hemorrhage. A statement of healthcare professionals from a special writing group of the stroke council, American Heart Association. Stroke 1999;30:905, with permission.
cerebral herniation occurring mainly during the first week after ICH.152 Mortality rate is related to the size, location, and underlying etiology of the hematoma, and age of the patient. The 30-day mortality rate is approximately 44%.153 The subgroup of patients with lobar hemorrhage fares significantly better, with a mortality rate estimated at 11%.101 The most consistent predictor of a poor outcome, described in various studies are large ICH, with clots less than 15 cc in volume commonly resulting in good outcome, and those larger than 60 cc almost always resulting in death or serious disability.132,154–157 Patient age, in particular those older than 60 to 70 years, is an independent predictor of mortality and poor outcome. Other co-dependent factors influencing outcome include neurologic condition at presentation (poorer outcome in comatose patients), and clot location (poorer outcome in deep compared to lobar bleeds, in posterior compared to anterior ganglionic bleeds, and in brainstem hemorrhage). Survival is not uncommon in younger patients who are often disabled and their lives and careers permanently altered by sequelae of ICH. Even patients with apparently good outcomes by gross criteria suffer cognitive, emotional, or other more subtle limitations affecting their quality of life.
Summary and Conclusions Outcome after Intracerebral Hemorrhage ICH accounts for 10% to 13% of strokes,151 but is associated with a case fatality rate of 50%.2 The main cause of death is
Hemorrhagic stroke is a catastrophic disease with numerous factors that have an impact on clinical outcome and secondary sequelae. A thorough understanding of every facet of
W9418-06.qxd
11/3/03 2:50 PM Page 177
Chapter 6
this illness and its potential complications has allowed a rational approach to its critical care management. Common aims are to minimize neurologic injury, prevent and treat systemic complications, enhance recovery, and prevent delayed recurrence and complications. Numerous interventions have been accredited by firm scientific evidence, while others depend on our best interpretation of current modalities and interventions. Much future research and rigorous documentation of outcome with various interventions will allow further refinements of the broad strategies outlined in this chapter. Current outcome predictions in hemorrhagic stroke, and the basis for clinical decisions, continue to be derived from
Hemorrhagic Cerebrovascular Disease
177
group or cohort statistical associations, while we lack good models for absolute outcome prediction in individual patients based on all factors known to affect the outcome of this disease. Such absolute outcome prediction modeling should be possible in the near future with emerging megadatabases and advanced statistical techniques. These would ideally include confidence intervals (measures of uncertainty) of outcome predictions, and modifiable and unmodifiable variables in a given clinical situation, to help guide treatment decisions and prognostic judgments in individual patients.
P earls 1. Hemorrhagic stroke, including intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH) constitutes 23% to 25% of all stroke cases. 2. The age-adjusted incidence of ICH is 10 to 15 per 100,000 per year, with a mean age of 65, and its incidence doubles with each decade of life above age 45. 3. The incidence of SAH has been estimated at 6 to 30 per 100,000 per year, with a mean age of 50 years. 4. Magnetic resonance imaging (MRI) is more sensitive than CT scan for detecting structural abnormalities such as tumors and vascular malformations, and it is used when these etiologies are suspected, as in ICH in younger patients. 5. The acute resuscitation of hemorrhagic stroke follows the modified “ABC” guidelines of airway, blood pressure, and cerebral perfusion. The airway should be cleared, and any patient with a Glasgow Coma Scale (GCS) score of 8 or less or unable to protect the airway should be intubated. 6. Hyponatremia is common in patients with hemorrhagic stroke. It may result from two mechanisms, the syndrome of inappropriate antidiuretic hormone (SIADH) with free water retention or inappropriate natriuresis (also known as cerebral salt wasting) mediated by the atrial natriuretic factor (ANF) and brain natriuretic peptide (BNP). Determining the likely cause is important because the two syndromes are managed differently. 7. The incidence of angiogram-negative SAH is estimated to be between 10% and 15%. A small fraction of these include an occult aneurysm with incomplete or misread angiogram, poor filling of aneurysm due to flow pattems or thrombosis, or parent vessel vasospasm. A second or third angiogram may reveal occult aneurysm. 8. The typical presentation of SAH is a sudden onset of severe headache, “thunderclap,” or the worst
9.
10.
11.
12.
13.
14. 15.
16.
headache the patient has ever experienced. The presentation could also be a seizure, loss of consciousness, or altered level of consciousness. Subarachnoid hemorrhage is frequently misdiagnosed with dire clinical consequences and a poorer outcome among misdiagnosed cases. Hence a concerted level of clinical suspicion should always accompany any clinical presentation suggestive of possible SAH. Among chronic headache sufferers, the headache of SAH is typically different and more severe than other previous headaches and may be associated with a sense of impending doom. The major cause of death in patients who survive an initial aneurysmal SAH is rebleeding. The timing of intervention should consider this risk, and there is a general consensus that good-grade patients should have early intervention to eliminate the aneurysm from circulation within the first 48 hours. While coiling may be favored in certain cohorts such as in the ISAT, or in aneurysms at certain locations, such as the basilar summit, and in older or sicker patients, there is no evidence that the introduction of coiling has improved overall outcome of aneurysm treatment at large neurovascular centers. Many survivors of SAH are left with persistent physical, cognitive, behavioral, and emotional changes that affect their day-to-day lives. The most common predictor of death or major disability after SAH is the patient’s clinical condition at presentation. The major risk factors for primary ICH are age, hypertension, and alcohol abuse. The most frequent sites of hypertensive ICH are the putamen (50%), thalamus (15%), pons (10% to 15%), and cerebellum (10%). The classic presentation of ICH is sudden onset of headache, decreased consciousness, a focal neuroContinued
W9418-06.qxd
11/3/03 2:50 PM Page 178
178
Section I
Introduction
logic deficit slowly progressing over hours, nausea, vomiting, and elevated blood pressure. The acute onset of thunderclap headache is typical of aneurysm rupture and is not documented classically with other etiologies. 17. The MRI/MRA can be useful in identifying ICH etiologies including AVMs, cavernomas, tumors, and aneurysms. Although MRI may miss small aneurysms and vascular malformations, it is superior to CT and angiography in detecting cavernous malformations. 18. Obstructive hydrocephalus may require ventriculostomy placement and external cerebrospinal fluid drainage. This can be a lifesaving maneuver and has contributed to many dramatic recoveries after intraventricular hemorrhage. 19. Lobar ICH in a patient neurologically stable and with a volume less than 20 cc should be treated conservatively with close clinical and radiographic follow-up,
while surgery is recommended in deteriorating young patients with large lobar hematomas. 20. Cerebellar hemorrhage is usually a surgical emergency because of the risk of brainstem compression and sudden death, and some nonrandomized studies have favored the evacuation of every cerebellar hematoma greater than 3 cm in diameter. 21. The most consistent predictor of a poor outcome, described in various studies, are large ICH, with clots less than 15 cc in volume commonly resulting in good outcome, and those larger than 60 cc almost always resulting in death or serious disability. 22. Survival is not uncommon in younger patients who are often disabled and their lives and careers permanently altered by sequelae of ICH. Even patients with apparently good outcomes by gross criteria suffer cognitive, emotional, or other more subtle limitations affecting their quality of life.
References 15. 1. Abdulrauf SI, Furlan AJ, Awad IA: Primary intracerebral hemorrhage and subarachnoid hemorrhage. J Stroke Cerebrovasc Dis 1999;8: 146. 2. Foulkes MA, Wolf PA, Price TR, et al: The stroke data bank: Design, methods and baseline characteristics. Stroke 1988;19:547. 3. Hamedani AG, Wells CK, Brass LM, et al: A quality-of-life instrument for young hemorrhagic stroke patients. Stroke 2001;32:687. 4. Furlan AJ, Whisnant JP, Elveback LR: The decreasing incidence of primary intracerebral hemorrhage: A population study. Ann Neurol 1979;5:367. 5. Sacco RL, Wolf PA, Bharucha NE, et al: Subarachnoid and intracerebral hemorrhage: Natural history, prognosis, and precursive factors in the Framingham study. Neurology 1984;34:847. 6. Wecht DA, Awad IA: Subarachnoid hemorrhage. In Grossman RG, Loftus CM (eds): Principles of Neurosurgery, 2nd ed. Philadelphia, Lippincott-Raven, 1999:297–309. 7. Mayberg MR, Batjer HH, Dacey R, et al: Guidelines for the management of aneurysmal subarachnoid hemorrhage. A statement of healthcare professionals from a special writing group of the stroke council, American Heart Association. Stroke 1994;25:2315. 8. Broderick JP, Adams HP, Barsan W, et al: Guidelines for the management of spontaneous intracerebral hemorrhage. A statement of healthcare professionals from a special writing group of the stroke council, American Heart Association. Stroke 1999;30:905. 9. Dul K, Drayer B: CT and MR imaging of intracerebral hemorrhage. In Kase CS, Caplan LR (eds): Intracerebral Hemorrhage, vol 5. Boston, Butterworth-Heinemann, 1994:73–93. 10. Ropper AH, King RB: Intracranial pressure monitoring in comatose patients with cerebral hemorrhage. Arch Neurol 1984;41:725. 11. Diringer MN: Intracerebral hemorrhage: Pathophysiology and management. Crit Care Med 1993;21:1591. 12. Broderick JP, Brott TG, Duldner JE, et al: Initial and recurrent bleeding are the major causes of death following subarachnoid hemorrhage. Stroke 1994;25:1342. 13. Inagawa T: Cerebral vasospasm in elderly patients treated by early operation for ruptured intracranial aneurysms. Acta Neurochir (Wien) 1992;115:79. 14. Haley EJ, Kassel NF, Torner JC: A randomized controlled trial of high dose intravenous nicardipine in aneurysmal subarachnoid hemor-
16.
17.
18. 19. 20.
21.
22.
23. 24.
25.
26.
27. 28.
rhage. A report of the Cooperative Aneurysm Study. J Neurosurg 1993;78:537. Haley EJ, Kassel NF, Torner JC: A randomized trial of nicardipine in subarachnoid hemorrhage: angiographic and transcranial Doppler ultrasound results. A report of the Cooperative Aneurysm Study. J Neurosurg 1993;78:548. Lanzino G, Kongable GL, Kassell NF: Electrocardiographic abnormalities after nontraumatic subarachnoid hemorrhage. J Neurosurg Anesthesiol 1994;6:156. Wells C, Cujec B, Johnson D, et al: Reversibility of severe left ventricular dysfunction in patients with subarachnoid hemorrhage. Am Heart J 1995;129:409. Kassell NF, Bourini DJ: Perioperative care of the aneurysm patient. Contemp Neurosurg 1984;6:1. Theodore J, Robin ED: Speculations on neurogenic pulmonary edema. Am Rev Respir Dis 1976;113:405. Wise BL: SIADH after spontaneous subarachnoid hemorrhage: A reversible cause of clinical deterioration. Neurosurgery 1978;3: 412. Wijdicks EFM, Schievink WI, Burnett JC: Natriuretic peptide system and endothelin in aneurysmal subarachnoid hemorrhage. J Neurosurg 1997;87:275. Vives KP, Maraire JN, Awad IA: Quality of life issues in vascular malformations of the central nervous system. In Jafar JJ, Awad IA, Rosenwasser RH (eds): Vascular Malformations of the Central Nervous System. Philadelphia, Lippincott, Williams and Wilkins, 1999:507–515. Wiebers DO, Whisnant JP, Sundt TM, et al: The significance of unruptured intracranial saccular aneurysms. J Neurosurg 1987;66:23. Wills S, Ronkainen A, Van der voet M, et al: Familial intracranial aneurysms: An analysis of 346 multiplex Finnish families. Stroke 2003;34:1370. Forget TR Jr, Benitez R, Veznedaroglu E, et al: A review of size and location of ruptured intracranial aneurysms. Neurosurgery 2001;49:1322. Wiebers DO, Whisnant JP, Huston J 3rd, et al: Unruptured intracranial aneurysms: Natural history, clinical outcome and risk of surgical and endovascular treatment. Lancet 2003;362:103. McCormick WF: The pathology of vascular (‘arteriovenous’) malformations. J Neurosurg 1966;24:807. Kopitnik TA, Samson DS: Management of subarachnoid hemorrhage. J Neurol Neurosurg Psychiatry 1993;56:947.
W9418-06.qxd
11/3/03 2:50 PM Page 179
Chapter 6 29. McCormick P, Spetzler R: Arteriovenous malformations of the brain. Curr Opinion Neurol Neurosurg 1991:71–75. 30. Ondra SL, Troupp H, George ED, et al: The natural history of symptomatic arteriovenous malformations of the brain: A 24-year follow up assessment. J Neurosurg 1990;73:387. 31. Kondziolka D, McLaughlin MR, Kestle JRW: Simple risk predictions for arteriovenous malformation hemorrhage. Neurosurgery 1995;37:851. 32. Crawford PM, West CR, Chadwick DW, et al: Arteriovenous malformations of the brain: Natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986;49:1. 33. Luessenhop AJ: Dural arteriovenous malformations. In Wilkins RH (ed): Neurosurgery. New York, McGraw-Hill, 1985:1473–1477. 34. Djindjian R, Merland JJ: Superselective Arteriography of the External Carotid Artery. New York, Springer Verlag, 1978. 35. Keynar MY, Wecht DA, Awad IA: Clinical spectrum and natural history of spinal vascular malformations. In Barrow DL, Awad IA (eds): Spinal Vascular Malformations. Elk Grove Village, IL, AANS, 1999:45–55. 36. Yamaura A: Nontraumatic intracranial arterial dissection: Natural history, diagnosis, and treatment. Contemp Neurosurg 1994;16:1. 37. Anson J, Crowell RM: Cervicocranial arterial dissection. Neurosurgery 1991;29:89. 38. Topcuoglu MA, Ogilvy CS, Carter BS, et al: Subarachnoid hemorrhage without evident cause on initial angiography studies: Diagnostic yield of subsequent angiography and other neuroimaging tests. J Neurosurg 2003;98:1235. 39. Muller-Forell W, Welschehold S, Kohler J, et al: Subarachnoid hemorrhage without aneurysm. Radiologe 2002;42:871. 40. Van Calenbergh F, Plets C, Goffin J, et al: Nonaneurysmal subarachnoid hemorrhage: Prevalence of perimesencephalic hemorrhage in a consecutive series. Surg Neurol 1993;39:320. 41. Rinkel GJE, Wijdicks EFM, Vermeulen M, et al: Outcome in perimesencephalic (nonaneurysmal) subarachnoid hemorrhage: A follow up study in 37 patients. Neurology 1990;40:1130. 42. Schwartz TH, Solomon RA: Perimesencephalic nonaneurysmal subarachnoid hemorrhage: Review of the literature. Neurosurgery 1996;39:433. 43. Wijdicks EFM, Schievnik WI, Miller GM: Pretruncal nonaneurysmal subarachnoid hemorrhage. Mayo Clin Proc 1998;73:745. 44. Tatter SB, Crowell RM, Ogilvy CS: Aneurysmal and microaneurysmal angiogram negative subarachnoid hemorrhage. Neurosurgery 1995;37:48. 45. Ellenbogen RG, Winston KR, Kupsky WJ: Tumors of the choroid plexus in children. Neurosurgery 1989;25:327. 46. Cardoso ER, Peterson EW: Pituitary apoplexy: A review. Neurosurgery 1984;14:363. 47. Broderick JP, Viscoli CM, Brott T: Major risk factors for aneurysmal subarachnoid hemorrhage in the young are modifiable. Stroke 2003; 34:137. 48. Kernan WN, Viscoli CM, Brass LM, et al: Phenylpropanolamine and the risk of hemorrhagic stroke. N Engl J Med 2000;343:1826. 49. Mayer PL, Awad IA, Todor R, et al: Misdiagnosis of symptomatic cerebral aneurysm. Prevalence and correlation with outcome at four institutions. Stroke 1996;27:1558. 50. Mayberg MR, Batjer HH, Dacey R: Guidelines for the management of aneurysmal subarachnoid hemorrhage. A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Circulation 1994;90:2592. 51. Chiang VLS, Claus EB, Awad IA: Toward more rational prediction of outcome in patients with high-grade subarachnoid hemorrhage. Neurosurgery 2000;46:28. 52. Yoshikai S, Nagata S, Ohara S, Yuhi F, et al: A retrospective analysis of the outcomes of patients with aneurysmal subarachnoid hemorrhages: A focus on the prognostic factors. No Shinkei Geka 1996;24:733.
Hemorrhagic Cerebrovascular Disease
179
53. Hunt WE, Hess RM: Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 1968;28:14. 54. Drake CG: Report of World Federation of Neurological Surgeons Committee on a Universal subarachnoid hemorrhage grading scale. J Neurosurg 1988;68:985. 55. Van Gijn J, Bromberg JE, Lindsay KW, et al: Definition of initial grading, specific events, and overall outcome, in patients with aneurysmal. subarachnoid hemorrhage. A survey. Stroke 1994;25: 1623. 56. Fisher CM, Kistler JP, Davis JM: Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by CT scanning. Neurosurgery 1980;6:1. 57. Crowell RM, Ogilvy CS, Gress DR, et al: General management of aneurysmal subarachnoid hemorrhage. In Ojemann RG, Heros RC, Crowell RM, Ogilvy CS (eds): Surgical Management of Neurovascular Disease, 3rd ed. Baltimore, William and Wilkins, 1996:111– 122. 58. International Subarachnoid Aneurysm Trial (ISAT) Collaborative Group: International subarachnoid aneurysm trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: A randomised trial. Lancet 2002;360:1267. 59. Brilstra EH, Rinkel GJE, van Der Graaf Y, et al: Treatment of intracranial aneurysms by embolization with coils: A systemic review. Stroke 1999;30:470. 60. Raymond J, Roy D: Safety and efficacy of endovascular treatment of acutely ruptured aneurysms. Neurosurgery 1997;41:1235. 61. Raymond J, Roy D, Bojanowski M, et al: Endovascular treatment of acutely ruptured and unruptured aneurysms of the basilar bifurcation. J Neurosurg 1997;86:211. 62. Nichols DA, Brown RD Jr, Thielen KR, et al: Endovascular treatment of ruptured posterior circulation aneurysms using electrolytically detachable coils. J Neurosurg 1997;87:374. 63. Anonymous: Guiglielmi Detachable Coil (GDC) U.S. Clinical Study Summary. Fremont, CA, Target Therapeutics, 1995. 64. Sturaitis MK, Rinne J, Chaloupka JC, et al: Impact of Guglielmi detachable coils on outcomes of patients with intracranial aneurysms treated by a multidisciplinary team at a single institution. J Neurosurg 2000;93:569. 65. Lee S, Huddle D, Awad IA: Which aneurysms should be referred for endovascular therapy? In Howard M, Elliot P (eds): Clinical Neurosurgery, vol. 47. Philadelphia, Lippincott, Williams and Wilkins, 2000:188–220. 66. Ng PY, Huddle D, Gunel M, et al: Intraoperative endovascular adjuncts in the microsurgical treatment of paraclinoid aneurysms of the internal carotid artery. J Neurosurg 2000;93:554. 67. Hassan D, Vermeulen M, Wijdicks EFM, et al: Management problems in acute hydrocephalus after subarachnoid hemorrhage. Stroke 1989;20:747. 68. Graff-Radford N, Torner J, Adams HP, et al: Factors associated with hydrocephalus after subarachnoid hemorrhage. Arch Neurol 1989;46:744. 69. Vermeij FH, Hasan D, Vermeulen M, et al: Predictive factors for deterioration from hydrocephalus after subarachnoid hemorrhage. Neurology 1994;44:1851. 70. Kusske JA, Turner PT, Ojemann GA, et al: Ventriculostomy for the treatment of acute hydrocephalus following subarachnoid hemorrhage. J Neurosurg 1973;38:591. 71. Van Gijn J, Hijdra A, Wijdicks EFM, et al: Acute hydrocephalus after aneurysmal subarachnoid hemorrhage. J Neurosurg 1985;63:355. 72. Bogdahn U, Lau W, Hassel W, et al: Continuous-pressure controlled, external ventricular drainage for treatment of acute hydrocephalusevaluation of risk factors. Neurosurgery 1992;31:898. 73. Rajshekhar V, Harbaugh RE: Results of routine ventriculostomy with external ventricular drainage for acute hydrocephalus following subarachnoid hemorrhage. Acta Neurochir (Wien) 1992;115:8.
W9418-06.qxd
11/3/03 2:50 PM Page 180
180
Section I
Introduction
74. Auer LM, Mokry M: Disturbed cerebrospinal fluid circulation after subarachnoid hemorrhage and acute aneurysm surgery. Neurosurgery 1990;26:804. 75. Proust F, Debono B, Gerardin E, et al: Angiographic cerebral vasospasm and delayed ischemic deficit on anterior part of the circle of Willis. Usefulness of transcranial Doppler. Neurochirurgie 2002;48:489. 76. Awad IA, Barnett GH: Acute management of subarachnoid hemorrhage. Neurosurgical Emergencies. Park Ridge, IL, American Association of Neurological Surgeons, 1994:137–149. 77. Barnwell SL, Higashida RT, Halbach VV, et al: Transluminal angioplasty of intracerebral vessels for cerebral arterial spasm: Reversal of neurological deficits after delayed treatment. Neurosurgery 1989;25:424. 78. Eskridge JM, Newell DW, Pendleton GA: Transluminal angioplasty for treatment of vasospasm. Neurosurg Clin North Am 1990;1:387. 79. Higashida RT, Halbach VV, Cahan LD, et al: Transluminal angioplasty for treatment of intracranial arterial vasospasm. J Neurosurg 1989;71:648. 80. Newell DW, Eskridge JM, Mayberg MR, et al: Angioplasty for the treatment of symptomatic vasospasm following subarachnoid hemorrhage. J Neurosurg 1989;71:654. 81. Kassell NF, Helm G, Simmons N, et al: Treatment of cerebral vasospasm with intra-arterial papaverine. J Neurosurg 1992;77:848. 82. Newell DW, Eskridge JM, Mayberg MR, et al: Angioplasty for the treatment of symptomatic vasospasm following subarachnoid hemorrhage. J Neurosurg 1989;71:654. 83. Linskey ME, Horton JA, Rao GR, et al: Fatal rupture of the intracranial carotid artery during transluminal angioplasty for vasospasm induced by subarachnoid hemorrhage. J Neurosurg 1991;74:985. 84. Fandino J, Schuknecht B, Yüksel C, et al: Clinical, angiographic, and sonographic findings after structured treatment of cerebral vasospasm and their relation to final outcomes. Acta Neurochir (Wien) 1999;141:677. 85. Ferch R, Pasqualin A, Barone G, et al: Surgical management of ruptured aneurysms in the eighth and ninth decades. Acta Neurochir (Wien) 2003;145:439. 86. Pinsker MO, Gerstner W, Wolf S, et al: Surgery and outcome for aneurysmal subarachnoid hemorrhage in elderly patients. Acta Neurochir Suppl 2002;82:61. 87. Lagares A, Gomez PA, Lobato RD, et al: Prognostic factors on hospital admission after spontaneous subarachnoid haemorrhage. Acta Neurochir (Wien) 2001;143(7):665. 88. Seifert V, Raabe A, Stolke D, et al: Management-related morbidity and mortality in unselected aneurysms of the basilar trunk and vertebrobasilar junction. Acta Neurochir (Wien) 2001;143(4):343–348; discussion 348. 89. Hijdra A, van Gijn J, Nagelkerke NJ, et al: Prediction of delayed cerebral ischemia, rebleeding, and outcome after aneurysmal subarachnoid hemorrhage. Stroke 1988;19:1250. 90. Lanzino G, Kassell NF, Germanson TP, et al: Age and outcome after aneurysmal subarachnoid hemorrhage: Why do older patients fare worse? J Neurosurg 1996;85(3):410. 91. B M Saciri BM, Kos N: Aneurysmal subarachnoid haemorrhage: Outcomes of early rehabilitation after surgical repair of ruptured intracranial aneurysms. J Neurol Neurosurg Psychiatry 2002;72:334. 92. O’Dell MW, Watanabe TK, De Roos ST, et al: Functional outcome after inpatient rehabilitation in persons with subarachnoid hemorrhage. Arch Phys Med Rehabil 2002;83:678. 93. Brott T, Thalinger K, Hertzberg V: Hypertension as a risk factor for spontaneous intracerebral hemorrhage. Stroke 1986;17:1078. 94. Juvela S, Hillbom M, Palomäki H: Risk factors for spontaneous intracerebral hemorrhage. Stroke 1995;26:1558. 95. Wakai S, Nagai M: Histological verification of microaneurysms as a cause of cerebral hemorrhage in surgical specimens. J Neurol Neurosurg Psychiatry 1989;52:595.
96. Newton TH, Potts DG, eds: Radiology of the skull and brain. St. Louis, Mosby, 1971. 97. Kaufman HH: Spontaneous intracerebral hematoma. In Grossman R (ed): Clinical Neurosciences, 2nd ed. New York, Raven, 1990. 98. Gilles C, Brucher JM, Khoubesserian P, et al: Cerebral amyloid angiopathy as a cause of multiple intracerebral hemorrhages. Neurology 1984;34:730. 99. Mandybur TI: Cerebral amyloid angiopathy: The vascular pathology and complications. J Neuropathol Exp Neurol 1986;45:79. 100. Vonsattel JP, Myers RH, Hedley-White ET, et al: Cerebral amyloid angiopathy without and with cerebral hemorrhages: A comparative histological study. Ann Neurol 1991;30:637. 101. Ropper AH, Davis KR: Lobar cerebral hemorrhages: Acute clinical syndromes in 26 cases. Ann Neurol 1980;8:141. 102. Gorelick PB, Kelly MA: Ethanol. In Feldman E (ed): Intracerebral Hemorrhage. Armork, NY, Futura, 1994:195–208. 103. Blackshear JL, Kopecky SL, Litin SC, et al: Management of atrial fibrillation in adults: Prevention of thromboembolism and symptomatic treatment. Mayo Clin Proc 1996;71:150. 104. Fihn SD, McDonell M, Martin D, et al: Risk factors for complications of chronic anticoagulation: A multicenter study. Ann Intern Med 1993;118:511. 105. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group: Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333:1581. 106. Aldrich MS, Sherman SA, Greenberg HS: Cerebrovascular complications of streptokinase infusion. JAMA 1985;253:1777. 107. Grines CL, Browne KF, Marco J, et al: A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993;328:673. 108. The Steering Committee of the Physician’s Health Study Group: Preliminary report: Findings from the aspirin component of the ongoing physician’s health study. N Engl J Med 1988;318:262. 109. Greenberg SM, Edgar MA: Hemorrhage in a 69-year old woman receiving warfarin. Case records of the Massachusetts General Hospital. N Engl J Med 1996;335:189. 110. Hartmann A, Mast H, Mohr JP, et al: Morbidity of intracranial hemorrhage in patients with cerebral arteriovenous malformation. Stroke 1998;29:931. 111. Nowak G, Schwachenwald D, Schwachenwald R, et al: Intracerebral hematomas caused by aneurysm rupture. Experience with 67 cases. Neurosurg Rev 1998;21:5. 112. Shende MC, Lourie H: Sagittal sinus thrombosis related to oral contraceptives: Case report. J Neurosurg 1970;33:714. 113. Symonds CP: Otitic hydrocephalus. Brain 1931;54:55. 114. Garcia RDJ, Baker AS, Cunningham MJ, et al: Lateral sinus thrombosis associated with otitis media and mastoiditis in children. Pediatr Infect Dis J 1995;14:617. 115. Estanol B, Rodriguez A, Conte G, et al: Intracranial venous thrombosis in young women. Stroke 1979;10:680. 116. Singh T, Chakera T: Dural sinus thrombosis presenting as unilateral lobar haematomas with mass effect: An easily misdiagnosed cause of cerebral haemorrhage. Australas Radiol 2002;46:351. 117. Kalfas IH, Little JR: Postoperative hemorrhage: A survey of 4992 intracranial procedures. Neurosurgery 1988;23:343. 118. Fukamachi A, Koizumi H, Nukui H: Postoperative intracerebral hemorrhages: A survey of computed tomographic findings after 1074 intracranial operations. Surg Neurol 1985;23:575. 119. Haines SJ, Maroon JC, Jannetta PJ: Supratentorial intracerebral hemorrhage following posterior fossa surgery. J Neurosurg 1978;49:881. 120. Harders A, Gilsbach J, Weigel K: Supratentorial space-occupying lesions following infratentorial surgery: Early diagnosis and treatment. Acta Neurochir (Wien) 1985;74:57. 121. Caplan LR, Skillman J, Ojemann R, et al: Intracerebral hemorrhage following carotid endarterectomy: A hypertensive complication. Stroke 1979;9:457.
W9418-06.qxd
11/3/03 2:50 PM Page 181
Chapter 6 122. Lowenstein DH, Collins SD, Massa SM, et al: The neurologic complications of cocaine abuse. Neurology 1987;37S1:195. 123. Levine S: Cocaine and stroke. Current concepts of cerebrovascular disease. Stroke 1987;22:25. 124. Harrington H, Heller A, Dawson D, et al: Intracerebral hemorrhage and oral amphetamines. Arch Neurol 1983;40:503. 125. Freeze TE, Miotto K, Reback CJ: The effects and consequences of selected club drugs. J Substance Abuse Treat 2002;23:151. 126. Levine SR, Brust JCM, Futrell N, et al: Cerebrovascular complications of the use of the crack form of alkaloid cocaine. N Engl J Med 1990;323:699. 127. Kernan WN, Viscoli CM, Brass LM: Phenylpropanolamine and the risk of hemorrhagic stroke. N Engl J Med 2000;343:1826. 128. Hornig CR, Dorndorf W, Agnoli AL: Hemorrhagic cerebral infarction: A prospective study. Stroke 1986;17:179. 129. Scott M: Spontaneous intracerebral hematoma caused by cerebral neoplasm. J Neurosurg 1975;42:338. 130. Dublin AB, Norman D: Fluid-fluid level in cystic cerebral metastatic melanoma. J Comput Assist Tomogr 1979;3:650. 131. Weir B, MacDonald N, Mielke B: Intracranial vascular complications of choriocarcinoma. Neurosurgery 1978;2:138. 132. Broderick JP, Brott TG, Duldner JE, et al: Volume of intracerebral hemorrhage: A powerful and easy-to-use predictor of 30-day mortality. Stroke 1993;24:987. 133. Kothari RU, Brott TG, Broderick JP, et al: The ABCs of measuring intracerebral hematoma volumes. Stroke 1996;27:1304. 134. Adams HP Jr, Brott TG, Furlan AJ, et al: Guidelines for thrombolytic therapy for acute stroke: A supplement to the guidelines for the management of patients with acute ischemic stroke: A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Circulation 1996;94:1167. 135. Broderick JP, Brott TG, Tomsick T, et al: Ultraearly evaluation of intracerebral hemorrhage. J Neurosurg 1990;72:195. 136. Waidhauser E, Hamburger C, Marguth F: Neurosurgical management of cerebellar hemorrhage. Neurosurg Rev 1990;13:211. 137. Ogilvy CS, Stieg PE, Awad IA: Recommendation for the management of intracranial arteriovenous malformations. A Statement for Healthcare Professionals From a Special Writing Group of the Stroke Council, American Stroke Association. Stroke 2001;32:1458. 138. Kase C, Mohr J, Caplan R, et al: Intracerebral hemorrhage. In Barnett H, Mohr J, Stein B, Yatsu F (eds): Stroke: Pathophysiology, Diagnosis, and Management. New York, Churchill Livingstone, 1992:561–616. 139. Broderick J, Brott T, Zuccarello M: Management of intracerebral hemorrhage. In Batjer H (ed): Cerebrovascular Disease. Philadelphia, Lipincott-Raven, 1996:1–18. 140. Auer LM, Auer T, Sayama I: Indications for surgical treatment of cerebellar hemorrhage and infarction. Acta Neurochir 1986;79:74. 141. Kanaya H: All Japan cooperative study on the treatment of hypertensive intracerebral hemorrhage. Jpn J Stroke 1990;12:509.
Hemorrhagic Cerebrovascular Disease
181
142. Batjer HH, Reisch JS, Allen BC, et al: Failure of surgery to improve outcome in hypertensive putaminal hemorrhage: A prospective randomized trial. Arch Neurol 1990;47:1103. 143. Auer L, Deinsberger W, Niederkorn K, et al: Endoscopic surgery versus medical treatment for spontaneous intracerebral hematoma: A randomized study. J Neurosurg 1989;70:530. 144. McKissock W, Richardson A, Taylor J: Primary intracerebral hemorrhage: A controlled trial of surgical and conservative treatment in 180 unselected cases. Lancet 1961;2:222. 145. Fayad PB, Awad IA: Surgery for intracerebral hemorrhage. Neurology 1998;51S:69. 146. Montes JM, Wong JH, Fayad PB, et al: Stereotactic computed tomographic-guided aspiration and thrombolysis of intracerebral hematoma: Protocol and preliminary experience. Stroke 2000;31:834. 147. Miller DW, Barnett GH, Kormos DW, et al: Stereotactically guided thrombolysis of deep cerebral hemorrhage: Preliminary results. Cleve Clin J Med 1993;60:321. 148. Schaller C, Rhode V, Meyer B, et al: Stereotactic puncture and lysis of spontaneous intracerebral hemorrhage using recombinant tissueplasminogen activator. Neurosurgery 1995;36:328. 149. Tzaan WC, Lee ST, Lui TN: Combined use of stereotactic aspiration and intracerebral streptokinase infusion in the surgical treatment of hypertensive intracerebral hemorrhage. J Formos Med Assoc 1997;96:962. 150. Lippitz BE, Mayfrank L, Spetzger U, et al: Lysis of basal ganglia hematoma with recombinant tissue plasminogen activator (rTPA) after stereotactic aspiration: Initial results. Acta Neurochir (Wien) 1994;127:157. 151. Ojemann RG, Heros RC: Spontaneous brain hemorrhage. Stroke 1983;14:468. 152. Poungvarin N, Bhoopat W, Viriyavejakul A, et al: Effect of dexamethasone in primary supratentorial intracerebral hemorrhage. N Engl J Med 1987;316:1229. 153. Broderick JP, Brott TG, Tomsick T, et al: Intracerebral hemorrhage more than twice as common as subarachnoid hemorrhage. J Neurosurg 1993;78:188. 154. Hallevy C, Ifergane G, Kordysh E: Spontaneous supratentorial intracerebral hemorrhage. Criteria for short-term functional outcome prediction. J Neurol 2002;249:1704. 155. Daverat P, Castel JP, Dartigues JF, et al: Death and functional outcome after spontaneous intracerebral hemorrhage. A prospective study of 166 cases using multivariate analysis. Stroke 1991;22:1. 156. Poungvarin N, Viriyavejakul A: Spontaneous supratentorial intracerebral hemorrhage: A prognostic study. J Med Assoc Thai 1990;73:206. 157. Tuhrim S, Dambrosia JM, Price TR, et al: Prediction of intracerebral hemorrhage survival. Ann Neurol 1988;24:258.
W9418-07.qxd 10/31/03 3:39 PM Page 183
Chapter 7 Basic Endovascular Neurosurgery and Neuroradiology Matthew V. Burry, MD and Robert A. Mericle, MD
Introduction The endovascular approach to cerebrovascular lesions is becoming an increasingly attractive treatment option. Endovascular techniques and technologies continue to evolve and improve rapidly, pushing the capabilities of the neuroendovascular surgeon further. Today, endovascular neurosurgery is an essential component of any comprehensive neurosurgical service and an exciting area for subspecialization. It is essential that neuro-critical care physicians understand the specific risks, complications, and other postoperative issues of patients treated by a neuroendovascular approach. In this chapter, the current state of endovascular neurosurgery is reviewed with particular emphasis on the details most pertinent for critical care physicians who may care for these patients postoperatively. The chapter is organized by disease and covers the nine diseases most commonly treated by endovascular neurosurgeons. Each disease section is divided into three parts. The first part broadly reviews the natural history of the condition and discusses the indications for the endovascular procedures. The second part describes some of the pertinent detail of the endovascular approaches for each disease. Finally, each section ends with a discussion of the possible intraoperative and postoperative complications from the endovascular procedures. Because this field is rapidly evolving, with many new technologies and approaches every year, a completely current review is difficult to accomplish. Every effort was made to include the most current information and references.
A team approach is optimal for providing the best health care for these challenging patients. Neurocritical care physicians are an essential component of the neurosurgical team. This chapter details the most commonly treated neurovascular diseases in the depth and breadth necessary for a neuro-critical care physician to increase his or her understanding of the disease process, endovascular indications, approaches, and to anticipate and manage most complications that can occur when treating these difficult diseases.
Endovascular Surgery for Acute Thromboembolic Stroke Review Stroke is the second most common cause of mortality worldwide, causing an estimated 5.1 million deaths annually.1 At least 80% of ischemic strokes are caused by acute thromboembolic cerebral artery occlusion.2 The arterial distribution and duration of local cerebral ischemia determine how large an infarct will become.3 The goal of acute stroke treatment is to preserve tissue affected by potentially reversible ischemia.4 Currently, the only U.S. Food and Drug Administration (FDA)-approved treatment for stroke is the emergent administration of intravenous (IV) recombinant tissue plasminogen activator (r-tPA) for thrombolysis of clots that have been symptomatic for less than 3 hours. Effective therapeutic alternatives and longer treatment windows are needed. Morbidity and mortality from 183
W9418-07.qxd 10/31/03 3:39 PM Page 184
184
Section I
Introduction
“routine” treatment of acute middle cerebral artery (MCA) occlusion are extremely high. This mortality can be approximated by the placebo groups of two of the recent randomized clinical trials on stroke treatment.5,6 The National Institute of Neurological Disorders and Stroke (NINDS) study group and the Intra-arterial prourokinase for acute ischemic stroke (PROACT II) study demonstrated that the mortality rate at 3 months after MCA occlusion without thrombolytic therapy was 21% and 27%, respectively.5,6 Initially, IV r-tPA seemed to hold great promise. The NINDS trial showed clinical benefit at 90 days when intravenous r-tPA was administered less than 3 hours after symptom onset. The fact that there was no significant benefit seen at 24 hours suggests that the large MCA clot did not dissolve immediately, but rather there was an improvement in smaller, nearby occluded arteries that saved some of the penumbra of the evolving stroke. Unfortunately, the efficacy of intravenous r-tPA was found to be very time-dependent. Six other randomized clinical trials using intravenous thrombolytic therapy on patients after 3 hours of symptom onset demonstrated an increased rate of symptomatic intracerebral hemorrhage (ICH), and subsequently, no benefit from the intravenous treatment.7–12 There has not yet been a randomized clinical trial demonstrating IV r-tPA to be beneficial in patients with large vessel occlusions. Secondary analysis of data from the NINDS randomized clinical trial showed that patients with a dense MCA sign on computed tomography (CT), which is indicative of thrombus within this vessel, have an extremely poor response after IV thrombolysis, with only one of 18 demonstrating a positive outcome.13 Other studies have shown IV r-tPA alone does not open major arterial occlusion during the first few hours.14 Because IV r-tPA has a limited therapeutic window and limited efficacy in treating MCA trunk occlusions, there have been investigations into local, intra-arterial infusion of thrombolytic agents at the site of the cerebral arterial occlusion.15,16 Local intra-arterial thrombolysis fully or partially recanalizes occluded cerebral arteries in approximately 50% to 85% of patients.5,15,16 The rate of symptomatic intracerebral hemorrhage after intra-arterial thrombolysis has been acceptable in these studies. The two largest randomized clinical trials of intra-arterial thrombolysis are the PROACT I and PROACT II studies, which compared intra-arterial recombinant pro-urokinase plus IV heparin to placebo treatment plus IV heparin.5,15 These studies doubled the effective treatment window for thrombolysis from 3 to 6 hours, when the thrombolytic is given intra-arterially by a catheter embedded in the clot. Despite inclusion of patients with more severe and disabling strokes, the PROACT I showed a recanalization rate of 58% after 2 hours of infusion of pro-urokinase and heparin compared with 14% after an infusion of placebo and heparin. PROACT II had recanalization rates in the pro-urokinasetreated group of 66% compared to 18% in the control group.
Excellent neurologic outcomes were more common after intra-arterial thrombolysis. In PROACT II, 40% of the experimental group achieved a modified Rankin score of two or less (excellent neurologic outcome). This was significantly higher than the control group, where only 25% of the patients achieved that outcome. The rates of symptomatic ICH and mortality were not significantly different between the experimental and control groups.5 Intra-arterial thrombolysis appears to achieve higher rates of recanalization than IV r-tPA, with hemorrhage rates that are similar to those in the NINDS stroke trial. It has been suggested by many experts that mechanical clot disruption may increase the recanalization rate and improve neurologic outcome compared to thrombolysis alone.17–19 Randomized controlled trials testing mechanical clot disruption should begin in the near future. Other ongoing clinical trials are testing new thrombolytic drugs. The most promising of these is the Fab antibody fragment Abciximab (ReoPro, Eli Lilly, Indianapolis, IN).20 Endovascular Approaches An important principle affecting the outcome of stroke patients is emergent detection and treatment. All physicians and lay people should know both the signs of acute stroke and that early intervention can save lives. All critical care physicians should know the specific risk factors for thromboembolic stroke in all their patients. Some of the most significant risk factors for stroke are smoking, hypertension, coronary artery disease, hypercholesterolemia, diabetes mellitus, and obesity. Additionally, any patient undergoing any cerebrovascular or neuroendovascular procedure is at increased risk for stroke perioperatively. The physician taking care of neuro-critical care patients should be aware of specific periprocedural risk factors in their patients. Recently placed aneurysm clips can cause thrombotic stroke. Any recent endovascular procedure can cause stroke in the early postoperative period. This is especially true if detachable coils, embolic agents, or stents were used. With any neurologic change in a patient, the possibility of thromboembolic stroke must be considered, and this event should be treated emergently. A CT scan and neurologic/neurosurgical consultation should be obtained; if a thromboembolic stroke seems likely, computed tomographic angiography (CTA), magnetic resonance angiography (MRA), or catheter cerebral angiography is necessary for an accurate anatomic diagnosis of acute thromboembolism. If the patient is considered a good candidate for intra-arterial thrombolysis, catheter cerebral angiography is performed emergently after the initial noncontrast CT scan. If an acute clot is found on catheter angiography, various endovascular interventions are possible. As mentioned previously, the most thoroughly studied technique is to place the tip of a microcatheter into the proximal clot and inject
W9418-07.qxd 10/31/03 3:39 PM Page 185
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
185
thrombolytics through the catheter and directly into the clot (Fig. 7-1) Although pro-urokinase is the most carefully studied thrombolytic agent for intra-arterial use, it is not currently commercially available. Therefore, we currently use r-tPA for this procedure. At our institution, we slowly inject 2 mg over 10 to 20 minutes. This dose may be repeated every 20 minutes if the major arterial occlusion persists until a maximum of 20 mg or until the patient has exceeded the
6-hour limit after the onset of symptoms. We also commonly use abciximab (ReoPro) as an adjunct to thrombolysis. Our protocol for abciximab is an intra-arterial bolus of 0.25 mg/kg followed by an intravenous infusion of 0.125 mcg/kg/min for a total of 12 hours. If recanalization is not achieved with pharmacologic agents, we frequently use mechanical disruption of the clot with wires, catheters, balloons, or saline injection devices (Fig. 7-2).
A
B
Figure 7-1. Digital subtraction angiography in the anteroposterior view of a patient who presented with left-sided hemiplegia. A, Initial angiographic imaging revealed complete occlusion of the right middle cerebral artery (MCA) and the right anterior cerebral artery (ACA). B, A microcatheter was implanted into the occluding clot of the MCA and intra-arterial thrombolytics were infused, resulting in excellent MCA recanalization. Note the persistent ACA occlusion. C, The microcatheter was moved to the ACA and intra-arterial thrombolysis was performed with an excellent result. The patient’s neurologic examination improved almost immediately and was intact to gross neurologic examination several days later.
C
W9418-07.qxd 10/31/03 3:39 PM Page 186
186
Section I
Introduction
B A Figure 7-2. Oblique view super-selective digital subtraction angiography of a patient who suffered a thromboembolic stroke. A, Selective middle cerebral artery (MCA) injection angiography showing an occlusion of two branches in the distal MCA territory. A microwire is in place in preparation for mechanical thrombolysis. B, After mechanical thrombolysis, partial recanalization of the anterior M3 branch and improvement in the posterior M3 branch was achieved.
Complications The most dreaded complication from endovascular treatment of acute stroke is an ICH, either from a hemorrhagic transformation of an infarct or vessel rupture during the procedure. Symptomatic ICHs occurred in 15% of PROACT patients and 10% of PROACT II patients who received intra-arterial pro-urokinase.5,15 To properly care for patients suspected of having an ICH, it is imperative that the critical care physician knows the technical and pharmacologic details of the procedure. The treating physician must have a clear sense of the patient’s current thrombolytic and coagulative state to effectively treat any possible hemorrhagic complications. The timing of the onset of the stroke, the onset of therapy, and any neurologic improvement or deterioration should be noted.
Endovascular Surgery for Intracranial Aneurysms Review Tremendous advances have been made during the past two decades in the endovascular treatment of intracranial aneurysms. With the introduction of electrolytically detachable platinum coils in 1991, an effective and safe method of endovascular aneurysm embolization became available. The
coils, Guglielmi detachable coil (GDC) (Target Therapeutics [Boston Scientific], Fremont, CA), are FDA approved and have rapidly spread to worldwide use.21–23 The endovascular placement of GDC is by far the most popular endovascular treatment of intracranial aneurysms. GDC placement is now the primary alternative for patients with intracranial aneurysms who are not good candidates for craniotomy and microsurgical clipping. Surgical clipping is still the gold standard for many intracranial aneurysms, but with further refinement of endovascular materials and techniques, the number of aneurysms treated with the minimally invasive endovascular approaches will continue to increase. Several other companies now have released their own version of endovascular detachable coils approved for embolization of intracranial aneurysms. It is essential for critical care physicians to understand coiling techniques, effectiveness, and potential complications. GDCs are thin platinum microcoils that are attached to a stainless steel introducing wire. They are advanced through a microcatheter that has been previously placed in the aneurysm. The coil is advanced into the aneurysm and it assumes a shape that is dependent both on the shape of the lumen of the aneurysm and the intrinsic shape of the coil. Several coil shapes are available, including standard, twodimensional and three-dimensional designs. While the coil is being placed into the aneurysm, it remains attached to the stainless steel introducing wire. Once placed to the satisfac-
W9418-07.qxd 10/31/03 3:39 PM Page 187
Chapter 7
tion of the endovascular surgeon, the coil can be electrolytically detached by placing a low voltage current (1 to 2 mA) that dissolves the attachment site of the coil to the wire introducer. This occurs because the attachment site undergoes electrolysis, but the stainless steel introducing wire and the platinum coil resist electrolysis. Multiple coils of varying size, stiffness, and shape can be sequentially placed in the aneurysm. The aneurysm is gradually filled from the outside in. Ideally, the entire aneurysm lumen can be occluded with the coils and associated thrombus (Fig. 7-3). The use of intra-operative anticoagulation is at the physician’s discretion. At our institution, all patients with unruptured aneurysms are given a 100 U/kg bolus of heparin before the placement of the first coil. In ruptured aneurysms, we administer heparin at the same dose only after the successful placement of the first coil. This may reduce the risk of aneurysm rupture, which is more common during the first coil placement. Also, the first coil can help improve the initiation of thrombosis in the aneurysms, if it is placed before systemic anticoagulation. Unfortunately, GDC placement does not achieve complete, durable occlusion of all intracranial aneurysms. One recent study has shown a 20% rate of incomplete occlusion and an 8.6% rate of delayed recanalization at 3 years’ follow-
Figure 7-3. Digital subtraction angiography in the anteroposterior view of a patient who has had two ophthalmic aneurysms treated with Guglielmi detachable coils (GDCs). The patient’s aneurysms were discovered secondary to visual loss. One aneurysm is occluded with GDCs, with no contrast entering the aneurysm lumen (right). The other aneurysm is not yet completely occluded with the GDCs. Notice the contrast entering the aneurysm lumen between the coils and at the dome of the aneurysm (left).
Basic Endovascular Neurosurgery and Neuroradiology
187
up.22 Despite these incomplete aneurysm embolizations, there is likely a protective effect from the incomplete coiling. Although this has not been proven in a clinical trial, few incompletely occluded aneurysms have hemorrhaged after coiling.23 Endovascular Approaches The decision to treat an aneurysm with either microsurgical clipping or endovascular coiling is complex and controversial. A complete discussion is beyond the scope of this chapter, but a few critical points are important for critical care physicians to understand. The decision to coil or clip an aneurysm is based primarily on four factors: 1. Aneurysm Geometry. Often, the decision to coil or clip an aneurysm is made on the basis of aneurysm geometry. The endovascular placement of coils favors an aneurysm shape that has a high fundus-to-neck ratio. The fundus is defined as the widest part of the body of the aneurysm. Most coils prefer a spherical shape and thus, the more closely an aneurysm approximates a spherical shape, the more natural fit for most coils. Aneurysms with complex shapes or wide necks are more difficult to manage, and once placed, the coils are less stable. Loops of coils can extrude out of the aneurysm into the parent vessel. This is highly undesirable and can lead to thromboembolic complications.24 Endovascular techniques have been developed to improve the coiling of difficult aneurysms. These include balloon assisted and stent assisted coiling.25,26 Briefly, these techniques involve the temporary placement of a balloon or permanent placement of a stent into the parent artery during coiling to keep the coil mass stable in the aneurysm lumen. These additional procedures add to the complexity and risk of the operation. 2. Aneurysm Size. Size decreases the safety and efficacy of coil embolization at both of the extremes of measurement. Very small aneurysms (< 3 mm) and giant aneurysms (≥ 25 mm) are much more difficult to occlude with endovascular coiling. Very small aneurysms are dangerous to catheterize, because there is an extremely limited safety margin for placement of the intraluminal catheter and the coils. Giant aneurysms are difficult to treat with endovascular coiling because of the very large intraluminal volume and frequent association with intraluminal thrombus. Many coils are required, and often it is not possible to fill the entire aneurysm. Also, long-term follow-up has not been as favorable for giant aneurysms. Late recanalization of the aneurysmal dome is more common in giant aneurysms and late re-bleeding rates are higher.27 3. Aneurysm Location. The risk of endovascular coiling versus microsurgical clipping varies with aneurysm location. Currently at our institution, most basilar tip
W9418-07.qxd 10/31/03 3:39 PM Page 188
188
Section I
Introduction
aneurysms, small paraclinoid internal carotid artery aneurysms, and other posterior circulation aneurysms are treated with coiling. These locations are generally believed to have increased risk associated with microsurgical clipping. Conversely, there are other anatomical regions where coiling is more difficult. Aneurysms that are particularly difficult for endovascular coils are most middle cerebral artery aneurysms and some anterior communicating artery aneurysms. 4. Clinical Status of Patient. In deciding on the appropriate therapy for aneurysm patients, it is essential to consider the overall neurologic status and the general health of the patient. Several grading systems have been developed to standardize the clinical classification of patients after subarachnoid hemorrhage. The commonly used grading systems are the Hunt and Hess grading system and the World Health Organization (WHO) system. These are covered more fully in the chapter of this text on subarachnoid hemorrhage. The general assumption is that in patients with a poor clinical grade, endovascular surgery is preferred over craniotomy with microsurgical clipping. There is considerable variation from institution to institution concerning the importance of these four factors in the choice of treatment depending on the particular expertise and experience of the treating physicians. In all institutions, the decision should be made by an interdisciplinary neurovascular team with expertise in endovascular and microsurgical aneurysm treatment. Complications The most dreaded intraoperative complication with endovascular surgery is aneurysmal rupture. This complication is far more likely to occur in previously ruptured aneurysms than in an unruptured aneurysm. Intraprocedural rupture can be either a spontaneous rupture unassociated with aneurysmal manipulation or it can be secondary to perforation of the aneurysm with a microwire, microcatheter, or coil. Aneurysmal rupture should be suspected with any sudden increase in systolic blood pressure, decrease in heart rate, increase or decrease in respirations, or deterioration in neurologic status. In patients with external ventricular drainage, the cerebrospinal fluid will often change to bright red. Conscious patients will complain of the worst headache of their life and possibly deteriorate neurologically. Angiographically, contrast material extravasation from the aneurysm lumen is seen (Fig. 7-4A). If the rupture is caused by perforation from one of the endovascular instruments, the offending microwire or microcatheter should not be withdrawn until the aneurysm is secured with either further coil deployment or rapid injection of N-butyl cyanoacrylate glue (NBCA) (Cordis, Miami Lakes, FL) or other embolic agent (Fig. 7-4B).
An intraoperative rupture is an immediate, lifethreatening emergency. If heparin has been administered, it should be immediately and completely reversed with protamine sulfate (Eli Lilly, Indianapolis, IN). At least 1 mg of protamine sulfate should be given for each 100 units of heparin for reversal. Next, the hemorrhage must be controlled by emergently embolizing blood flow to the aneurysms. This is usually most quickly performed by emergently placing two or three more coils (Fig. 7-4A and B) In cases of neurologic deterioration without external ventricular drainage, an emergent CT scan and ventriculostomy should be performed as soon as the bleeding has been controlled. Thromboembolism is a risk both during coil placement and in the postoperative period. This is the most frequent complication, occurring in 2.5% to 5% of GDC procedures.22,23,28 Thromboembolism can occur at any time during the endovascular procedure, but is especially likely while manipulating the coil or while withdrawing a misplaced coil (Fig. 7-5) It is important to remember that there is a small risk of thromboembolism from the coil mass for several weeks after the procedure. This is very uncommon, however, unless coils have herniated into the parent artery.24 If thromboembolism has occurred, intra-arterial thrombolysis should be performed as described in the section of this chapter on the endovascular treatment of acute stroke. The clot may be treated with r-tPA, ReoPro, or mechanical clot disruption with microwires, catheters, balloons, or saline injection devices (Fig. 7-6). Extreme care should be taken in the setting of a thromboembolus with a still unsecured ruptured aneurysm, because the thrombolytic agents could lead to a severe repeat hemorrhage. In this case, the aneurysm should be quickly secured with coils until it is adequately protected for the emergent thrombolytic procedure. Although a complete discussion of endovascular coil placement strategies and complications is beyond the scope of this chapter, knowledge of a few key points is important for critical care physicians treating postoperative coiling patients. First, coils may sometimes break or migrate from the coil mass. Broken or migrated coils can cause potentially serious ischemic events. Coils that have broken or migrated from the aneurysm body may be retrieved with endovascular snares, surgical removal, or stenting the coil against the artery wall. After being stented against the vessel wall, these coils presumably will become endothelialized to the wall.29 Sometimes a significant portion of the coil mass can protrude into the parent vessel. Attempts can be made to push this mass back into the aneurysm with angioplasty or stent placement, but success varies. Often with these complications, despite salvage efforts, one is left with a less than ideal situation in which some part of a coil is exposed in the parent artery. Most experts initiate long-term anticoagulant or antiplatelet therapy in this situation. All physicians managing the postoperative care of these patients should be aware of the details of the procedure and any complications, and the efforts to correct them.
W9418-07.qxd 10/31/03 3:39 PM Page 189
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
B A Figure 7-4. Towne’s view digital subtraction angiography (DSA) of a patient who presented with a subarachnoid hemorrhage from a basilar tip aneurysm. During the coiling procedure, acute severe hypertension developed with sudden bright red blood from the ventriculostomy. A, DSA confirmed extravasation outside of the aneurysm lumen. B, After emergently packing two more coils in the aneurysm lumen, repeat DSA shows no further contrast extravasation. Notice the difference between these two figures to appreciate the extraluminal contrast.
B A Figure 7-5. Lateral view digital subtraction angiograpyh of a patient with an MCA aneurysm who suffered an embolic complication after the removal of a coil during Guglielmi detachable coil (GDC) treatment. A, Note the clot in an inferior and posterior M3 branch of the MCA. B, The clot was treated by mechanical thrombolysis with a microwire. Notice improved blood flow to the area that previously was occluded.
189
W9418-07.qxd 10/31/03 3:39 PM Page 190
190
Section I
Introduction
B
A
C
Figure 7-6. Anteroposterior view digital subtraction angiography (DSA) of a patient with a giant internal carotid artery bifurcation aneurysm. A, During Guglielmi detachable coil treatment, the patient had a middle cerebral artery (MCA) occlusion secondary to an extruded intraluminal thrombus from the coil mass. B, The occlusion was treated with microwire mechanical thrombolysis. C, DSA after MCA angioplasty. Notice the improved flow to the M1 segment, but still no distal MCA filling. The two markers in the MCA represent the ends of a compliant angioplasty balloon and excellent distal flow.
Endovascular Surgery for Cerebral Vasospasm Review Cerebral vasospasm is a frequent and devastating condition following subarachnoid hemorrhage (SAH). Delayed ischemic neurologic deficits secondary to cerebral vasospasm will develop in as many as one third of patients who survive the initial hemorrhage.30 Cerebral vasospasm
has been shown to lead to a 1.5- to threefold increase in mortality at 2 weeks following SAH.31 Cerebral vasospasm occurs most often between 4 and 12 days after the initial hemorrhage.32 During this period, vasospasm often presents with focal neurologic deficits, decreased level of consciousness, confusion, headache, and meningismus. In patients with secured aneurysms, after the initial subarachnoid hemorrhage, cerebral vasospasm is the leading cause of death and disability.33 Recent MRI studies have shown that vasospasm might also be the cause of many strokes that do not cause
W9418-07.qxd 10/31/03 3:39 PM Page 191
Chapter 7
gross neurologic dysfunction detectable during the patient’s initial stay in the intensive care unit, but are a cause of the more subtle memory and personality changes frequently seen in these patients.34 Vasospasm is the narrowing of the cerebral arteries in delayed response to blood products in the subarachnoid space. The primary arteries affected are usually the large conductive arteries at the skull base: the supraclinoid segments of the internal carotid arteries, the proximal MCAs, anterior cerebral arteries, posterior cerebral arteries, the basilar artery, and the intradural vertebral arteries. Additionally, there has been a reported 10% incidence of focal distal vasospasm.35 The stenosis from cerebral vasospasm leads to decreased focal cerebral blood flow. This effect is predicted by the Hagen-Poiseuille equation: Q=
DPpr 4 8Lh
In this equation, Q is blood flow, P is the pressure in the artery, r is the radius or the artery, L is the length of the artery, and h is the viscosity of the blood. Blood flow varies with the fourth power of the radius; hence, even modest changes in vessel diameter can have profound effects on cerebral blood flow. Cerebral vasospasm has been shown in several randomized clinical studies to lead to cerebral infarction causing increased morbidity and mortality.36 Intensive basic and clinical research over the past 50 years has led to many advances in the detection and treatment of cerebral vasospasm. The diagnosis of vasospasm and the numerous medical therapies, including nimodipine, socalled triple H therapy (hypertension, hypervolemia, and hemodilution) and their complications, are beyond the scope of this chapter and are covered elsewhere in this book. Instead, the remainder of this chapter will cover the endovascular treatment of cerebral vasospasm, endovascular complications, and postoperative care. The two most important endovascular therapies will be discussed: pharmacologic and mechanical balloon angioplasty. Endovascular Approaches Two endovascular procedures for cerebral vasospasm are currently used: local intra-arterial infusion of a vasodilating drug and intracranial transluminal angioplasty. Both procedures aim to improve blood flow to areas of the brain that are suffering decreased blood flow secondary to severe vasospasm. These procedures are usually reserved for patients who have symptoms with neurologic deficits; therefore, these procedures are usually considered emergent. This factor of emergency timing cannot be overemphasized. Patients who have symptomatic vasospasm are in danger of irreversible cerebral infarction. The quicker blood flow is improved, the better the outcome. This has been demonstrated by a recent study showing a 2-hour window for blood
Basic Endovascular Neurosurgery and Neuroradiology
191
flow restoration for maximum improvement in clinical outcome.37 In a patient thought to be symptomatic from vasospasm, maximal medical therapy is instituted. If the patient’s deficit is refractory, the patient should be taken emergently to the endovascular suite for digital subtraction angiography and possible endovascular therapy. Pharmacologic vasodilation is performed by the local, intra-arterial infusion of a vasodilating drug. The vasodilating drug most frequently used is papaverine. Papaverine is a cyclic nucleotide phosphodiesterase inhibitor that increases cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) and leads to smooth muscle relaxation and arterial vasodilation. The reported half-life of papaverine is 45 to 60 minutes, although some sources think that it is closer to 24 hours.38 Following diagnostic angiography confirmation of vasospasm, a microcatheter is placed in the affected artery. Papaverine is usually injected in 60-mg increments up to 300 mg per major artery distribution, including the MCA, anterior cerebral artery, and posterior cerebral artery distributions. The medication given is usually diluted to 300 mg/100 cc administered over 30 to 60 minutes. This protocol has been shown to consistently dilate proximal and distal cerebral arteries.39 There are numerous difficulties with the intra-arterial administration of papaverine. When mixed with blood, both a 3% and a 0.3% papaverine solution sometimes precipitate out of solution. This nearly always occurs when the papaverine is mixed with heparinized saline. If any significant precipitation occurs, the papaverine crystals could lodge in the distal arteries being treated and an iatrogenic ischemic stroke could result. Thus, papaverine must be kept away from solution containing heparin and it must be infused very slowly to keep its concentration low enough to prevent crystallization.40 The side effects of papaverine are numerous. Dosedependent hypotension and bradycardia have been reported. Papaverine-induced thrombocytopenia has also been reported. Mydriasis is also a frequent side effect secondary to the smooth muscle dilation of the iris. Because of this, most operators ensure that the microcatheter is distal to the origin of the ophthalmic artery, the principal arterial supply to the globe. Numerous side effects of treatment are thought to be secondary to microemboli from papaverine that has precipitated out of solution. Transient focal deficits have been reported in up to 7% of patients, probably secondary to microemboli. Additionally, in cases involving treatment of the vertebrobasilar system, several cases of respiratory depression have been reported.36 Despite the frequent robust vasodilation seen after the infusion of papaverine (Fig. 7-7), the clinical efficacy of papaverine is unproven. One report showed immediate improvement in the neurologic examination and transcranial Doppler (TCD) readings following intra-arterial papaverine infusion, but no difference in the 3-month clinical outcome when compared to controls.41 Other reports
W9418-07.qxd 10/31/03 3:39 PM Page 192
192
Section I
Introduction
A
B
Figure 7-7. A, Digital subtraction angiography: lateral view showing severe distal spasm of the anterior cerebral artery in a patient with a subarachnoid hemorrhage. B, After treatment with intra-arterial papaverine, the vessel diameter is much improved.
have shown the frequent need for re-treatment after an apparently successful papaverine treatment.39 The reason for the lack of long-term efficacy is likely the short half-life of papaverine, leading to poor durability of vessel dilation. The future role of intra-arterial papaverine for cerebral vasospasm is unclear. The current indications at our institution are (1) extreme distal vasospasm not suitable for balloon angioplasty, and (2) temporary vasodilation of proximal vessels to facilitate the placement of a micro-balloon for more definitive intracranial transluminal angioplasty. Intracranial transluminal balloon angioplasty involves the placement of a micro-balloon into stenotic proximal cerebral arteries either through flow directed or “over the wire” techniques. The micro-balloons are gently inflated with a mixture of saline and angiographic contrast material to allow direct fluoroscopic visualization of the balloon inflation. Predictable angiographic dilations with consistent blood flow increases have been reported.42 Clinical studies have shown that up to 70% of patients treated demonstrated clinical benefit.32 The angiographic results with balloon angioplasty are more durable than the results from intraarterial papaverine infusion. It is extremely rare to see further vasospasm in an artery that has been treated with intracranial transluminal angioplasty (Fig. 7-8).40,43 The ease of balloon angioplasty continues to improve with each new generation micro-balloon. Complications The most serious complication with intracranial balloon angioplasty is arterial rupture secondary to inadvertent overinflation of the balloon. The currently available compliant balloons suitable for vasospasm angioplasty have maximum
inflation diameters of 3.5 to 4 mm, greater than vessels to be angioplastied (usually 2 to 3 mm). Great care must be taken to slowly inflate the balloons under good fluoroscopic visualization to prevent accidental overinflation. Vessel dissection or rupture from balloon angioplasty usually has devastating consequences. In one large series, 4% of patients undergoing balloon angioplasty died as a result of intraoperative vessel rupture.43 The routine use of systemic heparinization during balloon angioplasty has potential complications. Many of the patients treated for cerebral vasospasm will only be a few days postcraniotomy if their aneurysm was clipped. Physicians treating these patients after the procedure should monitor for signs of intracranial mass effect and ICH. Patients should be monitored for signs of ischemic neurologic events, which can occur from clots forming on the various endovascular instruments and from emboli or thrombus that can form from iatrogenic vessel injury and dissection.
Endovascular Surgery for Arteriovenous Malformations Review Arteriovenous malformations (AVMs) are congenital vascular lesions of the brain, which are rare, but with an unclear prevalence.44 These malformations consist of a dysplastic nidus that contains vascular channels connecting arteries and veins without intervening capillaries. This low resistance mass of vessels acts as a high flow shunt from the arterial feeders to the draining veins. The combined high flow and
W9418-07.qxd 10/31/03 3:39 PM Page 193
Chapter 7
Figure 7-8. Digital subtraction angiography (DSA) of a patient who became aphasic and developed a right hemiplegia several days after subarachnoid hemorrhage. A, DSA of the left internal carotid artery (ICA) shows severe vasospasm of the supraclinoidal ICA, proximal middle cerebral artery, and anterior cerebral artery. B, After balloon angioplasty, all vessels had improved diameters. The patient’s neurologic condition markedly improved.
Basic Endovascular Neurosurgery and Neuroradiology
A
high pressure leads to dilation of the vascular channels of the nidus as well as the afferent arteries and efferent veins. The combination of high-pressure arterialized blood and thinned, dilated vessel walls can lead to hemorrhage. Fifty percent of symptomatic AVMs present with an intracranial hemorrhage.45 The next most common symptoms at presentation are seizures and focal neurologic deficits. Seizures and focal deficits are thought to be more likely related to cerebral steal phenomena secondary to the low-pressure, high-flow arteriovenous shunt through the nidus, which can decrease the perfusion of the brain parenchyma surrounding the AVM. Reports have shown that up to 7% of cerebral arteriovenous malformations have at least one associated aneurysm.46 These aneurysms often occur on large feeding arteries and are likely formed by the pathologic high flow through artery feeding the nidus, and are referred to as “flow-related” arterial aneurysms. AVM-associated aneurysms have a greatly increased rate of subarachnoid hemorrhage. Studies show that the hemorrhage rate of an AVM is 2% to 4% per year and that the mortality rate of each of these hemorrhages is 10% to 15%.47 Historically, AVMs have been extremely difficult to treat. Early surgical experience to remove the nidus had a high mortality rate from intraoperative hemorrhage. To better predict the surgical risks according to the angiographic features of the AVM, a grading system was developed by Spetzler and Martin (Table 7-1).48 This grading system stratified the surgical risk of removing AVMs according to size, proximity to eloquent brain tissue, and the pattern of venous
193
B
drainage. The points are totaled, giving a score range of 1 to 5. The higher the grade, the higher the perioperative major morbidity and mortality. This grading system has been shown to adequately predict surgical morbidity and mortality in a prospective study.49 This study reported no major or minor neurologic deficits in patients who had a grade 1 AVM removed, while in patients having grade 5 lesions removed, 19% had a minor deficit and 12% had a major deficit. Other therapies have emerged to address this difficult pathology. Stereotactic radiation surgery has been extremely successful in treating many AVMs. During stereotactic radiation surgery, a single fraction, high-dose, finely targeted dose of radiation is given to the AVM. Secondary to dangers
Table 7-1 Spetzler Martin Grading of AVMs Feature
Point Assigned
Size of AVM Small (6 cm)
1 2 3
Eloquence of surrounding brain Noneloquent Eloquent
0 1
Pattern of venous drainage Superficial only Deep
0 1
W9418-07.qxd 10/31/03 3:39 PM Page 194
194
Section I
Introduction
of increasingly large irradiated volumes, the maximum diameter of AVM able to be safely treated is generally considered to be less than 3.5 cm.50 The radiation-induced injury that occurs during radiation surgery causes a gradual occlusion of the AVM over 2 to 3 years. AVM cure rates at centers treating large numbers of these malformations exceed 80% at 3 years.51 Unfortunately, the AVMs for which radiation surgery are efficacious are often the same, small AVMs that could safely be removed with microsurgery. Endovascular techniques were developed to serve as either a primary or adjuvant therapy for AVMs. It was hoped that particle or glue embolization would cure many large or deep AVMs. However, a cure using endovascular surgery exclusively is uncommon. In a recent analysis by Wikholm and Lundqvist of long-term follow-up of 150 patients with embolized AVMs, only 19 (13%) of the AVMs were totally obliterated with glue embolization.52 Thus, endovascular embolization is more often performed as an adjunct with a goal of making either microsurgery or radiation surgery more safe and efficacious. This goal is usually achieved by making the AVM smaller with much less abnormal blood flow (Fig. 7-9). Partial embolization has also been performed as a sole therapy by other investigators with the theory that a smaller AVM has a safer natural history. The Wikholm and Lundqvist long-term study found that protective effects of partial embolization only occurred if greater than 90% of the AVM was embolized.52 The goal of preoperative endovascular embolization is to make the dangerous, inoperable AVM into a safer, technically less difficult lesion (Fig. 7-10). This usually requires multiple endovascular procedures. Embolization is directed at reducing the flow through the nidus, eliminating deep arterial feeders that would be difficult to access surgically, and decreasing the overall size of the nidus. After serial embolizations, the patient has microsurgical removal of a lower flow, less dangerous lesion. The goals of pre-radiation surgery embolization are slightly different from those of preoperative embolization. The principal goal is to make a large AVM small enough to be safely and adequately treated with radiation surgery. This generally requires a volume of less than 25 cm3. Instead of attempting to remove large central feeders as in the preoperative embolizations, it is instead important to occlude peripheral feeders and nidus. Also, any dural component is also targeted with embolization because these respond less well to radiation surgery.53 The utility of pre-radiation surgery embolization has not yet been firmly established. Also, the glue casts left after embolization sometimes make the necessary stereotactic imaging and localization for radiation surgery more difficult. Further research is needed to investigate improvements for pre-radiation surgery embolization. Endovascular Approaches The two most common categories of materials used to embolize AVMs are particles and glue.54 Currently, the wide-
spread particulate material used is polyvinyl alcohol particles (PVA). These particles are available in a variety of size ranges. The particles are released into the AVM and become lodged into the nidus where they cause blood to clot and cause an inflammatory reaction. Unfortunately, this material is associated with frequent recanalization of the nidus and the effect of embolization is not durable. PVA is most effective as an embolic agent if the AVM is resected within a few days after embolization to decrease the amount of AVM recanalization. Glue embolic agents were developed because of the lack of permanent occlusion with particulate materials. A number of glues have been used. Recent FDA approval of NBCA has led to its widespread use. NBCA is a liquid polymer, which begins to polymerize on contact with an ionic environment, including blood. The rate of polymerization can be controlled through various dilutions with Ethiodol—an oil-based, radiopaque material. By appropriately diluting the glue and carefully controlling the rate of injection, NBCA can reliably be placed into the nidus of the AVM, causing permanent occlusion of the embolized section. A recent randomized trial has shown NBCA to be safer and at least as effective as PVA in the immediate embolization. There is also considerable data confirming NBCA has a more durable effect than PVA.54 To increase the safety of endovascular embolization, preembolization provocative testing can be performed. This testing involves the injection of a short-acting barbiturate drug (i.e., methohexital [Jones Medical Industries, St. Louis MO]) through the microcatheter before embolization. A neurologic examination is performed before and after the barbiturate injection to ascertain if the artery feeds a portion of normal brain, which would cause a deficit if embolized. All relevant higher cortical function is tested, including alertness, orientation, language production and comprehension, in addition to visual field examination, and a four-extremity motor and sensory examination. If there is a possibility that the arterial supply of a cranial nerve could also be embolized from the arterial pedicle, then it is mandatory that provocative testing also be performed with lidocaine because of differences in cellular neurophysiology between central and peripheral neurons. After confirming that no apparent neurologic deficit will occur with the occlusion of the artery, embolization proceeds. Sometimes, depending on surgeon preference, systemic heparinization is administered before the injection of glue. Usually, multiple vascular pedicles can be embolized during a single procedure. Complications The most dreaded complication from AVM embolization is intracerebral hemorrhage. This complication occurs most often secondary to the complex hemodynamic changes in and around the AVM that occur during and after embolization. Knowledge of these hemodynamic changes is essential for physicians taking care of these patients postoperatively.
W9418-07.qxd 10/31/03 3:39 PM Page 195
Chapter 7
A
Basic Endovascular Neurosurgery and Neuroradiology
B
D C Figure 7-9. A 22-year-old man presented with left hemiparesis and seizures. He was subsequently diagnosed with a very large arteriovenous malformation (AVM). A, Contrast-enhanced magnetic resonance imaging shows part of the right cerebral AVM. B, Digital subtraction angiography (DSA) of the right internal carotid artery shows the large AVM. Notice the absence of visualized distal middle cerebral artery and anterior cerebral artery (ACA) because of the high arteriovenous shunting. C, DSA after several glue embolizations showing markedly decreased AVM size and flow. Notice the wellvisualized normal ACA branches that were previously not visualized. D, Unsubtracted fluoroscopic view of glue cast after several glue embolizations. All of this glue material is lodged within the nidus of the AVM, preventing flow through their respective channels.
195
W9418-07.qxd 10/31/03 3:40 PM Page 196
196
A
Section I
Introduction
B
C
As each vascular pedicle is embolized, abrupt changes in the flow patterns and filling pressures of the AVM occur. With embolization of part of the nidus, redirected flow through other parts of the nidus causing a sudden change in filling pressure in the nidus of the AVM can lead to vessel rupture and intracerebral hemorrhage. Hemorrhage can also occur secondary to venous thrombosis. Thus, it is extremely important to understand and monitor the venous drainage of the AVM during the embolization procedure. The venous drainage of cerebral
Figure 7-10. Patient with arteriovenous malformation (AVM) that was embolized in two sessions. A, Digital subtraction angiography (DSA) anteroposterior view of the left carotid artery showing the AVM. B, DSA lateral view of the AVM. C, DSA lateral view after two glue embolizations. Notice the AVM was dramatically reduced in size.
AVMs is often characterized by large venous varices dilated from the pathologically high flow. Sometimes, after a large amount of the afferent arterial blood flow to the nidus is removed, either through embolization or surgery, the flow in these enlarged draining veins will become stagnant and the veins are at high risk to thrombose. Sudden thrombosis of the venous drainage system leads to on outflow restriction and resultant passive hyperemia.55 This vascular congestion places the patient at extremely high risk for infarction and intracerebral hemorrhage in the peri-procedural period.
W9418-07.qxd 10/31/03 3:40 PM Page 197
Chapter 7
Changes in the venous flow pattern are monitored during the procedure and the presence of venous stagnation during the procedure should be an indication for procedure termination and the consideration of anticoagulation therapy to prevent thrombosis or urgent surgical removal of the AVM. In a patient thought to be at high risk for postprocedural vascular thrombosis, it is recommended to maintain good hydration and ensure good cerebral perfusion pressure.56 Another cause of intracerebral hemorrhage following AVM embolization is the phenomenon of normal perfusion pressure breakthrough. This phenomenon is controversial but is commonly believed to occur from the sudden redistribution of blood that would normally transit through the AVM into the arteries of the nearby brain parenchyma.57 This brain tissue has chronically been exposed to low flow secondary to the nearby AVM and the resultant cerebral steal from its high flow shunt. The sudden, large increase in local blood flow that occurs when the nearby AVM is embolized can cause vessel rupture and intracerebral hemorrhage. The fundamental principle in decreasing risk of all of the preceding causes of intra- or postembolization intracerebral hemorrhage is a gradual, staged treatment of these lesions with multiple embolizations. Care must be taken not to decrease the blood flow too much and too quickly during a single embolization procedure because this can have a disastrous result on the hemodynamics of these lesions. Staging the procedures ensures that the hemodynamic changes the AVM is exposed to are as gradual as possible. Thus, it is commonly recommended that less than one third of an AVM be embolized during any single procedure. If patients are thought to be at high risk for an intracerebral hemorrhage from having greater than one third of their AVM removed during a single procedure, the risk of hemorrhage can be minimized with strict postoperative blood pressure control. This blood pressure can be slowly normalized over the next 3 days as the brain around the AVM gradually becomes accustomed to the increased flow. Ischemic stroke is another serious complication of AVM embolization. Ischemic stroke is a constant threat during all endovascular procedures where intravascular manipulations occur. In embolization procedures, there is further risk from the particles or glue used during the procedure. Preembolization provocative testing, as described previously, is helpful in predicting the safety of embolizing a specific artery, but accidental reflux of embolic agents can occur. All embolizations should proceed slowly and under fluoroscopic guidance. As an embolization proceeds and part of the nidus is occluded, the local hemodynamics can change and the embolic agents can take unexpected vascular routes. Sometimes these inadvertent arterial occlusions involve arteries that were not tested during provocative testing. Unexpected neurologic deficits can occur.
Basic Endovascular Neurosurgery and Neuroradiology
197
Endovascular Surgery for Dural Arteriovenous Fistulas Review Dural arteriovenous fistulas (DAVFs) are abnormal, acquired arteriovenous shunts that occur in the dura of the central nervous system. These uncommon lesions account for 10% to 15% of intracranial vascular malformations.58 These fistulae are usually multiple with tens or hundreds of branching arterial feeders all terminating on a major dural venous sinus. Any dural sinus can be involved, and DAVFs are classified according to the dural venous sinus that forms the primary outflow of the fistula. These lesions have an association with precedent venous thrombosis. The venous congestion caused by a venous sinus thrombosis could lead to the enlargement of normally microscopic arteriovenous shunts. The causes of the antecedent intracranial venous thrombosis leading to DAVF formation are numerous. Venous sinus thrombosis secondary to trauma, infection, surgical intervention, or a hypercoagulable state have all been shown to be associated with the DAVFs.59 Classification of DAVFs by anatomy is extremely important in understanding the symptoms, natural history of disease, and treatment. Several classification systems exist. The most frequently used is the system proposed by Cognard and colleagues.60 This system, like most systems, classifies these lesions according to the pattern of the venous outflow (Table 7-2). It is the pattern of venous outflow that has been shown to be most predictive of patient outcome. The symptoms, natural history and treatments vary significantly according to draining venous anatomy. Type I lesions, with normal antegrade venous drainage, have a benign course and pose little risk of future serious neuro-
Table 7-2 Classification of Dural Arteriovenous Fistulas Grade
Pattern of Venous Flow
I
Venous drainage into a sinus, normal antegrade flow Venous drainage into a sinus, with insufficient antegrade flow and reflux Retrograde venous drainage into a sinus only Retrograde venous drainage into a cortical vein only Retrograde venous drainage into a sinus and cortical veins Venous drainage into a cortical vein with ectasia Venous drainage directly into a cortical vein with venous ectasia larger than 5 mm diameter and three times larger than the diameter of the draining vein
II A B A+B III IV
W9418-07.qxd 10/31/03 3:40 PM Page 198
198
Section I
Introduction
logic events such as infarction or hemorrhage. These lesions often present with headaches or bruits. Sometimes the presenting symptoms are dependent on the anatomic location, for example, pulsatile tinnitus caused by highflow transverse sinus DAVFs, or retrobulbar pain and ocular motility disorders caused by high flow and venous congestion of a cavernous sinus DAVF. Numerous studies have shown that these lesions do not pose a serious neurologic risk.61–64 Treatment for these lesions is generally only performed if the symptoms are particularly bothersome to the patient. Patients with a type IIA lesion, sinus stenosis without cortical venous drainage, have a slightly more dangerous natural history.60 These patients have higher flow fistulas, and symptoms are likely to develop secondary to intracranial hypertension. This intracranial hypertension can lead to bilateral papilledema, loss of vision, and diplopia. Because these lesions do not have venous drainage that refluxes into the cortical venous system, they do not cause intracerebral hemorrhage. The goal of treatment for these lesions is to decrease flow to control symptoms and normalize intracranial pressure. Complete obliteration of the fistula is not usually necessary. Patients with venous drainage reflux into the cortical veins, types IIb, a + b, III-IV, do have significant risk for intracerebral hemorrhage. In a large series, it was reported that 10% of type IIB, 40% of type III, and 65% of type IV lesions presented with intracranial hemorrhage.60 A recent study that followed a subset of untreated DAVFs with cortical venous drainage found an annual mortality rate of 10.4%, with an annual rate of intracranial hemorrhage of 8.1%, and rate of nonhemorrhagic neurologic deficit of 6.9%. Unlike patients with type I and IIa, the treatment of these patients is not only to alleviate symptoms but to protect the patient from significant future risk of ICH and progressive neurologic deficits. The goal of endovascular or surgical treatment of these lesions is the elimination of the cortical venous reflux. Endovascular Approaches Several treatment approaches exist for DAVFs. The specific treatment performed depends on the lesion anatomy and goals of therapy. Carotid artery-jugular vein compression is a potential therapy for motivated patients with low-flow DAVFs. It is appropriate for DAVFs with primary flow from the meningeal branches of the internal carotid artery or dural branches of the external carotid artery. Theoretically, the flow stagnation produced by this compression of the carotid artery and jugular vein could lead to thrombosis of the fistula. Patients are instructed to compress the ipsilateral carotid artery in the neck with the contralateral hand for 30-minute sessions. This therapy enables a complete cure in 22% of cases, with clinical improvement in 33% of
patients.65 This technique is only appropriate for patients with small, low-flow fistulas. Arterial compression of the carotid artery is contraindicated in patients with ipsilateral atherosclerotic disease secondary to risk of embolic complications. Transarterial embolization with either PVA particles or NBCA glue can also be effective in the treatment of DAVFs. Please see the section of this chapter on embolization of intracerebral AVMs for a brief description of the various available embolic agents. It is important to understand that as with AVMs, transarterial embolization is rarely curative for DAVFs. These lesions are often extremely complex with numerous small arterial feeders too small to be selectively embolized. Instead, the goals of therapy are either reduction of flow for symptomatic relief or reduction of flow in anticipation of a curative procedure (microsurgical removal or transvenous embolization; Fig. 7-11). The procedure involves the super-selective catheterization of a DAVF feeder, usually an external carotid artery branch. Before embolization with PVA or glue, the super-selective angiogram is analyzed to assess for potential dangerous external carotid to internal carotid or vertebral artery anastomoses. Potential anastomoses to the ophthalmic artery or cranial nerves must also be assessed. Transvenous embolization is the most definitive endovascular procedure for DAVFs. The exact site of the fistula must be eliminated to cure these lesions. Because the fistula is frequently in multiple locations along the dural venous sinus, curing the fistula generally requires a venous approach. This can be accomplished by a transvenous embolization procedure, or by an open microsurgical skeletonization of the dural venous sinus. The transvenous approach can be via a transfemoral, transfacial, transjugular, or trans-superior ophthalmic vein. Transvenous embolization involves the endovascular occlusion of the venous sinus draining the fistula (Fig. 7-12). Careful analysis of the preoperative angiogram is important in establishing if the patient can tolerate the occlusion of the draining sinus. Occlusion of a transverse or sigmoid sinus is especially dangerous if the contralateral transverse and sigmoid sinus are absent or not robust. The adequacy of collateral venous drainage is important to note. This drainage can be tested during the procedure with temporary sinus occlusion with a nondetachable compliant balloon. In performing transvenous embolization, a microcatheter is placed in the venous sinus and the occlusion can be performed with glue or coils. As the sinus is packed with embolic materials, the embolization is monitored with intermittent angiograms from a catheter placed in a feeding artery. Embolization proceeds until no flow exists through the DAVF. This usually requires complete occlusion of the involved dural sinus. Often, to ensure the remaining sinuses and venous outflow are maintained during the transition to the new flow pattern, systemic anticoagulation is maintained for 24 hours after the procedure.
W9418-07.qxd 10/31/03 3:40 PM Page 199
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
A
B
Figure 7-11. Digital subtraction angiography of a patient with a dural arteriovenous fistula (DAVF) that presented with an audible bruit. A, Lateral view selective internal carotid artery (ICA) injection demonstrates that the DAVF is fed by dural branches of the meningohypophyseal trunk from the cavernous segment of the ICA. B, Lateral view selective external carotid artery (ECA) injection shows several dural branches that feed the fistula. Note there is cortical venous reflux at the left inferior part of the image. C, After N-butyl cyanoacrylate glue embolization of the ECA feeders. Notice a dramatic decrease in flow to the fistula. There were residual feeders from the meningohypophyseal trunk that could not be embolized safely. These will require either surgical or radiosurgical obliteration, or transvenous embolization.
C
Complications Because of potential anastomoses between dural arteries and the arteries supplying the cranial nerves and retina, loss of cranial nerve function and vision is a potential complication. The use of large PVA particles (>150 mm) can decrease the likelihood of this complication, because these are less likely to cause capillary level occlusion. Another potential complication of transarterial embolization is the inadvertent placement of glue past the fistula and into the venous drainage. Sudden venous occlusion can shift the venous drainage through new pathways. Sudden increased flow into a supe-
199
rior ophthalmic vein or cortical vein can lead to visual loss, intracranial hemorrhage, and neurologic deficit. Sometimes a relatively safe lesion with normal, antegrade venous drainage can be made into a more dangerous lesion with retrograde venous reflux into cortical veins. The most dangerous complications occur secondary to venous occlusions. With shifting of the venous drainage secondary to the loss of the treated sinus, symptoms of venous congestion with occasional intracerebral hemorrhage are possible. Postoperatively, patients should be monitored for the development of focal deficits. Venous congestion can affect various parts of the brain including
W9418-07.qxd 10/31/03 3:40 PM Page 200
200
Section I
Introduction
A
B
Figure 7-12. Lateral view digital subtraction angiography (DSA) of a patient with a preembolization DAVF who presented with proptosis, chemosis, and diplopia. A, Preembolization DSA showing a DAVF of the ophthalmic artery draining via the superior ophthalmic vein. B, Postembolization DSA of the same internal carotid artery injection. The glue embolization was delivered through a transvenous approach, and the lesion was cured.
cranial nerves, the cerebral cortex, and structures of the posterior fossa. It is imperative for all critical care physicians who take care of post-embolization patients to understand various special complications that can occur with these patients.
can include direct or indirect connections with the internal carotid artery (ICA) and/or external carotid artery (ECA) and the cavernous sinus. In understanding the etiology, natural history, and treatment of these lesions, it is important to classify them by their arterial anatomy and supply. The most common classification system used is the one proposed by Barrow and associates:66
Endovascular Surgery for CarotidCavernous Fistulas
A. Direct intracavernous ICA to cavernous sinus. B. Indirect dural ICA branches (i.e., meningohypophyseal trunk) to cavernous sinus. C. Indirect dural ECA branches to cavernous sinus. D. Indirect dural ICA and ECA branches to cavernous sinus.
Review Carotid cavernous fistulas (CCFs) are abnormal arteriovenous shunts between the carotid artery and the cavernous sinus. These lesions can occur spontaneously or they can be acquired secondary to trauma or venous thrombosis. A CCF
Type A fistulas are also called direct fistulas because they involve a direct connection between the ICA and the cav-
W9418-07.qxd 10/31/03 3:40 PM Page 201
Chapter 7
ernous sinus without intervening dural branches. The other three types (types B to D) are also called indirect fistulas because the arterial supply to the fistula involves dural branches (Fig. 7-13). Indirect CCFs are similar to other DAVFs in their pathogenesis and treatment and are considered a subset of DAVF based on their location at the cavernous sinus (see Figs. 7-11 to 7-13). Therefore, this section will only focus on Barrow type A direct CCFs, because they represent a unique pathologic entity with special treatment options. The reader is referred to the section in this chapter on the endovascular treatment of dural AVF for further information on the natural history and treatment of indirect carotid cavernous fistulas. Type A CCFs are most frequently traumatic in nature. The most common cause is motor vehicle accidents, followed by falls and penetrating injuries.67 The incidence of CCF is increased in patients with basilar skull fractures. In all of these cases, a rent forms in the cavernous segment of the ICA allowing direct, high-flow arteriovenous shunting into the cavernous sinus. Spontaneous CCFs also occur. This situation can occur following the rupture of an intracavernous ICA aneurysm. Collagen deficiency syndromes, such as Ehlers-Danlos syndrome and pseudoxanthoma elasticum, can predispose patients to spontaneous CCF.68,69 As with DAVFs, the venous drainage pattern of the CCF is extremely important in the clinical presentation and the natural history of the patient. The cavernous sinus normally
A
Basic Endovascular Neurosurgery and Neuroradiology
201
receives venous inflow from the orbit via the superior ophthalmic vein and from the brain via the superficial middle cerebral veins and the sphenoparietal sinus. Venous outflow from the cavernous sinus is to the transverse and sigmoid sinuses via the superior and inferior petrosal sinuses and basilar plexus. There is also an anteroinferior venous outflow via emissary veins to the pterygoid plexus. The route that the pulsatile, high flow blood takes from the arterialized cavernous sinus is variable and can change. Studies with large series of patients have shown a higher risk of intracranial hemorrhage in patients who present with retrograde cortical outflow through the sphenoparietal sinus into the cortical veins.70 The clinical presentations of CCFs are varied. The increased venous pressure through the superior ophthalmic vein leads to tissue congestion through the orbit. The affected eye is usually proptotic and chemotic. Diplopia can result from cranial nerve dysfunction or a restrictive ophthalmoplegia secondary to the swollen proptotic orbital contents. Patients often show evidence of decreased visual acuity and optic nerve dysfunction if the lesion is not promptly treated. The visual loss can be secondary to retinal ischemia from arterial steal and venous congestion, from glaucoma secondary to increased intraocular pressure secondary to venous congestion, and from exposure keratitis if the eye is too proptotic for proper lid closure or if the eye is anesthetic secondary to trigeminal dysfunction.
B Figure 7-13. A young boy presented after an all-terrain vehicle accident with proptosis, chemosis, and diplopia. The patient was found to have a Barrow type D carotid cavernous fistula (CCF). The CCF had fistulous filling of the cavernous sinus by very small, indirect dural branches of the internal carotid artery (A), and large, high-flow, indirect dural branches of the external carotid artery (B).
W9418-07.qxd 10/31/03 3:40 PM Page 202
202
Section I
Introduction
Patients often complain of unilateral, pulsatile tinnitus or bruit, or retro-orbital headache at presentation. This pain is likely secondary to distention of the dural and orbital content from venous congestion. Symptoms from cerebral ischemia secondary to steal are quite rare, but can occur if there is a combination of inadequate flow past the CCF into the distal supraclinoid carotid artery and inadequate collateral flow from the circle of Willis (isolated hemisphere). It is important to note that the affected eye is not always ipsilateral to the CCF. Intracavernous sinus flow through the circular sinus can lead to a bilateral presentation of ocular symptoms, or rarely a contralateral ocular presentation. The clinical indications for urgent treatment include increased intraocular pressure, progressive visual field loss, exposure keratitis, rapidly progressive proptosis, cerebral ischemia, and cortical venous outflow. Endovascular Approaches Carotid compression therapy, as described in the section on DAVFs, is rarely effective with high-flow type A CCFs. Although uncommon, occasional cures with few complications have been reported with this approach.71 It is very important that only the patient perform this maneuver, not family members, and that only the patient’s contralateral hand is used. This would automatically stop the occlusion if a hemiparesis occurs secondary to cerebral ischemia. Also, it is important to exclude severe atherosclerosis in the ipsilateral cervical carotid to prevent iatrogenic embolic events occurring from manual compression and disruption of plaque. The most common and most effective treatment is transarterial occlusion of the fistula with detachable balloons. This procedure entails passing a flow-directed detachable balloon into the affected carotid artery. The partially inflated balloon is allowed to pass through the fistula and into the cavernous sinus. The balloon is then pulled back into the fistula (i.e., the rent in the carotid artery wall) and inflated. Repeat angiography is performed. If the fistula is closed and the ICA is patent, the balloon can be detached. Sometimes the fistula is slowed but not occluded (Fig. 7-14). This is usually due to the hole in the artery being larger than the inflated diameter of the balloon. In this case, multiple balloons can be positioned into the fistula. These additional balloons are more difficult to place because the previous balloon or balloons are partially blocking the fistula. In a large study, 88% of traumatic type A fistulas were successfully treated with detachable balloons with preservation of the parent artery.72 Sometimes using detachable balloons alone in the occlusion of the fistula is not possible. In this case, transarterial placement of endovascular detachable coils into the fistula can be attempted either alone or in conjunction with detachable balloons. Transvenous placement of coils or other embolic agents into the cavernous sinus can also be attempted. If all of these therapies fail, it is sometimes
necessary to occlude the parent ICA both proximally and distally to the fistula with detachable balloons. Adequate collateral blood flow is essential (Fig. 7-14C). A temporary balloon occlusion test can be performed distal to the fistula to test the patient’s tolerance of an ICA occlusion. A balloon occlusion test is unnecessary if the pretreatment angiogram shows steal from the intracranial circulation down into the fistula. In this case, the patient has already tolerated having no cerebral flow from the affected artery and the flow will actually increase because there will no longer be cerebral steal after the artery is occluded. Most operators perform this procedure with the patient under systemic heparinization. Complications Complications following endovascular therapy for CCF are relatively uncommon. Most are minor in nature. Often following a successful balloon occlusion of the fistula, the patient will complain of worsening chemosis, proptosis, or headache. This is probably secondary to acute cavernous sinus thrombosis. These symptoms are usually limited and short-lived. It is also possible to develop a palsy or worsening function of cranial nerves III, IV, V, or VI (the intracavernous cranial nerves) secondary to mass effect, balloon compression, or cavernous sinus thrombosis. One very serious complication that can occur is balloon migration. This can occur during the procedure or can be a delayed complication. Detachable balloons can be difficult to place and control. They are mechanically detached and at times can detach inadvertently or can move inappropriately by the detaching process. These balloons can also deflate gradually and cause late ischemic complications or reestablishment of the fistula. Any misplaced detached balloon can move into the cerebral circulation and cause ischemic sequelae. Delayed, gradual balloon deflation can be a problem because of osmotically driven diffusion across the balloon, which is a semipermeable membrane. This phenomenon can be essentially eliminated if the balloons are inflated with an isosmotic, isotonic contrast agent. Another important complication is an extension of the flow dynamics discussed previously. In attempting to occlude the CCF, the venous flow can sometimes be changed to a much more dangerous pattern. During an attempt to treat a patient with predominant venous flow anteriorly through the superior ophthalmic vein, it is possible to inadvertently change the arterialized venous outflow posteriorly through the sphenoparietal sinus and superior and inferior petrosal sinuses. This can cause acute venous hypertension of the brain with the associated risks of cerebral edema or intracranial hemorrhage. In an article by Higashida, 4.9% of patients undergoing treatment for direct type A fistulas had major complications. Half of these patients had transient ischemic symptoms and the other half had thromboembolic strokes.72 It is important for all physicians treating these patients postoperatively to
W9418-07.qxd 10/31/03 3:40 PM Page 203
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
A
Figure 7-14. Digital subtraction angiography (DSA) of a patient who presented after a motor vehicle accident with proptosis and chemosis of the right eye. A, Lateral view of the right internal carotid artery (ICA) showing a type A carotid cavernous fistula (CCF) with enlargement of the superior ophthalmic vein and significant cortical venous reflux. This fistula was treated with balloon packing. B, Lateral DSA view of the CCF with a single balloon in the fistula. Note that there is less fistulous flow. Because of this, now the middle cerebral artery and anterior cerebral artery are filling intracranially, which were not previously visualized. Also notice that there is no further intracranial retrograde cortical venous flow as seen previously and increased arterial cerebral filling. To cure this lesion, the ICA had to be sacrificed. C, Anteroposterior view of the contralateral (left) ICA, which demonstrates excellent collateral flow through the anterior communicating artery of the circle of Willis.
understand the procedural details and any adverse technical events to understand better the postoperative risks to the patients.
Endovascular Surgery for Extracranial Stenosis Review Angioplasty and stenting can be safely performed on a variety of brachiocephalic atherosclerotic lesions. The vast majority of brachiocephalic lesions are atherosclerotic
203
B
C
lesions of the cervical carotid bifurcation. Additional vessels routinely treated are the origins of the innominate, subclavian, carotid, and vertebral arteries. Carotid bifurcation revascularization will be discussed first, followed by a discussion of angioplasty and stenting of the other brachiocephalic vessels. Atherosclerotic disease at the carotid bifurcation has been studied extensively. The natural history of the disease and treatments have been well delineated in several large randomized clinical trials. Carotid bifurcation atherosclerosis causes 20% of all ischemic strokes and transient ischemic attacks (TIAs).73 The natural history of the lesion is most dependent on the degree of luminal stenosis and on whether it is associated with symptoms of TIAs.
W9418-07.qxd 10/31/03 3:40 PM Page 204
204
Section I
Introduction
Asymptomatic carotid stenosis is a risk factor for ipsilateral cerebral infarction; and with a stenosis of greater than 50%, a 2.4% annual stroke risk exists.74 The asymptomatic lesion with a 60% to 99% stenosis has been shown to have a 3.2% annual risk of ipsilateral stroke.75 The proper treatment for asymptomatic carotid stenosis is an area of debate. The largest study examining the role of carotid endarterectomy (CEA) in these lesions is the Asymptomatic Carotid Atherosclerosis Study (ACAS). This randomized study of patients with lesions of at least 60% stenosis found a significant benefit from surgery. The 5-year risk of stroke or death was 5% in patients who had a CEA versus 11% in the population treated with medical therapy only. Of note, the benefit of CEA was only significant in men and when the total perioperative complications were less than 3%. The benefit of CEA in asymptomatic lesions is modest and is dependent on keeping the surgical complication rates low. Symptomatic carotid stenosis is defined as a stenotic lesion associated with an ipsilateral stroke, ipsilateral hemispheric TIA, or ipsilateral amaurosis fugax (i.e., transient monocular blindness). These lesions have a much worse natural history, and therefore patients with symptomatic lesions receive a more substantial benefit from CEA.76 In the North American Symptomatic Carotid Endarterectomy Trial (NASCET), patients with symptomatic stenosis greater than 70% were found to have an ipsilateral stroke risk of 26% when treated with medical therapy versus 9% when treated with CEA. Patients with moderate carotid stenoses (50% to 69%) also benefit from CEA compared to medical treatment, but the difference was less dramatic. Those treated with CEA had a 5-year stroke rate of 15.7% while the medically treated group had a 5-year stroke rate of 22.2%.77 Based on these studies, CEA is now the recommended treatment for symptomatic lesions with a stenosis greater than 70% and for selected asymptomatic lesions with a stenosis greater than 60% and symptomatic patients with stenosis greater than 50%. The perioperative surgical risks of major stroke and death must be less than 3% in asymptomatic patients, and less than 6% in symptomatic patients, for many of the benefits from CEA to be realized.76 Endovascular carotid revascularization with angioplasty and stenting has become increasingly common as an alternative to CEA (Fig. 7-15). Originally, the procedure was developed as an alternative therapy for patients who did not fit the inclusion criteria of the study populations in the previously mentioned trials, and those patients for whom the risks of CEA are deemed excessive. Endovascular revascularization held the promise of treating these high-risk patients with a potentially lower risk procedure, requiring no anesthesia, less operative time, and less systemic stress. With increased experience, the indications for endovascular therapy are expanding at many centers. Many neurosurgeons will recommend endovascular revascularization for patients if they have certain factors that increase the technical difficulty of the CEA. These include but are not limited to bifur-
cation disease in patients with an unusually high cervical bifurcation, patients with recurrent stenosis who have undergone ipsilateral CEA, patients with radiation-induced stenosis, patients with occlusions of the contralateral carotid artery, or patients with one or more vertebral artery occlusions. The efficacy of endovascular therapy for cervical carotid stenosis as a therapy for all patients including those who do fulfill the NASCET criteria is very controversial and is currently being studied in several randomized clinical studies.73,78 One of these trials, Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS), was recently published.73 This trial randomized patients with symptomatic or asymptomatic cervical carotid stenosis to either CEA or endovascular therapy. This study found that the overall morbidity and mortality at 30 days was the same in each group. The perioperative stroke and death rate was similar between the two groups (surgery 9.9% vs. angioplasty and stenting 10%). Although the morbidity and mortality for the two groups were the same in this study, these rates were considerably higher than the NASCET study, which had a rate of 5.8%.73,79 Another important finding in the CAVATAS study was a significantly higher rate of restenosis at 1 year in the endovascular group. This was not, however, associated with increase in clinical symptoms at 1 year. Most of the patients in CAVATAS treated by endovascular means underwent angioplasty alone without stenting. Most carotid atherosclerotic lesions are now treated with stenting in addition to angioplasty. It is thought that the stent will decrease emboli during the procedure, decrease symptomatic plaque dissections, and improve long-term restenosis rates. Until further large, randomized trials with large numbers of stented patients are published, endovascular therapy for carotid atherosclerosis cannot be recommended for patients who are candidates for carotid endarterectomy. Balloon angioplasty and stenting are currently reserved for patients who are poor candidates for surgery secondary to comorbidities, complex anatomy, or a history of CEA or radiation-induced stenosis. Other brachiocephalic vascular lesions frequently treated with endovascular surgery are stenoses of the origins of the carotid and vertebral arteries, innominate artery stenosis, and subclavian stenosis. The indications for these procedures typically include intermittent neurologic symptoms referable to the area of the brain supplied by the diseased artery. Although these spells can be embolic in nature, similar to stenosis of the carotid bifurcation, symptoms in the proximal vertebral, carotid, innominate, and subclavian arteries are usually secondary to hypoperfusion caused by the stenotic blockage (Fig. 7-16). One notable symptom complex occurs in the case of subclavian steal. These patients often experience posterior fossa ischemic symptoms, including dizziness, diplopia, visual disturbances, and unilateral and bilateral weakness when they use one of their arms. The anatomic substrate is a subclavian stenosis or occlusion that
W9418-07.qxd 10/31/03 3:40 PM Page 205
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
A
Figure 7-15. Digital subtraction angiography (DSA) of a patient with transient ischemic attacks of the right cerebral hemisphere. A, View of right cervical internal carotid artery showing severe stenosis. The patient was treated with angioplasty and stenting. B, DSA showing the crossing of the lesion with the undeployed stent after preliminary angioplasty. C, DSA showing deployed stent. Notice that the diameter of the carotid artery lumen has dramatically increased.
B
C
205
W9418-07.qxd 10/31/03 3:40 PM Page 206
206
Section I
Introduction
A B Figure 7-16. Digital subtraction angiography (DSA) of a patient who presented with posterior fossa transient ischemic attacks. A, Anteroposterior view of the left subclavian artery showing severe stenosis of the left vertebral artery origin. Angioplasty and stenting were performed. B, After angioplasty and stent placement, DSA shows much improved vessel diameter.
is proximal to the origin of the ipsilateral vertebral artery. The perfusion demand of the arm causes a reversal of flow through the vertebral artery and steals blood flow from the intracranial circulation. Sometimes the patient will only experience symptoms when the ipsilateral arm is active (Fig. 7-17).80
Endovascular Approaches The goal of endovascular therapy for the preceding lesions is to reduce the level of stenosis and restore normal anatomy. This goal will increase cerebral blood flow and decrease sludging and turbulence of flow, thereby reducing embolic events. During balloon angioplasty, the atherosclerotic plaque is stretched and cracked and the three layers of the vessel are dilated. To decrease the distal embolization of plaque, vessel dissection, and long-term restenosis, a metallic stent is placed at the site of the stenosis. The stent is placed before angioplasty if the lesion can be safely crossed without preceding angioplasty. This, of course, is not possible in
extremely tight lesions where the remaining lumen is no bigger than the undeployed stent. These lesions require angioplasty before crossing the lesion with the stent. Endovascular therapy of these lesions is almost always done with the patient receiving systemic anticoagulants. Some operators pretreat the stenotic lesion with abciximab or r-tPA to clear any thrombus from the lesion. The lesion is crossed with a wire over which the angioplasty balloon or stent will be passed. If the stenosis is not extremely tight, then primary stenting is performed. The stent is deployed over the area of greatest stenosis with some overlap of normal artery to prevent stent migration. Balloon angioplasty can be performed after stenting if there is a “waist” in the stent, indicating residual stenosis. This is often referred to as “poststent remodeling,” performed after stent deployment with slow dilation of the balloon up to a predetermined diameter. The result is checked with an angiogram. The angioplasty can be repeated until there is no residual stenosis identifiable. The patient’s vital signs are constantly monitored during balloon inflation for bradycardia or hypotension. Following the procedure, the patient is gener-
W9418-07.qxd 11/21/03 12:42 PM Page 207
A
C
B
D Figure 7-17. Digital subtraction angiography (DSA) of a patient who presented with subclavian steal syndrome. A, Anteroposterior DSA showing severe stenosis of the innominate artery. B, A left vertebral artery injection shows retrograde flow “steal” down the right vertebral artery. C, Angioplasty and stenting DSA showing an excellent angiographic result. D, A left vertebral artery injection after the innominate artery was angioplastied and stented shows normal antegrade flow into the basilar artery system, and therefore an angiographic cure of the “subclavian steal.”
W9418-07.qxd 10/31/03 3:40 PM Page 208
208
Section I
Introduction
ally maintained on heparin drip for 24 hours, clopidrogel for 3 months, and lifelong enteric-coated aspirin.
Endovascular Surgery for Intracranial Stenosis
Complications
Review
Major intraoperative complications of extracranial angioplasty and stenting include bradycardia and asystole, stroke, and perfusion breakthrough. During the balloon dilation of the atherosclerotic lesion, the carotid bulb is stretched. Stretch receptor afferents in the carotid bulb travel to the brainstem via cranial nerve IX. These signals can lead to a profound bradycardia or asystole during the carotid bulb manipulation. The operator should monitor the patient’s vital signs continuously during the dilation process. Early in our experience, prophylactic transvenous pacing, external pacing, or atropine were used. We discovered that these are rarely necessary unless the bradycardic response is severe and sustained, which is exceedingly rare. The bradycardia induced by angioplasty resolves after prompt balloon deflation, especially if the balloon is deflated at the first sign of decreased heart rate, before brachycardia occurs. If symptomatic bradycardia occurs, atropine can be given. Ischemic stroke occurs as a peri-procedural complication of endovascular cervical carotid revascularization in up to 10% of patients, according to the recently published CAVATAS trial.73 The causes of these strokes are embolic material dislodged during the balloon angioplasty or from emboli from iatrogenic carotid dissection. Numerous techniques and devices are used to help protect the distal circulation from embolic materials. Some operators balloon occlude the internal carotid artery distal to the stenosis, flush out debris produced during the angioplasty, and then remove the distal balloon.81 New, investigational distal protection devices, which catch the embolic debris with a filter, are currently being studied in clinical trials with early promising results.82,83 Patients who have a severe stenosis with poor distal perfusion of the ipsilateral hemisphere are at risk for ICH related to the phenomenon of normal perfusion pressure breakthrough, caused by the sudden increase in the transmitted blood pressure normally blocked by the stenosis.84,85 This brain tissue, which has been accustomed to very low flow, has arterioles and capillaries that maximally dilate to increase blood flow to the parenchyma. The sudden hyperemia following treatment of the stenosis can lead to headache, seizures, focal deficits, vessel rupture, and resultant ICH. The incidence of this complication has been reported to be as high as 5%.86 This complication can be minimized by strict postoperative blood pressure control, sometimes to 70% of the preoperative systolic blood pressure. The blood pressure is then slowly normalized over the next several days, as the normally dilated vessels gradually acquire autoregulatory capacity and regain the ability to constrict in response to the increased flow.
Stroke secondary to thromboembolic events from intracranial stenoses has been underestimated. Studies have shown that approximately 8% of transient ischemic attacks are secondary to intracranial stenosis.87 Autopsy studies have shown an incidence of these atherosclerotic lesions of 6.5%.88 The natural history of these atherosclerotic lesions is poorly understood. Various studies have suggested that intracranial stenosis carries a risk of stroke of greater than 8% in 1 year.89,90 One reason the natural history of intracranial stenosis has been difficult to define is that it varies according to the anatomic location of the stenosis. For example, while MCA stenosis may have an annual stroke rate of 7.8%, stenosis of the basilar artery has an annual stroke rate of 11%.91,92 There has not been convincing data in the literature correlating the natural history of intracranial stenosis with the degree and severity of stenosis. There is general consensus that hemodynamic significance, and therefore a possible dangerous natural history, begins to increase with stenosis greater than 50%.89 The clinical presentation of patients with intracranial stenosis is variable, depending on the anatomic location. Description of the various ischemic syndromes is beyond the scope of this chapter. Recurrent TIAs, although much more common in extracranial disease, can indicate intracranial disease. Recurring, fluctuating symptoms might indicate the low-flow hemodynamics that occur as a result of these lesions versus the sudden, maximal deficit associated with thromboembolism. This low-flow state is important in assessing these patients. Patients with clinically asymptomatic lesions are sometimes considered at higher risk if brain blood flow imaging, including single photon emission computed tomography, positron emission tomography, and other radiographic diagnostic testing show poor perfusion in the territory of the involved artery. Previous treatments for these lesions include antiplatelet medications, anticoagulation, and extracranial-tointracranial bypass. The best medical therapy is not yet clear. Currently, a large, randomized clinical trial is underway comparing warfarin and aspirin in patients with symptomatic intracranial stenosis.89 A large randomized study of extracranial-to-intracranial arterial bypass was performed in 1985 to assess efficacy in preventing stroke in patients with extracranial carotid occlusion, distal carotid occlusive disease, and MCA stenosis.93 Subgroup analysis showed no benefit in the prevention of future strokes in patients with carotid siphon or MCA stenoses when compared with medical therapy. Patients with severe (>70%) stenosis of the MCA actually did worse in the extracranial-to-intracranial bypass study group.
W9418-07.qxd 10/31/03 3:40 PM Page 209
Chapter 7
With the rapid advances in endovascular technology, new therapies are available. Micro-balloons are now available that are small and flexible enough to be maneuvered into the intracranial circulation. Many intracranial stenoses are now treated with angioplasty with or without stenting. To date, no randomized controlled trial has been completed, although several are underway. The best treatment for these lesions is under intense study but at this time remains unclear.
Endovascular Approaches Endovascular treatment of intracranial stenosis is most often attempted after the patient has failed maximal medical therapy. This includes patients who are symptomatic despite appropriate antiplatelet therapy and anticoagulation, or have unacceptable side effects from these medications. Endovascular treatment originally involved balloon angioplasty alone. With the recent introduction of coronary microballoons with improved flexibility, intracranial stenting has been performed along with the angioplasty (Fig. 7-18). Currently, these stents are metallic mesh, deployed by balloon inflation. In the near future, biodegradable, self-expanding stents and covered micro-stents will be available for the intracranial circulation.
Basic Endovascular Neurosurgery and Neuroradiology
209
In angioplasty or stenting, the balloon or balloondeployed stent are placed across the lesion in an over-thewire manner. Systemic heparinization with a bolus is usually administered immediately before the placement of the balloon or balloon stent. After satisfactory placement of the device, the balloon is inflated to a predetermined pressure depending on the desired balloon diameter using a balloon compliance chart supplied by the manufacturer. Once the desired inflation is performed, the balloon is deflated and removed from the lesion to a more proximal location. If a stent was used, it should remain in the lesion. An angiogram is obtained to assess vessel patency and distal arterial filling (Fig. 7-19). The patient’s vital signs and neurologic examination, if general anesthesia is not used, are continuously monitored during this critical portion of the procedure. Depending on the preference of the operator and the final anatomic result, systemic anticoagulation can be continued after the procedure. Many operators are using perioperative abciximab (ReoPro) instead of heparin when performing intracranial angioplasty and stenting. Complications One complication from the endovascular treatment of intracranial stenosis is ischemic stroke. These strokes can be secondary to occlusion or thrombosis of the treated vessel or
A B Figure 7-18. Digital subtraction angiography (DSA) of a patient with posterior fossa transient ischemic attacks who was found to have a severe, symptomatic basilar artery stenosis. The patient also had a large basilar artery aneurysmal dissection that was asymptomatic. A, Lateral view DSA showing severe basilar artery stenosis with a large fusiform basilar aneurysm. The stenosis was treated with angioplasty and stenting. B, Same lateral view DSA after angioplasty and stenting of the stenosis with an excellent angiographic result.
W9418-07.qxd 10/31/03 3:40 PM Page 210
210
Section I
Introduction
B A Figure 7-19. Digital subtraction angiography (DSA) of a patient having recurrent transient ischemic attacks secondary to severe carotid siphon stenosis. A, Lateral view DSA showing severe stenosis of the carotid siphon. This lesion was treated with angioplasty and stenting. B, Same lateral view after angioplasty and stenting of the carotid siphon with increased vessel diameter. The patient was no longer symptomatic after treatment.
perforating branches, or from embolic occlusion of distal branches. This complication has been reported to occur in 8% to 11% of patients during the procedure.94–96 More recent, but smaller studies have reported a 0% rate for intraprocedural ischemic stroke.97,98 Patients with acute thromboembolic events should be treated with intra-arterial thrombolysis and clot disruption as described in the section of this chapter on the endovascular treatment of acute stroke. The physician treating these patients postoperatively should understand that these patients are at risk for hemorrhage or further ischemic injury in the territory of the treated vessel and should be closely monitored, because postoperative hemorrhage, TIAs, stroke, and restenosis have been reported in most studies. Intracranial hemorrhage is another devastating complication that uncommonly occurs during or after intracranial angioplasty or stenting, especially in the basilar artery (Fig. 7-20). The vessels being treated are small, severely atherosclerotic, and fragile. Any manipulation from
the various wires, catheters, or balloon can cause a vessel perforation or hemorrhagic dissection. The most dangerous portions of the procedure are balloon inflation and stent deployment. Careful angiographic analysis is performed to determine the vessel diameter. These determinations can be difficult in diseased vessels, and the margin for error is small. Especially in the context of systemic anticoagulation, hemorrhages are extremely dangerous and mortality is high. Rapid reversal of anticoagulation is mandatory with the appropriate agent. Sometimes dissections can occur during the procedure, which then hemorrhage in the postoperative period. Because of this, as with all procedures, the critical care physicians treating these patients should always know the anticoagulation and antiplatelet medications that have been given to the patient, along with the dosages and the time the medication was given. Patient lives can be saved by rapid reversal of anticoagulation. Therefore, constant monitoring for possible hemorrhagic complications is imperative.
W9418-07.qxd 10/31/03 3:40 PM Page 211
Chapter 7
Basic Endovascular Neurosurgery and Neuroradiology
A
B
C
Figure 7-20. Digital subtraction angiography (DSA) of an elderly patient with a severe basilar artery stenosis who presented with posterior fossa transient ischemic attacks that were refractory to medical therapy. A, Anteroposterior (AP) view showing the severe basilar stenosis. The lesion was treated with angioplasty and stenting. B, Same AP view showing the vessel after angioplasty and stenting with an improvement in the stenosis. During the procedure the patient complained of a sudden headache. The patient was found to have a subarachnoid hemorrhage immediately after the procedure. C, Noncontrastenhanced head computed tomography after the angioplasty and stenting procedure shows subarachnoid hemorrhage and contrast material extravasation. The patient had immediate reversal of his anticoagulation and an external ventricular drain placed. Several hours later, the patient experienced another episode of hemorrhage and died.
211
W9418-07.qxd 10/31/03 3:40 PM Page 212
212
Section I
Introduction
P earls 1. The NINDS trial showed clinical benefit at 90 days when intravenous r-tPA was administered less than 3 hours after symptom onset. The fact that there was no significant benefit seen at 24 hours suggests that the large MCA clot did not dissolve immediately. 2. The two largest randomized clinical trials of intraarterial thrombolysis are the PROACT I and PROACT II studies, which compared intra-arterial recombinant pro-urokinase plus intravenous heparin to placebo treatment plus intravenous heparin.5,15 These studies doubled the effective treatment window for thrombolysis from 3 to 6 hours, when the thrombolytic agent is given intra-arterially by a catheter embedded in the clot. 3. Symptomatic intracerebral hemorrhages occurred in 15% of PROACT patients and 10% of PROACT II patients who received intra-arterial pro-urokinase. 4. Unfortunately, GDC placement does not achieve complete, durable occlusion of all intracranial aneurysms. One recent study has shown a 20% rate of incomplete occlusion and an 8.6% rate of delayed recanalization at 3 years follow-up. 5. The most dreaded intraoperative complication with endovascular surgery is aneurysmal rupture. This complication is far more likely to occur in previously ruptured aneurysms than in an unruptured aneurysm. 6. Thromboembolism is a risk both during coil placement and in the postoperative period. This is the most frequent complication, occurring in 2.5% to 5% of GDC procedures.
7. Delayed ischemic neurologic deficits secondary to cerebral vasospasm will develop in as many as one third of patients who survive the initial hemorrhage.30 Cerebral vasospasm has been shown to lead to a 1.5to threefold increase in mortality at 2 weeks following SAH. 8. Recent MRI studies have shown that vasospasm might also be the cause of many strokes that do not cause gross neurologic dysfunction detectable during the patient’s initial stay in the intensive care unit, but are a cause of the more subtle memory and personality changes frequently seen in these patients. 9. Patients who have symptomatic vasospasm are in danger of irreversible cerebral infarction. The quicker blood flow is improved, the better the outcome. This has been demonstrated by a recent study showing a 2-hour window for blood flow restoration for maximum improvement in clinical outcome. 10. Vessel dissection or rupture from balloon angioplasty usually has devastating consequences. In one large series, 4% of patients undergoing balloon angioplasty died as a result of intraoperative vessel rupture. 11. Fifty percent of symptomatic AVMs present with an intracranial hemorrhage.45 The next most common symptoms at presentation are seizures and focal neurologic deficits. 12. Type A CCFs are most frequently traumatic in nature. The most common cause is motor vehicle accidents, followed by falls and penetrating injuries.
References 1. Gorelick PB: Stroke prevention therapy beyond antithrombotics: Unifying mechanisms in ischemic stroke pathogenesis and implications for therapy: An invited review. Stroke 2002;33:862–875. 2. del Zoppo GJ, Poeck K, Pessin MS, et al: Recombinant tissue plasminogen activator in acute thrombotic and embolic stroke. Ann Neurol 1992;32:78–86. 3. Rosner G, Heiss W: Survival of cortical neurons as a function of residual flow and duration of ischemia. J Cereb Blood Flow Metab 1983;3:s393–s394. 4. Overgaard K: Thrombolytic therapy in experimental embolic stroke. Cerebrovasc Brain Metab Rev 1994;6:257–286. 5. Furlan A, Higashida R, Wechsler L, et al: Intra-arterial prourokinase for acute ischemic stroke. The PROACT II study: A randomized clinical trial. Prolyse in Acute Cerebral Thromboembolism. JAMA 1999;282:2003–2011. 6. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group: Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333:1581–1587. 7. Clark WM, Wissman S, Albers GW, Jhamanadas JH, Madden KP, Hamilton S: Recombinant tissue-type plasminogen activator (Alteplase) for ischemic stroke 3 to 5 hours after symptom onset. The
8.
9.
10.
11.
12.
13.
ATLANTIS Study: A randomized controlled trial. Alteplase Thrombolysis for acute Noninterventional Therapy in Ischemic Stroke. JAMA 1999;282:2019–2026. Donnan GA, Davis SM, Chambers BR, et al: Streptokinase for acute ischemic stroke with relationship to time of administration: Australian Streptokinase (ASK) Trial Study Group. JAMA 1996;276: 961–966. Hacke W, Kaste M, Fieschi C, et al: Randomised double-blind placebocontrolled trial of thrombolytic therapy with intravenous alteplase in acute ischemic stroke (ECASS II). Second European-Australasian Acute Stroke Study Investigators. Lancet 1998;352:1245–1251. Hacke W, Kaste M, Fieschi C, et al: Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA 1995;274:1017–1025. Multicentre Acute Stroke Trial-Italy (MAST-I) Group: Randomised controlled trial of streptokinase, aspirin, and combination of both in treatment of acute ischaemic stroke. Lancet 1995;346:1509–1514. Multicentre Acute Stroke Trial-Europe Study Group: Thrombolytic therapy with streptokinase in acute ischemic stroke. N Engl J Med 1996;335:145–150. Tomsick T, Brott T, Barsan W, et al: Prognostic value of the hyperdense middle cerebral artery sign and stroke scale score before ultraearly thrombolytic therapy. Am J Neuroradiol 1996;17:79–85.
W9418-07.qxd 10/31/03 3:40 PM Page 213
Chapter 7 14. Wolpert SM, Bruckmann H, Greenlee R, Wechsler L, Pessin MS, del Zoppo GJ: Neuroradiologic evaluation of patients with acute stroke treated with recombinant tissue plasminogen activator. The rt-PA Acute Stroke Study Group. Am J Neuroradiol 1993;14:3–13. 15. del Zoppo GJ, Higashida RT, Furlan AJ, Pessin MS, Rowley HA, Gent M: PROACT: A phase II randomized trial of recombinant prourokinase by direct arterial delivery in acute middle cerebral artery stroke. Stroke 1998;29:4 –11. 16. Gönner F, Remonda L, Mattle H, et al: Local intra-arterial thrombolysis in acute ischemic stroke. Stroke 1998;29:1894–1900. 17. Ueda T, Sakaki S, Nochide I, Kumon Y, Kohno K, Ohta S: Angioplasty after intra-arterial thrombolysis for acute occlusion of intracranial arteries. Stroke 1998;29:2568–2574. 18. Ringer AJ, Qureshi AI, Fessler RD, et al: Angioplasty of intracranial occlusion resistant to thrombolysis in acute ischemic stroke. Neurosurgery 2001;48:1282–1288. 19. Becker KJ, Monsein LH, Ulatowski J, et al: Intraarterial thrombolysis in vertebrobasilar occlusion. Am J Neuroradiol 1996;17:255–262. 20. The Abciximab in Ischemic Stroke Investigators: Abciximab in acute ischemic stroke: A randomized, double-blind, placebo- controlled, dose-escalation study. Stroke 2000;31:601–609. 21. Guglielmi G, Vinuela F, Sepetka I, Macellari V: Electrothrombosis of saccular aneurysms via endovascular approach. Part 1: Electrochemical basis, technique, and experimental results. J Neurosurg 1991;75:1–7. 22. Thomas JE, Armonda RA, Rosenwasser RH: Endosaccular thrombosis of cerebral aneurysms: Strategy, indications, and technique. Neurosurg Clin North Am 2000;11:101–121. 23. Connors JJ, Wojak JC: Intracranial aneurysms: General considerations. In Connors JJ, Wojak JC (eds): Interventional Neuroradiology. Philadelphia, WB Saunders, 1999. 24. Derdeyn CP, Cross DT 3rd, Moran CJ, et al: Postprocedure ischemic events after treatment of intracranial aneurysms with Guglielmi detachable coils. J Neurosurg 2002;96:837–843. 25. Mericle RA, Lanzino G, Wakhloo AK, Guterman LR, Hopkins LN: Stenting and secondary coiling of intracranial internal carotid artery aneurysm: Technical case report. Neurosurgery 1998;43:1229–1234. 26. Mericle RA, Wakhloo AK, Rodriguez R, Guterman LR, Hopkins L: Temporary balloon protection as an adjunct to endosaccular coiling of wide-necked cerebral aneurysms: Technical note. Neurosurgery 1997;41:975–978. 27. Hayakawa M, Murayama Y, Duckwiler GR: Natural history of the neck remnant of a cerebral aneurysm treated with the Guglielmi detachable coil system. J Neurosurg 2000;93:561–568. 28. Lempert TE, Malek AM, Halbach VV, et al: Endovascular treatment of ruptured posterior circulation cerebral aneurysms. Stroke 2000;31: 100–110. 29. Fessler RD, Ringer AJ, Qureshi AI, Guterman LR, Hopkins LN: Intracranial stent placement to trap an extruded coil during endovascular aneurysm treatment: Technical note. Neurosurgery 2000;46:248–251. 30. Kassel NF, Torner JC, Haley EC, Jane JA, Adams HP, Kongable GL: The International Cooperative Study on the Timing of Aneurysm Surgery. Part 1: Overall management results. J Neurosurg 1990;73:18–36. 31. Biller J, Godersky JC, Adams HP Jr: Management of aneurysmal subarachnoid hemorrhage. Stroke 1998;19:1300–1305. 32. Treggiari-Venzi MM, Suter PM, Romand JA: Review of medical prevention of vasospasm after aneurysmal subarachnoid hemorrhage: A problem of neurointensive care. Neurosurgery 2001;48:249–261. 33. Yalamanchili K, Rosenwasser RH, Thomas JE, Liebman K, McMorrow C, Gannon P: Frequency of cerebral vasospasm in patients treated with endovascular occlusion of intracranial aneurysm. Am J Neuroradiol 1998;19:553–558. 34. Shimoda M, Takeuchi M, Tominaga J, et al: Asymptomatic vs. symptomatic infarcts from vasospasm in patients with subarachnoid hemorrhage: Serial magnetic resonance imaging. Neurosurgery 2001;49: 1341–1348. 35. Song JK, Eskridge JM: Intracranial angioplasty and thrombolysis. Neurosurg Clin North Am 2000;11:49–65.
Basic Endovascular Neurosurgery and Neuroradiology
213
36. Macdonald RL, Weir B: Medical aspects of vasospasm, cerebral vasospasm. San Diego, Academic Press, 2001. 37. Rosenwasser RH, Armonda RA, Thomas JE, Benitez RP, Gannon PM, Harrop J: Therapeutic modalities for the management of cerebral vasospasm: Timing of endovascular options. Neurosurgery 1999;44:975–980. 38. Cook P, James I: Drug therapy: Cerebral vasodilators (first of two parts). N Engl J Med 1981;305:1508–1513. 39. Milburn JM, Moran CJ, Cross DT 3rd, Diringer MN, Pilgram TK, Dacey RG Jr: Increase in diameters of vasospastic intracranial arteries by intraarterial papaverine. J Neurosurg 1998;88:38–42. 40. Connors JJ, Wojak JC: Endovascular therapy of postsubarachnoid hemorrhage vasospasm. In Connors JJ, Wojak JC (eds): Interventional Neuroradiology. Philadelphia, WB Saunders, 1999. 41. Polin RS, Hansen CA, German P, Chadduck JB, Kassel NF: Intraarterially administered papaverine for the treatment of symptomatic cerebral vasospasm. Neurosurgery 1998;42:1256–1267. 42. Firlik AD, Kaufmann AM, Jungreis CA, Yonas H: Effect of transluminal angioplasty on cerebral blood flow in the management of symptomatic vasospasm following aneurysmal subarachnoid hemorrhage. J Neurosurg 1997;86:830–839. 43. Eskridge JM, McAuliffe W, Song JK, et al: Balloon angioplasty for the treatment of vasospasm: Results of first 50 cases. Neurosurgery 1998;42:510–516. 44. Berenstein A, Lasjaunias P: Surgical Neuroangiography. Berlin, Springer, 1991. 45. Drake CG: Cerebral arteriovenous malformations: Considerations for and experience with surgical treatment in 166 cases. Clin Neurosurg 1979;26:145–208. 46. Crawford PM, West CR, Chadwick DW, et al: Arteriovenous malformations of the brain: Natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986;49:1–10. 47. Brown RD, Wiebers DO, Forbes G, et al: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 1988;68:352–357. 48. Spetzler RF, Martin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–483. 49. Hamilton MG, Spetzler RF: The prospective application of a grading system for arteriovenous malformations. Neurosurgery 1994;34:2–6. 50. Friedman WA, Bova FJ: Linear accelerator radiosurgery for arteriovenous malformations. J Neurosurg 1992;77:832–841. 51. Pollock BE, Lunsford LD, Kondziolka D, Maitz A, Flickinger JC: Patient outcomes after stereotactic radiosurgery for “operable” arteriovenous malformations. Neurosurgery 1994;35:1–8. 52. Wikholm G, Lundqvist C, Svendsen P: The Göteborg cohort of embolized cerebral arteriovenous malformations: A 6-year follow-up. Neurosurgery 2001;49:799–806. 53. Pan DH, Chung WY, Guo WY, Wu H, Liu KD, Shiau CY, Wang LW: Stereotactic radiosurgery for the treatment of dural arteriovenous fistulas involving the transverse-sigmoid sinus. J Neurosurg 2002;96: 823–829. 54. The n-BCA Trial Investigators: N-butyl cyanoacrylate embolization of cerebral arteriovenous malformations: Results of a prospective, randomized, multi-center trial. Am J Neuroradiol 2002;23:748–755. 55. al Rodhan NR, Sundt TM Jr, Piepgras DG, et al: Occlusive hyperemia: A theory for the hemodynamic complications following resection of intracerebral arteriovenous malformations. J Neurosurg 1993;78: 167–175. 56. Wilson CB, Hieshima G: Occlusive hyperemia: A new way to think about an old problem. J Neurosurg 1993;78:165–166. 57. Spetzler RF, Wilson CB, Weinstein P, et al: Normal perfusion pressure breakthrough theory. Clin Neurosurg 1978;25:651–672. 58. Newton TH, Cronqvist S: Involvement of dural arteries in intracranial arteriovenous malformations. Radiology 1969;93:1071–1078. 59. Sundt TM, Jr., Piepgras DG: The surgical approach to arteriovenous malformations of the lateral and sigmoid dural sinuses. J Neurosurg 1983;59:32–39.
W9418-07.qxd 10/31/03 3:40 PM Page 214
214
Section I
Introduction
60. Cognard C, Gobin YP, Pierot L, et al: Cerebral dural arteriovenous fistulas: Clinical and angiographic correlation with a revised classification of venous drainage. Radiology 1995;194:671–680. 61. Awad IA, Little JR, Akarawi WP, Ahl J: Intracranial dural arteriovenous malformations: Factors predisposing to an aggressive neurological course. J Neurosurg 1990;72:839–850. 62. van Dijk JM, terBrugge KG, Willinsky RA, Wallace MC: Clinical course of cranial dural arteriovenous fistulas with long-term persistent cortical venous reflux. Stroke 2002;33:1233–1236. 63. Lasjaunias P, Chiu M, ter Brugge K, Tolia A., Hurth M, Bernstein M: Neurological manifestations of intracranial dural arteriovenous malformations. J Neurosurg 1986;64:724–730. 64. Vinuela F, Fox AJ, Pelz DM, Drake CG: Unusual clinical manifestations of dural arteriovenous malformations. J Neurosurg 1986;64:554–558. 65. Halbach VV, Higashida R, Hieshima GB, Goto K, Norman D, Newton TH: Dural fistulas involving the transverse and sigmoid sinuses: Results of treatment in 28 patients. Radiology 1987;163:443–447. 66. Barrow DL, Spector RH, Braun IF, Landman JA, Tindall SC, Tindall GT: Classification and treatment of spontaneous carotid-cavernous sinus fistulas. J Neurosurg 1985;62:248–256. 67. Larsen D, Higashida R, Connors JJ: Treatment of carotid-cavernous sinus fistulae. In Connors JJ, Wojak JC (eds): Interventional Neuroradiology. Philadelphia, WB Saunders, 1999. 68. Farley MK, Clark RD, Fallor MK, Geggel HS, Heckenlively JR: Spontaneous carotid-cavernous fistula and the Ehlers-Danlos syndrome. Ophthalmology 1983;90:1337–1342. 69. Koo AH, Newton TH: Pseudoxanthoma elasticum associated with carotid rete mirabile. Am J Roentgenol Radium Ther Nucl Med 1972;116:16 –22. 70. Halbach VV, Hieshima GB, Higashida RT, Reicher M: Carotid cavernous fistulae: Indications for urgent treatment. AJR 1987;149: 587–593. 71. Higashida RT, Hieshima GB, Halbach VV, Bentson JR, Goto K: Closure of carotid cavernous sinus fistulae by external compression of the carotid artery and jugular vein. Acta Radiol Suppl 1986;369:580–583. 72. Higashida R, Halbach VV, Tsai FY, et al: Interventional neurovascular treatment of traumatic carotid and vertebral artery lesions: Results in 234 cases. AJR 1989;153:577–582. 73. CAVATAS Investigators: Endovascular vs. surgical treatment in patients with carotid stenosis in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS): A randomised trial. Lancet 2001; 357:1729–1737. 74. Hobson RW, Weiss DG, Fields WS, et al: Efficacy of carotid endarterectomy for asymptomatic carotid stenosis. The Veterans Affairs Cooperative Study Group 5. N Engl J Med 1993;328:221–227. 75. Inzitari D, Eliasziw M, Gates P, Sharpe BL, Chan RK, Meldrum HE, Barnett HJ: The causes and risk of stroke in patients with asymptomatic internal-carotid artery stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 2000;342:1693–1700. 76. Sacco RL: Clinical practice. Extracranial carotid stenosis. N Engl J Med 2001;345:1113–1118. 77. Barnett HJ, Taylor DW, Eliasziw M, et al: Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 1998;339:1415–1425. 78. Hobson RW: Update on the Carotid Revascularization Endarterectomy versus Stent Trial (CREST) protocol. J Am Coll Surg 2002;194:S9–14.
79. North American Symptomatic Carotid Endarterectomy Trial Collaborators: Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med 1991;325: 445–453. 80. Eisenhauer AC: Subclavian and innominate revascularization: Surgical therapy versus catheter-based intervention. Curr Intervent Cardiol Rep 2000;2:101–110. 81. Theron JG, Payelle GG, Coskun O, Huet HF, Guimaraens L: Carotid artery stenosis: Treatment with protected balloon angioplasty and stent placement. Radiology 1996;201:627–636. 82. Henry M, Amor M, Henry I, et al: Carotid stenting with cerebral protection: First clinical experience using the PercuSurge GuardWire system. J Endovasc Surg 1999;6:321–331. 83. Parodi JC, La Mura R, Ferreira LM, et al: Initial evaluation of carotid angioplasty and stenting with three different cerebral protection devices. J Vasc Surg 2000;32:1127–1136. 84. Schoser BG, Heesen C, Eckert B, et al: Cerebral hyperperfusion injury after percutaneous transluminal angioplasty of extracranial arteries. J Neurol 1997;244:101–104. 85. McCabe DJ, Brown MM, Clifton A: Fatal cerebral reperfusion hemorrhage after carotid stenting. Stroke 1999;30:2483–2486. 86. Meyers PM, Higashida RT, Phatouros CC, et al: Cerebral hyperperfusion syndrome after percutaneous transluminal stenting of the craniocervical arteries. Neurosurgery 2000;47:335–343. 87. Gorelick PB, Caplan LR, Langenberg P, et al: Clinical and angiographic comparison of asymptomatic occlusive cerebrovascular disease. Neurology 1988;38:852–858. 88. Borozan PC, Schuller JJ, LaRosa MP, et al: The natural history of isolated carotid siphon stenosis. J Vasc Surg 1984;1:744. 89. The Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) Study Group: Prognosis of patients with symptomatic vertebral or basilar artery stenosis. Stroke 1998;29:1389–1392. 90. Bogousslavsky J, Barnett HJM, et al: Atherosclerotic diseases of the middle cerebral artery. Stroke 1986;17:1112–1120. 91. Craig DR, Meguro K, Watridge C, Robertson JT, Barnett HJ, Fox AJ: Intracranial internal carotid artery stenosis. Stroke 1982;13:825– 828. 92. Chimowitz MI, Kokkinos J, Strong J, et al: The Warfarin-Aspirin Symptomatic Intracranial Disease Study. Neurology 1995;45:1488–1493. 93. EC/IC Bypass Study Group: Failure of extracranial-intracranial arterial bypass to reduce the risk of ischemic stroke: Results of an international randomized trial. N Engl J Med 1985;313:1191–2000. 94. Hyodo A, Matsumaru Y, Anno I, et al: Percutaneous transluminal angioplasty for atherosclerotic stenosis of the intracranial cerebral arteries: Results with more than one year follow-up (abstract). Intervent Neuroradiol 1997;3:38. 95. Clark WM, Barnwell SL, Nesbit G, O’Neill OR, Wynn ML, Coull BM: Safety and efficacy of percutaneous transluminal angioplasty for intracranial atherosclerotic stenosis. Stroke 1995;26:1200–1204. 96. Connors JJ, III, Wojak JC: Percutaneous transluminal angioplasty for intracranial atherosclerotic lesions: Evolution of technique and shortterm results. J Neurosurg 1999;91:415–423. 97. Mori T, Kazita K, Chokyu K, et al: Short-term arteriographic and clinical outcome after cerebral angioplasty and stenting for intracranial vertebrobasilar and carotid atherosclerotic occlusive disease. Am J Neuroradiol 2000;21:249–254. 98. Marks MP, Marcellus M, Norbash AM, et al: Outcome of angioplasty for atherosclerotic intracranial stenosis. Stroke 1999;30:1065–1069.
W9418-08.qxd 10/31/03 3:41 PM Page 215
Chapter 8 Multiple Organ System Injuries Resulting from and Critical Care of Isolated Severe Central Nervous System Trauma John L. D. Atkinson, MD, FACS and Jack E. Wilberger, Jr., MD
Introduction
Pathophysiology of Head Injury
Head injury remains a serious public health problem, occurring at a rate of 150 per 100,000 population per year in the United States. The most current data indicate that head injury accounts for more than 20,000 deaths and 50,000 permanent disabilities each year.1,2 With an understanding of the mechanisms of secondary injury and the development of appropriate critical care treatment strategies, there has been a significant decline in head injury mortality over the past two decades. Indeed, contemporary multicenter studies are reporting mortality rates as low as 17%.3 In this regard, two important questions remain unanswered: Can such favorable outcomes be achieved outside rigorous scientific studies in routine clinical practice? Can critical care treatment strategies be developed to further improve outcomes? It is likely that adherence to the management principles enumerated in the Guidelines for the Management of Severe Head Injury coupled with the concept of targeted therapy and improved brain monitoring will result in a sustained lowering of mortality after head injury. Such principles and technologies can now be readily applied in most current neurointensive care practice settings. For the future, however, improved outcomes may require refocusing attention on neuro-protective agents and identification of the genetic factors resulting in repair and recovery after head injury.
Patients with severe traumatic central nervous system injury may experience immediate alterations in cerebral and systemic physiology within the first 10 minutes of the postinjury period. These alterations not only determine life or death at the scene of injury, but set into motion critical physiologic and biochemical cascades that influence medical care and determine outcome. This critical phase of injury encompasses the pathophysiologic sequelae of apnea and catecholamine surge responsible for multiple organ injuries early post–central nervous system trauma. The following sections discuss the pathophysiology and treatment along with potential future therapeutic options. Critical Phase of Head Injury Apnea and Catecholamine Surge The critical phase of head injury is arbitrarily defined in this manuscript as the first 10 minutes after the onset of severe head injury, as patients will live or die at the scene based on the pathophysiology that occurs during this period. The phases of severe head injury outlined in Figure 8-1 and Table 8-1 are extrapolated from JD Miller’s review of head injury4 and Overgaard and Tweed’s5 summary comments. These authors note that significant ischemic and hypoxic brain injury occurs before hospital admission, and they emphasize the importance of this critical phase in patient outcome. 215
W9418-08.qxd 10/31/03 3:41 PM Page 216
216
Section I
Introduction
Figure 8-1. Outline of the phases of head injury. The first 10 minutes are aptly termed the “critical phase” because if respirations do not resume (spontaneously or otherwise), death will ensue due to hypoxic cardiovascular collapse.
Other phases in addition to this “critical phase” are depicted as well, and listings of various other mechanisms of injury during severe head injury management are outlined. There are two immediate pathophysiologic events occurring with the onset of severe head injury that markedly affect subsequent outcome: head injury–induced apnea, and a stress-related massive sympathetic discharge (Fig. 8-2). In combination, the effects of hypoxia, hypercarbia, acidosis, and blood pressure surge, as well as the direct effects of catecholamines on tissue, all lead to a synergistic injury effect in the host. The extent of the catecholamine surge and apnea occurring after severe head injury is directly related to the amount
Table 8-1 Phases of Head Injury Critical phase Apnea: Always occurs with concussive head injury; the more energy delivered to the brain, the longer the subsequent apnea, and the poorer the respiratory recovery. • No respiratory recovery; dead at the scene unless resuscitate early. • Long apnea and poor respiratory recovery fosters hypoxia and hypercarbia leading to early, massive brain swelling from increased cerebral blood volume due to vasodilatation from hypercarbia; profound hypoxia may take minutes to occur but markedly effects the neuronal environment. • Early intervention may dramatically alter course. Catecholamine surge: Markedly elevated blood pressure occurs immediately due to massive sympathetic discharge; this augments hypercarbia-induced cerebrovascular dilatation and promotes early vasogenic edema, endothelial injury, BBB disruption, and progressively increased ICP occurs depending on the magnitude of epinephrine release and CO2 retention. Intense vasoconstriction of other susceptible vascular beds produces ischemic gastric mucosal ulceration (stress ulcers, previously known as Cushing’s ulcers), and neurogenic pulmonary edema, along with catecholamine tissue injury such as myocardial necrosis. Primary brain injury: Fracture or tearing of bones, meninges, brain parenchyma, and blood vessels occurs to varying degrees as forces converge through various vectors. Secondary brain injury mechanisms initiated: Vasogenic edema; astrocytic swelling with altered EAA uptake and K+ alterations, etc.; hemorrhage; neuronal and oligodendroglial ischemia; thromboplastin release with altered coagulation; SAH and progressive ICP, various molecular cascades, etc. Exponential phase Respiratory recovery frequently leads to hyperventilatory drive; any respiratory recovery may prolong life; early ventilation intervention may alter outcome. • Catecholamine surge abates and blood pressure falls to mid- or high-normal levels. However, brain may be massively swollen due to hypercarbic-induced vasodilatation and subsequent BP surge, and any hemorrhage may have been augmented; ischemic injury to gastric mucosa, myocardium, and neurogenic pulmonary edema may present complications. • Molecular cascades continue to progress such as buildup of excess excitatory amino acids, lipid peroxidation, possible apoptosis of select cell populations, etc. • Hemorrhage progresses or injured vessels thrombose; ICP may continue to increase from progressive edema, mass effect, SAH; nonviable or ischemic brain undergoes cellular swelling; marginally compensated parenchymal cells live or die depending on cellular milieu; diffuse axonal injury matures; seizures may augment cerebral blood volume and ischemic cascade. Plateau phase ICP stabilizes or may increase due to gradual transition from vasogenic edema to cellular edema; delayed intracerebral hematoma may develop from dead or injured brain and blood vessels; dead parenchyma promotes edema; SAH and RBC lysis may precipitate vascular injury which may lead to vasospasm; molecular cascades may slow or stop; 75% of deaths occur in the first 48 to 72 hours. Resolution phase Collagen and glial repairs progress; SAH-induced vasospasm may evoke ischemia and infarction; previously infarcted parenchyma maximally swells and resolves; continued risk for 1 to 2 weeks of delayed intracerebral hematoma; cellular edema becomes greater component of swelling vs. vasogenic edema and may slowly subside; post-traumatic hydrocephalus may evolve short or long term. BP, blood pressure; EAA, excitatory amino acid; ICP, intracranial pressure; RBC, red blood cell; SAH, subarachnoid hemorrhage. Data from Miller JD: Head injury and brain ischaemia—Implications for therapy. Br J Anaesth 1985;57:120–129; and Overgaard J, Tweed WA: Cerebral circulation after head injury. Part 4: Functional anatomy and boundary-zone flow deprivation in the first week of traumatic coma. J Neurosurg 1983;59:439– 446.
W9418-08.qxd 10/31/03 3:41 PM Page 217
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
217
Figure 8-2. Severe head injury sequelae of disturbance in medullary-driven diaphragmatic ventilation and massive sympathetic stimulation. The magnitude and severity of both events are directly proportional to the magnitude of energy transmitted to the brainstem.
of energy transmitted to the brainstem. As a result, a significant number of severely head-injured patients live or die at the scene based on whether or not there is resumption of breathing. Prolonged apnea-induced hypoxic brain and cardiac injury, augmented by markedly elevated stress catecholamines, will increase mortality and, in survivors, morbidity. Therefore, the substrate from which medical and surgical therapy must begin is determined at the scene before hospital-based medical assistance is rendered. The modern aggressive care delivered in hospital and directed at optimizing cerebral perfusion, while scientifically grounded, may in many cases be too little, too late. The clear implication here is that the prehospital care provided by the emergency medical services system, as well as bystander witnesses to the injury, will significantly influence outcome for better or worse (see Chapter 14, Prehospital Care of the CNS Injured Patient). The following sections will critically examine apnea and catecholamine surge induced in the critical phase of head injury. From this analysis, conclusions for clinical management will be drawn. Apnea. Apnea resulting from head concussive injury has been recognized experimentally for longer than a century. Koch and Filehne6 reported in 1874 that repeated small
blows to the head of animals led to death by respiratory paralysis without any visible structural abnormality in the brain. Polis7 showed in 1894 that concussive head injury in the cat, dog, and rabbit were followed by respiratory arrest and a significant increase in mean arterial blood pressure. If respirations did not recover, the animal died, even though there were no gross anatomic lesions in the brains. In 1896, Kramer8 also reported that animals receiving a blow to the head experienced respiratory paralysis. In 1927 Miller9 repeated Polis’ work with identical findings. However, it was Denny-Brown10 in 1941 who clearly revealed that immediate death from most severe experimental head injuries was due to respiratory failure. Using respirometric methods, he demonstrated that increasing degrees of energy delivered to the brain produced increasing duration of apnea. A light blow produced a respiratory gasp, a moderate blow produced varying degrees of apnea with respiratory recovery, and heavy blows produced respiratory arrest and subsequent death due to hypoxic cardiovascular collapse. There were no observable lesions in the brains when sectioned. DennyBrown concluded that the response was brainstemmediated, as it was noted to occur in decerebrate preparations of animals. He also thought that it was energydependent, based on the force and rate of change (acceleration) transmitted to the brainstem, and had nothing to do
W9418-08.qxd 10/31/03 3:41 PM Page 218
218
Section I
Introduction
with physical disruption of the brain or development of space-occupying hematomata. In 1944, Walker11 followed Denny-Brown’s work with a classic report on the physiologic basis of concussion in a number of different animal species. Using hammer, weight drop, and projectile techniques, he confirmed that the respiratory response was energydependent and characterized by a respiratory gasp at low energy, and apnea of long duration with higher energy. Walker also demonstrated that this response was brainstemmediated, and not related to elevated intracranial pressure. Sullivan and Becker12 showed that animals subjected to fluid percussion injuries exceeding the lethal threshold for apnea could recover normal function by both clinical and electroencephalographic (EEG) parameters if they added respiratory support. Gennarelli13 developed a headaccelerating machine and found that in primates, increasing the magnitude of acceleration caused apneustic changes in respiration, and at higher forces, the animals died without ventilatory support. In short, the more energy delivered to the brainstem, the more likely the animal was to die. Carey14 concluded the same with a projectile mechanism of brain injury in cats. Respiratory arrest was a constant physiologic response, even though the missile did not injure the brainstem directly. The duration of apnea was directly proportional to the energy transmitted to the brain. He concluded that the integrity of the medullary respiratory center after head injury determines, very early, whether the animal will live or die, regardless of the degree of other parenchymal injury. Carey further noted that if respiratory support were delayed, secondary hypoxic brain and cardiac damage ensued, resulting in marked morbidity or death. Anderson and associates15 combined mechanical brain injury with controlled postinjury hypoventilation followed by resumption of normal respiratory parameters in an attempt to simulate a more realistic model of head injury. These investigators found marked alterations in cerebral blood flow, cerebral metabolic rate of glucose, and brain ischemia not seen with either head injury or hypoxic injury alone. This was further delineated by Ito and associates,16 showing that weight drop brain injury–induced hypoventilation and hypotension in laboratory animals produced marked brain ischemia. Diffusion weighted magnetic resonance imaging (MRI) identified aberrations that were not visualized if only head injury, without hypoventilation or hypotension, occurred. Kim and associates17 and Levasseur and associates18 produced graded respiratory responses to fluid percussion head injury, and showed a marked increase in fatalities in ethanol pretreated animals who were subjected to this model of injury. They postulated that increased alcohol-related traumatic fatalities were due to synergistic respiratory depression as a result of head injury combined with the well-known depressant effect of ethanol. Others have also substantiated the increased morbidity and mortality resulting from head injury when combined with
ethanol.19–21 Both Kim and associates17 and Levasseur and colleagues18 verified that mechanical ventilation after head injury is essential in reducing acute mortality after experimental head injury. To date, irrespective of the methods used, all experimental head injury in a spontaneously breathing model has consistently produced apnea.22 The degree of respiratory paralysis and recovery has repeatedly been shown to be directly related to the amount of energy delivered to the brain (Fig. 8-3).22 In fact, the lethality of the respiratory response after head injury is used as a benchmark for the evaluation of new techniques in experimental brain injury model.23 Despite overwhelming experimental evidence of apnea and dysfunctional respiration immediately after head injury, there is less corroborating evidence derived from clinical studies. Johnson and associates24 examined nonaccidental closed head injury (i.e., child abuse) in 28 children. These investigators found 57% had a verifiable history of apnea before hospitalization. They concluded that trauma-induced apnea causes cerebral hypoxia and ischemia, which proves to be more fundamental to outcome in these patients than the mechanism of primary brain injury itself (i.e., subdural hematoma, subarachnoid hemorrhage, diffuse axonal injury, or contusions). Severe closed head injury–associated hypoxia (19%), hypotension (24%), or both (7%) were reported in the Traumatic Coma Data Bank as strong predictors of morbidity and mortality.25–27 These percentages are certainly under-representative, as they were recorded on admission to the emergency department (ED) following ambulanceassisted resuscitation, and do not reflect changes that might have been noted at the scene of head injury. However, the most powerful testimony to the association of apnea with clinical head injury results from eyewitness accounts recorded by physicians who were at the scene of injury.28 In two separate patients with head injury, both were rescued at the scene by early ventilation. Both patients initially had Glasgow Coma Scale scores of 3, with fixed dilated pupils, no corneal reflexes, and no respirations or pulse. Both patients made uneventful recoveries. The authors summarized that without early ventilatory assistance, these two patients would almost certainly have died or suffered serious brain damage due to respiratory failure as a result of their nonlethal head injuries. Catecholamine Surge. It is irrefutable that catecholamine surge is a “stress response” to head injury, and as such represents an important and neglected contribution to head injury morbidity and mortality. Marked elevations in blood pressure and heart rate are universal responses to head injury and have been documented for longer than a century. For example, Polis, in 1894,7 described marked elevation in blood pressure and heart rate with experimental head injury. Furthermore, blood pressure and heart rate elevation is discussed as a concomitant response of all previously discussed studies regarding head injury–associated apnea.7,10–23
W9418-08.qxd 10/31/03 3:41 PM Page 219
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
219
Figure 8-3. Rat model of fluid percussion head injury with increasing magnitudes of energy delivered to the brain and corresponding responses of blood pressure, spirometry measured tidal volume, and intracranial pressure. A, 1.15 atm pressure delivered to the brain with 45 seconds of apnea followed by rapid resumption of normal respiratory pattern and no significant change in the mean arterial blood pressure or ICP. B, 2.2 atm pressure delivered to the brain with 45 seconds of apnea followed by very slow resumption of normal respiratory pattern with immediate marked elevation in blood pressure and rapid (1–3 minutes) elevation in ICP to 10x baseline level. C, 3.5 atm pressure delivered to the brain with immediate apnea never followed by any organized respiratory effort with immediate and marked elevation in blood pressure and ICP. Animals in this group will die from hypoxic cardiovascular collapse, but can be salvaged if mechanically ventilated. There were no space-occupying hematomas identified in any of the groups. (From Atkinson JLD, Anderson RE, Murray MJ: The early critical phase of severe head injury: Importance of apnea and dysfunctional respiration. J Trauma 1998;45:941–945, with permission.)
The blood pressure and dynamic response of the heart to head injury was labeled early in the past century as the “Cushing response”29 because of Cushing’s early work with experimentally induced intracranial mass lesions, itself a continuation of Spencer and Horsely’s previously performed work.30 However, the Cushing response, although mediated by a catecholamine surge, is a response to increased intracranial pressure and the diminution of cerebral perfusion. Subsequent early studies concentrated on this sympathetic discharge resulting from acutely elevated intracranial pressure (ICP). Grimson and asscociates31 reported in 1937 that the increase in blood pressure from increasing intracranial pressure could be abolished by total sympathectomy.
Freeman and Jeffers32 repeated this experiment by creating sudden increased intracranial pressure by injection of saline under pressure into the cisterna magna of dogs. They discovered that the systemic blood pressure response was prevented by sympathectomy.32 Roozerkrans and Van Zwieten33 were able to block the “Cushing response” by adrenergic blockade using phentolamine in a manner similar to Cushing’s blockade of the splanchnic response of the intestines to blood pressure surge by cocainization of the medullary centers.29 The previous studies demonstrate that elevated ICP results in activation of the sympathetic nervous system with subsequent marked catecholamine release. However, it has
W9418-08.qxd 10/31/03 3:41 PM Page 220
220
Section I
Introduction
become clear that massive catecholamine surge is an immediate brainstem mediated response to severe head injury in the absence of elevated ICP, as vividly described by Walker11 in brainstem preparation animals. Massive sympathetic discharge and subsequent marked elevation in mean arterial blood pressure and heart rate has proven to be a consistent immediate response in any experimental severe head injury model, such as fluid percussion injury,12 acceleration,13 projectile,14 or weight drop techniques.22 The benchmark experimental study conducted by Rosner in Becker’s laboratory at Medical College of Virginia34 confirmed what Beckman and Iams had concluded earlier.35 They documented as much as a 500-fold increase in plasma epinephrine and 100fold increase in plasma norepinephrine with severe head injury. The conclusion was that both epinephrine and norepinephrine plasma levels increased in direct response to the amounts of energy delivered to the brain and paralleled injury severity. The systolic arterial blood pressure directly correlated with the level of circulating catecholamines, was an instantaneous sequela to all but the most lethal of head injuries sequelae, and did not correlate with elevated ICP. In the clinical setting, patients often continue to have increased circulating catecholamines for many days after isolated head injury. Hörtnagl and associates36 found elevated plasma epinephrine and norepinephrine levels in 15 severely head-injured patients, and suggested that longstanding overactivity of the sympathetic nervous system was a characteristic feature in the clinical course of head injury. Sculte Am Esch and associates37 also found a hyperdynamic state after head injury due to over-activity of the sympathetic nervous system. Clifton and colleagues38 reported that, in patients with head injury alone, circulating catecholamine concentrations were significantly elevated, and blood pressure was directly proportional to the plasma concentration of norepinephrine. Subsequent clinical work from the same institution has shown that increased circulating catecholamines were common in head-injured patients, but did not correlate with either the GCS score or increased intracranial pressure.39,40 The investigators suggested that rational treatment in head-injured patients would include beta-blockade both to decrease hypertension as well as the catabolic effects of this hyperdynamic state.41 Increased circulating catecholamines have been confirmed in other stress conditions as well, such as subarachnoid hemorrhage,42,43 shock,44 or severe thermal injury,45 in addition to having been reconfirmed in clinical head injury.46,47 Serum catecholamine levels have even been used as a marker to assess the severity of head injury in the setting of alcohol-intoxicated patients in the ED.48 Recommended treatment of this hyperdynamic, sympathetically mediated response to head injury46,47 is betablockade.49–51 Many studies suggest that the early elevated ICP seen after severe head injury, in the absence of a space occupying mass, is due to a marked increase in cerebrovascular volume and
breakdown of the blood-brain barrier.52–56 It is logical to suggest that an apneic patient with profound hypercarbic and hypoxic cerebrovascular dilatation, in the setting of probable impaired autoregulation, and subjected to massive blood pressure elevation, would experience a rapid increase in cerebral blood volume, with resultant increased intracranial pressure. It may well be that in many patients without space-occupying hematomata, elevated ICP noted on admission to the ED has been fostered and augmented by these immediate post–head injury conditions. Figure 8-3 reveals that increasing energy transmitted to the brain yields different physiologic results. In Figure 8-3B, 2.2 atm of pressure is delivered with immediate marked elevation of blood pressure. However, the ICP elevation closely parallels hypoventilation-induced CO2 elevation and subsequent gradual cerebrovasodilatation with increased cerebrovascular volume. Figure 8-3C represents a rapid elevation of and significant elevation in ICP that is graphically representative of absent autoregulation in the setting of a massive blood pressure surge with a resultant marked increase in cerebral blood volume. There were no space-occupying hematomata in any of the experimental animals.22 In summary, head injury induces an instantaneous stress response via massive sympathetic discharge. The response depends on the “energy dose” applied to the brain, and clinically this hyperdynamic state may last for hours or even days in many patients with severe head injury. Negative ramifications of this catecholamine surge include stressinduced hyperglycemia, and injuries to the gastroduodenum, heart, and lungs, as will be discussed in the following sections. Anatomic Abnormalities Primary Neural Injury. Primary neural injury of the brain occurs at impact, resulting in the shearing and destruction of neurons, glial, and vascular structures by the mechanical forces imposed by the impact. This injury is irreversible and the neurologic deficits created by the primary injury have little potential for recovery. When epidural or subdural hematomata or brain contusions occur, the significance of primary neural injury may be unclear. Prompt and appropriate surgical and medical therapy can frequently improve the patient’s condition in these settings. Primary neural injury rarely results in immediate mortality, thus providing the opportunity for affecting survival through postinjury management. The degree and extent of primary neural injury are fixed at the time of the accident, and depend partially on the mechanism of injury. The only defense against the occurrence of primary neural injury is prevention. Extensive efforts at injury prevention have occurred in recent years. However, the success of programs such as decreased speed limits, mandatory seat belt laws, and passive restraints (air bags) have yet to be confirmed. In Pennsylvania, which has an extensive trauma registry program, there was a 30%
W9418-08.qxd 10/31/03 3:41 PM Page 221
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
decrease in severe traumatic brain injury in 1996 as compared to 1995 (unpublished data). The reason(s) for the significant decrease have not yet been analyzed. Ongoing educational programs such as Think First, are aimed at increasing the awareness of the association of risk-taking behaviors and the potential occurrence of trauma brain injury and its consequences. Secondary Neural Injury. Singular advancements in the crit-
ical care management of head injury have been based on the understanding of the concept that a large number of patients die not because of the initial injury to the brain but entirely because of additional secondary insults that occur following injury.3,25,57 Several clinical and multiple biochemical events occur following primary neural injury. These events cause secondary neural injuries that can convert a potentially recoverable traumatic brain injury into one resulting in either mortality or significant long-term disability. Hypotension and hypoxia are the most consistent predictors of poor outcome in head injury, presumably because of their role in facilitating the processes that lead to secondary neural injury. More than 30% of patients with severe traumatic brain injury present to the ED with significant hypotension (systolic blood pressure < 90 mm Hg) or hypoxia (PaO2 < 60 mm Hg). The occurrence of hypotension or hypoxia alone increases morbidity and mortality of traumatic brain injury by up to 50%.25,57 When hypoxia and hypotension occur together, the risk of a poor outcome increases to more than 60%. Adequate prehospital and ED resuscitation is of critical importance. Concern over the effects of hypotension underlies one of the cornerstones of head injury management: adequate volume resuscitation and maintenance of positive fluid balance. Given the concern that currently used intravenous fluids can potentially exacerbate cerebral edema especially in areas of blood-brain barrier breakdown, other resuscitative and maintenance fluids are under active investigation. Because of the small volumes required (approximately 8 mL/kg) to maintain an adequate blood pressure, hypertonic saline solution has been used as a resuscitative fluid in experimental studies.58 The biomechanical substrates of secondary injury have been extensively investigated. A cascade of events is triggered shortly after injury, having a variable time course and effect. This cascade is multifactorial, interrelated, and, to some extent, time dependent.59–62 More recently, investigators have focused on the genetic response to head injury. The research so far has been primarily descriptive in nature, however, a number of genes have been found to be up or down regulated following ischemia and trauma.3,63,64 While the concept of secondary injury is valid and generally applicable in a given head injury patient, it must be borne in mind that head injury is structurally, physiologically, metabolically, and biochemically heterogeneous. Additionally, the factors that may be operative in any individual patient may vary from one time to another. For example,
221
autoregulation may be present, absent, or altered in the same head-injured patient at various intervals of assessment.65 Such information is vitally important as some forms of current critical care treatment take advantage of the presence—i.e., cerebral perfusion pressure management—or absence—i.e., barbiturates—of autoregulation. Thus, the heterogeneity of head injury and the patient-dependent features of secondary injury provide a further challenge to develop more sophisticated monitoring systems and to be able to modify treatments based on patient specific parameters. Central Nervous System Trauma–Associated Pathologies Hyperglycemia and Brain Injury. It is known that the head injury–induced “stress response” results in a marked hyperglycemia mediated both by catecholamine surge and, likely, stress-induced cortisol release.34,66 It is also known that hyperglycemia existing before ischemic brain injury significantly worsens outcome.67–69 Serum glucose levels correlate directly with the magnitude of head injury and can rise to as high as 400 mg/dL by 1 hour postinjury in experimental animals.34 This relatively slow rise in glucose levels may be too late to adversely effect immediate postinjury brain hypoxic ischemia, but may significantly worsen any secondary brain injury ischemic damage, such as focal ischemia derived from evolving space-occupying lesions. In an animal model, Cherian and associates70,71 have convincingly demonstrated that cortical impact injury followed by ischemic insult in the presence of hyperglycemia significantly increases brain ischemic volume, contusion volume, and mortality, and decreases functional outcome in survivors. Hyperglycemia remains to be adequately assessed in the clinical head-injury setting. Based on the preceding conclusions, elevated glucose levels (>200 mg/dL) in patients at risk for central nervous system ischemia, due to oligemic blood flow or elevated ICP, may warrant insulin management. Maintenance of serum glucose between 80 and 150 mg/dL (“normal range”) should be strongly considered. Avoidance of glucose solutions would be prudent.69 Gastric Mucosal Ulceration. Gastroduodenal mucosal injury
frequently develops after severe head injury, and frank ulceration is common as well. Esophageal gastroduodenoscopy in patients soon after severe head injury reveals these lesions in as many as 90% of patients.72–74 These ulcerations occur very early after head injury, and are similar to catecholamine stress ulcers produced experimentally,72,75,76 or by other systemic stress injuries such as burns or sepsis.77,78 The mechanism may be a combination of factors, such as increased vagal activity resulting in increased gastric acidity or slowed gastric emptying, or increased circulating pancreatic polypeptide levels.79,80 However, the predominant theory of
W9418-08.qxd 10/31/03 3:41 PM Page 222
222
Section I
Introduction
gastric ulcerogenesis in head-injured patients is that catecholamine-induced vasoconstriction causes decreased splanchnic blood flow with resultant ischemic mucosal injury.81,82 Regardless of the mechanism, pharmacologic prophylaxis of further mucosal erosion should be considered in these patients.72–74 Prophylaxis of further mucosal erosions differs from the small percentage of these patients who actually suffer active gastrointestinal bleeding. Fortunately, most hemorrhages are minor. The first line of treatment is volume replacement. H2blockers, proton-pump inhibitors, or sucralfate should be used prophylactically. If significant blood loss occurs, or if bleeding continues despite therapy, then general surgery or gastroenterology specialists should be consulted early for endoscopic assessment and management.72 Myocardial Injury. Isolated head trauma-induced cardiac
injury is well documented in the pathology literature, and may be found at autopsy in up to two thirds of patients dying from acute head injury alone.83–85 The catecholamine surge accompanying head injury,86–88 spinal cord injury,89 or subarachnoid hemorrhage90–93 can precipitate these cardiac lesions. The injuries vary from disorganization, clumping, and necrosis of myocardial cells, to extensive areas of necrosis with hemorrhage, particularly in the ventricular septal region.84,85,94 The most common clinical manifestations are electrocardiographic abnormalities such as S-T segment depression or elevation, Q-T interval prolongation, and inversion of T waves.90,92,94,95 The lesions, if sufficiently extensive in their acute form, can contribute to, or cause, death.84,95,96 There is evidence that pretreatment with catecholamine blocking agents will markedly reduce the cardiac injury induced by stress catecholamine surge and help prevent further extension of myocardial damage.97,98 Any posttraumatic ECG changes warrant serial electrocardiographs, myocardial enzymes, and troponin-I evaluation. Transthoracic or transesophageal echocardiography is helpful in delineating wall motion abnormalities. Most cardiac arrhythmias are transient, and often do not require therapeutic intervention. However, enzyme elevation suggesting myocardial death or persistent arrhythmias should be assessed by an intensivist or cardiologist. Use of betareceptor blocking agents in patients at risk will diminish the extent of cardiac injury.91,97 Neurogenic Pulmonary Edema. A retrospective review of
the Traumatic Coma Data Bank suggests that pulmonary lesions, resulting from factors other than direct pulmonary trauma, are common.99,100 Although neurogenic pulmonary edema (NPE) may be mediated by a variety of mechanisms, the prevailing theory is that it is caused by a massive catecholamine surge.101 Marked pulmonary congestion and edema have been noted in other head injury autopsy series102 (Fig. 8-4) and are known to be a response in animals to massive autonomic discharge.103 This theory
Figure 8-4. Anteroposterior chest radiograph of a 46-year-old woman with severe head injury 2 hours after a single car rollover ejection from the vehicle. The radiograph shows marked neurogenic pulmonary edema.
suggests acute cardiac failure results from a catecholamineinduced increase in cardiac preload due to venoconstriction, and increased cardiac afterload by massive arterial constriction. In conjunction with catecholamine-induced cardiac injury, these mechanisms combine to produce edema in the lungs by increasing pulmonary capillary pressures. Early experiments showed that pulmonary edema could be prevented by adrenergic blocking agents or cordotomy, and that vagotomy by itself had little, if any, effect.103–106 The conclusions of the classic Vietnam war-era head injury series were that acute pulmonary edema after severe head injury occurred rapidly, resulted from a massive sympathetic discharge that produced a fluid shift from the periphery to the lungs, and was augmented by transient left ventricular failure and loss of left ventricular compliance from catecholamine-induced myocardial injury.101 Since that landmark article,101 evidence has grown in support of catecholamine surge as the major pathway in neurogenic pulmonary edema. Any cause of massive stressinduced autonomic discharge such as seizures, head injury, subarachnoid hemorrhage, or major stroke can produce NPE.107–109 Experimental models of head injury or subarachnoid hemorrhage continue to show massive sympathetic discharge as the primary mediator of neurogenic pulmonary edema,110–114 and this etiology permeates the literature in clinical and summary articles as well.113,115–117 Fulminant neurogenic pulmonary edema after severe head injury is uncommon. The most common cause of hypoxemia in the intensive care unit management of these patients is ventilation-perfusion mismatch due to atelectasis or aspiration. The diagnosis of NPE is supported by marked hypoxemia, appropriate chest radiograph findings (see Fig. 8-4), and exclusion of a cardiogenic cause for pulmonary
W9418-08.qxd 10/31/03 3:41 PM Page 223
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
edema. Maximizing oxygen delivery at the lowest inspired fraction of oxygen (FiO2) setting with the addition of positive end-expiratory pressure (PEEP; 10 to 15 cm H2O) if tolerated, and augmented by dobutamine (5 to 15 mg/kg/min), may help in refractory cases.107,108,115
Therapeutic Options for Postinjury Apnea MacIver and associates118 pleaded 40 years ago that the staggering morbidity and mortality of severe head injury could be markedly reduced simply by ventilating these patients at the accident scene. When this simple measure was undertaken, the 90% mortality in these patients was dramatically decreased to 40%. It can be seriously argued that the most significant reduction of trauma morbidity and mortality in this country came about with the 1966 publication of “Accidental Death and Disability: The Neglected Disease of Modern Society,” and the subsequent legislative development and continued refinement of emergency medical services (EMS).119 It comes as no surprise, when isolated head injury is reviewed within trauma resuscitation systems, that morbidity and mortality have continued to decline over the past decade almost exclusively due to improvement in the rapid response of EMS which, in turn, facilitates timely medical and surgical intervention.120 Even when comparing neurosurgical trauma care and patient outcome in countries as medically diverse as the United States and India, it again should come as no surprise that it is timely care in the field and expeditious transport to trauma centers that are the most important criteria separating the two countries’ head injury morbidity and mortality statistics.121 Brain ischemia remains a major focus of brain injury research. This investigative work has been stimulated by the neuropathologic findings reported by Graham and colleagues,122,123 who discovered that ischemic brain lesions were a common autopsy finding in humans who died following traumatic brain injury. Over half of the ischemic brain lesions found in head injury fatalities were of the arterial boundary zone type. Such lesions are most often seen in patients with a known clinical episode of hypoxemia or hypotension due to hypoxic cardiac failure. The explanation proposed by the investigators for this type of ischemic brain injury was that the pathophysiologic events likely occurred immediately after head injury, but before arrival of medical personnel to the scene of trauma. Other autopsy series have clearly documented the same ischemic brain injury findings, substantiating the proposition that factors occurring during the critical phase of head injury are the initiating events.83,124 These same pathologic cascades are believed to initiate injury to the gastrointestinal tract, heart, and lungs.83 It is evident that clinical and laboratory research should be directed toward lowering patient morbidity and mortality, and improving outcome of traumatic head injury.
223
However, the vast majority of this care is directed toward patients who survive long enough to arrive in the emergency department, while the “critical phase” has essentially been ignored in the current spectrum of treatment.125 This is exemplified by the fact that 21 clinical head-injury trials targeting therapies directed toward patients on their arrival to hospital have all failed to statistically improve outcome (Ross Bullock, MD, PhD; personal communication).126 The purpose of this chapter is to draw attention to the critical phase of head injury and emphasize the importance of immediate post–head-injury events in determining patient outcome. Apnea and catecholamine surge clearly dictate the patient substrate from which medical or surgical intervention must begin. Therefore, it is imperative that treatment options be directed toward traumatic headinjured patients in the prehospital setting. Treatment Strategies The treatment principles of the 1980s were based on a concept of preventing and minimizing cerebral edema through fluid restriction, neural protection using prophylactic hyperventilation and steroids, and minimizing secondary injury by monitoring and treating elevated ICP. A better understanding of the elements of secondary injury has led to a virtually complete revolution in basic treatment tenets: adequate fluid resuscitation to prevent hypotension, ICP monitoring as a means of maximizing cerebral perfusion pressure (CPP), and avoidance of hyperventilation are now first-line treatment strategies. Nevertheless, it is critical to continually reevaluate our treatment approaches as recent articles have questioned the effectiveness of CPP management at the levels currently recommended. Alternative resuscitative fluid such as hypertonic saline and mannitol solutions are being aggressively studied, and the reportedly deleterious effects of hyperventilation are being re-examined.127–129 A codification of current treatment principles can be found in the Guidelines for the Management of Severe Head Injury. Guidelines for the Management of Severe Head Injury The Guidelines for the Management of Severe Head Injury were developed using an evidence-based approach.130 All pertinent clinical literature for the past 20 years was assessed, reviewed, and classified based on the following methods: class I evidence—prospective randomized controlled clinical trials; class II evidence—clinical studies with prospective data collections such as case control studies, cohort studies, or retrospective analyses based on reliable data; and class III evidence—retrospective data collections such as clinical series, databases, case reviews, and expert opinion. Each article assessed in this process was then carefully studied with respect to design and method to ascertain the
W9418-08.qxd 10/31/03 3:41 PM Page 224
224
Section I
Introduction
reliability of its findings. This evidence was then weighted to determine the level of certainty that could be determined that a particular treatment or intervention in question would positively affect patient outcome. This level of certainty was then expressed as a standard, guideline, or option with respect to patient management strategies: standard— accepted principles of patient management that reflect a high degree of clinical certainty; guideline—recommendations that reflect a particular strategy or range of management strategies with a moderate degree of clinical certainty; option—all remaining strategies for patient management for which there is unclear clinical certainty. Generally, standards can only be supported by high-quality class I evidence. The lack of such studies in current neurotrauma literature is reflected in the fact that the guidelines propagate only three standards in the entire document. The guidelines have attempted to comprehensively address the clinical management of adult closed-head injury from initial resuscitation to the critical care phase of management. Fifteen separate sections address individual treatment considerations and pertinent examples are provided for each. Each recommendation is supported by detailed explanation of the available scientific evidence, evidentiary tables, and recommendations for future research. The guidelines were revised in 2000 to reflect new scientific literature that had become available since their publication in 1996. Additional chapters that are in the process of
being added include pediatric head injury, penetrating head injury, and surgical management of head injury. Prehospital Care Improved Emergency Medical Services Response A faster response of EMS would be extremely beneficial for any traumatically injured patient. Specifically, unwitnessed motor vehicle accidents (i.e., single car) or accidents in rural areas would all benefit from immediate notification and response of EMS to the exact location of injury. Endeavors such as the Mayday experiment at the Mayo Clinic and other areas are such an attempt to improve response time (Fig. 8-5), but other ideas need to be generated as well. In an effort to refine EMS care during the critical phase of head injury, there must be efforts to gather information at the scene of injury. Such field measurement might include analysis of arterial blood, including PaO2, PaCO2, pH, and level of circulating catecholamines, in an effort to determine the magnitude of energy the brain has received as well as the presence of hypoxic injury before arrival at the hospital. This information will give the treating neurosurgeon and trauma surgeon in the emergency department valuable information as to the severity of injury during the critical phase, and how best to manage and target therapy to the brain and other organs when the patient arrives. Data gathered at the scene might also spur laboratory development of treatments that
Figure 8-5. The Mayday trial uses a motor vehicle mounted sensor detecting collision events strongly suggestive of occupant injury and sends an immediate notification to the nearest EMS for rapid response deployment of service to the exact satellite depicted location of the accident.
W9418-08.qxd 10/31/03 3:41 PM Page 225
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
could be expeditiously initiated in the field rather than on arrival to the trauma care facility. Treatment of apnea in the field will be difficult to address. Patients who recover ventilation early after head injury are not at significant apneic hypoxic risk, but may have contributing aspiration or other pulmonary injury risks. The patient who is apneic for some period followed by a dysfunctional but poorly coordinated respiratory effort may sustain life on their own for a longer period of time awaiting EMS arrival. Patients who do not resume any respirations after the initial apnea will die unless ventilated early in the course of their head injury. These patients may benefit most from first responder delivered cardiopulmonary resuscitation (CPR). Continued education that CPR at the scene may be lifesaving in head-injured patients should be emphasized. For the patient with prolonged apnea, it is doubtful whether EMS response, regardless of how rapid, will make a significant difference in outcome. Ventilatory support must be provided early by the first responders at the scene of injury or multisystem hypoxic injury will occur. Target Therapy There is significant evidence that early generalized increased intracranial pressure and marked brain swelling on initial CT scans of head injured patients reflect an increased cerebral blood volume. In many cases, head injury may be characterized as a pathological state in which the cerebrovasculature is maximally vasodilated by hypercarbia and hypoxia, in the setting of head injury–induced dysfunctional vascular autoregulation. Thus, when a catecholamineinduced blood pressure surge occurs, the already increased cerebral blood volume of the brain will markedly increase. As the pressure surge enters a maximally dilated, unresponsive cerebrovascular bed, the result may lead to markedly elevated ICP with or without blood-brain barrier disruption.52–56 Better methods of triaging injured patients, based on information gathered at the scene regarding the critical phase of head injury, should help to select patients who merit maximal resuscitation efforts versus those with little or no likelihood of survival. There is no question that imaging is a powerful tool in determining prognosis, and newer MRI techniques may improve our diagnostic and prognostic capability.131,132 Target therapy could then be directed to foster recovery in predicted survivors, and as clinical information of the pathophysiology during this phase of head injury accumulates, potentially redirect future therapy to be rendered at the scene of head injury. The Critical Care of Head Injury Monitoring Modalities As noted previously, head injury has heterogeneous clinical manifestations and, more importantly, the pathophysiology in any given patient may change from one time epoch to the
225
next.129,133 Thus, it is vitally important to establish appropriate monitoring of these patients to react to pathophysiologyinduced changes. At the most fundamental level, continuous arterial blood pressure monitoring is essential to rapidly react to episodes of hypotension. While there is still some controversy over the utilization of intracranial pressure monitoring with regard to its direct impact on patient outcome, it is clear that prolonged periods of intracranial hypertension increase head injury mortality. It has been clearly shown that if intracranial pressure is greater than 20 mm Hg and cannot be controlled by any measure, head injury mortality approaches 100%.134 Thus, it is important to consider intracranial pressure monitoring in any comatose patient with head injury or in patients who meet the criteria described by Narayan and associates,135 that is, normal head CT, older than 40 years of age, episode(s) of hypotension, brainstem dysfunction as evidenced by posturing. The exact method of ICP monitoring is, perhaps, not as important as the information obtained therefrom. Many prefer intraventricular monitoring because of the simultaneous ability to use cerebrospinal fluid drainage as a therapeutic modality. Nevertheless, there is reasonable and appropriate concern about the incidence of infection with these types of monitors. Various parenchymal monitors have proven useful; however, one must be aware of the limitations associated with drift of the monitors that can be as high as 2 mm Hg per day. Most authors consider subdural or epidural monitors to be relatively insensitive to levels of increased intracranial pressure in the setting of severe head injury (see Chapter 25, Elevated ICP). Monitoring of hemodynamic function by central venous pressure or pulmonary artery (PAC) lines are generally recommended to guide fluid resuscitation and to minimize the occurrence of hypotensive episodes. If intensive therapy, such as barbiturates, is used in the treatment of patients, then pulmonary artery catheter monitoring is mandatory because of the potential deleterious effects of these medications on cardiac function, even in young patients. Recently, critical care interest in the treatment of head injury has focused on more specific monitors of the perfusion and metabolic requirements of the injured brain. Perhaps the best known of these are the oximetric jugular venous oxygen saturation monitors. When the brain is ischemic, it will extract more oxygen and, thus, less is present in the jugular venous outflow. Robertson and associates136 have made great use of these types of monitors in the clinical treatment of severe head injury and have established clear correlation between jugular venous blood desaturation and outcome. Nevertheless, jugular venous oximetry has not been widely used because of its propensity for falsepositive readings and the need to frequently cross check the accuracy and reliability of the device. The use of cerebral perfusion pressure (mean arterial pressure minus intracranial pressure) has been advocated as an important parameter to not only monitor, but also aggressively treat to prevent
W9418-08.qxd 10/31/03 3:41 PM Page 226
226
Section I
Introduction
cerebral ischemia occurring after severe head injury. Rosner and colleagues have been the most outspoken proponents of this monitoring (and treatment) modality.137,138 A CPP of greater than 70 mm Hg has been thought to be necessary to insure adequate brain perfusion. Obviously, to determine the CPP, one must have in place an appropriate blood pressure monitoring device as well as an intracranial pressure monitor. More recently, commercial monitors of focal cerebral oxygenation, glucose levels, pH and CO2 levels have become available. While several studies have indicated that brain PO2 levels correlate with outcome, there is no reliable information on how to specifically treat brain PO2 levels or how to use brain glucose, pH, and CO2 levels in the overall management strategy for patients with head injury.139–141 A critique of these monitors is that they only monitor tissue of several cubic millimeters in size and, thus, may have no correlation whatsoever with the overall perfusion and metabolic situation of the injured brain, unless the injury is diffuse. While microdialysis in head-injured patients has been studied, and clear and significant elevations in a number of different compounds—most prominently, the excitatory neurotransmitter glutamine—found, these technologies are available only experimentally. Critical Care Treatment Guidelines Once appropriate monitoring of the head-injured patient has been accomplished, it is then possible to use this information in conjunction with careful repeated clinical examination to establish treatment guidelines based on the principles enumerated in the Guidelines for the Management of Severe Head Injury. As noted previously, a number of secondary insults can potentially occur throughout the patient’s initial hospitalization—especially in the critical care phase—that require immediate attention. The first and foremost of these are the prevention and/ or early recognition and aggressive treatment of hypotension or hypoxia. There are multiple sources for hypotension in the severely head-injured patient. The least likely source is a terminal event associated with cerebral herniation and bilateral pupillary dilatation. Thus, it is often possible to intervene appropriately to maintain an adequate systolic and mean arterial blood pressure. Initially, the intravascular volume is best assessed utilizing a central venous pressure monitor or a pulmonary artery catheter. Hypovolemia can be effectively treated with infusions of crystalloid, colloids or, when appropriate, blood. If adequate volume replacement does not provide for sufficient maintenance or restoration of blood pressure, then inotropic or pressor agents should be used. In head-injured patients, a well-tolerated inotropic agent is dobutamine, while norepinephrine provides optimal peripheral vasoconstriction. It should always be borne in mind, however, that when fluid or pressors are being used to restore or maintain blood pressure, the source
of the underlying hypotension must be determined and treated. Blood loss from chest or abdominal injuries, cardiac contusion or tamponade, or tension pneumothorax must always be considered and investigated. Spinal shock from spinal cord injury should also be considered in individuals who are difficult to evaluate from a clinical neurologic standpoint. Hypoxia is another dangerous secondary insult to the brain and is generally treated by adequate artificial ventilation. In most patients, increasing the FiO2 improves oxygenation. However, many of these patients suffer from pulmonary contusions or develop acute respiratory distress syndrome (ARDS) and may require the addition of positive end-expiratory pressure (PEEP). Concerns, however, have arisen over the effects of PEEP on intracranial pressure. Levels of PEEP in excess of 8 cm H2O may be associated with a rise in intracranial pressure and a decrease in cerebral perfusion pressure. Usually, however, simply elevating the patient’s head reduced the effects of PEEP on ICP. Further, in noncompliant lungs, airway pressure is minimally transmitted, so no increase in ICP would be seen. Thus, if PEEP is required to maintain adequate oxygenation, there should be little concern over the use of this therapy. The primary focus of critical care management to this point in time has been maintenance of normal or near normal intracranial pressure. There are a number of general measures that can be undertaken to affect some degree of ICP control. These include utilization of sedatives and analgesics. However, one must remember that the use of these medications may result in a drop in blood pressure and their use must be titrated appropriately and monitored on a continuous basis. The primary agents currently utilized include midazolam at a dose of 0.025 to 0.35 mg/kg or through a continuous infusion at 0.05 to 5 mg/kg/min. The primary narcotics include morphine at a dose of 2 to 5 mg/hour or fentanyl at a dose of 50 to 100 mg/hour. Other agents such as a propofol and etomidate have been studied. Propofol does provide for smooth sedation and can be reversed quickly; however, the expense of the agent may limit its use. Infusions of Etomidate are inappropriate because the drug blocks 11-betahydroxylase and may lead to adrenal insufficiency. When these general measures fail to maintain an appropriate ICP, pharmacologic paralysis is often instituted. Such drug therapy prevents agitation, posturing, or coughing that can increase ICP to excessive levels. Unfortunately, when these agents are used, the neurologic examination is lost and consideration needs to be given to more frequent CT scan monitoring. In addition, prolonged use of pharmacologic paralysis can increase pulmonary complications, prolong stay in the ICU, and lead to iatrogenic neuromuscular disorders. The use of neuromuscular blocking agents should be limited to the shortest time possible. One popular agent in use at this time is the nondepolarizing agent vecuronium. It is given as a bolus at a dose of 0.1 to 0.2 mg/kg ideal body
W9418-08.qxd 10/31/03 3:41 PM Page 227
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
weight, and then run at a maintenance infusion of 1 to 10 mg/hour, titrated to one to two twitches of a train of four, stimulated by supramaximal electrical impulses at the ulnar nerve. The depolarizing agent succinylcholine is not recommended, except in an emergency, after the first several days of severe head injury because of its potential potassium elevating effect. Cerebrospinal fluid drainage is the simplest and most direct route to lowering intracranial pressure. This, however, requires ventriculostomy placement and as discussed in the previous section, it is not always possible to accomplish in patients with extremely small ventricles secondary to elevated intracranial pressure. Additionally, when there is intraventricular blood, the catheter may intermittently malfunction. Finally, when large amounts of cerebrospinal fluid are withdrawn, the ventricles normally collapse, thus limiting the utilization of this approach. Osmotic diuretics have long been utilized in the treatment of elevated intracranial pressure. Mannitol is typically given in a bolus dose of 0.25 to 1 g/kg and will maximally reduce the ICP within 15 minutes. This reduction generally lasts for 3 to 4 hours. Additional doses can be given as needed so long as there is careful monitoring of serum sodium and osmolality to prevent a hyperosmolar state. Oftentimes, mannitol is combined with furosemide at a dose of 20 mg given several times a day. The combination of these medications has been felt to be synergistic, although there are no definitive studies to indicate that such is indeed the case. Certainly, if mannitol and furosemide are to be used concomitantly, then even greater vigilance must be maintained for the development of a hyperosmolar state. A subsequent step available in controlling intracranial pressure when other measures have failed is the use of optimized hyperventilation. As noted previously, discussion related to the use of hyperventilation has significantly changed over the past several decades. Prolonged prophylactic hyperventilation, not being used to control intracranial pressure, is contraindicated because of the potentially significant decrease in cerebral blood flow associated with vasoconstriction. It is well known that for each 1 mm Hg drop in the PaCO2, there is a 3% decrease in cerebral blood flow and in the first 24 hours after severe head injury, the cerebral blood flow may be as low as 50% of normal. Nevertheless, when intracranial pressure is refractory to other modes of treatment, hyperventilation may be very effective. When hyperventilation is utilized, it is generally recommended that the PaCO2 be kept above 25 mm Hg. If more extreme levels of hyperventilation are required, then alternative monitoring is undertaken in order to document whether or not the brain is being made ischemic by the degree and extent of vasoconstriction associated with the hyperventilation. Such methods include xenon cerebral blood flow studies, jugular venous oxygen saturation monitoring, and the new technology of brain PO2 sensors.
227
If the preceding measures have failed, the use of barbiturates is generally considered. Barbiturates suppress cerebral metabolism and, thus, decrease the demand for glucose and oxygen to theoretically tide the brain through periods of inadequate perfusion. When barbiturates are utilized, however, there must be exceedingly careful hemodynamic monitoring, including the placement of a pulmonary artery catheter. Barbiturates are associated with hypotensive complications and in many cases, have been the source of a patient’s demise in this setting. The most frequently used barbiturate is pentobarbital. The typical loading dose is 5 to 10 mg/kg given over 30 minutes with the subsequent administration of 5 mg/kg to obtain therapeutic levels. A maintenance infusion is then established at 1 to 2 mg/kg/hour. Various authors have suggested that following the levels of barbiturates to ensure that they are “therapeutic” provides the best information as to the establishment of appropriate treatment with this modality. However, therapeutic levels are sometimes difficult to obtain and may take many hours to return. Thus, the use of continuous electroencephalography (EEG) monitoring is real time and exceedingly sensitive to “therapeutic” levels of barbiturates. The end-point of treatment with barbiturates is 90% burst suppression on the two-channel EEG. There are other important management issues in the critical care of these patients, which include avoidance or correction of electrolyte disturbances and of hyperthermic states. It should be borne in mind that severe head-injured patients may be susceptible to cerebral vasospasm because of associated traumatic subarachnoid hemorrhage; this possibility should be considered in patients who are otherwise not responding to treatment. This occurrence can be documented either by the use of transcranial Doppler or, if necessary, angiography. Corticosteroids, which have been used extensively in the treatment of severe head injury in the past, are not indicated for the treatment of intracranial pressure or to improve the outcome after severe head injury. Extensive studies including meta-analysis of various studies have shown that corticosteroids have no effect whatsoever in this regard and are associated with a level of complications that are not acceptable. Targeted Therapy A logical extension of the Guidelines for the Treatment of Severe Head Injury is the concept of targeted therapy— treating a specific patient’s specific pathophysiology. Such an approach requires pertinent and contemporaneous information based on systemic and cerebral monitoring. An excellent example of this approach is directed management to maintain a CPP greater than 70 mm Hg. Such a goal requires ICP and systolic blood pressure monitoring. However, once established, a variety of targeted approaches can be brought into play. Various studies suggest that this approach improves head-injury outcome and it is, indeed,
W9418-08.qxd 10/31/03 3:41 PM Page 228
228
Section I
Introduction
one of the most widely used strategies for the treatment of head injury today.137,138 However, this treatment strategy has frequently been taken to extremes in raising the CPP above 90 or 100 mm Hg to counteract extremely elevated ICPs. A recent randomized controlled clinical trial was reported by Robertson and associates136 in which they studied patients who were treated based on CPP alone versus those who were treated based on ICP alone. This study found no significant difference between neurologic outcomes in either group of patients. The authors, however, did find statistically significant increases in medical complications—particularly those related to pulmonary dysfunction (i.e., ARDS) in patients who were treated with large volumes of fluid and pressers in order to maintain elevated CPPs.136 Another example of targeted therapy is the avoidance of impaired cerebral oxygenation either due to a decrease in oxygen delivery or an increase in oxygen consumption. This approach currently requires jugular venous oxygen saturation monitoring in addition to a variety of other systemic variables. There are, however, now available commercial surface PO2 sensors, the reliability of which is currently being validated. The identification of oxygen desaturation allows for targeted therapy, depending on its source. Several studies have indicated that optimization of blood flow and metabolism through monitoring and normalization of cerebral oxygenation influences outcome.133,140–143 Fluid Management On a more fundamental level of targeted therapy, recent attention has been focused on the type of resuscitation and maintenance fluids used in head-injured patients. As noted previously, the concepts of fluid management in headinjured patients have been revolutionized over the past several decades. Nevertheless, aggressive fluid management may have its drawbacks as was discussed in relationship to CPP management. With these concerns in mind, a number of investigators have studied the use of hypertonic saline solution in traumatic brain injuries, both as a resuscitative and a maintenance fluid. It has been suggested that hypertonic saline solution not only improves hemodynamic parameters and modulation of intracranial hypertension after head injury, but also may possess osmotic, vascular, neurochemical, and immunologic effects. Numerous animal models support the use of hypertonic saline solution in the presence of traumatic brain injury, especially in the presence of hypotension. Worthley and associates144 reported two patients with traumatic injury and intractable intracranial hypertension successfully treated with 20 mL of 29.2% hypertonic saline solution. Einhaus and associates145 reported a similar experience with a patient with traumatic brain injury (TBI) who had intracranial hypertension refractory to mannitol. Simma and colleagues146 were the first to prospectively evaluate hypertonic saline solution, randomizing severely
head-injured pediatric patients to receive either 1.7% hypertonic saline or lactated Ringers solution as maintenance fluid for the first 72 hours following admission. They observed that patients receiving hypertonic saline solution had lower ICP values and required fewer interventions to manage ICP elevations. Qureshi and associates147 evaluated the effects of a continuous infusion of hypertonic saline solution in eight patients with intracranial hypertension of various causes. Patients were given 3% saline solution to raise their serum sodium to 145 to 155 mEq/L. An inverse relationship between serum sodium and ICP was observed in patients with TBI or postoperative edema but not with subarachnoid hemorrhage or ischemia. Patients receiving hypertonic saline solution also had less edema and mass effect on serial CT scans. Neuroprotection Considerable effort has been expended in the past decade to identify and organize clinical trials of pharmacologic agents which potentially hold great promise in blocking or ameliorating various components of the secondary injury biochemical cascade.148 In spite of extremely encouraging experimental animal studies to date, no human trial has proven successful. 149–151 Currently there is an ongoing trial at the Medical College of Virginia and the University of Florida evaluating the use of cyclosporin A to prevent secondary injury in TBI. A recent trial of a selective NMDAsubunit antagonist has been completed but results were not available at the time of this writing. There is speculation as to the reasons for the failure to identify a clinically active compound. The most pertinent concerns are an inability to appropriately target the population of patients most likely to benefit from such an intervention and the unrealistic projections of the magnitude of benefit that these agents might provide. Nevertheless, if a patient could be identified who clearly had a specific biochemical alteration associated with the head-injury pathophysiology—that is, excess glutamate release in association with a cerebral contusion—he or she could then be targeted for treatment with the appropriate pharmacologic agent in the appropriate time frame. Unless such strategies become clinically practical, however, it is unlikely that the field of neuroprotection will advance significantly beyond its present state. Gene Therapy The ultimate substrate of the CNS response to injury and repair is genetically mediated152 (see Chapters 31 and 32). There are ongoing discoveries related to the CNS genetic response to trauma and ischemia. While most studies have been descriptive in nature and their relevance to recovery, if any, unknown, increasing efforts are being expended not only to quantify this response but to manipulate it as well. Teasdale and associates153 recently reported preliminary data suggesting a relationship between apolipoprotein E and the brain’s response to injury. Individuals with a specific allele of
W9418-08.qxd 10/31/03 3:41 PM Page 229
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
this gene appear to have an increased vulnerability to head injury. Similarly, Sorbi and colleagues154 have suggested a possible association between another allele of the apolipoprotein E gene and poor outcome after severe head injury. The ultimate treatment strategy, therefore, would be to downregulate those genes responsible for the harmful biochemical cascade occurring after injury and to upregulate those genes responsible for repair and recovery. Given current research, such seems an obtainable goal within the next decade.
Strategies for the Millennium Marked improvements have occurred in the recognition and management of severe head injury with concomitant
improvements in outcome. The majority of these accomplishments have evolved through greater attention to the details of critical care management of this patient population. More sophisticated monitoring technology has further enhanced the timely identification and reaction to evolving pathophysiology and metabolic derangements. These advancements have been codified in the Guidelines for the Management of Severe Head Injury. The concept of targeted therapy is a logical extension of the guidelines to provide for more patient specific intervention. Ultimately, identification of a patient’s specific biochemical and genetic responses to injury will allow for complete understanding of the secondary injury process with a concomitant opportunity to provide the ultimate form of neuroprotection from head injury.
P earls 1. Head injury remains a serious public health problem, occurring at a rate of 150 per 100,000 population per year in the United States. The most current data indicate that head injury accounts for over 20,000 deaths and 50,000 permanent disabilities each year. 2. This critical phase of injury encompasses the pathophysiologic sequelae of apnea and catecholamine surge responsible for multiple organ injuries in the early period after central nervous system trauma. 3. There are two immediate pathophysiologic events that occur with severe head injury onset that markedly effect subsequent outcome: head injury–induced apnea, and a stress-related massive sympathetic discharge. In combination, the effects of hypoxia, hypercarbia, acidosis, and blood pressure surge, as well as the direct effects of catecholamines on tissue, all lead to a synergistic injury effect in the host. 4. To date, irrespective of the method used, all experimental head injury in a spontaneously breathing model has produced apnea as a consistent response. 5. . . . trauma-induced apnea causes cerebral hypoxia and ischemia which proves to be more fundamental to outcome in these patients than the mechanism of primary brain injury itself (i.e., subdural hematoma, subarachnoid hemorrhage, diffuse axonal injury, or contusions). 6. Massive sympathetic discharge and subsequent marked elevation in mean arterial blood pressure and heart rate has proven to be a consistent immediate response in any experimental severe head injury
229
7.
8.
9.
10.
11.
model, such as fluid percussion injury, acceleration, projectile, or weight drop techniques. The degree and extent of primary neural injury is fixed at the time of the accident, and depends partially on the mechanism of injury. The only defense against the occurrence of primary neural injury is prevention. Singular advancements in the critical care management of head injury have been based on the understanding of the concept that a large number of patients die not because of the initial injury to the brain but entirely because of additional secondary insults that occur following injury. The occurrence of hypotension or hypoxia alone increases morbidity and mortality of traumatic brain injury by up to 50%. When hypoxia and hypotension occur together, the risk of a poor outcome increases to more than 60%. Given the concern that currently used intravenous fluids can potentially exacerbate cerebral edema especially in areas of blood brain barrier breakdown, other resuscitative and maintenance fluids are under active investigation. Because of the small volumes required (approximately 8 mL/kg) to maintain an adequate blood pressure, hypertonic saline solution has been used as a resuscitative fluid in experimental studies. Gastroduodenal mucosal injury frequently develops after severe head injury, and frank ulceration is common as well. Esophageal gastroduodenoscopy in patients soon after severe head injury reveals these lesions in as many as 90% of patients. These Continued
W9418-08.qxd 10/31/03 3:41 PM Page 230
230
Section I
Introduction
ulcerations occur very early after head injury, and are similar to catecholamine stress ulcers produced experimentally, or by other systemic stress injuries such as burns or sepsis. 12. Isolated head trauma–induced cardiac injury is well documented in the pathology literature, and may be found at autopsy in up to two thirds of patients dying from acute head injury alone. 13. There is evidence that pretreatment with catecholamine blocking agents will markedly reduce the cardiac injury induced by stress catecholamine surge and help prevent further extension of myocardial damage. 14. . . . cardiac failure results from a catecholamineinduced increase in cardiac preload due to venoconstriction, and increased cardiac afterload by massive
References 1. CDC Demographic Data, 1999: Reported in Brain Injury Association Bulletin, Sept. 1999. 2. Graves E: Detailed diagnosis and procedures. National hospital discharge survey, 1990. Vital Health Statistics 1992;13:113. 3. Marshall LF: Head injury: Recent, past, and future. Neurosurgery 2000;47:546–561. 4. Miller JD: Head injury and brain ischaemia—implications for therapy. Br J Anaesth 1985;57:120–129. 5. Overgaard J, Tweed WA: Cerebral circulation after head injury. Part 4: Functional anatomy and boundary-zone flow deprivation in the first week of traumatic coma. J Neurosurg 1983;59:439–446. 6. Koch W, Filehne W: Belträge zur experimentellen chirurgie. Arch Klin Chir 1874;17:190–231. 7. Polis A: Recherches expérimentales: Sur la commotion cérébrale. Rev Chir 1894;14:645–730. 8. Kramer SP: A contribution to the theory of cerebral concussion. Ann Surg 1896;23:163–173. 9. Miller GG: Cerebral concussion. Arch Surg Chicago 1927;14:891-916. 10. Denny-Brown D, Russell WR: Experimental cerebral concussion. Brain 1941;64:93–164. 11. Walker AE, Kollros JJ, Case TJ: The physiologic basis of concussion. J Neurosurg 1944;1:103–116. 12. Sullivan HG, Martinez J, Becker DP, et al: Fluid-percussion model of mechanical brain injury in the cat. J Neurosurg 1976;45:520–534. 13. Genarelli TA: Head injury in man and experimental animals: clinical aspects. Acta Neurochir Suppl (Wien) 1983;32:1–13. 14. Carey ME, Sarna GS, Farrell JB, et al: Experimental missile wound to the brain. J Neurosurg 1989;71:754–764. 15. Andersen BJ, Unterberg AW, Clarke GD, et al: Effect of posttraumatic hypoventilation on cerebral energy metabolism. J Neurosurg 1988;68:601–607. 16. Ito J, Marmarou A, Barzo P, et al: Characterization of edema by diffusion-weighted imaging in experimental traumatic brain injury. J Neurosurg 1996;84:97–103. 17. Kim HJ, Levasseur JE, Patterson JL Jr, et al: Effect of indomethacin pretreatment on acute mortality in experimental brain injury. J Neurosurg 1989;71:565–572. 18. Levasseur JE, Patterson JL Jr, Ghatak NR, et al: Combined effect of respirator-induced ventilation and superoxide dismutase in experimental brain injury. J Neurosurg 1989;71:573–577.
arterial constriction. In conjunction with catecholamine-induced cardiac injury, these mechanisms combine to produce edema in the lungs by increasing pulmonary capillary pressures. 15. It is evident that clinical and laboratory research should be directed toward lowering patient morbidity and mortality, and improving outcome of traumatic head injury. However, the vast majority of this care is directed toward patients who survive long enough to arrive in the emergency department, while the “critical phase” has essentially been ignored in the current spectrum of treatment. This is exemplified by the fact that 17 clinical head-injury trials targeting therapies directed toward patients on their arrival to hospital have all failed to statistically improve outcome.
19. Kelly DF: Alcohol and head injury: an issue revisited. J Neurotrauma 1995;12:883–890. 20. Kelly DF, Lee SM, Pinanong PA, et al: Paradoxical effects of acute ethanolism in experimental brain injury. J Neurosurg 1997;86:876– 882. 21. Zink BJ, Feustel PJ: Effects of ethanol on respiratory function in traumatic brain injury. J Neurosurg 1995;82:822–828. 22. Atkinson JLD, Anderson RE, Murray MJ: The early critical phase of severe head injury: Importance of apnea and dysfunctional respiration. J Trauma 1998;45:941–945. 23. Marmarou A, Abd-Elfattah Foda MA, Van Den Brink W, et al: A new model of diffuse brain injury in rats. Part 1: Pathophysiology and biomechanics. J Neurosurg 1994;80:291–300. 24. Johnson DL, Boal D, Baule R: Role of apnea in nonaccidental head injury. Pediatr Neurosurg 1995;23:305–310. 25. Chesnut RM, Marshall LF, Klauber MR, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 1993;34:216–222. 26. Chesnut RM, Marshall SB, Piek J, et al: Early and late systemic hypotension as a frequent and fundamental source of cerebral ischemia following severe brain injury in the Traumatic Coma Data Bank. Acta Neurochir Suppl (Wien) 1993;59:121–125. 27. Marshall LF, Gautille T, Klauber MR, et al: The outcome of severe closed head injury. J Neurosurg 1991;75(Suppl):S28–S36. 28. Levine JE, Becker DP, Chun T: Reversal of incipient brain death from head-injury apnea at the scene of accidents. Letter to the Editor. New Engl J Med 1979;301:109. 29. Cushing H: Concerning a definite regulatory mechanism of the vasomotor centre which controls blood pressure during cerebral compression. Bull Johns Hopkins Hosp 1901;12:290–292. 30. Spencer W, Horsley V: On the changes produced in the circulation and respiration by increase of the intracranial pressure or tension. Philos Trans 1892;182(Br):201–254. 31. Grimson KS, Wilson H, Phemister DB: The early and remote effects of total and partial paravertebral sympathectomy on blood pressure. Ann Surg 1937;106:801–825. 32. Freeman NE, Jeffers WA: Effect of progressive sympathectomy on hypertension produced by increased intracranial pressure. Am J Physiol 1940;128:662–671. 33. Roozerkrans NTP, Van Zwieten PA: The role of the adrenergic nervous system in the pressor response due to acute elevation of intracranial pressure (Cushing’s response). A pharmacologic analysis. Arch Int Pharmacodyn 1977;228:79–91.
W9418-08.qxd 10/31/03 3:41 PM Page 231
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
34. Rosner MJ, Newsome HH, Becker DP: Mechanical brain injury: The sympathoadrenal response. J Neurosurg 1984;61:76–86. 35. Beckman DL, Iams SG: Circulating catecholamines before and after lethal head injury. Proc Soc Exp Biol Med 1979;160:200–202. 36. Hörtnagl H, Hammerle AF, Hackl JM, et al: The activity of the sympathetic nervous system following severe head injury. Intensive Care Med 1980;6:169–177. 37. Sculte AM, Esch J, Murday H, Pfeifer G: Haemodynamic changes in patients with severe head injury. Acta Neurochir 1980;54:243– 250. 38. Clifton GL, Ziegler MG, Grossman RG: Circulating catecholamines and sympathetic activity after head injury. Neurosurgery 1981;8: 10–14. 39. Clifton GL, Robertson CS, Kyper K, et al: Cardiovascular response to severe head injury. J Neurosurg 1983;59:447–454. 40. Robertson CS, Clifton GL, Grossman RG: Oxygen utilization and cardiovascular function in head-injured patients. Neurosurgery 1984;15:307–314. 41. Robertson CS, Clifton GL, Taylor AA, et al: Treatment of hypertension associated with head injury. J Neurosurg 1983;59:455–460. 42. Dilraj A, Botha JH, Rambiritch V, et al: Levels of catecholamine in plasma and cerebrospinal fluid in aneurysmal subarachnoid hemorrhage. Neurosurgery 1992;31:42–51. 43. Kasuya H, Kawashima A, Namiki K, et al: Metabolic profiles of patients with subarachnoid hemorrhage treated by early surgery. Neurosurgery 1998;42:1268–1275. 44. Benedict CR, Grahame-Smith DG. Plasma noradrenaline and adrenaline concentrations and dopamine—hydrosylase activity in patients with shock due to septicaemia, trauma, and hemorrhage. Q J Med 1978;47:1–20. 45. Wilmore DW, Long JM, Mason AD Jr, et al: Catecholamines: mediator of the hypermetabolic response to thermal injury. Ann Surg 1974;180:653–669. 46. Koiv L, Merisalu E, Zilmer K, et al: Changes of sympatho-adrenal and hypothalamic-pituitary-adrenocortical system in patients with head injury. Acta Neurol Scand 1997;96:52–58. 47. Shiozaki T, Taneda M, Kishikawa M, et al: Transient and repetitive rises in blood pressure synchronized with plasma catecholamine increases after head injury. Report of two cases. J Neurosurg 1993;78:501–504. 48. Levitt MA, Cook LA, Simon BC, et al: Biochemical markers of cerebral injury in patients with minor head trauma and ethanol intoxication. Acad Emerg Med 1995;2:675–680. 49. Rise IR, Kirkeby DJ: Effect of alpha-adrenergic blockade on the cerebrovascular response to increased intracranial pressure after hemorrhage. J Neurosurg 1998;89:454–459. 50. Simard JM, Bellefleur M: Systemic arterial hypertension in head trauma. Am J Cardiol 1989;63:32C–35C. 51. Tietjen CS, Hurn PD, Ulatowski JA, et al: Treatment modalities for hypertensive patients with intracranial pathology: Options and risks. Crit Care Med 1996;24:311–322. 52. Bouma GJ, Muizelaar JP: Cerebral blood flow, cerebral blood volume, and cerebrovascular reactivity after severe head injury. J Neurotrauma 1992;9(Suppl):S333–S348. 53. Engelborghs K, Verlooy J, VanReempts J, et al: Temporal changes in intracranial pressure in a modified experimental model of closed head injury. J Neurosurg 1998;89:796–806. 54. Kimelberg HK: Current concepts of brain edema. Review of laboratory investigations. J Neurosurg 1995;83:1051–1059. 55. Rosner MJ, Becker DP: Origin and evolution of plateau waves. Experimental observations and a theoretical model. J Neurosurg 1984;60:312–324. 56. Tanno H, Nockels RP, Pitts LH, et al: Breakdown of the blood-brain barrier after fluid percussion brain injury in the rat: Part 2: Effect of hypoxia on permeability to plasma proteins. J Neurotrauma 1992;9:335–347.
231
57. Pietropaloi JA, Rogers FB, Shackford ST, et al: The deleterious effects of intraoperative hypotension on outcome in patients with severe head injuries. J Trauma 1992;33:403–407. 58. Gunnar WD, Merlotti GJ, Jonasson O, et al: Resuscitation from hemorrhagic shock: Alterations of intracranial pressure after normal saline, 3% saline, and dextran-40. Ann Surg 1986;204:686–692. 59. Bacher A, Wei I, Grafe MR, et al: Serial determinations of cerebral water content by magnetic resonance imaging after an infusion of hypertonic saline. Crit Care Med 1998;26:108–114. 60. Bullock R, Zauner A, Myseros JS, et al: Evidence for prolonged release of excitatory amino acids in severe human head trauma: Relationship to clinical events. Ann NY Acad Sci 1995;17:290–298. 61. Bullock R, Zauner A, Woodward JJ: Factors affecting excitatory amino acid release following human head injury. J Neurosurg 1998;89:507– 518. 62. Palmer AM, Marion DW, Botscheller ML, et al: Increased transmitter amino acid concentration in human ventricular CSF after brain trauma. Neurol Rep 1994;6:153–156. 63. Dekosky ST, Goss JR, Miller PD, et al: Up regulation of nerve growth factor following cortical trauma. Exp Neurol 1994;130:173–177. 64. Dutcher SA, Underwood BD, Walker PD, et al: Patterns of heat and shock protein -70 biosynthesis following human traumatic brain injury. J Neurotrauma 1998;15:411–419. 65. Younger EC, Newell DW, Grant GA, et al: Cerebral autoregulation following minor head injury. J Neurosurg 1997;86:425–432. 66. Sapolsky RM, Pulsinelli WA: Glucocorticoids potentiate ischemic injury to neurons: Therapeutic implications. Science 1985;229:1397– 1400. 67. deCourten-Myers GM, Kleinholz M, Wagner KR, et al: Normoglycemia (not hypoglycemia) optimizes outcome from middle cerebral artery occlusion. J Cereb Blood Flow Metab 1994;14:227–236. 68. Ginsberg MD, Welsch FA, Budd WW: Deleterious effect of glucose pretreatment on recovery from diffuse cerebral ischemia in the cat. I. Local cerebral blood flow and glucose utilization. Stroke 1980;11:347– 354. 69. Wass CT, Lanier WL: Glucose modulation of ischemic brain injury: Review and clinical recommendations. Mayo Clin Proc 1996;71:801– 812. 70. Cherian L, Goodman JC, Robertson CS: Hyperglycemia increases brain injury caused by secondary ischemia after cortical impact injury in rats. Crit Care Med 1997;25:1378–1383. 71. Cherian L, Hannay HJ, Vagner G, et al: Hyperglycemia increases neurologic damage and behavioral deficits from post-traumatic secondary ischemic insults. J Neurotrauma 1998;15:307–321. 72. Brown TH, Davidson PF, Larson GM: Acute gastritis occurring within 24 hours of severe head injury. Gastrointest Endosc 1989;35:37–40. 73. Kamada T, Fusamoto H, Kawano S, et al: Acute gastroduodenal lesions in head injury. Am J Gastroenterol 1977;68:249–253. 74. Larson GM, Koch S, O’Dorisio TM, et al: Gastric response to severe head injury. Am J Surg 1984;147:97–105. 75. Hayase M, Takeuchi K: Gastric acid secretion and lesion formation in rats under water immersion stress. Dig Dis Sci 1986;31:166–171. 76. Menguy R, Masters YF: Gastric mucosal energy metabolism and “stress ulceration.” Ann Surg 1974;180:538–548. 77. LeGall JR, Mignon FC, Rapin M, et al: Acute gastroduodenal lesions related to severe sepsis. Surg Gynecol Obstet 1977;142:377–380. 78. McAlhaney JC, Czaja AJ, Pruitt BA: Antacid control of complications from acute gastroduodenal disease after burns. J Trauma 1976;16:645– 649. 79. McArthur CJ, Gin T, McLaren IM, et al: Gastric emptying following brain injury: Effects of choice of sedation and intracranial pressure. Intensive Care Med 1995;21:573–576. 80. Ott L, Young B, Phillips R, et al: Altered gastric emptying in the head injured patient: relationship to feeding intolerance. J Neurosurg 1991;74:738–742.
W9418-08.qxd 10/31/03 3:41 PM Page 232
232
Section I
Introduction
81. Kincaid EH, Miller PR, Meredith JW, et al: Enalaprilat improves gut perfusion in critically injured patients. Shock 1998;9:79–83. 82. Pastores SM, Katz DP, Kvetan V: Splanchnic ischemia and gut mucosal injury in sepsis and the multiple organ dysfunction syndrome. Am J Gastroenterol 1996;91:1697–1710. 83. Clifton GL, McCormick WF, Grossman RG: Neuropathology of early and late deaths after head injury. Neurosurgery 1981;8:309–314. 84. Connor RCR: Myocardial damage secondary to brain lesions. Am Heart J 1969;78:145–148. 85. McCormick WF: Trauma. In: Schochet SS Jr (ed). Clinical Neurosciences, vol 3. New York: Churchill-Livingstone, 1983, pp 241–283. 86. Burch GE, Sun SC, Colcolough HL, et al: Acute myocardial lesions following experimentally-induced intracranial hemorrhage in mice: A histological and histochemical study. Arch Pathol 1967;84:517–521. 87. Hawkins WE, Clower BR: Myocardial damage after head trauma and simulated intracranial hemorrhage in mice: The role of the autonomic nervous system. Cardiovasc Res 1971;5:524–529. 88. Heinrich D, Müller W: Focal myocardial necrosis in cases of increased intracranial pressure. Eur Neurol 1974;12:369–376. 89. Guha A, Tator CH: Acute cardiovascular effects of experimental spinal cord injury. J Trauma 1988;28:481–490. 90. Greenhoot JH, Reichenbach DD: Cardiac injury and subarachnoid hemorrhage. A clinical, pathological, and physiological correlation. J Neurosurg 1969;30:521–531. 91. Marion DW, Segal R, Thompson ME: Subarachnoid hemorrhage and the heart. Neurosurgery 1986;18:101–106. 92. Offerhaus J, Van Gool J: Electrocardiographic changes and tissue catecholamines in experimental subarachnoid hemorrhage. Cardiovasc Res 1969;3:433–440. 93. Yuki K, Kodama Y, Onda J, et al: Coronary vasospasm following subarachnoid hemorrhage as a case of stunned myocardium. Case report. J Neurosurg 1991;75:308–311. 94. Prichard BN, Owens CW, Smith CC, et al: Heart and catecholamines. Acta Cardiol 1991;46:309–322. 95. Reichenbach DD, Benditt EP: Catecholamines and cardiomyopathy: The pathogenesis and potential importance of myofibrillar degeneration. Human Pathol 1970;1:125–150. 96. Jiang JP, Downing SE: Catecholamine cardiomyopathy: Review and analysis of pathogenic mechanisms. Yale J Biol Med 1990;63:581–591. 97. Cruickshank JM, Neil-Dwyer G, Degaute JP, et al: Reduction of stress/catecholamine-induced cardiac necrosis by Beta-1-Selective blockade. Lancet 1987;2:585–589. 98. McNair JL, Clower BR, Sanford RA: The effect of reserpine pretreatment on myocardial damage associated with simulated intracranial hemorrhage in mice. Eur J Pharmacol 1970;8:1–6. 99. Atkinson JLD: Acute lung injury in isolated traumatic brain injury [letter]. Neurosurgery 1997;41:1214–1215. 100. Bratton SL, Davis RL: Acute lung injury in isolated traumatic brain injury. Neurosurgery 1997;40:707–712. 101. Simmons RL, Martin AM, Heisterkamp CA, et al: Respiratory insufficiency in combat casualties: II. Pulmonary edema following head injury. Ann Surg 1969;170:39–44. 102. Agar JM: The medical complications of the early management of head injury in the adolescent. Med J Aust 1966;2:1182–1183. 103. Ducker TB, Simmons RL: Increased intracranial pressure and pulmonary edema. II. The hemodynamic response of dogs and monkeys to increased intracranial pressure. J Neurosurg 1968;28:118–123. 104. Berman IR, Ducker TB: Changes in pulmonary, somatic, and splanchnic perfusion with increased intracranial pressure. Surg Gynecol Obstet 1969;128:8–14. 105. Berman IR, Ducker TB: Pulmonary, somatic and splanchnic circulatory response to increased intracranial pressure. Ann Surg 1969;169:210–216. 106. MacKay EM: Experimental pulmonary edema. IV. Pulmonary edema accompanying trauma to the brain. Proc Soc Exp Biol Med 1950;74:695–697.
107. Chen HI: Hemodynamic mechanisms of neurogenic pulmonary edema. Biol Signals 1995;4:186–192. 108. Pender ES, Pollack CV Jr: Neurogenic pulmonary edema: case reports and review. J Emerg Med 1992;10:45–51. 109. Theodore J, Robin ED: Pathogenesis of neurogenic pulmonary oedema. Lancet 1975;2(7938):749–751. 110. Hoff JT, Nishimura M, Garcia-Uria J, et al: Experimental neurogenic pulmonary edema. Part 1: The role of systemic hypertension. J Neurosurg 1981;54:627–631. 111. Lang SA, Maron MB: Oxygen consumption after massive sympathetic nervous system discharge. Am J Physiol 1989;256:E345–E351. 112. Maron MB: Pulmonary vasoconstriction in a canine model of neurogenic pulmonary edema. J Appl Physiol 1990;68:912–918. 113. Maron MB, Holcomb PH, Dawson CA, et al: Edema development and recovery in neurogenic pulmonary edema. J Appl Physiol 1994;77:1155–1163. 114. Millen JE, Glauser FL: Low levels of concussive brain trauma and pulmonary edema. J Appl Physiol 1983;54:666–670. 115. Deehan SC, Grant IS: Haemodynamic changes in neurogenic pulmonary oedema—effect of dobutamine. Intensive Care Med 1996;22:672–676. 116. Mayer SA, Fink ME, Homma S, et al: Cardiac injury associated with neurogenic pulmonary edema following subarachnoid hemorrhage. Neurology 1994;44:815–820. 117. Weir BK: Pulmonary edema following fatal aneurysm rupture. J Neurosurg 1978;49:502–507. 118. MacIver IN, Frew IJC, Matheson JG: The role of respiratory insufficiency in the mortality of severe head injuries. Lancet 1958;1:390–393. 119. National Academy of Sciences—National Research Council: Accidental Death and Disability: The Neglected Disease of Modern Society. Washington, National Academy of Sciences—National Research Council, 1966. 120. Klauber MR, Marshall LF, Toole BM, et al: Cause of decline in headinjury mortality rate in San Diego County, California. J Neurosurg 1985;62:528–531. 121. Colohan ART, Alves WM, Gross CR, et al: Head injury mortality in two centers with different emergency medical services and intensive care. J Neurosurg 1989;71:202–207. 122. Graham DI, Adams JH, Doyle D: Ischemic brain damage in fatal non-missile head injuries. J Neurol Sci 1978;39:213–234. 123. Graham DI, Ford I, Adams JH, et al: Ischemic brain damage is still common in fatal non-missile head injury. J Neurol Neurosurg Psychiatry 1989;52:346–350. 124. Freytag E: Autopsy findings in head injuries from blunt forces. Statistical evaluation of 1,367 cases. Arch Pathol 1963;75:402–413. 125. Marshall LF, Marshall SB, Klauber MR, et al: A new classification of head injury based on computerized tomography. Report on the Traumatic Coma Data Bank. J Neurosurg 1991;75(Suppl):S14–S20. 126. Doppenberg EM, Bullock R: Clinical neuroprotection in traumatic brain injury: Lessons from previous studies. J Neurotrauma 1997;14:71–80. 127. Battistella F, Wisner D: Combined hemorrhagic shock and head injury: effects of hypertonic saline (7.5%) resuscitation. J Trauma 1991;31:182–188. 128. Freshman S, Battistella F, Matteumli M, Wisner D: Hypertonic saline (7.5%) vs. mannitol: A comparison for treatment of acute head injuries. J Trauma 1993;35:344–348. 129. Jones PA, Andrews PJD, Midgely S, et al: Measuring the burden of secondary insults in head injured patients during intensive care. J Neurosurg Anesthesiol 1994;6:4–14. 130. Bullock RM, Chestnut RM, Clifton GL, et al: Management and prognosis of severe traumatic brain injury—Guidelines for the management of severe head injury. J Neurotrauma 2000;17:449–553. 131. Firsching R, Woischneck D, Diedrich M, et al: Early magnetic imaging of brainstem lesions after severe head injury. J Neurosurg 1998;89:707–712.
W9418-08.qxd 10/31/03 3:41 PM Page 233
Chapter 8
Multiple Organ System Injuries and Isolated Severe Central Nervous System Trauma
132. Tomida M, Muraki M, Uemura K, et al: Postcontrast magnetic resonance imaging to predict progression of traumatic epidural and subdural hematomas in the acute stage. Neurosurgery 1998;43:66–71. 133. Messeter K, Nordstrom CH, Sundbarg G, et al: Cerebral hemodynamics in patients with acute severe head trauma. J Neurosurg 1986;64:231–237. 134. Chestnut RM, Marshall LF, Marshall SB: Medical management of intracranial pressure. In Cooper PR (ed): Head Injury. Baltimore: Williams & Wilkins, 1993, pp 225–246. 135. Narayan RK, Greenberg RP, Miller JD, et al: Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning and intracranial pressure. J Neurosurg 1981;54:751–762. 136. Robertson CS, Valadka AB, Hannay HJ, et al: Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999;27:2086– 2095. 137. Rosner M: Cerebral perfusion pressure: Link between intracranial pressure and systemic circulation. In Wood JH (ed): Cerebral Blood Flow. New York: McGraw-Hill, 1987, pp 425–448. 138. Rosner MJ, Coley IB: Cerebral perfusion pressure, intracranial pressure, and head elevation. J Neurosurg 1986;65:636–641. 139. Chan KH, Dearden NM, Miller JD, et al: Multimodality monitoring as a guide to treatment of intracranial hypertension after severe brain injury. Neurosurgery 1993;32:547–553. 140. Cruz J, Miner ME, Allen J, et al: Continuous monitoring of cerebral oxygenation in acute brain injury: Assessment of hemodynamic reserve. Neurosurgery 1991;29:743–749. 141. Unterberg AW, Klening KL, Hartl R, et al: Multimodal monitoring in patients with head injury: evaluation of the effects of treatment on cerebral oxygenation. J Trauma 1997;42:532–537. 142. Gopanath SP, Robertson CS, Contant CF, et al: Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 1994;57:717–723. 143. Obrist WD, Langfitt T, Yaggi J, et al: Cerebral blood flow and metab-
144.
145.
146.
147.
148.
149.
150.
151.
152. 153.
154.
233
olism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg 1984;61:241–253. Worthley LI, Cooper DJ, Jones N: Treatment of resistant intracranial hypertension with hypertonic saline—Report of two cases. J Neurosurg 1988;68:478–481. Einhaus S, Croce M, Watridge C, et al: The use of hypertonic saline for the treatment of increased intracranial pressure. J Tennessee Med Assoc 1996;89:381–382. Simma B, Burger R, Falk M, et al: A prospective randomized controlled study of fluid management in children with severe head injury: Lactated ringer solution vs. hypertonic saline. Crit Care Med 1998;26:1265–1270. Qureshi A, Surez J, Bhardwaj A, et al: The use of hypertonic (3%) saline/acetate infusion in the treatment of cerebral edema: Effective control of intracranial pressure and lateral displacement of the brain. Crit Care Med 1998;26:440–446. McIntosh TK: Novel pharmacologic therapies in the treatment of experimental traumatic brain injury—a review. J Neurotrauma 1993;10:215–222. Marshall LF, Maas AIR, Marshall SB, et al: A multi-centered trial on the efficacy of using tirilazad mesylate in cases of head injury. J Neurosurg 1998;89:519–525. Murray GD, Teasdale GM, Schmitz H: Nimodipine in traumatic subarachnoid hemorrhage: A reanalysis of the HIT1 and HIT2 trials. Acta Neurochir 1996;138:1163–1167. Young B, Runge JW, Wilberger JE, et al: Effect of pegorgotein on neurological outcome of patients with severe injury. A multi-centered randomized control trial. JAMA 1996;276:538–543. Yakolev AG, Faden AI: Molecular biology of CNS injury. J Neurotrauma 1995;12:767–777. Teasdale GM, Nicoll JAR, Murray G, et al: Association of apolipoprotein E polymorphism with outcome after head injury. Lancet 1997;350:1069–1071. Sorbi S, Nacimias B, Piacenti S, et al: Apo E as a prognostic factor for post traumatic coma. Natl Med 1995;1:852.
W9418-09.qxd 11/3/03 2:53 PM Page 235
Chapter 9 Spinal Neurosurgery Trent L. Tredway, MD, Lorenzo F. R. Muñoz, MD, Robin L. Wellington, PhD, and Richard G. Fessler, MD, PhD
The spine, through its various segments, serves to protect the neural structures it encloses while at the same time providing flexibility of movement. As a result, this bony structure is prone to degenerative changes that result from the dynamic forces of “wear and tear” to which it is constantly exposed. Moreover, the spinal column is also the end target of trauma, degenerative disorders, systemic disease, congenital malformations, and primary or metastatic tumors. Diseases of the spine can manifest as neurologic deficits, pain, and cosmetic deformities. Once identified, surgical clinical judgment defines the operation that is suitable for the pathologic process. As a result of the complexity of spinal anatomy, there are a myriad of procedures, approaches, and types of instrumentation that can be used. Subsequently, there are also a number of complications that can occur. Detailed management of these complications is beyond the scope of this chapter. However, a comprehensive review of the epidemiology and natural history of the most common disease processes will be addressed in conjunction with common complications.
Cervical Spine Cervical Spine Trauma According to the National Spinal Cord Injury Registry, when spine trauma occurs the cervical segment is the most often affected location. Approximately 50% of these injuries are caused by motor vehicle accidents and approximately 25% are the result of falls. Additionally, sports and recreation–
related events account for 10% of these injuries with pool diving being the most common recreational cause. The incidence of traumatic spinal cord injury is five times greater in males than in females.1 The cervical spine is the region most frequently injured in car accidents. National studies have found a one in 300 incidence of severe neck injury in those vehicular accidents severe enough to have the vehicle towed from the scene. That incidence increases to one in 14 for cases in which the patient has been ejected from the car.2 The C2 vertebral level is the most frequently injured with fractures of C5 and C6 being the next most frequently injured levels. These injuries are most common in the third decade of life and decrease with advancing age.
Atlas Fractures (C1) The incidence of atlas fractures peaks in the second decade of life, with men being affected almost twice as often as women. Atlas fractures account for 4% to 15% of all cervical fractures. The most common mechanism of injury is motor vehicle accidents. When isolated, the incidence of neurologic deficits is between 4% and 17%. Atlas fractures occur in association with axis fractures in approximately half of the cases.3–5 Atlas fractures can be classified as: (1) anterior arch fractures, (2) posterior arch fractures, (3) simple lateral mass fractures, (4) comminuted lateral mass fractures, (5) multiple ring burst (Jefferson) fractures, and (6) anteroposterior ring fractures (Fig. 9-1).6 235
W9418-09.qxd 11/3/03 2:53 PM Page 236
236
Section I
Introduction
A
C
C1 burst fractures carry the eponym of “Jefferson fracture” after Sir Geoffrey Jefferson’s first systematic description of this injury in 1920.7 A “pure” Jefferson fracture is a fourpart ring fracture as a result of the lateral displacement of the lateral masses secondary to compression by the occipital condyles. The eponym is often used to describe most types of C1 fractures regardless of the specific fracture anatomy involved. The biomechanics of axial loading are thought to be the most common factor in the genesis of atlas fractures. The injury can occur with different neck positions so that multiple force vectors come into play. Therefore, it is imperative for the pathology of the fracture to be properly identified to assess if there has been an injury to the transverse ligament. When assessing the integrity of the transverse ligament, an open mouth or anteroposterior radiograph may be obtained. If the sum total of the overhang of both C1 lateral masses on C2 is greater than or equal to 7 mm, then the transverse ligament is probably disrupted. This measurement is commonly known as the “Rule of Spence.”8 This injury requires
B
Figure 9-1. Atlas fracture. A, Anterior arch fracture; B, Posterior ring fracture; C, Jefferson fracture.
immediate rigid immobilization and eventual surgical fusion. Management of Atlas Fractures Nondisplaced atlas fractures can be treated with semirigid external orthoses such as the Philadelphia collar or sternooccipital-mandibular immobilizer brace. More complex fractures, such as comminuted, or fractures with widely displaced fragments, must be treated with a Halo brace. In fact, more than 95% of isolated atlas fractures and 75% of C1-C2 combination fractures can be treated with an orthosis.6 There are two types of atlas fractures that are considered unstable. These types occur when the transverse liagment has become disrupted or when its insertion into the bony tubercle has been fractured. In either case, surgery is indicated because external immobilization will not promote healing of these structures. The goal of surgery is fixation of the unstable segment with preservation of normal motion. When the posterior arch of C1 is intact, posterior cervical wiring may be a rea-
W9418-09.qxd 11/3/03 2:53 PM Page 237
Chapter 9
sonable option.9 If the posterior arch is incompetent to sustain such wiring, then posterior atlantoaxial facet screws may be indicated. However, this procedure can only be perfomed if intact C2 pedicles and intact C1 lateral masses are present.10 If C1 cannot be directly fixated, an occipitocervical fusion may become necessary. When assessing the treatment for C1 and C2 combination fractures, there are essentially two instances in which internal fixation is warranted: (1) when there is a disruption of the transverse liagment, and (2) when the associated odontoid type II fracture has greater than 6 mm of displacement between the dens fragments.
B
237
Axis Fractures (C2) There are essentially three types of axis fractures: (1) odontoid, the most common (55%); (2) “hangman’s” fracture (23%); and (3) those classified as “other” (22%) but most commonly being fractures of the body of C2 (Fig. 9-2).11 In 1974, Anderson and D’Alonzo described subtypes of odontoid fractures.12 Type I fractures (0% to 5%) consist of a fracture through the tip of the dens. This occurs secondary to avulsion of the apical ligament. Although long considered to be a stable injury, they may not occur as an isolated fracture and may be a manifestation of atlanto-occipital disloca-
A
Figure 9-2. Axis fracture. A, Odontoid fracture; B, Hangman’s fracture.
Spinal Neurosurgery
W9418-09.qxd 11/3/03 2:53 PM Page 238
238
Section I
Introduction
tion.13 Type II fractures, the most common (37% to 80%), consist of a fracture through the base of the dens where it attaches to the body of C2. Type II fractures are considered unstable; however, controversy exists on whether rigid immobilization or operative fusion is the treatment of choice. The degree of displacement and the age of the patient are important factors in predicting nonunion and should be considered when treating this type of odontoid fracture. Type III fractures extend into the cancellous portion of the body of C2 and are usually considered stable fractures. Approximately 90% heal with external immobilization if adequately maintained for 8 to 14 weeks.14 A hangman’s fracture is a bilateral fracture through the pars interarticularis (the bridge of bone between the superior and inferior facets) or through the pedicles of the axis. Hangman’s fractures typically are the result of hyperextension-axial loading injuries. These fractures have been classified into three types based on the amount of subluxation, angulation, and disruption of the facet capsules. Type I fractures have a subluxation of C2 on C3 of less than 3 mm, whereas type II fractures have a greater subluxation with angulation of greater than 11 degrees. Type III fractures involve disruption of the facet capsules with significant subluxation and spinal canal compromise (Fig. 9-3). Closed reduction of this injury may lead to further deterioration in the neurologic examination.15–17 Open reduction with internal fixation is recommended in this highly unstable type of C2 fracture. The fractures characterized as “other” are mostly fractures of the body of C2. Benzel and co-workers have classified these fractures mostly based on the orientation of the fracture (e.g., coronally, sagittal).18 Management of Axis Fractures Almost 100% of type I odontoid fractures heal without treatment as long as there has been no C1 ligamentous injury. Type II odontoid fractures are a management challenge because of their high rate for nonunion. Hadley and colleagues described a 67% nonunion rate for those fractures displaced 6 mm or more and a 26% union rate for those less than 6 mm.14 These nonunion rates are applicable even with adequate Halo brace external immobilization. The Halo brace remains the mainstay of treatment for these fractures in the initial stages. When deciding whether an open fixation is needed, other factors such as age, ligamentous injuries, systemic disease, and so on, must be taken into consideration. Type II odontoid fractures can be addressed with either an anterior or posterior approach. The anterior approach involves the standard exposure for the anterior cervical spine. Then, with the assistance of biplanar fluoroscopy, a lag screw is inserted through the body of C2 into the fractured odontoid segments (Fig. 9-4A). This procedures allows for instant fixation of the odontoid while preserving the rotational axis of the C1/C2 complex. The posterior approach deals mostly with wiring or screw fixation and fusion (arth-
Figure 9-3. Hangman’s fracture types I, II, and III.
odesis) of the posterior arch of C1 with C2 (Fig. 9-4B). However, the preservation of rotation that is seen with odontoid screws is eliminated with posterior stabilization techniques. For wiring to be feasible, the posterior arch of C1 needs to be intact. If the posterior arch is not intact, then C1C2 transarticular screws can be placed. Alternatively, the occiput could be incorporated into the fusion using lateral mass plates or rods (Fig. 4C and D). Type III odontoid fractures have a higher union rate then Type II. The nonunion rate varies from 0% to 17%.19 External immobilization is the treatment of choice with fusion rates near 100%.11 An internal fixation might be required in those more complex fractures were there are comminuted fragments or severe ligamentous injuries. In these cases, a C1/C2 wiring with arthrodesis is recommended. Complications of Axis Fractures The complications of anterior odontoid screw fixation are similar to other anterior cervical approaches. The main difference pertains to the actual screw placement. It is imperative that biplanar fluoroscopy guide the experienced surgeon in the screw trajectory. Malposition of the screw can result
W9418-09.qxd 11/3/03 2:53 PM Page 239
Chapter 9
A
B
C
Figure 9-4. A, Odontoid screw fixation; B, Transarticular screw fixation; C, C1/C2 lateral mass screw fixation; D, Occipitocervical fusion.
D
Spinal Neurosurgery
239
W9418-09.qxd 11/3/03 2:53 PM Page 240
240
Section I
Introduction
in dural penetration with cerebrospinal fluid (CSF) leak, catastrophic neurologic deficits, or vertebral artery injury. As with other instrumentation, screw migration or screw fractures can occur over time. For this reason frequent followup cervical spine radiographs are recommended. Another complication of odontoid screw fixation is nonunion. Apfelbaum describes a nearly 100% rate of union in those fractures nonunited less than 6 months.20 In contrast, the rate of succesful bony union was just 27% in those chronically nonunited fractures (more than 6 months). However, successful stabilization, defined as bony plus fibrous union, was approximately 60% in nonunited fractures. Posterior atlantoaxial wiring and arthrodesis provides excellent translational and rotational stability. Several techniques for wiring and bone graft have been proposed. The interspinous method for atlantoaxial arthrodesis, as proposed by Sonntag, has a fusion rate of 86%.21 Extreme caution must be used when making the sublaminar wire passes in order to avoid dural tears or damage to the underlying cervical spinal cord. The bone grafts can migrate and the wires can break or erode through the bone increasing that rate of nonunions. Often, this is a function of the health of the bone undergoing fusion. For this reason, many surgeons will use a halo brace for several months after surgery. Cervical Subaxial Fractures Trauma also affects the subaxial (C3-C7) spine. Injuries to this region of the spine most commonly occur as a result of flexion-rotation or hyperflexion injuries. Ligamentous injuries may occur that allow for subluxation of vertebrae. Cadaveric studies have been performed, and demonstrate that a horizontal subluxation of greater than 3.5 mm of one vertebral body on another, or greater than 11 degrees angulation of one vertebral body relative to the next indicates ligamentous instability.22 Severe ligamentous injuries caused by hyperflexion may result in “locked-facets” (Fig. 9-5A). Patients with unilateral locked facets may present with subluxation greater than 3.5 mm and are generally neurologically intact. Bilateral lockedfacets occur less frequently and are associated with severe hyperflexion injuries. These patients present with subluxation of greater than 50% and also demonstrate signs of spinal cord or nerve root injury. Locked-facets require reduction that may be attempted through a closed technique using incremental traction or by open surgical reduction. In either case, surgical intervention with instrumentation and arthrodesis is required for definitive treatment. Teardrop fractures, as originally described by Schneider, are unstable fractures resulting from hyperflexion.23 These injuries are so named because of the chip of bone just beyond the anterior inferior edge of the vertebral body seen on a lateral cervical spine radiograph (Fig. 9-5B). This injury should be differentiated from a simple avulsion fracture. With teardrop fractures, a fracture through the sagittal plane of the vertebral body may be observed. Patients with
teardrop fractures are often quadriplegic, although some may be intact. True teardrop fractures are unstable and require surgical stabilization. The extent of injury usually requires a combined anterior and posterior surgical procedure. A more common and less threatening injury seen in the subaxial spine is the Clayshoveler’s fracture that was first described in Australia (Fig. 9-5C). In this injury, the spinous process, usually C7, is avulsed. The fracture is stable and does not require surgical intervention. Flexion-extension radiographs should be performed to assess if any occult fractures have made this segment of the spine unstable. Congenital Abnormalities of the Craniovertebral Junction Congenital abnormalities of the craniovertebral junction are rare conditions that are often latent until late childhood or early adulthood. Many of these conditions are recognized secondary to traumatic or degenerative processes. These conditions often require surgical intervention to halt the progression of neurologic deficits, decrease intractable pain, or stabilize the occipital-C1-C2 complex. Basilar invagination is the most common congenital anomaly of the atlanto-occipital junction. The anomaly results from malformation of all three parts of the occipital bone (basiocciput, exocciput, and squamous occipital bone) (Fig. 9-6).24 This anomaly may compromise the space available within the foramen magnum. Patients with symptomatic basilar invagination often complain of paresthesias and weakness in the limbs. Vertebral artery anomalies may accompany basilar invagination; therefore, symptoms of vertebral artery insufficiency can occur.25 The diagnosis of basilar invagination is based on radiographic imaging. A series of reference lines have been described to help define this anomaly. McRaes’ line is an imaginary line across the foramen magnum connecting the basion with the opisthion. The length of this line should exceed 19 mm and no part of the odontoid should be above this line. If any part of the odontoid is above this line, then basilar invagination is present.26 A second reference line, Chamberlain’s line, connects the hard palate to the opisthion. Less than 3 mm of the dens should be above this line.27 A third line, McGregor’s baseline, connects the posterior margin of the hard palate to the most caudal point of the occiput. No more than 4.5 mm of the dens should be above this line.28 A fourth reference line, Wackenheim’s clivus-canal line, joins the dorsum sellae to the tip of the clivus. The odontoid should be tangential to or below this line. A fifth reference line, Fishgold’s digastric line, connects the digastric notches on an anteroposterior radiograph. The odontoid should lie below this imaginary line.29 Basilar impression is often associated with basilar invagination; however, the former describes the acquired form of basilar invagination secondary to softening of the occipital bone. This condition is prevalent in rheumatoid arthritis,
W9418-09.qxd 11/3/03 2:53 PM Page 241
Chapter 9
Figure 9-5. Subaxial fracture. A, Bilateral locked facets; B, Teardrop fracture; C, Clayshoveler’s fracture.
B
Spinal Neurosurgery
241
A
C
as well as in Paget’s disease, hyperparathyroidism, achondroplasia, and osteogenesis imperfecta.24 These patients often require surgical intervention consisting of decompression and stabilization. Another commonly seen congenital anomaly of the craniovertebral junction is assimilation of the atlas. In this condition, a failure of segmentation between the fourth occipital and the first spinal sclerotomes occurs. It occurs in approximately 0.25% of the population, and is often associated with other anomalies such as cleft palate, basilar invagination, and Klippel-Feil syndrome.30–33
The onset of symptoms usually occurs between the ages of 20 to 40 years. Symptoms include weakness, spasticity, gait disturbances, and occasionally cranial nerve dysfunction.34 Furthermore, evidence of cerebellar dysfunction may also be present. Surgical stabilization and fusion may be necessary in these patients. Degenerative Disorders Degenerative processes of the cervical spine most frequently present as spondylosis, osteophytic compression, and disk
W9418-09.qxd 11/3/03 2:53 PM Page 242
242
Section I
Introduction
Figure 9-6. Basilar invagination.
herniation.35 Cervical radiculopathy is a common occurrence. It has been reported that individuals who lift heavy objects as well as those who smoke are predisposed to acute cervical disk disease. Operators of vibrating equipment and individuals riding in cars for extended periods also have an increased frequency of disk herniations.36 Repetitive subclinical trauma probably influences the onset and rate of progression of cervical spondylosis.37 There appear to be no gender differences in the occurrence of spondylosis; however, women are less severely affected. The incidence of cervical disk herniations peaks in the fourth decade. This peak incidence in disk herniations appears to be the result of the combination of the maximum expansile strength of the disk (i.e., when the disk is the most hydrated) and the peak incidence of annular tears. Thus, after the age of 50, the dehydrated disks are less prone to herniate even when the incidence of annular tears increases. Studies revealed 61% to 68% of patients exhibit preoperative motor weakness with disk herniation.38 However, the recovery of motor function after decompressive surgery is excellent. Cervical spondylosis as defined by Clarke and Robinson in 1956 refers to “chronic degenerative changes due primarily to intervertebral disc decay, which are probably universally present in elderly persons” (Fig. 9-7).39 Some authors extend Clarke and Robinson’s definition to include degenerative changes in the facet joints, longitudinal ligaments, and ligamentum flavum. The osteophytes of spondylosis are associated with degeneration of the intervertebral disk.40 The genesis of the spondylotic ridge lies in the bulging annulus fibrosus, which in turn elevates the periosteum triggering subperiosteal bone formation (osteophytes). The deterioration of the disk actually starts in childhood when children begin to walk. This has been associated with a loss of blood supply, which might be secondary to the increased axial loading pressures of upward posture.41
Although disk herniations may cause significant pain, the etiology of the pain is rather elusive. For one, there seems to be no uniform explanation as to why some patients with clear evidence of disk herniation do not have the typical radicular pain distribution.42,43 Moreover, there also is no universal explanation for the sharp pain experienced as a result of the nerve root compression caused by the extruded fragment. For instance, usually paresthesias and numbness are all that is felt with entrapment neuropathies such as carpal tunnel syndrome. The difference may be due to traction on the nerve root instead of compression from the disk fragment.44 Other theories postulate that the extruded disk fragments elicit reactions in the epidural space that cause a cascade of enzymatic reactions that in turn hydrolyze the extruded material.45 This breakdown of products may produce the nerve root irritation. The most common sites of herniation in the cervical spine are at C5-C6 and C6-C7, causing compression of the C6 and C7 root, respectively. Involvement of the C5 nerve root is perhaps the most disabling. The deltoid muscle weakness that ensues makes it difficult to adduct the arm over 20 degrees. The C6 radicular pain radiates down the lateral aspect of the forearm to include the first two fingers. The biceps muscle is affected with loss of its reflex and motor weakness. The C7 radicular pain radiates down the posterolateral aspect of the arm going on to the middle finger. Its characteristic motor involvement is the triceps muscle. Early loss of the triceps reflex and weakness are commonly seen.
Figure 9-7. Degenerative cervical spine/spondylosis.
W9418-09.qxd 11/3/03 2:53 PM Page 243
Chapter 9
Cervical spondylotic changes are often the cause for cervical myelopathy. Classic signs in cervical spondylotic myelopathy (CSM) include Lhermitte’s sign (electrical shock sensation associated with neck flexion), hand clumsiness, distal weakness, generalized hyperreflexia, Hoffman’s sign (contraction of the thumb and index finger upon flickering of middle finger), and spastic gait.46,47 Most cases of cervical radiculopathy will resolve with conservative management.48 In fact, almost 80% of those patients with cervical radiculopathy who are treated conservatively will have partial or total relief of their symptoms.49 However, much less is known about the natural history of spondylotic cervical myelopathy. Onset is typically in the sixth decade of life with males more commonly affected. CSM has a slow onset, and once present, complete reversal is rare.50 The treatment options for cervical spine disease are divided between nonoperative and operative management. It is reasonable to attempt nonoperative treatment if there is improvement in pain relief without deterioration in motor function. Surgery is recommended when there has been neurologic deterioration or nonoperative treatment has failed. Nonsurgical Management Immobilization of the cervical spine is a primary goal in nonoperative management. With this in mind, numerous braces have been designed. Most soft collars do not limit cervical motion.51 These collars serve as a reminder for the patient to protect the neck against excessive motion. The more rigid orthoses (Philadelphia collar, sternooccipital-mandibular immobilizer brace, etc.) provide reasonable immobilization of the mid-cervical segments for flexion and extension but fail to immobilize against lateral flexion. Other orthoses, such as the Minerva body vest, can provide significant immobilization. One of the most significant limitations in the use of orthoses is patient compliance and tolerance. Bed rest and cervical traction can often be used in the nonoperative management of cervical spine disease while a combination of muscle relaxants, nonsteroidal anti-inflammatory drugs, steroids, and analgesics can provide reasonable relief to many patients. Antidepressants can be used in some patients with chronic pain.48 Following pain control, physical therapy becomes a key component to return the patient to full activities. Surgical Management There are many approaches and operations in the treatment of cervical spine disease. The best approach is the one that allows the most direct and safe decompression of the offending pathology. Each patient’s individual (Fig. 9-8) pathology must be scrutinized so that the operative plan can be tailored to the specific disease process. As with any other surgical
Spinal Neurosurgery
243
interventions, there are complications associated with these different procedures that vary with the anatomy. Anterior Cervical Diskectomy with Interbody Fusion The anterior approach allows direct visualization of the intervertebral interspace. Thus, it allows direct access for decompression of the anterior aspect of the cervical spinal cord. This approach does not only enable the surgeon to effectively treat soft disks, but it also allows for the resection of osteophytes. In the anterior approach to the cervical spine, thorough knowledge of the anatomic landmarks of the neck is important in planning the incision. The hyoid bone is approximately at the level of C3, the thyroid cartilage corresponds to the level of C4-C5, and the cricoid cartilage corresponds to the C5-C6 interspace. A transverse incision is often chosen over a vertical one because it is cosmetically more acceptable. Once the platysma muscle is opened and undermined for further access, the plane medial to the sternocleidomastoid muscle is bluntly dissected. The carotid sheath is displaced laterally with gentle retraction and the prevertebral fascia is opened until the longus colli muscles are found attached to the anterior aspect of the vertebral bodies. A combination of monopolar and bipolar coagulation is used to separate the muscles from the vertebral bodies along their medial borders. The longus colli muscles are retracted laterally with self-retaining retractors. The disk level is then checked with the help of a spinal needle and a cross table radiograph. Once the level is confirmed, the anterior longitudinal ligament and the annulus are incised and the anterior aspect of the disc is removed. A high-speed hand drill can be used to remove bone from the two adjacent vertebral bodies to better decompress the spinal canal and gain access to the posterior longitudinal ligament. Osteophytic processes may also be carefully removed with the drill or with rongeurs. Next, the posterior longitudinal ligament is opened and the epidural space is inspected for loose fragments. After an adequate decompression has been confirmed, a bone graft is placed in the interspace with the help of gentle interbody retraction (Fig. 9-8C). Instrumentation may also be used in this approach with little complication.52 Complications. The incidence of increased neurologic deficits after surgery is less than 1%.53 The most common neurologic complication following anterior cervical diskectomy is persistence of neurologic signs and symptoms. Persistent or worsening symptoms have a 5% incidence and are most likely secondary to the incomplete removal of the bony osteophytes compressing the neural elements.54 The possible complications associated with this approach start from the time the patient is intubated and positioned. Excess flexion or extension can produce neurologic damage in the congenitally or degeneratively narrowed canal. In patients with narrowed canals, fiberoptic intubation is often
W9418-09.qxd 11/3/03 2:53 PM Page 244
244
Section I
Introduction
B
A
C
Figure 9-8. Cervical herniated nucleus pulposus. A, Sagittal MRI; B, Axial MRI; C, Anterior cervical diskectomy with fusion radiograph.
chosen when placing these patients under anesthesia. Once anesthetized, that patient’s neck should be carefully protected from extreme positioning. The patient may be placed in gentle traction to help avoid this problem. Transient dysphagia secondary to tissue swelling is common. Direct esophageal injury has also been reported.55 This complication can be decreased when the esophagus is identified intraoperatively after insertion of a nasogastric tube. Pneumothorax is a possible complication when a low anterior approach (C7-T1) is performed. At this level, the dome of the parietal pleura is at risk for injury.56
Hoarseness, secondary to anterior cervical spine surgery, as reported by Riley and colleagues, has a 4% incidence.57 Permanent recurrent laryngeal nerve injury and CSF leak complications have a 1% incidence. Horner’s syndrome may result from manipulation of the sympathetic ganglia located on the lateral surface of the longus colli. Injury to the spinal cord and nerve roots may lead to devastating consequences. Graft extrusion is seen in 2% to 8% of cases.58 Patients could very well be asymptomatic after graft extrusion; however, a retropulsed graft can cause neurologic problems
W9418-09.qxd 11/3/03 2:53 PM Page 245
Chapter 9
Spinal Neurosurgery
245
Complications. The complications of corpectomies are not different from those associated with the anterior diskectomy approach insofar as the anterior neck anatomy is at risk for injury. Specifically, the esophagus, cranial nerves X and XII, the vertebral arteries, the trachea and the vessels of the carotid sheath are at risk. Corpectomies tend to be lengthy procedures that, even in the best of hands, usually take a few hours. Injuries to the vascular structures may be caused by the sharp surfaces of the retractor blades. These lesions, if identified intraoperatively, can often be repaired primarily.63 Postoperative paratracheal soft tissue swelling can compromise the airway and thus require prolonged intubation.64 Postoperative hoarseness is often present. However, it is most commonly related to the endotracheal intubation or soft tissue dissection. The incidence of voice changes secondary to recurrent laryngeal nerve injury is less than 1%.65 However, because a true transection of the nerve is rare, phonation improves within weeks to months. Injury to the cervical sympathetic chain causing Horner’s syndrome can occur with excessive lateral retraction on the longus colli.66 Esophageal injuries can have disastrous consequences. If undetected, they can result in prevertebral abscesses and even mediastinitis. Again, the insertion of a nasogastric tube can help the surgeon palpate the esophagus when positioning the retractor blades to avoid undue retraction pressure. CSF leaks, when present, usually occur during removal of the Figure 9-9. Postoperative complication: failed instrumentation with broken screw.
as it impinges on the spinal cord. Usually, surgery is necessary to realign the malpositioned graft. Complications associated with the iliac bone graft harvest are infection, hemorrhage, and pain.59 For this reason, many surgeons prefer the use of allograft to eliminate these postoperative complications. Instrumentation failures may also occur (Fig. 9-9). Corpectomy for Cervical Spondylotic Myelopathy The advent of magnetic resonance imaging (MRI) (see Fig. 9-8) has enabled the sagittal visualization of the cervical spinal cord as well as provided another pathway to the treatment algorithm of CSM. Most neurosurgeons will agree that the approach to cervical spine disease (i.e., anterior vs. posterior) is largely dependent on two aspects: the anatomic location of the pathology and the prevalent curvature of the spine. When a diskectomy will not adequately decompress the spinal cord from in front, then a corpectomy may be indicated (Fig. 9-10). Laminectomy may be indicated if the cervical lordosis is well preserved.60 Saunders and associates found that 85% of those patients in the series had significant improvement of their CSM with the central corpectomy approach.61 In contrast with another series, a 70% improvement was reported with laminectomy or diskectomy.62
Figure 9-10. Anterior cervical corpectomy.
W9418-09.qxd 11/3/03 2:53 PM Page 246
246
Section I
Introduction
posterior longitudinal ligament. Given the very narrow surgical corridor, primary repair is virtually impossible. Many surgeons use a lumbar drain to reduce pressure and divert CSF during the acute healing phase. Posterior Approach for Cervical Disk Disease The posterior approach in the treatment of cervical disk disease may be indicated where the cervical nerve root compression is in a posterolateral location. In this approach, the medial aspect of the cervical facet is usually approached through a midline incision. Once the lamina/articular process complex has been exposed, the bony dissection proceeds. The next phase involves removing bone over the junction between the lamina and the articular process of the level in question. The ligamentum flavum and the venous plexus associated with the nerve root sheath will be seen. The bone removal is then carried out into the facet joint laterally and dorsal to the nerve root. To avoid instability, it is important to avoid drilling of more than half of the facet joint.67 Complications. Careful dissection should be performed to delineate the ligamentum flavum from the exiting nerve root. Furthermore, incomplete bony decompression will most certainly result in the persistence of the radiculopathy. Removal of a disk through a limited posterior approach can also result in spinal cord injury. Finally, as in most spinal procedures, it is imperative that the appropriate level of pathology be properly selected. Therefore, intraoperative radiographs are recommended to confirm the correct surgical level.
Posterior Lateral Mass Plate Fixation of the Subaxial Cervical Spine These systems, which have evolved from spinous process wire fixation into today’s titanium alloy lateral mass plates/screws, are used for the management of posterior cervical instability. The rigid internal fixation of the posterior cervical spine is most often performed with lateral mass plates. One of the greatest advantages of lateral mass plates is that they provide immediate stability to the treated segments of the posterior cervical spine (Fig. 9-11). However, given the persistent loading forces of the cervical spine, all instrumentation is destined to “fail.” For that reason, for the construct to be successful, a solid bony fusion is needed. The necessity for arthrodesis in conjunction with lateral mass plates is nevertheless a controversial issue because some authors report good results with plating alone. When performed, the facet joints are denuded of their cartilage in order to facilitate apposition of their bony surfaces. It is at this point that some surgeons choose to add some cancellous bone to the denuded joint. Generally, there are four indications for spinal stabilization: (1) to restore clinical stability to a spine in which the structural integrity has been compromised, (2) to maintain alignment after a deformity correction has been performed,
Figure 9-11. Posterior lateral mass plates with anterior cervical plating and fusion.
(3) to prevent progression of a deformity, and (4) to alleviate pain.68 Trauma is one of the most common indications for lateral mass plating.69 However, external fixation usually is adequate in treating those injuries in which there are no ligamentous injuries. Moreover, when extensive tumor resection mandates wide laminectomies with involvement of the facet joints or pedicles, lateral mass plating is also useful in the stabilization of the posterior cervical spine. This technique might suffice as long as the tumor does not involve the anterior and middle spinal column. Once the patient has been carefully positioned and intubated, proper reduction of the cervical spine is achieved and confirmed with a cross table radiograph. Reduction and proper alignment are often achieved with axial traction. However, internal reduction may be required. Once subperiosteal exposure as far laterally as the facet joints has been achieved, the field is then prepared for the bony fusion and for the lateral mass plate implantation. Complications. Sawin and Traynelis70 subdivide the compli-
cations associated with lateral mass plating into three categories: (1) wound complications, (2) neurovascular injuries, and (3) spinal/biomechanical complications. Problems such as hematomas, CSF leak, and infections were identified as causes for wound breakdown. Sizable hematomas can cause compression of the neighboring neural structures with catastrophic results. Even small
W9418-09.qxd 11/3/03 2:53 PM Page 247
Chapter 9
hematomas that do not cause neurologic problems can hinder wound healing and even promote infection. CSF leaks, as previously mentioned, should be repaired intraoperatively. Usually these leaks are not a result of the instrumentation per se but rather are caused from a combination of the pathology at hand and the surgical technique. Persistent pseudomeningoceles can contribute to wound breakdown and even persistent infections. Surgical reexploration and lumbar CSF drainage are at times required to address this problem. When infection occurs, surgical debridement of the site may be warranted. At first, the instrumentation should be spared from removal with aggressive antibiotic treatment. However, if the infection persists, then the instrumentation should be removed. When placing lateral mass screws, the spinal cord and the vertebral artery are at risk for injury. However, the incidence of injury to these structures is low. Acute neural compromise may occur as a result of misguided drilling and screw placement. Immediate deficits would be recognized in this setting. It is possible that late onset deficits could occur in situations where the cervical spine alignment is altered. Cadaveric studies indicate that neither the spinal cord nor the vertebral artery is in real danger when lateral mass screws are placed along the standard trajectories.71 A less than 1% to a 3.6% incidence of inadvertent nerve root involvement with screw placement has been reported.71,72 In a review of 490 cases, Traynelis reports only one case of vertebral artery trauma as a result of this technique.72 Postoperative spinal complications resulting from hardware misplacement can occur. Suboptimal screw placement may lead to further deformities over time. Loosening of screws can occur even months after the procedure. Although otherwise rare, these complications are more commonly encountered with osteoporotic bone. Heller and colleagues reported a 1.3% incidence of plate breakage over an average follow-up of 1.5 years. A 0.1% incidence of screw fracture and 0.9% incidence of screw loosening were also reported during this period.73 Laminectomy for Cervical Spondylotic Myelopathy Laminectomy has been performed for years for the treatment of cervical spondylotic myelopathy. It is generally indicated for patients with a compressive myelopathy and an associated lordosis. Benzel has reported an increased effectiveness with sectioning of the dentate ligament in those patients with progressive or more severe myelopathy symptoms.62 If the laminectomy extends far laterally past the medial one fourth or one third of the facet joint, instability may ensue. In these cases, fusion and instrumentation may be indicated. Ossification of the Posterior Longitudinal Ligament Ossification of the posterior longitudinal ligament (Fig. 9-12) appears in approximately 2% of the cervical spine
Spinal Neurosurgery
247
Figure 9-12. Ossification of posterior longitudinal ligament.
radiographs in the Japanese population. Autopsy studies from Japan demonstrate an incidence of 20% in subjects older than 60 years of age.74 The pathogenesis for this disease remains unclear. Patients with this rare disease usually present with myelopathic findings. Because the process is usually a long, progressive disease, many patients remain asymptomatic. However, patients may deteriorate and require surgical intervention in the form of decompression to enlarge the spinal canal. Paget’s Disease Paget’s disease is a metabolic bone disorder that commonly affects the spinal column. The cause of this disease is unknown, but a viral origin has been postulated. Areas of bone resorption and new bone deposition with the cumulative effect of a net positive bone balance characterize the disease. Patients with Paget’s disease may present with pain, paresthesias, or neurologic deficits secondary to neural compromise. The overgrowth of bone may exacerbate foraminal and canal stenosis. Surgical decompression may be necessary in some individuals; however, surgery does not halt the disease process.
W9418-09.qxd 11/3/03 2:53 PM Page 248
248
Section I
Introduction
Rheumatoid Arthritis Rheumatoid arthritis is a systemic disease of unknown etiology that primarily affects the synovium of the feet, hands, and spine. This disease is more commonly observed in women and may result in severe neurologic problems. The cervical spine is frequently involved and may result in atlantoaxial subluxation, subaxial subluxation, and vertical basilar impression of the odontoid process. The cervical spine abnormalities are a result of destruction in the joints, ligaments, and bone by synovitis.75 Atlantoaxial subluxation represents the most common manifestation of rheumatoid involvement of the cervical spine.76 The most common clinical finding in these patients with atlantoaxial involvement is severe and persistent pain. The pain may be exacerbated with neck movement and may not respond to conservative management. Furthermore, as the disease progresses, patients may demonstrate signs and symptoms of myelopathy. The degree of myelopathy may not correspond to the degree of bony impingement, as often the culprit is a large pannus formation (Fig. 9-13). Indications for surgery include worsening of neurologic examination findings, progressive myelopathy, and intractable pain. Neural decompression with fusion of C1-C2 or occipitocervical fusion is usually performed. Subaxial subluxation of the cervical spine secondary to the rheumatoid process may lead to myelopathic findings as well. These patients often are osteoporotic and therefore,
Figure 9-14. Ankylosing spondylitis with pathologic fracture.
surgical intervention with instrumentation may lead to a higher complication rate when compared to the nonrheumatoid patient. Basilar impression secondary to pannus formation may cause myelopathy as well as lower cranial nerve palsies. The vertebral artery may also be compromised from the disease process and patients can present with signs and symptoms of vertebrobasilar insufficiency.77 Ankylosing Spondylitis Ankylosing spondylitis is an inflammatory disorder that affects synovial and cartilaginous joints, especially the joints of the spinal column. The etiology of this disorder remains unknown; however, individuals possessing the HLA-B27 histocompatibility subtype are at higher risk.78 Furthermore, there is a 3 : 1 to 8 : 1 male predominance.79 The age of onset is between 15 and 30 years, with less than 5% presenting after age 50.80 The prevalence in the U. S. population is approximately 0.1%.81 Inflammation of the ligamentous attachments, enthesopathy, occurs along the spinal ligaments. Intervertebral disks and vertebrae erode with new bone formation contributing to ankylosis. The bone is stiffer and more prone to fracture (Fig. 9-14) and subluxation after trivial trauma.82 In plain radiographs, the classic appearance of a “bamboo spine” may be observed. Patients with ankylosing spondylitis may require surgical intervention in the form of decompression and fusion or correction of severe kyphosis with osteotomy and fusion.
The Thoracic Spine Anatomic Considerations Figure 9-13. Rheumatoid arthritis with C1 pannus.
The thoracic spine is protected from injury by the paraspinal muscles and the thoracic cage; however, the spinal canal
W9418-09.qxd 11/3/03 2:53 PM Page 249
Chapter 9
Spinal Neurosurgery
249
diameter is narrow throughout the thoracic spine and may lead to neurologic deficits with further narrowing of the canal. Furthermore, the lower cervical/upper thoracic region has a tenuous blood supply that may complicate injuries to this region. The blood supply to the upper thoracic cord relies on an anastomotic supply from the thyrocervical and costocervical trunks as well as the radicular arteries. In the middle region, T4-T8, a single thoracic radicular artery supplies this vulnerable area. In trauma cases or when hypotension occurs, this region of the spinal cord is the first to be affected. The middle and lower thoracic region has a more robust blood supply that relies on the great radicular artery of Adamkiewicz. This artery usually arises on the left at the T10 to T12 levels in 75% of patients. The anatomy of this region needs to be taken into consideration when performing thoracic spinal surgery. In addition, general and vascular surgeons should keep this anatomic relationship in mind when performing surgery on abdominal aortic aneurysms because cross-clamping of the aorta proximal to the great radicular artery takeoff may leave patients with postoperative paraplegia if the collateral blood flow is insufficient during the procedure. Congenital: Scoliosis and Kyphosis Congenital lesions of the thoracic spine include congenital scoliosis and congenital kyphosis (Fig. 9-15). In the former, abnormal curvature in the coronal plane develops secondary to anomalous vertebrae present at birth. Individuals with these hemivertebrae anomalies do not present until childhood or early adulthood. Clinical presentation of these lesions encompasses a wide spectrum. Many present as an
Figure 9-15. Thoracic spine scoliosis.
Figure 9-16. Scheuermann’s disease.
asymptomatic finding during a routine physical examination; however, some patients present with pain, neurologic deficits, and occasionally rapidly progressive scoliosis resulting in severe morbidity. In general, 25% of congenital scoliosis patients do not progress, 50% progress slowly, and 25% progress rapidly.83 Congenital kyphosis is an uncommon sagittal plane deformity that is caused by formation segmentation failure. Both congenital scoliosis and kyphosis may require combined anterior and posterior instrumentation procedures. Scheuermann first described progressive dorsal kyphosis of adolescent children in 1920.84 The deformity is a fixed thoracic kyphosis that does not correct with hyperextension, thus differentiating it from a postural kyphosis. Typically, Scheuermann’s disease involves the midthoracic spine, most commonly the T7 and T8 vertebrae.85 A mild scoliosis is present in 20% to 30% of patients.86 The characteristic features of this disease of ventral wedging of 5 degrees or more of at least three adjacent vertebrae were described by Sorenson.87 Other characteristics include kyphosis of greater than 40 degrees, vertebral body end plate irregularity, and disk space narrowing (Fig. 9-16).88 The prevalence of the disease ranges from 0.4% to 8% and occurs predominantly in males.87 Initial treatment consists of rigid bracing, exercise, and regular clinical and radiographic examination. Surgical treatment of patients with Scheuermann’s disease is contro-
W9418-09.qxd 11/3/03 2:53 PM Page 250
250
Section I
Introduction
versial and reserved for patients exhibiting advanced kyphosis, failure of conservative bracing, or neurologic deficits.89 Degenerative or Herniated Nucleus Pulposus Thoracic disk herniation (Fig. 9-17) accounts for 0.25% to 0.57% of all symptomatic disk herniations of the spine.90 The symptoms from herniated disks usually present as localized back pain; however, the pain may be represented as a bandlike radiculopathy in the region of the particular dermatome level. Motor deficits, sensory deficits, and bowel or bladder dysfunction may also occur. The surgical approaches to thoracic disks vary. Traditionally, dorsal decompression via laminectomy was performed with poor results as reviewed by Logue in 1952.91 Many of the patients undergoing laminectomy were paraplegic after the surgery. This complication was a stimulus for safer surgical approaches to the thoracic spine. Transpedicular, lateral extracavitary, lateral parascapular extrapleural, and costotransversectomy approaches have reduced the complication rate of paraplegia. Furthermore, ventral approaches including transthoracic thoracotomy, transthoracic thorascopy, and retropleural thoracotomy approaches have allowed the neurosurgeon access to the ventral thoracic spine to address ventral pathology more safely. Pathologic Processes of the Thoracic Spine Although disk herniation is quite rare, pathologic processes involving the thoracic spine are common. These processes include traumatic vertebral fractures, spinal metastases, bacterial and tuberculous infections, primary bone tumors, meningeal tumors, vascular malformations, primary bone disease, and connective tissue or skeletal disorders. These disorders commonly cause a compressive myelopathy or radiculopathy. Many of these disorders require decompres-
sion of the neural elements with spinal reconstructive stabilization for treatment of the symptoms. Trauma Thoracic fractures and thoracolumbar fractures are usually associated with high-speed motor vehicle accidents or traumatic falls. The thoracolumbar region (T10-L5) has greater mobility than that of the thoracic spine (T1-T9) and is, therefore, more commonly associated with traumatic injury. Thoracic fractures may be divided into three categories: (1) compression or wedge fractures, (2) burst fractures, and (3) fracture-dislocation (Fig. 9-18).92 Wedge fractures result from severe flexion and are generally stable. If contiguous vertebral levels are involved, then severe kyphosis can occur and stabilization may be necessary. Burst fractures result from axial compression with varying degrees of flexion. The result is compression and failure of both anterior and posterior cortices of the vertebral body. These fractures may be associated with spinal canal compromise secondary to retropulsion of the bony elements. Fracture-dislocation injuries are the most severe of the fracture types and usually occur as a result of lap belt injuries. In this fracture type, fractures occur and then distraction occurs resulting in severe ligamentous injury. This type of fracture requires surgical stabilization. The three-column model as described by Denis (Fig. 9-19) may aid in determining the instability of fractures.93 In this widely acknowledged system, the vertebral body and the posterior elements are classified into columns: anterior, middle, and posterior. The anterior column consists of the anterior longitudinal ligament and the anterior half of the vertebral body and disk. The middle column consists of the posterior half of the vertebral body and disk as well as the posterior longitudinal ligament. The posterior column consists of the posterior arch with facets, supraspinous and interspinous ligaments as well as the ligamentum flavum. In the Denis model, if two or more columns are disrupted instability exists. Indications for surgical treatment of thoracolumbar fractures include progressive neurologic deficit, spinal cord compression, dural laceration, and instability. Instability may include fracture-dislocation injuries; anteroposterior or lateral translocation; and severe wedge or burst fractures with canal compromise, and progressive angulation as assessed on serial radiologic examinations. Spinal Metastases
Figure 9-17. Thoracic disk herniation.
Cancer is the second leading cause of death in the United States with approximately 1.3 million new cases per year.94,95 The spinal column, especially the thoracic region, is the most frequent site of bony metastasis.96 Most patients with metastases report a history of back pain; however, epidural compression has been reported to be the presenting symptom in approximately 8% to 10% of patients with metastatic
W9418-09.qxd 11/3/03 2:53 PM Page 251
Chapter 9
A
Spinal Neurosurgery
251
B
C
D
Figure 9-18. Traumatic fracture of the spine. A, Compression/wedge; B, Burst; C, Fracture/dislocation; D, Chance.
disease.97 At diagnosis, more than 50% will have a paraparesis or bowel/bladder disturbance.98 Surgical intervention for metastatic disease to the spine remains controversial. Experience with different surgical approaches has allowed for easier decompression and stabilization. Indications for surgical intervention include biopsy for histopathologic analysis, decompression for progressing neurologic deficit, and stabilization. It is important to assess the prognosis of the systemic disease before undertaking a long and laborious operation. Generally, a life expectancy of less than 6 months may help guide intervention. Furthermore, many of the metastatic tumors are radiation-sensitive and chemosensitive and these treatment options should be considered before any surgical intervention. Prognostic
factors and individual differences influence outcome, although, in general, extended survival may be commonly observed in patients with breast, prostate gland, thyroid gland, or renal carcinomas. Patients with other types of tumors, such as adenocarcinoma of the lung, are associated with relatively short survival times. Primary Tumors of the Spine Multiple myeloma (Fig. 9-20) accounts for 45% of all malignant bone tumors.99 It is primarily a disease of the sixth and seventh decades with a predilection for the thoracic spine, followed by the lumbar and rarely the cervical spine.100 The 5-year survival rate is 18%.101 On occasion, isolated plasma-
W9418-09.qxd 11/3/03 2:53 PM Page 252
252
Section I
Introduction
ISL LF C
Figure 9-19. Three-column model, as per Denis.
cytomas may occur and have a better prognosis when compared to multiple myeloma. Other malignant bone tumors include chondrosarcomas, chordomas, and lymphomas. Chondrosarcomas affects the spine in only 6% of cases.102 Chondrosarcomas frequently occur between the second and third decades of life. Histologically, the appearance of these tumors lies between a benign chondroma and a malignant sarcoma thus making them difficult to diagnose. Chordomas are thought to originate from the notochord remnant and are locally invasive tumors most commonly found in the sacrum or clivus. Age of onset can vary, but is most common between the third and fifth decades of life. Treatment of chordomas usually consists of surgical resection; however, these tumors have a high recurrence rate.103 Chordomas do not respond well to traditional radiation therapy, but recent studies suggest that these lesions may respond to stereotactic radiation surgery and proton beam therapy. Lymphomas involve the spine in approximately 10% of extranodal lymphoma and usually respond well to steroids and radiation therapy.94 Benign tumors of the spinal column include osteochondromas (the most common benign bone tumor), giant cell tumors, osteomas, osteoblastomas, aneurysmal bone cysts, hemangiomas, and eosinophilic granulomas. Often, these tumors require surgical intervention with removal, decompression of the neural elements, and spinal reconstruction as the mainstay of therapy. Spinal Cord Tumors Spinal cord tumors may present with varied symptoms. Back pain is the most common initial complaint, which is usually diffuse and unrelated to mechanical activity. Patients may also present with myelopathic findings, motor deficits, paresthesias, and even bowel or bladder problems. Radiologic imaging with MRI and computed tomography (CT) may
suggest the location of the tumor; however, CT myelography often will demonstrate whether the lesion is intradural or extradural. Tumors may be intramedullary, intradural extramedullary, or extradural. Intramedullary spine neoplasms represent 2% to 4% of all central nervous system tumors.104 The intramedullary tumors include astrocytomas, ependymomas, and hemangioblastomas. Often the spinal cord is enlarged with edema present on T2-weighted MR images. Cysts or syrinxes may be also be observed. Astrocytomas (Fig. 9-21) of the spinal cord portend a poor prognosis; however, compared to the brain astrocytomas, these lesions usually are of lower grade. Ependymomas of the spinal cord are the second most common type of intramedullary tumor and are often seen in the conus. In this region, a variant known as a myxopapillary ependymoma is the most common and is seen in the adolescent/young adult population. Surgical removal of these lesions provides a more favorable prognosis. Hemangioblastomas are the third most common intramedullary tumor and are often associated with von Hippel-Lindau disease. Complete resection of these tumors is often curative; however, if remnants of these tumors are inadvertently missed, recurrence is likely. Intradural extramedullary tumors may be best imaged with CT myelography. With this technique, a meniscus may be seen with a paucity of dye in the region of the mass. Schwannomas and meningiomas (Fig. 9-22) are the most common types of intradural extramedullary lesions. Neurofibromas may also be found in the intradural extramedullary space and are more often associated with the phacomatoses (neurofibromatosis type 1). Many patients present with pain secondary to nerve root irritation, but some may present with myelopathy from spinal cord compression. Treatment consists of exploration and excision of tumor as well as radiation therapy for lesions that recur. Extradural lesions are usually secondary to metastases. However, meningiomas, schwannomas, and neurofibromas may consist of only extradural involvement. These lesions are usually readily resectable. Lipomatous masses causing spinal cord compression may also occur in patients with Cushing’s disease. Infection Infections of the spine include intramedullary abscesses, subdural and epidural abscesses, and osteomyelitis. Intramedullary spinal cord abscesses are uncommon. The pathogenesis of this type of abscess is thought to be secondary to infection in an area of infarction or septic emobolization. Treatment usually includes surgical exploration and drainage combined with long-term antibiotic therapy. Spinal epidural abscesses usually present with severe neck or back pain and progressive neurologic deficit that may occur rapidly. Approximately one half of epidural abscesses result from hematogenous spread to the epidural space. This type of abscess is more common in intravenous drug abusers. The
W9418-09.qxd 11/3/03 2:53 PM Page 253
Chapter 9
A
Spinal Neurosurgery
253
B
Figure 9-20. Multiple myeloma. A, Sagittal MRI; B, Sagittal CT; C, Axial CT.
C
treatment usually consists of surgical debridement and evacuation followed with antibiotic therapy. The neurologic sequelae may be secondary to direct compression from the abscess or may be due to infarction and necrosis of the spinal cord.105 Spinal subdural abscesses are extremely rare, but are treated in a similar fashion with evacuation and antibiotic therapy as the mainstay of treatment. The thoracic and lumbar regions are common sites of osteomyelitis of the spine. More than half of the patients presenting with vertebral osteomyelitis are older than 50 years and the time from initial presentation to diagnosis ranges from 2 weeks to 5 months with a mean of 6 to 8 weeks.106,107 Patients may present with pain and muscle spasm, paresthesias, or neurologic deficits.
Evidence of pyogenic osteomyelitis of the spine has been found in Egyptian mummies and was described by Hippocrates. Although physicians knew of this problem for many years, antibiotic therapy finally facilitated adequate treatment of this disease. The most frequent route of osteomyelitis is via hematogenous spread. The most common infections of the spinal column include pyogenic organisms (Staphylococcus aureus and coliform bacilli), infections caused by fungi (Actinomycetes and Blastomycetes), and Pott’s disease (Mycobacterium tuberculosis).108 In the adult, the vertebral body is a site favorable for seeding due to its rich vascular supply. The infection involves the body, but commonly does not respect the vertebral body/disk interface. This is an important clue when differ-
W9418-09.qxd 11/3/03 2:53 PM Page 254
254
Section I
Introduction
Figure 9-21. Spinal cord astrocytoma.
A
entiating infection from tumor when reviewing radiographs and CT/MRI examinations (Fig. 9-23). These infections usually present with subacute back pain and commonly intense paraspinal muscle spasms. Treatment may consist of biopsy with antibiotic therapy. Often, thoracolumbosacral orthoses are used to decrease the rate of kyphotic and scoliotic deformities. If severe bony destruction is present, spinal stabilization above and below the involved vertebral levels may need to be performed. Pott’s disease, a tuberculous infection of the spinal column, is a common problem in economically underdeveloped countries; however, epidemiology suggests resurgence in more wealthy countries. Extrathoracic involvement increased from 8% to 18% with an estimated 20,000 cases per year in the United States with the thoracolumbar spine as the most commonly affected site (Fig. 9-24).109 Patients may present with radiculopathy, cauda equina, or chronic back pain and muscle spasm secondary to paraspinal muscle involvement. Purified protein derivative testing should be performed on suspected cases; however, some patients may exhibit anergy secondary to the infectious process. Open surgical exploration and evacuation of the granulomatous tissue is often required and allows for specimens to be obtained and sent for microbiologic analysis. Dorsal decompression is usually sufficient, but many cases may require instrumentation and stabilization if the bone has been destroyed.
B
Vascular Malformations Vascular malformations of the spine are rare entities but have been classified into four types. The most common type of spinal arteriovenous malformation (AVM) is the dural, or type I, AVM. Most patients with type I dural AVMs are between 40 and 70 years of age; 80% are male and no familial tendency has been identified.110,111 The most common symptom associated with dural AVMs is pain, which may be
Figure 9-22. Nerve sheath tumor. A, Schwannoma; B, Meningioma.
local, radicular, or nonspecific. Most patients will also complain of lower extremity weakness and sensory changes by the time of diagnosis. Because of the gradual clinical course, many patients go undiagnosed until the appropriate studies are performed, namely MRI and selective angiography (Fig. 9-25). These lesions may be treated surgically; however, with advancing technology in endovascular procedures, options other than surgery may exist. Type II, or glomus, spinal AVMs are intramedullary AVMs with a true compact nidus.112 Type III spinal AVMs, also known as juvenile AVMs, are less common. These lesions are usually more extensive and involve both intramedullary and extramedullary spaces over more than one spinal level.112 The final type is type IV and consists of intradural extramedullary AV fistulas as classified by Heros.113 Type IV lesions are fed from the anterior spinal artery or, less commonly, from the posterior spinal
W9418-09.qxd 11/3/03 2:53 PM Page 255
Chapter 9
A
Spinal Neurosurgery
255
B Figure 9-23. Osteomyelitis/diskitis. A, MRI; B, CT sagittal reconstruction.
artery. These lesions lie outside of the spinal cord and its pia mater and vary in size and flow.114
The Lumbosacral Spine Congenital Disorders and Spondylolisthesis Excluding spinal dysraphism (myelomeningoceles), congenital lesions of the lumbar spine are rare. Congenital
Figure 9-24. MRI of Pott’s disease.
spinal stenosis occurs in a very small number of patients who present with spinal stenosis. Congenital or dysplastic spondylolisthesis is a more common entity. Spondylolisthesis refers to the slippage of all or part of one vertebra onto another. Wiltse and colleagues have proposed the most widely accepted classification of spondylolisthesis.115 They divided spondylolisthesis into five types: type I, dysplastic or congenital; type II, isthmic; type III, degenerative; type IV, traumatic; and type V, pathologic. Congenital spondylolisthesis accounts for 14% to 21% of the cases of spondylolisthesis with a 2 : 1 female-to-male ratio.116,117 This type of spondylolisthesis is characterized by structural anomalies of the lumbosacral junction including dysplasia of the lamina and facet joints. The defects allow for slippage to occur, compromising the neural foramina. Individuals may present with hamstring spasm, back or leg pain, or neurologic deficits. Spondylolisthesis may be graded regardless of the etiology. In this system, grade I (Fig. 9-26) refers to subluxation of the superior vertebral body on the inferior vertebral body of less than 25% its anteroposterior diameter. Grade II refers to slippage of 25% to 50%. Grade III refers to slippage of 50% to 75%. Grade IV refers to slippage greater than 75%. Grade V, or spondyloptosis, refers to complete subluxation of a vertebral body on another. Grade I spondylolisthesis may be followed with serial radiographs; however, many patients will have significant pain unrelieved by
W9418-09.qxd 11/3/03 2:53 PM Page 256
256
Section I
Introduction
A
B Figure 9-25. Spinal cord AVM. A, MRI; B, Angiogram.
conservative therapy. Many patients with higher grade spondylolisthesis will require decompressive surgery with fusion. Degenerative changes in the disk and vertebral column occur over time and may be accelerated due to increased loading forces, vascular compromise, and anatomic changes. The result of these ongoing processes leads to disk herniation, spondylosis, and compression of the neural elements, which ultimately result in radiculopathy. The intervertebral disk is comprised of an annulus fibrosus, the nucleus pulposus, and the cartilaginous endplates (Fig. 9-27). The nucleus pulposus is a remnant of the notochord and is contained within the tough outer annulus. Radial tears in the annulus will allow the nucleus pulposus to extrude into the spinal canal or into the neural foramen, leading to radicular symptoms. The cartilaginous endplates are also affected by the degenerative changes and may allow for herniation into adjacent vertebral bodies. The disk is composed of proteoglycans and has high water content. During the aging process, the disks lose their water content and may appear as blackened disks on MRI. With the loss of water and the changes that occur within the disks, the elastic properties of the disk change. The changes in the disks can also lead to changes in the facet joints in the form of hypertrophy. The combination of disk disease and
facet hypertrophy may also contribute to instability of the spine. Epidemiology of Low Back Pain Low back pain (LBP) is extremely prevalent; it is the second most common reason for people to seek medical attention.118 LBP accounts for approximately 15% of all sick leaves, and is the most common cause of disability for persons younger than 45 years of age.119 The estimated lifetime prevalence is 60% to 90% with an annual incidence of approximately 5%.120 Because of its prevalence and demand on health care, back pain is one of the most expensive medical problems. The natural history of back pain is very favorable; most episodes resolve in 10 to 30 days.121 Of the patients seeking medical attention, only 1% of patients will have nerve root symptoms and only 1% to 3% will have lumbar disk herniations. More than 50% of patients who have an acute episode of low back pain will have another episode within 1 year.121 Occupational factors associated with increased risk of low back pain include heavy physical work, frequent bending and twisting, lifting, pushing, pulling, repetitive work, static work postures, vibrations, and psychologic and psychosocial
W9418-09.qxd 11/21/03 2:49 PM Page 257
Chapter 9
Spinal Neurosurgery
257
factors.123 Individual factors associated with low back pain include age and gender, with men having a higher risk. The highest prevalence rate for back pain occurs in the 35- to 55year age range for men, whereas women have a somewhat higher prevalence rate later in life. Surgery is seldom necessary for low back pain. The lifetime incidence of surgery ranges from 1% to 3%.121 Back pain is initially treated with nonsurgical conservative therapy. This therapy usually consists of analgesics, muscle relaxants, physical therapy, education, and occasionally epidural injections. A course of this type of therapy should be continued for approximately 4 to 6 weeks. Most individuals (85%) will demonstrate an improvement in pain relief and do not require surgical intervention. However, some individuals may experience progression and require surgery. Indications for surgery in the lumbar region include failure of conservative management, cauda equina syndrome, progressive motor deficits, and intolerable pain. Imaging of the lumbosacral spine includes plain radiographs of the lumbosacral spine, CT, or MRI. The correlation between images and clinical findings improves outcome. Because there are a number of pathologic processes that could contribute to low back pain, the surgery performed should address the demonstrated pathology.
Figure 9-27. Intervertebral disk anatomy.
Lumbar Herniated Disks
Figure 9-26. Grade I spondylolisthesis.
A herniated disk in the lumbosacral spine (Fig. 9-28) is a very common finding among healthy individuals. When the disks protrude and cause neural compression or irritation, radiculopathies occur. Patients harboring a herniated nucleus pulposus may complain of back pain, radicular pain, weakness, paresthesias, and occasionally bowel and bladder problems. These symptoms are typically managed with conservative therapy; however, some patients may require surgical intervention in the form of a diskectomy (see Fig. 9-28). The L5-S1 disk is the most common region of disk herniation in the lumbar spine followed by L4-L5 (40%) and then L3-L4 (3% to 10%). An L5-S1 disk herniation with compression of the S1 nerve root usually causes pain in the buttock with radiation to the posterior leg and often to the ankle. Weakness may be observed on plantar flexion, which
W9418-09.qxd 11/3/03 2:53 PM Page 258
258
Section I
Introduction
B
A
Figure 9-28. Lumbar herniated nucleus pulposus. A, Sagittal MRI; B, Axial MRI; C, Free fragment.
C
tests the strength of the gastrocnemius muscle. Furthermore, an ankle jerk reflex may be diminished or absent and decreased sensation may be assessed over the lateral malleolus and lateral foot. The L4-L5 disk usually compresses the L5 nerve root and may cause pain in the posterolateral leg with decreased sensation in the web space of the large toe and dorsum of the foot. Weakness may be assessed in the extensor hallucis longus (EHL) and tibialis anterior. A foot drop may be the presenting sign. There is no reliable reflex to assess an L5 nerve root compression. The L3-L4 disk usually compresses the L4 nerve root and may manifest as pain in the anterior thigh with radiation to the medial malleolus and medial foot. Decreased sensation may also be observed in this particular distribution. The quadriceps muscle is commonly affected causing weakness in knee extension and a decreased patellar reflex.
Lumbar laminectomy with microdiskectomy is one of the most common procedures performed by the spinal surgeon. The overall risk of mortality in large series is 0.06%.122 During this procedure, partial removal of the lamina is necessary to obtain exposure to the disk space and neural foramina. Removal of the ligamentum flavum and medial retraction of the thecal sac is necessary to expose the nerve root and the protruding disk. The disk space is then entered and its contents removed. During this procedure, numerous complications can occur. Dural tears may occur during the removal of the bony elements or when removing the disk. The risk of a CSF fistula requiring operative repair is approximately 10 in 10,000.122 Nerve root injury may occur in the form of retraction injury as well as direct trauma from the instruments used during the procedure. Worsened motor deficits are seen in approximately 1% to 8% of cases.123 When entering the disk space,
W9418-09.qxd 11/3/03 2:53 PM Page 259
Chapter 9
special attention to the depth of the instruments is important because deep penetration past the anterior vertebral body can lead to retroperitoneal injury. Many surgeons mark the instruments at 3 cm distal to the tip to ensure that the instrument does not penetrate through the anterior longitudinal ligament. Hemorrhage during and after diskectomy is very rare; however; if significant hypotension and hypovolemia occur intraoperatively, a vascular injury should be suspected. The mortality of a vascular injury during a lumbar diskectomy can be as high as 50%.124 Lumbar Stenosis Lumbar stenosis refers to a degenerative disease with associated osteophyte formation, hypertrophied ligamentum flavum, and hypertrophied facet joints. In lumbar stenosis, narrowing of the anteroposterior diameter of the spinal canal occurs. This narrowing causes compromise of the neural elements and/or blood supply to the spinal cord and nerve roots. Lumbar stenosis is found on plain radiographs in 95% of men and 80% of women older than 65.125 It is more common in the lumbar region where the majority of the load share is transmitted (Fig. 9-29). Patients with lumbar stenosis often present with back pain and radicular complaints. A syndrome known as neurogenic claudication is common. In this syndrome, unilateral or bilateral buttock, hip, thigh, or leg pain that is precipitated by standing or walking and characteristically relieved by a change in posture to sitting, squatting, or recumbency is observed. Once again, this syndrome is thought to arise from ischemia of the lumbosacral roots secondary to an increased metabolic demand during exercise. Surgical treatment for lumbar stenosis includes dorsal decompression of the neural elements via laminectomies
A
Spinal Neurosurgery
259
and foraminotomies. Complications of decompressive lumbar spinal surgery include dural tears, neurologic injuries, and worsening pain syndromes. Cauda Equina Syndrome Cauda equina syndrome (CES) is a complex of symptoms and signs including low back pain, sciatica, motor weakness, sensory changes, and bowel or bladder incontinence. The nerve root compression may be due to trauma, metastatic tumors of the spine, spinal infections, and severe spinal stenosis. However, acute CES most commonly presents secondary to intervertebral disk prolapse. The incidence of CES has been estimated to range from 1.2% to 6%.126–128 CES is an indication for urgent surgical intervention. The onset of bladder paralysis is an important indicator for urgent surgery. Although the prognosis for CES is good, there is a significant difference in the outcome of cases operated on within 24 hours of bladder paralysis compared to those operated on after this period.129 Thus, if CES occurs postoperatively, it is usually reversible if recognized early. These early postoperative causes may be secondary to hematoma or improper graft placement. A delayed cause may be secondary to abscess. Fractures of the Lumbosacral Spine Fractures of the lower lumbar spine and lumbosacral junction are encountered infrequently.130 These fractures usually result from severe flexion and compression and occasionally rotation. In these injuries, CES may result. These injuries are usually treated with decompression of the neural elements and stabilization with transpedicular screw systems or rod systems combined with posterolateral bone grafting.
B Figure 9-29. Lumbar stenosis. A, MRI sagittal; B, MRI axial.
W9418-09.qxd 11/3/03 2:53 PM Page 260
260
Section I
Introduction
Spinal Fusion Procedures Posterior Lumbar Interbody Fusion Procedures Ralph Cloward, a neurosurgeon living in Hawaii during World War II, pioneered and popularized posterior lumbar interbody fusions (PLIF). In this procedure, a subtotal diskectomy is performed along with decompression of the neural elements (Fig. 9-30). The interbody space is then fused with autologous bone. The indications for a PLIF procedure include persistent low back pain, recurrent disk disease, spondylolisthesis, and symptomatic spinal stenosis with or without degenerative scoliosis or spondylolisthesis.131 Patient selection remains the most critical factor for surgical success.132 Results of the PLIF procedure are usually determined by clinical improvement and by fusion rate. Cloward reported a clinical success rate of 87% to 92% and a fusion success rate of 92% in his 40-year experience performing the operation.133 Others including Gill and Blumenthal, Stefee and Brantigan, and Ray have subsequently supported these findings.134–136 The procedure is technically demanding and tragic mishaps may occur in the hands of inexperienced surgeons. The surgical complications observed in the dorsal decompression of the lumbar spine are inherent in this procedure; however, PLIF-related complications may also be observed.
A
In the immediate postoperative period, complications secondary to technical error may occur. New or increased neurologic deficits occur postoperatively in 0.5% to 4.0% of the patients after a PLIF.131,134,137–140 These injuries may be a result of excessive traction placed on the nerve roots while performing the procedure. Retraction of the sacral nerve roots may also lead to postoperative urinary retention.133,134 Increased bleeding may occur secondary to bleeding from epidural veins or the vertebral bodies. Infection rates after PLIF range from 0.2% to 7%.131,134,138,141,142 The infections may be superficial in nature or may lead to osteomyelitis. Delayed complications, usually 3 to 6 months postoperatively, occur secondary to instability and strain problems. Delayed complications include disease at adjacent levels and pseudoarthrosis. A motion segment adjacent to the fused spine may undergo accelerated degeneration and lead to new-onset symptoms. Spondylosis and spondylolisthesis has also been reported after PLIF procedures.143 These complications may present with new-onset pain or neurologic deficit. Finally, pseudoarthrosis after a PLIF procedure may lead to failed back syndrome. Anterior Lumbar Interbody Fusion Procedures Anterior lumbar interbody fusions (ALIFs) (see Fig. 9-30) reconstruct the anterior column of the spine and improve
B
Figure 9-30. Posterior lumbar interbody fusion and anterior lumbar interbody fusion. A, Lateral radiograph; B, Anteroposterior radiograph.
W9418-09.qxd 11/3/03 2:53 PM Page 261
Chapter 9
sagittal plane alignment. Operative indications for this procedure include degenerative disk disease with chronic back pain as well as degenerative spondylolistheses. A failure of a previous dorsal lumbar surgery may be an indication for ALIF. Chronic low back pain with failure of conservative therapy may also be an indication for ALIF. In this procedure, a transabdominal or retroperitoneal approach is performed to allow access to the ventral lumbar spine. With these approaches, the neural elements are avoided. However, risk of injury to the peritoneal structures may occur. Injuries to the bowel may occur and should be detected and treated with direct repair. Injuries to the major blood vessels are rare; however, injuries to the iliac arteries or veins, the inferior vena cava, and the aorta can occur. These injuries should be addressed quickly and may require the experience of a general or vascular surgeon. Postoperative hernias may also be present if meticulous attention to fascial closure is not adhered. A serious complication of ALIF procedures in male patients is retrograde ejaculation. This occurs when the autonomic nerves are injured, usually at the L5-S1 level. This problem occurs in 0.5% to 2.0% of all ALIF procedures performed on males.144–147 Urinary retention may also result from ALIF surgery, but it is usually temporary. Endoscopic Procedures With advancing technology, the endoscope has allowed many standard spinal surgical procedures to be performed with minimal invasiveness (Fig. 9-31). This burgeoning field is advantageous for numerous reasons. First, with the use of dilators, tissue damage while approaching the spine can be
Figure 9-31. Minimally invasive surgery diagram.
Spinal Neurosurgery
261
reduced. This technique can result in less postoperative pain. Second, these procedures may be performed on an outpatient surgery format with decreased hospitalization stay. Also, blood loss is usually less when compared to that with open surgery. Although minimally invasive surgery may be a useful technique in the spinal surgeon’s armamentarium, there is a steep learning curve. In inexperienced hands, complications can arise because surgical procedures are being performed with a two-dimensional image. Neural element damage, CSF leaks, and hemorrhage can occur and may require an open surgical procedure to address the complications. Medical Complications of Spinal Surgery In spinal surgery, medical complications in the postoperative period can lead to significant morbidity and mortality. Many of the complex spinal cases are routinely observed in the intensive care unit to limit these complications. Moreover, preventative measures may also be taken to lessen the amount of morbidity in these patients. Thromboembolic disease is a serious complication in the spinal surgery patient. Rates of acute deep venous thrombosis (DVT) may be as high as 14%.148 The incidence is significantly higher in spinal cord–injured patients. DVTs may lead to pulmonary embolism, which has been reported in up to 8% of spinal surgery patients.149 Recommendations for DVT prophylaxis vary; however, gradient pressure stockings, intermittent pneumatic compression devices, mini-dose heparin, low-molecular-weight heparin, and low-dose warfarin may be used. Another alternative to prevent pulmonary embolism is placement of an inferior vena cava filter.
W9418-09.qxd 11/3/03 2:53 PM Page 262
262
Section I
Introduction
The cardiovascular system may be adversely affected by spinal cord injury. With cervical and high thoracic injuries, the sympathetic outflow may be completely or partially compromised. Initially, the patient may experience what has been labeled as spinal cord shock. A block in the sympathetic outflow with an unopposed parasympathetic stimulus below the level of the injury leads to spinal cord shock. The effect leads to vascular pooling of blood, hypotension, and bradycardia. It is important to recognize spinal shock because overaggressive fluid resuscitation may cause the patient to become fluid-overloaded and pulmonary edema may ensue. Therefore, evaluation of central venous pressure, blood pressure, and heart rate are extremely important for proper management. Intravenous crystalloid fluids as well as colloid fluids may be used initially for resuscitation; however, dopamine and dobutamine may be required to help maintain adequate blood pressure and heart rate. Another effect of spinal cord injury on the cardiovascular system includes dysrhythmias that may occur in the absence of prior cardiac disease. A careful cardiac history should be ascertained and an echocardiogram should be obtained on admission. A third well-described phenomenon observed in spinal cord–injured patients is autonomic dysreflexia. In patients with injuries above T6, uncontrolled sympathetic reflex response to noxious stimuli may cause intense headache, flushing, excessive diaphoresis, and episodes of extreme hypertension. The stimuli may include distention of the bowel or bladder and aggressive movement of the patient. The treatment for autonomic dysreflexia involves removal of the inciting stimuli. The hypertension associated with this phenomenon may require acute treatment with nitroprusside and/or hydralazine. Patients with spinal cord injuries often have serious, lifethreatening pulmonary complications. The major causes of death in the acute phase in spinal cord–injured patients are pulmonary failure and shock.150 Approximately 35% of spinal cord–injured patients will have a major pulmonary complication.151 The diaphragm is innervated by the ventral roots of C3, C4, and C5 and therefore, patients with spinal cord lesions of C4 or higher will need immediate intubation and mechanical ventilation. Patients with lower cervical/high thoracic cord injuries may initially have normal respiration, but may progress to require mechanical ventilation. In patients with impaired external intercostal muscle function as seen with thoracic cord injuries, the activity of the diaphragm combined with the inactivity of the external intercostal muscles causes a decrease in the anteroposterior diameter of the chest. This type of breathing has been referred to as paradoxical breathing. If this type of breathing is observed in a patient, close monitoring is critical because these patients often succumb to respiratory collapse. The position of the patient is also an important concept concerning the respiratory status of a spinal cord–injured patient. The upright position allows for gravity to affect the
abdominal viscera, which allows for better respiratory performance. The upright position may also aid in pulmonary toileting and thus decrease the incidence of respiratory complications and infections. Gastrointestinal complications may also be prevalent in the spinal cord–injured patient. Stress ulcerations may result from a complicated surgery, the use of high-dose steroids, or a concomitant brain injury. Prophylaxis with H-2 blockers and gastric mucosal protective agents should be used. Adynamic ileus is a well-known complication of spinal surgery and also occurs in patients with acute spinal cord injury. An ileus may lead to increased abdominal distention with compromise of diaphragmatic excursion. This may further complicate pulmonary function. The treatment of an ileus should include decreasing oral intake, and the use of laxatives, bowel stimulants, and enemas. Occasionally, a nasogastric tube may be fed into the small bowel to help decompress the distended bowel. Neostigmine, at a dose of 0.5 to 2 mg, has been used successfully for treatment of this complication. Olgivie’s syndrome, or pseudo-obstruction of the colon, has been reported in patients after lumbar spinal surgery.152 This syndrome is characterized by abdominal distention with an enlarged cecum (>9 cm). Nausea, vomiting, constipation, and diarrhea may be present in patients suffering from this syndrome. Initial treatment consists of nasogastric tubes, insertion of rectal tubes, and decreasing oral intake. If patients do not respond to initial therapeutic intervention, then laparotomy and placement of a cecostomy tube may be required. The mortality rate for patients treated conservatively with colonoscopy is 13%; however, in patients undergoing laparotomy with cecostomy tube placement, the mortality rate is approximately 30%.153 Genitourinary complications after spinal cord injury may lead to significant morbidity. Initially, spinal cord–injured patients may exhibit detrusor muscle inactivity, decreased bladder sensation, and compromised sphincter activity. Uninhibited reflex activity of the detrusor and sphincter gradually returns and during this time, the patient may exhibit varying dysfunction. Distention of the bladder or bowel can lead to autonomic dysreflexia in patients with spinal cord injuries above T6 as previously described. Therefore, it is important that a catheter regimen be enforced in these patients to limit morbidity. Frequent bladder catheterizations as well as aggressive bowel programs should be implemented. Patients with injuries to the lower thoracic spinal cord usually have impairment of bladder sensation, detrusor hyperreflexia, and sphincteric dyssynergia. These impairments lead to incomplete bladder emptying and elevated bladder pressures. Renal damage may ensue due to hydroureteronephrosis. In-dwelling catheters may be used for genitourinary problems associated with spinal cord injury; however, these catheters are subject to infection. Therefore, scheduled bladder catheterizations may be associated with less morbidity when compared to indwelling catheters.
W9418-09.qxd 11/3/03 2:53 PM Page 263
Chapter 9
Summary With the advancing age of our society, the number of patients seeking surgical intervention for spinal disease will increase. This increased demand will place added pressure on hospitals, doctors, nurses, and ancillary medical services. Therefore, a sound understanding of the epidemiology, natural history, and various treatment modalities will be necessary to facilitate this shift in workforce. A multidisciplinary approach consisting of primary care physicians, neurologists, neurosurgeons, internists, anesthe-
Spinal Neurosurgery
263
siologists, and rehabilitation specialists may be required to provide comprehensive care for the patient with a spinal disorder. Perhaps the most challenging problem we will be faced with in the new millennium is not how to maintain the quality of care for the patient with a spinal disorder, but how to advance our understanding of the disease processes and develop novel treatment modalities. The exciting and rapidly shifting disciplines of molecular biology, biomedical engineering, and minimally invasive surgery may offer some answers to these perplexing problems.
P earls 1. National studies have found a 1 in 300 incidence of severe neck injury in vehicular accidents severe enough to have the vehicle towed from the scene. That incidence increases to 1 in 14 for cases in which the patient has been ejected from the car. 2. When assessing the integrity of the transverse ligament, an open mouth or anteroposterior radiograph may be used. If the sum total of the overhang of both C1 lateral masses on C2 is greater than or equal to 7 mm, then the transverse ligament is probably disrupted. This measurement is commonly known as the “Rule of Spence.” 3. The complications of anterior odontoid screw fixation are similar to other anterior cervical approaches. The main difference pertains to the actual screw placement. It is imperative that biplanar fluoroscopy guide the experienced surgeon in the screw trajectory. Malposition of the screw can result in dural penetration with CSF leak, catastrophic neurologic deficits, or vertebral artery injury. 4. Trauma also affects the subaxial (C3-C7) spine. Injuries to this region of the spine most commonly occur as a result of flexion-rotation or hyperflexion injuries. Ligamentous injuries may occur that allow for subluxation of vertebrae. 5. Degenerative processes of the cervical spine most frequently present as spondylosis, osteophytic compression, and disk herniation.35 Cervical radiculopathy is a common occurrence. Individuals who lift heavy objects as well as those who smoke are predisposed to acute cervical disk disease. 6. The incidence of cervical disk herniation peaks in the fourth decade. This peak incidence in disk herniation appears to be the result of the combination of the maximum expansile strength of the disk (i.e., when the disk is the most hydrated) and the peak incidence of annular tears. Thus, after the age of 50, the dehydrated disks are less prone to herniate even when the incidence of annular tears increases.
7. Although disk herniation may cause significant pain, the etiology of the pain is rather elusive. 8. The most common sites of herniation in the cervical spine are at C5-C6 and C6-C7, causing compression of the C6 and C7 root, respectively. Involvement of the C5 nerve root is perhaps the most disabling. 9. Cervical spondylotic changes are often the cause for cervical myelopathy. Classic signs in CSM include Lhermitte’s sign (electrical shock sensation associated with neck flexion), hand clumsiness, distal weakness, generalized hyperreflexia, Hoffman’s sign (contraction of the thumb and index finger upon flickering of middle finger), and spastic gait.46,48 10. Corpectomies tend to be lengthy procedures that, even in the best of hands, usually take a few hours. Injuries to the vascular structures may be caused by the sharp surfaces of the retractor blades. These lesions, if identified intraoperatively, can often be repaired primarily.63 Postoperative paratracheal soft tissue swelling can compromise the airway and thus requires prolonged intubation.64 11. Given the persistent loading forces of the cervical spine, all instrumentation is destined to “fail.” For that reason, for the construct to be successful, a solid bony fusion is needed. 12. The blood supply to the upper thoracic cord relies on an anastomotic supply from the thyrocervical and costocervical trunks as well as the radicular arteries. In the middle region, T4-T8, a single thoracic radicular artery supplies this vulnerable area. In trauma cases or when hypotension occurs, this region of the spinal cord is the first to be affected. 13. The symptoms from herniated disks usually present as localized back pain; however, the pain may be represented as a band-like radiculopathy in the region of the particular dermatome level. Motor deficits, sensory deficits, and bowel or bladder dysfunction may also occur. Continued
W9418-09.qxd 11/3/03 2:53 PM Page 264
264
Section I
Introduction
14. Cancer is the second leading cause of death in the United States, with approximately 1.3 million new cases per year.94,95 The spinal column, especially the thoracic region, is the most frequent site of bony metastasis.96 15. Multiple myeloma accounts for 45% of all malignant bone tumors.99 It is primarily a disease of the sixth and seventh decades with a predilection for the thoracic spine, followed by the lumbar and rarely the cervical spine.100 The 5-year survival rate is 18%.101 16. Spinal epidural abscesses usually present with severe neck or back pain and progressive neurologic deficit that may be rapid. Approximately one half of epidural abscesses result from hematogenous spread to the epidural space. This type of abscess is more common in intravenous drug abusers. 17. The most common infections of the spinal column include pyogenic organisms (Staphylococcus aureus and coliform bacilli), infections caused by fungi (Actinomycetes and Blastomycetes), and Pott’s disease (Mycobacterium tuberculosis).108 18. Spondylolisthesis refers to the slippage of all or part of one vertebra onto another.
References 1. Yashon D: Spinal Injury. Norwalk, Conn., Appleton-Century-Crofts, 1986, pp 7–11. 2. Huelke DF, O’Day J, Mendlesohn RA: Cervical injuries suffered in automobile crashes. J Neurosurg 1981;54:316–322. 3. Hadley MN, Dickman CA, Browner CM, et al: Acute traumatic atlas fractures: Management and long term outcome. Neurosurgery 1988;23:31–35. 4. Landells CD, Van Peteghem PK: Fractures of the atlas: Classification, treatment, and morbidity. Spine 1988;13:450–452. 5. Fowler JL, Sandhu A, Fraser RD: A review of fractures of the atlas vertebra. J Spinal Dis 1990;3:19–24. 6. Dickman CA, Greene KA: Treatment of atlas fractures. In Menezes AH, Sonntag VK (eds): Principles of Spinal Surgery. New York,. McGraw-Hill, 1996, pp 855–866. 7. Jefferson G: Fracture of the atlas vertebra: Report of four cases, and a review of those previously recorded. Br J Surg 1920;7:407– 422. 8. Spence KF, Decker S, Sell KW: Bursting atlantal fracture associated with rupture of the transverse ligament. J Bone Joint Surg 1970;52A: 543–549. 9. Dickman CA, Papadopolous SM, Sonntag VKH, et al: The interspinous method of posterior atlantoaxial arthrodesis. J Neursosurg 1991;74:190–198. 10. Marcotte P, Dickman CA, Sonntag VKH, et al: Posterior atlantoaxial facet screw fixation. J Neurosurg 1993;79:234–237. 11. Hadley MN, Browner C, Sonntag VKH: Axis fractures: A comprehensive review of management and treatment in 107 cases. Neurosurgery 1985;17:281–290. 12. Anderson LD, D’Alonzo RT: Fractures of the odonoid process of the axis. J Bone Joint Surg 1974;56A:1663–1674.
19. Hemorrhage during and after diskectomy is very rare; however; if significant hypotension and hypovolemia occur intraoperatively, a vascular injury should be suspected. The mortality of a vascular injury during a lumbar diskectomy can be as high as 50%.124 20. CES is an indication for urgent surgical intervention. The onset of bladder paralysis is an important indicator for urgent surgery. Although the prognosis for CES is good, there is a significant difference in the outcome of cases operated on within 24 hours of bladder paralysis compared to those operated on after this period.129 21. Although minimally invasive surgery may be a useful technique in the spinal surgeon’s armamentarium, there is a profound learning curve. In inexperienced hands, complications can arise because surgical procedures are being performed with a two-dimensional image. 22. The major causes of death in the acute phase in spinal cord–injured patients are pulmonary failure and shock.150 Approximately 35% of spinal cord–injured patients will have a major pulmonary complication.151
13. Scott EW, Haid RW, Peace D: Type I fractures of the odontoid process: Implications for atlanto-occipital instability: Case report. J Neurosurg 1990;72:488–492. 14. Sonntag VKH, Hadley MN: Nonoperative management of cervical spine injuries. Clin Neurosurg 1988;34:630–649. 15. Levine AM, Edwards CC: The management of traumatic spondylolisthesis of the axis. J Bone Joint Surg 1985;67A:217–226. 16. Effendi B, Roy D, Cornish B, et al.: Fractures of the ring of the axis: A classification based on the analysis of 131 cases. J Bone Joint Surg 1981;63B:319–327. 17. Sonntag VKH, Dickman CA: Treatment of upper cervical spine injuries. In Rea GL, Miller CA (eds): Spinal Trauma: Current Evaluation and Management.Neurosurgical Topics. Park Ridge, Ill, American Association of Neurological Surgeons, 1993, pp 25–74. 18. Benzel EC, Hart BL, Ball PA, et al: Fractures of the C-2 vertebral body. J Neurosurg 1994;81:206–212. 19. Clark CT, Apuzzo MLJ: The evaluation and management of trauma to the odontoid process. In Cooper PR (ed): Management of Posttraumatic Spinal Instability. Park Ridge, Ill, American Association of Neurological Surgeons, 1990, pp 77–97. 20. Apfelbaum RI: Screw fixation of odontoid fractures. In Rengachary SS, Wilkins RH (eds): Neurosurgery, vol 2. New York, McGraw-Hill, 1996, pp 2965–2972. 21. Sonntag VKH, Dickman CA: Occipitocervical and high cervical stabilization. In Rengachary S, Wilkins R (eds): Neurosurgical Operative Atlas. Baltimore, Williams and Wilkins, 1991, pp 327–339. 22. White AA, Johnson R M, Panjabi MM, et al,: Biomechanical analysis of clinical stability in the cervical spine. Clin Orthop 1975;109: 85–96. 23. Schneider RC, Kahn EA, Arbor A: Chronic neurological sequelae of acute trauma to the spine and spinal cord. The significance of acute flexion or teardrop cervical fracture-dislocation of the cervical spine. J Bone Joint Surg 1956;38A.
W9418-09.qxd 11/3/03 2:53 PM Page 265
Chapter 9 24. Menezes AH: Developmental and acquired abnormalities of the craniovertebral junction.In Van-Gilder JC, Menezes AH, Dolan KD (eds): The Craniovertebral Junction and Its Abnormalities. New York, Karger, 1987, pp 109–158. 25. Bernini F, Elefante R, Smaltino F, et al: Angiographic study on the vertebral artery in cases of deformities of the occipitocervical joint, abstracted. Am J Roentgenol Radium Ther Nucl Med 1969;107: 526. 26. McRae DL: The significance of abnormalities of the cervical spine. AJR 1960;70:23–46. 27. Chamberlain WE: Basilar impression (platybasia): Bizarre developmental anomaly of occipital bone and upper cervical spine with striking and misleading neurologic manifestations. Yale J Biol Med 1939;11:487–496. 28. McGregor J: The significance of certain measurements of the skull in the diagnosis of basilar impression. Br J Radiol 1948;21:171–181. 29. Hinck VC, Hopkins CE, Savara BS: Diagnostic criteria for basilar impression. Radiology 1961;76:579. 30. VonTorkus D, Gehle W: The upper cervical spine. Regional anatomy, pathology and traumatology. In Georg, Thieme, Verlag (eds): A Systemic Radiological Atlas and Textbook. New York, Grune & Stratton, 1972, pp 2–77. 31. Bharucha EP, Dastur HM: Craniovertebral abnormalities (a report of 40 cases). Brain 1964;97:469–480. 32. McRae DL, Barnum AS: Occipitalization of the atlas. Am J Roentgenol 1953;70:23–46. 33. Spillane JD, Pallis C, Jones AM: Developmental abnormalities in the region of the foramen magnum. Brain 1957;80:11–48. 34. Hensinger RN: Congenital anomalies of the cervical spine—atlantooccipital fusion. In Rothman RH, Simeone FA (eds): The Spine. Philadelphia, WB Saunders, 1992, pp. 288–289. 35. Ball P, Benzel E: Pathology of Disk degeneration. In: Principles of Spinal Surgery. New York, McGraw-Hill, 1996, p 517. 36. Kelsey J, Githenss P, Walter S, et al: An epidemiological study of acute prolapse cervical intervertebral discs. J Bone Joint Surg 1984; 66A:907. 37. Braakman R: Cervical Spondylotic Myelopathy: Advances and Technical Standards in Neurosurgery. New York, Springer-Verlag, 1979, pp 137–170. 38. Henderson C, Henessy R: Posterolateral foraminotomy as an exclusive operative technique for cervical radiculopathy: A review of 846 consecutively operated cases. Neurosurgery 1983;13:504. 39. Clarke E, Robinson PK: Cervical myelopathy: A complication of cervical spondylosis. Brain 1956;79:483–510. 40. Parke WW: Correlative anatomy of cervical spondylotic myelopathy. Spine 1988;13:831–837. 41. Kramer J: Intervertebral Disc Disease. Causes, Diagnosis, Treatment, and Prophylaxis, 2nd ed. New York, Thieme, 1990. 42. Jensen MC, Brandt-Zawadski MN, Obuchowski N: Magnetic resonance imaging of the lumbar spine in people without back pain. N Engl J Med 1994;331:69–73 43. Boden SD, McCowin PR, Davis DO: Abnormal magnetic resonance scans of the cervical spine in asymptomatic subjects: A prospective investigation. J Bone Joint Surg 1990;72A:1178. 44. Smyth MJ, Wright V: Sciatica and the intervertebral disc: An experimental study. J Bone Joint Surg 1958;40A:1401–1418. 45. Brown M: Pathophysiology of disc disease. Orthop Clin North Am 1971;2:359–370. 46. Montgomery DM, Brower RS: Cervical spondylotic myelopathy. Orthop Clin North Am 1992;23:487–493. 47. Clark CR: Cervical spondylotic myelopathy. Spine 1988;13:347– 349. 48. Beck DW: Cervical spondylosis: Clinical findings and treatment. Contemp Neurosurg 1991;13:1–6. 49. Simeone FA, Rothman RH: Cervical Disc Disease: The Spine, 2nd ed. Philadelphia, Saunders, 1988, pp 440–499.
Spinal Neurosurgery
265
50. Zeidman SM, Ducker TB: Cervical disc diseases: Part I. Treatment options and outcomes. Neurosurg Q 1992;2:116–143. 51. Johnson RM, Hart DL, Simmons EF, et al: Cervical orthoses. J Bone Joint Surg 1977;59A:332–339. 52. Kaiser MG, Haid RW, Subach BR, et al: Anterior cervical plating enhances arthorodesis after discectomy and fusion with cortical allograft. Neurosurgery 2002;50:229–236. 53. Wilson DH, Campbell DD: Anterior cervical discectomy without bone graft. J Neurosurg 1977;47:551–555. 54. Jeffries RV: The surgical treatment of cervical myelopathy due to spondylosis and disc degeneration. J Neurol Neurosurg Psychiatry 1986;49:353–361. 55. Connolly E, Seymour R, Adams J: Clinical evaluation of anterior cervical fusion for degenerative cervical disc disease. J Neurosurg 1965;23:431–437. 56. Fielding W: Complications of anterior cervical disc removal and fusion. Clin Orthop Rel Res 1992;284:10–13. 57. Riley L, Robinson R, Johnson K, et al: The results of anterior interbody fusion of the cervical spine. J Neurosurg 1969;30:127–133. 58. Gore D, Sepic S: Anterior cervical fusion for degenerated of protruded discs. Spine 1984;9:667. 59. Bishop RC, Moore KA, Hadley MN: Anterior cervical interbody fusion using autogeneic and allogeneic bone graft substrate: A prospective comparative analysis. J Neurosurg 1996;85:206–210. 60. Maravilla KR, Hartling RP:Imaging decisions in degenerative spinal disease: A practical approach. MRI Decisions 1988;2:2–15. 61. Saunders RL, Bernini PM, Shirrefs TG: Central corpectomy for cervical spondylotic myelopathy: A consecutive series with long term follow up evaluation. J Neurosurg 1991;74:163–170. 62. Benzel EC, Lacon J, Kesterson L, Hadden T. Cervical laminectomy and dentate ligament section for cervical spondylotic myelopathy. J. Spinal Disord 1991;4:286–295. 63. Whitecloud TS II: Cervical spondylosis: The anterior approach. In Frymoyer JW,ed.: The Adult Spine: Principles and Practice. New York, Raven Press, 1978, pp 1165–1186. 64. Emery SE, Smith MD, Bohllman HH: Upper airway obstruction after multilevel cervical corpectomy for myelopathy. J Bone Joint Surg 1991;73A:544–551. 65. Bulger RF, Rejowski JE, Beatty RA: Vocal cord paralysis associated with anterior cervical fusion: Consideration for prevention and treatment. J Neurosurg 1985;62:657–661. 66. Tew JM, Mayfield FH: Complications of surgery of the anterior cervical spine. Clin Neurosurg 1976;23:424–434. 67. Raynor RB, Pugh J, Shapiro I: Cervical facetectomy and its effects on spine strength. J Neurosurg 1985;63:278. 68. White AA, Panjabi MM: Biomechanical considerations in the surgical management of the spine. In Wite AA, Panjabi MM (eds): Clinical Biomechanics of the Spine, 2nd ed. Philadelphia, Lippincott, 1990, pp 511–634. 69. Nazarian SM, Louis RP: Posterior internal fixation with screw plates in traumatic lesions of the cervical spine. Spine 1987;16S:64. 70. Sawin PD, Traynelis VC: Posterior articular mass plate fixation of the subaxial cervical spine. In Menezes AH, Sonntag VK (eds): Principles of Spinal Surgery. New York, McGraw Hill, 1996, pp 1099–1100. 71. Heller JG, Carlson GD, Abitbol JJ, Garfin SR: Anatomic comparison of the Roy-Camille and Magerl technique for screw placement in the lower cervical spine. Spine 1991;16S:552. 72. Traynelis VC: Anterior and posterior plate stabilization of the cervical spine. Neurosurg Q 1992;2:59. 73. Heller JG, Silcox H, Sutterlin CE: Complications of posterior cervical plating. Presented at the 22nd annual meeting of the cervical spine research society, Baltimore, December, 1994. 74. Tsuyama N: Ossification of the posterior longitudinal ligament of the spine. Clin Orthop Relat Res 1984;184:71–84. 75. Lipson SJ: Rheumatoid arthritis in the cervical spine. Clin Orthop 1989;239:121–127.
W9418-09.qxd 11/3/03 2:53 PM Page 266
266
Section I
Introduction
76. Simpson JM, An HS, Balderston RA: Complications of surgery of the spine in rheumatoid arthritis and ankylosing spondylitis. In Balderston RA, An HS (eds): Complications of Spinal Surgery. Philadelphia, WB Saunders, 1991, pp 169–175. 77. Grantham SA: Rheumatoid arthritis and other noninfectious inflammatory diseases: Atlantoaxial instability. In The Cervical Spine Research Society: The Cervical Spine. Philadelphia, Lippincott-Raven, 1983. 78. Moller P, Vinje O, Dale K, et al: Family studies in Bechterew’s syndrome (ankylosing spondylitis) I. Prevalences of symptoms and signs in relatives of HLAB27 positive proband. Scand J Rheumatol 1984;13:1–10. 79. Gran JT: The epidemiology of rheumatoid arthritis. Monogr Allergy 1987;21:162–196. 80. Blumberg B. Ragan C: The natural history of rheumatoid spondylitis. Medicine 1956;35:1. 81. Cardenosa G, Deluca SA: Ankylosing spondylosis. Am Fam Phys 1990;42:147–150. 82. Murray GC, Persellin RH: Cervical fracture complicating ankylosing spondylitis: A report of eight cases and review of the literature, review. Am J Med 1981;70:1033–1041. 83. Winter RB, Moe JH, Eilers VE: Congenital scoliosis: a study of 234 patients treated and untreated. J Bone Joint Surg 1968;50A:1–47. 84. Scheuermann H: Kyphosis dorsalis juvenilis. Ugeskr Laeger 1920;82: 385–393. 85. Rothman RH, Simeone FA: Scheuermann’s juvenile kyphosis. In The Spine, 3rd ed., Vol 1. Philadelphia, WB Saunders, 1980, pp 2380–2388. 86. Bradford DS, Moe JH, Montalvo FJ, Winter RB: Scheuermann’s kyphosis and roundback deformity: Results of Milwaukee brace treatment. J Bone Joint Surg 1974;56A:749. 87. Sorenson KH: Scheuermann’s Juvenile Kyphosis. Clinical Appearances, Radiography, Aetiology, and Prognosis. Copenhagen, Munksgaard, 1964. 88. Bradford DS: Juvenile kyphosis. Clin Orthop Relat Res 1977;128:45– 55. 89. Sturm PF, Dobson JC, Armstrong GWD: The surgical management of Scheuermann’s disease. Spine 1993;18:685–691. 90. Kumar R, Dunsker SB: Surgical management of thoracic disk herniations. In Schmidek, ed: Operative Neurosurgical Techniques, 4th ed. Vol 2. Philadelphia, WB Saunders, 2000, pp 2122–2131. 91. Logue V: Thoracic intervertebral disc prolapse with spinal cord compression. J Neurol Neurosurg Psych 1952;15:227–241. 92. Jelsma RK, Kirsch PT, Rice JF, Jelsma LF. The radiographic description of thoracolumbar fractures. Surgl Neurol 1982;18:230–236. 93. Denis F. The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 1983;8:817–831. 94. Masaryk TJ: Neoplastic disease of the spine. Radiol Clin North Am 1991;29:829–843. 95. O’Conner MI CB: Metastatic disease of the spine. Orthopedics 1992;15:611–620. 96. Malawer MM, Delaney TF: Treatment of metastatic cancer to bone. In DeVita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology, 4th ed. Philadelphia, Lippincott-Raven, 1993, p 2225. 97. Boogerd W JV: Diagnosis and treatment of spinal cord compression in malignant disease. Cancer Treat Rev 1993;19:129–150. 98. Gilbert RW, Kim JH, Posner JB: Epidural spinal cord compression from metastatic tumour: Diagnosis and treatment. Ann Neurol 1978;3:40–51. 99. Ansari A YD, Seymour JL: Acute pyogenic spondylodiscitis with epidural phlegmon. Diagnosis and management by MRI and multidisciplinary approach. Minn Med 1993;76:21–24. 100. Wolcott WP, Malik JM, Shaffrey CI, Shaffrey ME, Jane JA: Differential diagnosis of surgical disorders of the spine. In Benzel (ed): Spine Surgery: Techniques, Complication Avoidance, and Management. Philadelphia, Churchill Livingstone, 1999, pp 25–51.
101. Masaryk TJ: Neoplastic disease of the spine. Radiol Clin North Am 1991;29:829–843. 102. Abdelwahab IF Casden AM, Klein MJ, Spollman A: Chondrosarcoma of a thoracic vertebra. Bull Hosp J Dis Orthop Inst 1991;55:34–39. 103. Saeger W, Ludecke DK, Muller S, et al: Chordome des clivus: Histologie, Ultrastruktur und Klinik. Tumor Diagnostik Therapie 1983;4:74– 79. 104. Stein BM MP: Intramedullary neoplasms and vascular malformations. Clin Neurosurg 1992;39:361–387. 105. Hanley EN, Phillips ED: Profiles of patients who get spine infections and the type of infections that have a predilection for the spine. Semin Spine Surg 1990;2:257–267. 106. Sapico FL MJ: Vertebral osteomyelitis, review. Infect Dis Clin North Am 1990;4:539–550. 107. Correa AG EM, Baker CJ: Vertebral osteomyelitis in children, review. Pediatr Infect Dis J 1993;12:228–233. 108. Siddiqi SN, Fehlings MG: Ventral and ventrolateral spine decompression and fusion. In Benzel (ed): Spine Surgery: Techniques, Complication Avoidance, and Management. Philadelphia, Churchill Livingstone, 1999, pp 267–284. 109. Bloch AB SD: The epidemiology of tuberculosis in the United States. Clin Chest Med 1989;10:297–313. 110. Rosenblum B, Oldfield EH, Doppman JL, et al: Spinal arteriovenous malformations: A comparison of dural arteriovenous fistulas and intradural AVMs in 81 patients. J Neurosurg 1987;67:795–802. 111. Symon L, Kuyama H, Kendall B: Dural arteriovenous malformations of the spine: Clinical features and surgical results in 55 cases. J Neurosurg 1984;60:238–247. 112. Malis LI: Arteriovenous malformations of the spinal cord.In Youmans JR (ed): Neurological Surgery: A Comprehensive Reference Guide to the Diagnosis and Management of Neurosurgical Problems, 2nd ed. Philadelphia, WB Saunders, 1982, pp 1850–1874. 113. Heros RC, Debrun GM, Ojemann RG, et al: Direct spinal arteriovenous fistula: A new type of spinal AVM: Case report. J Neurosurg 1986;64:134–139. 114. Barrow DL, Colohan AR, Dawson R: Intradural perimedullary arteriovenous fistulas (type IV spinal cord arteriovenous malformations). J Neurosurg 1994;81:221–229. 115. Wiltse LL, Newman PH, Macnab I: Classification of spondylolysis and spondylolisthesis. Clin Orthop 1976;117:23–39. 116. Boxall D, Bradford DS, Winter RB et al: Management of severe spondylolisthesis in children and adolescents. J Bone Jt Surg 1979;61A:479–495. 117. Newman PH: Stenosis of the lumbar spine in spondylolisthesis. Clin Orthop 1976;115:116–121. 118. Cypress BK: Characteristics of physician visits for back symptoms: A national perspective. Am J Public Health 1983;73:389–395. 119. Cunningham LS, Kelsey JL: Epidemiology of musculoskeletal impairments and associated disability. Am J Public Health 1984;74:574–579. 120. Frymoyer JW: Back pain and sciatica. N Engl J Med 1988;318:291–300. 121. Anderson GB: Epidemiology. In Weinstein, Rydevik, and Sonntag (eds): Essentials of the Spine. New York, Raven Press, 1995, pp 1–10. 122. Ramirez LF, Thisted R: Complications and demographic characteristics of patients undergoing lumbar discectomy in community hospitals. Neurosurgery 1981;25:226–231. 123. Davis RA: A long-term outcome analysis of 984 surgically treated herniated lumbar discs. J Neurosurg 1994;80:415–421. 124. Smith DA, Cahill DW: Vascular and soft-tissue complications. In Benzel (ed): Spine Surgery: Techniques, Complication Avoidance, and Management. Philadelphia, Churchill Livingstone, 1999, pp 1407–1417. 125. Lawrence JS: Disc degeneration. Its frequency and relationship to symptoms. Ann Rheum Dis 1969;28:121. 126. Raaf J: Removal of protruded lumbar intervertebral discs. J Neurosurg 1970;32:604–611. 127. Spangfort EV: The lumbar disc herniation: A computer-aided analysis of 2,504 operations. Acta Orthop Scand Suppl 1972;142:1–95.
W9418-09.qxd 11/3/03 2:53 PM Page 267
Chapter 9 128. Gleave JR, MacFarlane R: Prognosis for recovery of bladder function following lumbar central disc prolapse. Br J Neurosurg 1990;4:205– 209. 129. Dinning TA, Schaeffer HR: Discogenic compression of the cauda equina: A surgical emergency. Aust NZ J Surg 1993;63:927–934. 130. Das De S, McCreath SW. Lumbosacral fracture-dislocations: A report of four cases. J Bone Joint Surg (Br) 1981;63B:58–60. 131. Hutter CG: Spinal stenosis and posterior lumbar interbody fusion. Clin Orthop 1985;193:103–114. 132. Gill K: Clinical indications for lumbar interbody fusion. In Lin PM, Gill K (eds): Lumbar Interbody Fusion. Rockville, Md, Aspen Publishers, 1989, pp 35–53. 133. Cloward RB: Posterior lumbar interbody fusion updated. Clin Orthop 1985;193:16–19. 134. Gill K, Blumenthal SL: Posterior lumbar interbody fusion. A 2 year follow-up of 238 patients. Acta Orthop Scand 1993;64[suppl 251]:108–110. 135. Steffee AD, Brantigan WJ: The VSP spinal fixation system. Report of a prospective study of 250 patients enrolled in FDA clinical trials. Proceedings of the North American Spine Society, 7th Annual Meeting, Boston, MA, 1992 (July 8–11). 136. Ray CD: Spinal interbody fusions: A review, featuring new generation techniques. Neurosurg Q 1997;7:135–156. 137. Blume HG: Unilateral posterior lumbar interbody fusion: Simplified dowel technique. Clin Orthop 1985;193:75–84. 138. Collins JS: Total disc replacement: A modified lumbar interbody fusion. Report of 750 cases. Clin Orthop 1985;193:64–67. 139. Lin PM: Posterior lumbar interbody fusion technique: Complications and pitfalls. Clin Orthop 1985;193:2–4. 140. Rish BL: A critique of posterior lumbar interbody fusion: 12 years’ experience with 250 patients. Surg Neurol 1989;31:281–289.
Spinal Neurosurgery
267
141. Branch CL, Branch CL Jr: Posterior lumbar interbody fusion with the keystone graft: Technique and results. Surg Neurol 1987;27:449–454. 142. Schechter NA, France MP, Lee CK: Painful internal disc derangements of the lumbosacral spine: Discographic diagnosis and treatment by posterior lumbar interbody fusion. Orthopedics 1991;14:447–451. 143. Brunet JA, Wiley JJ: Acquired spondylolysis after spinal fusion. J Bone Joint Surg 1984;66:720–724. 144. Gill K: Technique and complications of anterior lumbar interbody fusion.In Lin P, Gill K (eds.): Lumbar Interbody Fusion. Rockville, Md, Aspen Publishers, 1989, pp 95–106. 145. Goldner JL, Urbaniak JR, McCollum DE: Anterior disc excision and interbody spinal fusion for chronic low back pain. Orthop Clin North Am 1971;2:543–568. 146. Johnson RM, McGuire EJ: Urogenital complications of anterior approaches to the lumbar spine. Clin Orthop 1981;154:114–118. 147. Sacks S: Anterior interbody fusion of the lumbar spine. J Bone Joint Surg Br 1965;47B:211–223. 148. Rokito SE, Schwartz MC, Neuwirth MG: Deep vein thrombosis after major reconstructive spinal surgery. Spine 1996;21:853–859. 149. Ferree BA: Deep venous thrombosis following lumbar laminectomy. Orthopedics 1994;17:35–38. 150. Soden RJ, Walsh J, Middleton JW, et al: Causes of death after spinal cord injury. Spinal Cord 2000;38:604–610. 151. Reines HD, Harris RC: Pulmonary complications of acute spinal cord injuries. Neurosurg 1987;21:193–196. 152. Feldman RA, Karl RC: Diagnosis and treatment of Ogilvie’s syndrome after lumbar spinal surgery. Report of three cases. J Neurosurg 1992; 76:1012–1016. 153. Vanek VW, Al-Salti M: Acute pseudo-obstruction of the colon (Olgivie’s syndrome). An analysis of 400 cases. Dis Colon Rectum 1986;29:203–210.
W9418-10.qxd 10/31/03 3:41 PM Page 269
Chapter 10 Stereotactic and Functional Neurosurgery William A. Friedman, MD
Introduction The present state of stereotactic surgery is the consequence of more than 100 years of evolution in experimental neurology, neuroimaging modalities, engineering and, most recently, computer technology.1 The need for a method of exact intracranial localization and reproducible targeting was recognized long ago but, aside from the creation of several craniometric systems that intended to relate different brain structures to visible or palpable cranial reference points, little progress was made until early in this century. While doing basic research on anatomic networks, Sir Victor Horsley became disappointed with his ability to hit the deep cerebellar nuclei using a free-hand directed electrode. More often than not, his lesions fell far from the desired target. He recruited Robert H. Clarke, a young engineer with little previous background in experimental neurology, to help him find “a means of producing lesions of the cerebellar nuclei which should be accurate in position, limited . . . in extent, and involving as little injury as possible to other structures.”2 Clarke’s solution to the problem was simple, original, and enduring. He transformed the brain into a regular geometric body, dividing it with three imaginary intersecting spatial planes, orthogonal to each other: horizontal, frontal, and sagittal (Fig. 10-1). In this manner, each hemisphere was split into four segments, each having three deep planar walls and one curved wall corresponding to the brain surface. Any point within the brain could be specified by measuring its distance along the three intersecting planes (Fig. 10-2).
This concept, the brain as a geometric volume, is central to stereotaxis. The two other basic elements of stereotactic surgery involve the definition of suitable reference points in this geometric volume, and the construction of appropriate surgical instruments for operating on the targets thus identified. In this chapter, we will show how these three basic principles: geometry, reference points, and surgical instruments, have evolved, with the aforementioned developments in experimental neurology, imaging modalities, and computer technology, to produce the field of modern stereotactic and functional neurosurgery.
The Three Basic Elements of Stereotaxis The Brain in a Geometric (Cartesian) Coordinate System The place an object occupies in space is determined by its relative position with respect to a given point, which is arbitrarily defined as the reference. The addition of three orthogonal planes intersecting at the given reference point (defined as zero), establishes a system of axes. The location of any other point within the system requires the measurement of its distance from zero in the three planes of space, that is, x centimeters anterior, y centimeters to the left, z centimeters up. This concept, introduced by the French mathematician Rene Descartes in the 17th century, is intrinsic to modern geometry. However, its full application to intracranial 269
W9418-10.qxd 10/31/03 3:42 PM Page 270
270
Section I
Introduction
Figure 10-1. Clarke’s solution: Divide the brain along three orthogonal planes to generate a Cartesian coordinate system.
Figure 10-2. The stereotactic map is made by cutting brain slices parallel to the three planes of the selected Cartesian reference system. The slices are stained to enhance either gray nuclei or white fibers. They are photographed along with millimetric scales zeroed at the intersection of the reference planes.
W9418-10.qxd 10/31/03 3:42 PM Page 271
Chapter 10
localization had to await the seminal intervention of Robert Clarke. He based the new concept of stereotactic localization on the application of a Cartesian system of axes to the brain. This system requires the selection of suitable reference points within the skull or brain. The Reference Points There are two ways of defining the position of a given target in relation to the selected Cartesian reference system; the obvious one is to directly measure its distance from the zero point of the axes. This ideal method is feasible today with the interfacing of stereotactic frames and computed tomography (CT) or magnetic resonance imaging (MRI), as long as the target is visible (e.g., a tumor). For targets invisible even to these modern imaging modalities (e.g., the ventrolateral nucleus of the thalamus), an indirect localization method—the stereotactic map—has to be adopted. For its construction, fixed brain specimens are cut in regular spaced slices, parallel to the reference planes (sagittal, axial, frontal). The slices are stained to enhance either nuclei or fiber pathways. Then, each slice is photographed along with a millimetric ruler, and its relevant structures are identified and labeled. Each slice is numbered according to its distance from the corresponding zero reference plane. The reference planes must relate the invisible targets to structures that are visible on the imaging modality used (see the following discussion). All measurements are standardized from a large number of specimens, and the coordinates are depicted on a specimen representing a statistical average for normal subjects. It follows that map coordinates are only a good approximation to target localization in practice because the patient may not correspond to the standard. With these limitations in mind, map stereotactic coordinates are used to hit a target in clinical practice, once the reference points used in the map have been identified in the patient. Stereotaxis was born as a technical aid in experimental neurology. Clarke selected external skeletal points to define the planes of section in laboratory animals: a line through the auditory canals and the inferior margin of the orbits constituted the basal plane (actually, a parallel plane 1 cm above the former was used, bringing the basal plane closer to the center of the brain). A section passing through both external auditory meati and orthogonal to the basal plane (the interaural line), was selected as the frontal plane. The midline of the skull represented the sagittal plane. Skeletal points were the only reasonable reference available at that time and they were practical for use in the laboratory. Because the skull shape is remarkably constant in many animals, their external skull points were also reliable. Multiple animal stereotactic atlases developed since then, applying the previously mentioned reference points (starting with Horsley and Clarke’s study on the Macacus Rhesus brain), have proven their reproducibility.
Stereotactic and Functional Neurosurgery
271
Spiegel and Wycis introduced stereotaxis to clinical practice in 1947. They recognized that, given the wide variability in skull shape characteristic of the human being, external skeletal points would be useless as a reference frame for intracerebral targets. Consequently, they determined the need for intrinsic brain reference points. As the likelihood of statistical error increases with the distance between any two points (already noted by Horsley and Clarke), Spiegel and Wycis looked for intracerebral landmarks as close as possible to their potential targets. The first stereotactic applications were envisioned for the interruption of neural pathways in the thalamus to “reduce emotional reactivity,” mesencephalotomy for “interruption of the spinothalamic tract in certain types of pain” and “. . . production of pallidal lesions for involuntary movements.” Thus, most of the potential targets were anatomic structures surrounding the third ventricle. Consequently, the reference points selected by Spiegel and Wycis were periventricular structures (the pineal body when calcified, or the posterior commissure [PC]) consistently depicted by ventriculography or pneumoencephalography, the fundamental neuroimaging modalities at the time. Reids’ base line (a skeletal landmark), was relied on to define the angulation of the basal plane. The first human stereotactic map was published by Spiegel and Wycis in 1952,3–5 using the PC as the only intracerebral reference point. This system of axes had certain imperfections that prompted its abandonment a few years later. The calcified pineal body was acknowledged by Spiegel and Wycis as unreliable, because its inconstant calcification may lie at any point within its mass of 12 ¥ 8 ¥ 4 mm. The angulation of Reids’ line (or any other line based on cranial landmarks) has an unpredictable variability that also made it unsuitable to define the angulation of the cerebral basal plane. Talairach recognized the need for a completely intracerebral reference framework and introduced the anterior commissure/posterior commissure (AC/PC) system in his stereotactic atlas of the human brain.6 The line joining both commissures represented the basal plane, and two orthogonal lines at each commissure became the reference frontal planes. The sagittal plane corresponded with the midline. Two years later, Schaltenbrand and Bailey7 published their atlas based on a simplified AC/PC line reference system: the frontal plane was defined by a single line orthogonal to the basal plane, erected at the midpoint of the AC/PC (Fig. 10-3). Although other atlases were published later, with improvements in anatomical definition of certain structures, this reference system has endured until the present day. The Stereotactic Frame Although some exceptions exist, most stereotactic systems consist of two elements: the coordinate frame and the aiming
W9418-10.qxd 10/31/03 3:42 PM Page 272
272
Section I
Introduction
Figure 10-3. Because external skull landmarks are unreliable for determining intracerebral stereotactic coordinates in the human, a mapping system related to purely intracranial coordinates was devised. Modern systems rely on the mapped position of functional target relative to the anterior commissure and posterior commissure. These commissures were originally identified on ventriculography, then CT scans and, now, on MR images. The basal plane is defined by the AC-PC line. The coronal planes are perpendicular slices anterior and posterior to the midpoint of that line.
device. The frame is a rigid, metallic platform, which can be secured to the skull in such a way that no displacement is possible. This rigid attachment is in general achieved with three or four screws that tightly abut the outer table of the skull. The frame now becomes a foundation upon which a localizing system can be elaborated, or the aiming device can be attached. The latter is a rigid system of precision moving parts, bearing a probe holder. This can be moved in either
multiple angular or linear directions, so that it can be set to direct a probe to any target within the skull. Diverse geometric systems could potentially be used for construction of stereotactic guidance instruments. Due to practical constraints, however, only three of them have been actually selected for most of the frames currently available. These systems are polar coordinate, arc-radius, and focal point.8 The most popular stereotactic frames in use today are the Leksell frame9 and the Cosman-Roberts-Wells (CRW) frame, both of the arc-radius type. With this kind of system, once the target coordinates are set, the probe holder and arc can be moved to any entry point (Fig. 10-4).
Modern Stereotaxis
Figure 10-4. In arc-radius frames (the Leksell frame is shown here), the aiming arc can be moved along the three spatial planes (AP, lateral, and vertical), according to the obtained target coordinates. After this is completed, the focal point of the arc corresponds with the target. A probe equal in length to the arc’s radius will then hit the target regardless of its position along the arc, or the elevation of the arc from the horizontal plane, allowing the selection of virtually infinite trajectories for any target.
Before the revolution in neuroimaging brought about by the introduction of CT scanning in 1973 and MRI in the 1980s, stereotactic localization required conventional radiographs of the skull, supplemented with gas or dye ventriculography. Aside from being time-consuming and painful, exact orthogonal radiologic pairs and precise frame application were critical to avoid parallax and simplify the already cumbersome calculations needed to eliminate radiologic magnification. When CT became available, displaying normal and abnormal cerebral anatomy in undistorted, scaled axial slices, the scenario was set for a revolution in stereotactic localization. The Vertical Coordinate Problem Once stereotactic frames were constructed with low-artifact materials, it became possible to obtain undistorted CT scans with the frame secured to the patient’s head. For
W9418-10.qxd 10/31/03 3:42 PM Page 273
Chapter 10
the first time, the brain and the geometric system of reference could be seen together, in linear scale, without parallax or differential magnification. All that was needed was to mark the center of the frame on a CT slice and, considering it the zero point, anteroposterior (AP) and lateral coordinates for any visible intracranial target became directly obtainable. Determination of the vertical coordinate for the target was still a problem. Several years would elapse and many short lived methods would be proposed before effective solutions were worked out. In 1979, a medical student at the University of Utah applied a simple geometric principle and computerized trigonometric algorithms to derive a solution to the problem.10 His Lucite prototype became, with modifications, the first stereotactic system entirely designed to interface with CT: the Brown-Roberts-Wells frame. Three Nshaped arrays of carbon fiber rods are attached to the stereotactic base ring for CT localization. Each N produces three fiducial artifacts (for a total of nine) in any CT slice. The distance of the diagonal rod from the vertical rods allows calculation of the slice height (Fig. 10-5). Determination of the height at three points defines the spatial orientation of a plane through the frame and the patient’s skull, obviating a fixed relation between the frame and the CT gantry. This last feature is especially helpful in functional procedures: the CT gantry may be tilted as needed to include the anterior and posterior commissures in the same CT slice, substantially simplifying the data acquisition process. Magnetic Resonance Imaging Localization The introduction of MRI in the 1980s prompted the adoption of nonferromagnetic alloys for construction of stereotactic frames. The ability of MRI to directly image in the sagittal and coronal planes demanded the addition of localizing rods in proper planes, to specially designed localizer frames. Although MRI provides superior imaging quality for most brain lesions, especially in the posterior fossa, spatial distortion can be a problem. Nonlinear distortion of scanned structures increases in the vicinity of the magnet. As a consequence, central intracranial structures are less likely to be distorted but the stereotactic localizer itself, near the periphery of the lesion, is more likely to be distorted. Most importantly, MRI allows easy identification of the AC/PC plane, rendering it very useful as a functional surgical imaging modality. Currently two approaches are taken in the use of MRI in stereotactic surgery. First are systems that use MRIcompatible, nonferromagnetic frames. These systems allow the direct acquisition of MR images for use in stereotaxis but do not necessarily detect or correct for image distortion. Second are systems that acquire the MR images in a nonstereotactic format, some time before the actual procedure. On the day of surgery, a stereotactic CT scan is obtained and software is used to fuse the previously acquired MR images
Stereotactic and Functional Neurosurgery
273
to the CT scans. This approach has the advantage of reducing imaging time on the day of the procedure and detecting spatial distortion when MR images are warped compared to CT scans. This approach has the disadvantage of introducing another source of error: the fusion process itself (Fig. 10-6). Visible and Invisible Targets A visible target is any structure (normal or abnormal) falling within the resolution of a given imaging technique. As such, its coordinates may be obtained directly, without referring to a stereotactic map. High-field MRI is pushing the boundaries of visibility; more stable magnetic fields, changes in coil design, and different signal recovery techniques are constantly improving the spatial and tissue resolution of this modality so that, more than ever, the concept of visible target is a dynamic one. As previously noted, coordinates for visible targets are obtained directly from the scan, and referred exclusively to the Cartesian system represented by the stereotactic frame. For the invisible targets, such as specific thalamic nuclei, reliance on a stereotactic map is still necessary. A stereotactic procedure requiring the use of a map involves the use of two unrelated Cartesian systems: the stereotactic anatomic atlas and the stereotactic frame system. The two systems are tied together through the identification of common reference points such as the AC-PC line. When map coordinates for an invisible target are marked on the stereotactic scan, the target becomes visible (Figs. 10-7, 10-8). Its coordinates can then be identified as stereotactic frame coordinates. The identification of an invisible target, using a stereotactic map, is only an approximation for any individual patient, given normal variability in anatomy.11 However, it enables the design of an operative plan, the selection of an entry point, and a probe trajectory. Once the probe has been advanced to the tentative target as obtained by map coordinates, physiologic confirmation of its positioning is mandatory. This may be carried out by two methods, both of which are performed under local anesthesia: recording of spontaneous or evoked electrical activity; and/or electrical stimulation with either microelectrodes or macroelectrodes. The final position of the probe depends on the results of this physiologic testing.12 From Point to Volume The amazing progress of computer technology has turned three-dimensional reconstruction of CT scans and MR images into a real-time process. Constant reduction in hardware costs is making this technology cost-efficient for wider use as an aid in stereotactic surgery. Images generated by CT, MRI, angiography, and MR angiography, may be superimposed and reconstructed in different planes in stereotactic space, presenting the lesion geometry (and the surrounding anatomy) as it will be seen from the surgical trajectory.13
W9418-10.qxd 10/31/03 3:42 PM Page 274
274
A
B
Section I
Introduction
Figure 10-5. A, The BRW CT localizer has three N-shaped carbon fiber rods. In each N, the distance between the vertical rods is fixed and known. The distance of the diagonal rod from the posterior rod varies with the height of the CT slice. Its measurement allows the determination of the height above the base ring at which the rods have been imaged. B, The nine fiducial rods are visible on each CT slice as nine points. This computer screen shot is from a program called “CT process,” which automatically identifies the nine points and computes the vertical coordinate relative to the stereotactic head ring. In this manner the entire CT database, which may contain over 100 slices, can be quickly converted to a coherent stereotactic database wherein each CT pixel has an assigned anteroposterior, lateral, and vertical coordinate.
W9418-10.qxd 10/31/03 3:42 PM Page 275
Chapter 10
Stereotactic and Functional Neurosurgery
275
Figure 10-6. Using image fusion software, a nonstereotactic MR image is fused to a stereotactic CT scan. This computer screen is used to evaluate the alignment of the MR images and CT scans. The cursor slides across each screen, to dynamically change from CT (left) to MRI (right). The alignment of the ventricular surfaces, tumor edges, sulci, and vessels can be quickly determined.
Controlled changes in depth of the surgical field may be instantly correlated with concomitant modifications of the computer display. Current image-guided stereotactic programs are capable of directly porting stereotactic images to computer workstations in the operating room. Standard orthogonal, oblique, and probe’s-eye views are readily available. Any point on the scan is instantaneously converted to stereotactic coordinates. A trajectory for a biopsy or lesioning probe can be adjusted to produce the safest pathway to the target. Frameless Stereotaxis Although frame-based stereotaxis remains the most accurate and most popular method for performing stereotactic biopsy and most functional stereotactic procedures, frameless stereotaxis has become increasingly popular as a com-
puterized guidance method to facilitate craniotomy. Many commercial systems are now available. The basic method is as follows: First, markers are attached to the patient’s scalp (usually with glue). Then CT or MRI is performed. In surgery, the positions of the markers are registered by using a stereotactic device such as a mechanical wand or infrared probe. All of the acquired CT scans or MRI images are displayed on a computer in the operating room. After the registration process is complete, the wand or probe can then be used to display its position, in real time, on the imaging database. Thus, a linear skin incision can be made directly over a tumor, or a deep lesion can be found through a minimally invasive brain pathway. As computer technology and registration methods become faster and more transparent to the user, a greater percentage of neurosurgical procedures will undoubtedly be performed with this type of guidance.
W9418-10.qxd 10/31/03 3:42 PM Page 276
276
Section I
Introduction
Figure 10-7. This is the Shaltenbrand atlas axial plate, 2 mm above the AC-PC plane. The midline is on the right edge of the image. The grid displays 10-mm boxes running from PC (bottom) to AC (top). Vim is the desired target for thalamotomy and for thalamic stimulation. Its Cartesian coordinates relative to AC and PC are read from this map and then superimposed onto the stereotactic MR image for surgical targeting.
Common Stereotactic Procedures Stereotactic Biopsy A typical stereotactic biopsy procedure is as follows: The patient presents in the morning to preoperative holding. After the injection of local anesthetic, a stereotactic head ring is applied (Fig. 10-9). The patient is then transported to the CT scanner. There, a series of 1-mm-thick CT scan slices are taken from the top to the bottom of the head. These images are transferred, via Internet, to the stereotactic computer system, where each slice is quickly converted to a set of pixels, each of which has a defined AP, lateral, and vertical coordinate relative to the fixed head ring. A nonstereotactic MR image, obtained the day before the procedure, is then fused, with special software, to the stereotactic CT scan. Alternatively, many systems allow for the direct acquisition of stereotactic MR images while in an MRI-compatible head frame. The patient is then transported to the operating room. In the operating room, the stereotactic MRI is viewed on com-
puter. The desired target points and a precise trajectory, designed to avoid blood vessels and other danger spots, are computed. The target point is set up on a device called a “phantom.” The stereotactic frame is set to the desired coordinates and connected to the phantom to verify that no errors of setup have occurred. The skin is shaved and prepped over the small scalp area where the entry point is anticipated. The stereotactic frame is attached to the head ring. At the point where the biopsy probe touches the scalp, local anesthetic is injected and an incision made. A single burr hole is then placed (some prefer a smaller twist drill hole.). The dura is coagulated and opened. A biopsy needle is advanced through the burr hole to the target point and several biopsies taken (Fig. 10-10). A neuropathologist examines the tissue14; once a pathologic diagnosis is confirmed, the needle is withdrawn and the scalp wound closed in layers. The stereotactic frame is removed and the patient returned to the recovery room and, later, the hospital patient ward. The following morning the patient is discharged.
W9418-10.qxd 10/31/03 3:42 PM Page 277
Chapter 10
A
Figure 10-8. A, Axial T1-weighted MR image shows anterior commissure (cursor). This structure is seen clearly as a white matter pathway crossing from one hemisphere to the other. It is found several slices below the foramen of Monroe. B, Axial T1-weighted MR image shows posterior commissure (behind cursor). This white matter structure is less obvious and is found below the pineal gland. C, The computer has taken the indicated positions of the anterior and posterior commissures and has determined the Shaltenbrand atlas location of the posteromedial globus pallidus (cursor). The target position actually appears to be within the internal capsule, probably because the width of the third ventricle is relatively small. This reflects the anatomic variation commonly seen and the reason why final adjustment of the lesion position is done with reference to the actual MR image and with reference to intraoperative physiologic confirmation of correct target location.
Complications The most feared complication of stereotactic biopsy is hemorrhage. Fortunately symptomatic hemorrhage occurs in less than 2% of cases.15–17 Hemorrhage will usually be manifest in surgery, with blood out the biopsy needle and, sometimes, with the onset of new neurologic deficit. Usually such bleeding will stop spontaneously, with no deficit. If a deficit develops, however, rapid conversion of the procedure to a general anesthetic and open craniotomy may be necessary. Other re-
Stereotactic and Functional Neurosurgery
277
B
C
ported complications are seizure (rare), focal deficit without hemorrhage, infection, and lack of diagnostic tissue. Functional Stereotactic Lesions A variety of brain lesioning procedures have been used to control pain, psychiatric diseases, and movement disorders. By far the most commonly performed modern lesioning procedures are stereotactic thalamotomy and stereotactic
W9418-10.qxd 10/31/03 3:42 PM Page 278
278
Section I
Introduction
Figure 10-9. The stereotactic head ring is applied. After the injection of local anesthetic, metal-tipped pins are screwed into place. They rest tightly against the outer table of the skull, producing a rigid platform on which subsequent imaging and treatment can be performed.
pallidotomy. Thalamotomy is used primarily for the treatment of “intention” tremor disorders, which can be familial or secondary to trauma or multiple sclerosis. Intention tremors are usually mild or absent at rest, but become much more severe when the patient attempts to use that extremity. Severe intention tremor can interfere with eating and grooming, as well as all other fine motor activities. When
Figure 10-10. After computer planning, a stereotactic frame is attached to the head ring. This figure shows the BrownRoberts-Wells frame. In the BRW system, four angular coordinates and a depth are computed and the frame assembled. At the point where the drill touches the scalp, local anesthetic is injected. A twist drill hole or burr hole is placed. The biopsy needle is then inserted to the target point and a specimen of the lesion obtained. After frozen section confirmation of pathology, the small skin wound is sutured and the frame removed.
unresponsive to medication (primidone is best), the patient may be a candidate for thalamotomy. The thalamotomy procedure starts in an identical way to stereotactic biopsy, with ring application and data acquisition. The thalamic target is a so-called invisible target on MRI, so its location must be inferred from the position of the anterior commissure and posterior commissure, using a stereotactic atlas. Once identified on the computer, a probe is inserted to the target point. Under local anesthesia, the patient’s response to electrical stimulation is monitored. A good response is disappearance of tremor, with minimal sensory or motor effect. A radiofrequency heat lesion is then made to make the effect permanent. Some surgeons also perform microelectrode recording of the potential target to further refine its selection. Thalamotomy is a tried-and-true procedure. It carries a small risk of hemorrhage, paralysis, language problems, and memory disturbance. Bilateral thalamotomy is generally regarded as unsafe because of the risk of pseudobulbar side effects (drooling, difficulty speaking and swallowing, etc.). Pallidotomy is used to treat Parkinson’s disease. Parkinson’s disease is a degenerative disorder of the brain, which causes loss of dopamine-secreting neurons in the substantia nigra. This leads to resting tremor (not intention tremor), bradykinesia, and rigidity. Parkinson’s disease is usually effectively treated with medication. After 5 to 10 years of medical treatment, severe adverse effects to medications occur, the most common of which is the on-off effect. When “on,” patients have severe dyskinetic movements; when “off,” they are frozen. Such patients are regarded as candidates for surgical intervention, including pallidotomy. The pallidotomy procedure is very similar to thalamotomy except the target is the posteromedial globus pallidus (Fig. 10-11). The procedure has a remarkable effect on dyskinesias, with additional reduction of tremor, rigidity, and akinesia. Adverse effects can include hemorrhage, paralysis, partial visual loss, and so forth.18–20 Bilateral pallidotomies have been performed in large numbers of patients but are also thought to carry a risk of pseudobulbar complications. A recent review article looked at 1959 patients undergoing pallidotomy for Parkinson’s disease at 40 centers worldwide.21 There was a consensus on the benefits of pallidotomy for off period motor function and on period, drug-induced, dyskinesias. The overall mortality rate was 0.4% and persistent neurologic morbidity was estimated at 14%. Major adverse events, including intracerebral hemorrhage, hemiparesis, and visual field cut, occurred in 5% of cases. Limited data are available on the long-term outcome of this procedure. Deep Brain Stimulation The observation that electrical stimulation during pallidotomy leads to transient relief of tremor has given rise to the use of implanted stimulators for functional disorders, as opposed to heat lesions. An implantable stimulator can, at
W9418-10.qxd 10/31/03 3:42 PM Page 279
Chapter 10
Figure 10-11. Motor cortex activates the putamen, which then interacts with the globus pallidus via indirect and direct pathways. The globus pallidus output inhibits the thalamus, which then exerts inhibitory control on motor cortex. The object of pallidotomy and deep brain stimulation of the globus pallidus and subthalamic nucleus is to ultimately reduce the inhibitory feedback to motor cortex.
Stereotactic and Functional Neurosurgery
279
Figure 10-12. This lateral skull radiograph was taken postoperatively, after installation of a deep brain stimulator. The stimulating electrode is inserted stereotactically to the desired target point (such as the ventrolateral thalamus, for tremor). The final electrode position is adjusted to produce the desired clinical result (like tremor relief), then secured in place. The patient is then repositioned for insertion of the battery/ computer system in a subcutaneous pocket under the clavicle.
Radiosurgery least in theory, be turned off if the effect is undesirable, whereas a lesion’s effects are permanent. Deep brain stimulation (DBS) has primarily been used for chronic pain, with marginal success. Now DBS is increasingly used for movement disorders, especially in Europe. The deep brain stimulation equivalent of thalamotomy involves an identical targeting procedure. Instead of a lesioning probe, a stimulating electrode is introduced to the target point and tested under local anesthesia. The electrode is implanted under the skin and connected to a computer/ battery combination, which is installed under the clavicle (Fig. 10-12). The stimulator can be interrogated and adjusted through the skin with a special programmer. The patient can turn the stimulator on and off as desired. Unlike thalamotomy, bilateral stimulators appear both safe and effective, so severe bilateral tremor can be effectively treated. DBS has the same risks as lesioning procedures during the electrode insertion process, and also has the risk of equipment breakage or malfunction. Bilateral DBS of the globus pallidus and subthalamic nuclei has proven successful for Parkinson’s disease. If preliminary results are borne out by further studies, DBS may replace lesioning procedures in the treatment of Parkinson’s disease and other movement disorders.22
In 1951, Lars Leksell extended stereotactic techniques to the delivery of radiation to circumscribed targets in the brain.23 He described the concept of focusing multiple nonparallel beams of external radiation on a stereotactically defined intracranial target. The averaging of these crossfiring beams resulted in very high doses of radiation to the target volume, but much lower doses to nontarget tissues along the path of any given beam. Leksell coined the term “radiosurgery” to emphasize the precise destruction of a defined intracranial target—the focused radiation replacing the surgeon’s blade or probe by destroying a well-circumscribed volume of tissue while sparing surrounding structures. His research culminated in the development of the gamma knife, a system (described in the next section) that uses concentrically focused gamma rays from radioactive cobalt sources fixed in a hemispherical array (Fig. 10-13). All radiosurgery systems use this fundamental principle of intersecting beams to produce focal high-dose radiation and a steep dose gradient that spares nontarget structures. The development of radiosurgery presented the attractive prospect of administering a single, heavy dose of radiation to destroy any deep brain structure without the morbidity associated with open techniques. Radiosurgery was initially intended for use in functional neurosurgery for the section of deep fiber tracts or nuclei. This application has been
W9418-10.qxd 10/31/03 3:42 PM Page 280
280
Section I
Introduction be a limiting factor in certain cases. Recently developed techniques for the delivery of fractionated stereotactic radiation give up the convenience of single dose therapy, but combine the advantages of a well-circumscribed target volume with traditional exploitation of the differential radiation sensitivity between neoplastic lesions and healthy brain tissue. Three main systems have been developed for the stereotactic delivery of radiation to the brain. Their descriptions follow. The Gamma Knife System
Figure 10-13. The gamma knife is a hemispherical array containing 201 fixed cobalt sources. All sources are focused on one spot, producing a highly focused radiation treatment with very steep falloff.
limited by the inability to verify proper localization by stimulating or recording from a site before its destruction, as routinely practiced in open stereotactic lesioning methods. Over time, however, radiosurgery has proven effective for the treatment of selective vascular malformations, acoustic neuromas, meningiomas, pituitary adenomas, and other lesions not amenable to surgical resection. The noninvasive nature of radiosurgery is its obvious advantage. Disadvantages include the delay between treatment and therapeutic effect, and efficacy for only a limited size and spectrum of lesions. In comparison to conventional radiation therapy, radiosurgery does not rely on, or exploit, the higher radiation sensitivity of neoplastic lesions relative to normal brain (therapeutic ratio). Its selective destruction is dependent only on sharply focused high-dose radiation and a steep dose gradient away from the defined target. This allows treatment to be administered in a single dose and eliminates the inclusion of large amounts of healthy brain tissue in the field of radiation. A therapeutic ratio is not required, so traditionally radiation-resistant lesions can be treated. Because destructive doses are used, however, any normal structure included in the target volume is subject to damage. Advances in neuroimaging and the precise contouring of target volumes have minimized this problem, but in the case of pituitary tumors abutting the optic chiasm, this continues to
The gamma knife is the culmination of the design efforts of Leksell and associates. The modern gamma knife contains 201 cobalt-60 sources fixed in a hemispherical array. As the Co60 nuclei decay, they emit photons with an average energy of 1.25 MeV (gamma radiation). The radiation from each source is collimated initially so that all 201 beams are focused on a single point. Secondary collimation is achieved through the use of one of four collimator helmets (4, 8, 14, or 18 mm). The different sized collimator openings are used to vary the diameter of the dose distribution. The Leksell stereotactic frame is used to position the patient relative to the isocenter. The stereotactic head ring is fixed to the patient’s skull and attached to an adjustable assembly. The assembly is then set to the proper coordinates and the patient is slid into the gamma knife. The positioning assembly locks into a mechanical mount that accurately locates the frame relative to the dose isocenter. The dose delivered is determined by the duration of the irradiation (10 minutes per isocenter on average). Multiple isocenter treatments are performed by withdrawing the patient, repositioning the head ring assembly, choosing the appropriate collimator helmet, and administering the desired dose of radiation to that isocenter. Particle Accelerator Systems At a few institutions, mainly where high-energy physics research is conducted, charged particle irradiation is used as an alternative to standard photon radiation for radiosurgery.24 These systems use a synchrocyclotron to generate beams of energetic (100 to 200 MeV) nuclei of lowmolecular-weight atoms such as protons, or helium nuclei. Particle beam radiation has some advantages for application to radiosurgery, such as less beam scatter than x-rays or gamma rays, increased biologic effectiveness over photon radiation, and a favorable depth-dose distribution called the “Bragg peak.” Particle beams lose energy uniformly (“plateau”) until the particle nears the end of its range. At this point, the particle decelerates rapidly, depositing a well-defined maximum dose two to four times greater than the path dose. This region of increased dose at the end of the beam is the Bragg peak. The depth at which the Bragg peak occurs can be varied by interposing extracranial absorbers to change the entrance energy of the particle beam. Absorbers are also used to tailor the
W9418-10.qxd 10/31/03 3:42 PM Page 281
Chapter 10
width of the Bragg peak to match the target width. Because there is little exit dose of radiation beyond the target and the entry path dose is low, exploitation of the Bragg peak allows for very steep dose gradients with relatively few convergent beams. The primary limitation of this system is that a synchrocyclotron is required as the radiation source. In the United States, these facilities are currently available at only two sites: Harvard and Loma Linda universities.
Linear Accelerator Systems Linear accelerators (LINACs) are devices that use microwave power to accelerate electrons to high energies. These electrons, traveling at nearly the speed of light, are focused onto a heavy metal target. When the electrons collide with the target, their kinetic energy is converted into heat and photon radiation called x-rays. The x-ray beam thus generated has an effective energy equal to approximately one third of the maximum energy of the LINAC, so a 6-MeV LINAC produces an x-ray beam with an average energy of approximately 2 MeV. This 2-MeV x-ray beam is comparable to the 1.25-MeV gamma ray beam generated by the decay of radioactive cobalt in the gamma knife. Both beams are photon radiation and they differ only in their sources. The LINAC was developed in the 1950s and over the ensuing decades LINACs have become the favored treatment device for conventional radiation therapy. In 1984, Betti and colleagues described a radiosurgery system using a LINAC as the radiation source.25 Colombo and associates described such a system in 1985, and LINACs have subsequently been modified in various ways to achieve the precision and accuracy required for radiosurgery applications.26 In 1986, a team composed of neurosurgeons, radiation physicists and computer programmers began development of the University of Florida radiation surgery system.27 All LINAC radiosurgery systems rely on the following basic paradigm (Fig. 10-14): A collimated x-ray beam is focused on a stereotactically identified intracranial target. The gantry of the LINAC rotates around the patient, producing an arc of radiation focused on the target. The patient couch is then rotated in the horizontal plane and another arc performed. In this manner, multiple noncoplanar arcs of radiation intersect at the target volume and produce a high target dose, with minimal radiation to surrounding brain. This dose concentration method is exactly analogous to the multiple intersecting beams of cobalt radiation in the gamma knife.
Radiosurgical Paradigm Although the details of radiosurgery treatment techniques differ somewhat from system to system, the basic paradigm is quite similar everywhere. Following is a detailed description of a typical radiosurgery treatment at the University of Florida.
Stereotactic and Functional Neurosurgery
281
Almost all radiosurgery procedures in adults are performed on an outpatient basis. The patient reports to the neurosurgical clinic the day before treatment for a detailed history and physical examination, as well as an in-depth review of the treatment options. If radiosurgery is deemed appropriate, the patient is sent to the radiology department for volumetric MRI. The next morning, the patient arrives at 7:00 am. A stereotactic head ring is applied under local anesthesia. No skin shaving or preparation is required. Subsequently, stereotactic CT scanning is performed. One-millimeter slices are obtained throughout the entire head. The patient is then transported to an outpatient holding area for breakfast and relaxation until the treatment-planning process is complete. The stereotactic CT scan, as well as the nonstereotactic volumetric MR image are transferred via Internet to the treatment-planning computer. The CT scans are quickly processed so that each pixel has an anteroposterior, lateral, and vertical stereotactic coordinate, all related to the head ring previously applied to the patient’s head. Using image fusion software, the nonstereotactic MR image is fused, pixel for pixel, with the stereotactic CT scan. Dosimetry then begins and continues until the neurosurgeon, radiation therapist, and radiation physicist are satisfied that an optimal dose plan has been developed (Fig. 10-15). A variety of options are available for optimizing the dosimetry. The basic goal is to deliver a radiation field that is precisely conformal to the tumor shape, while delivering a minimal dose of radiation to all surrounding neural structures. The dosimetric options include arc-weighting, arc-tilting, and multiple isocenters. A detailed review of dosimetry is beyond the scope of this chapter. As soon as dosimetry is complete, the radiosurgery device is attached to the LINAC. The patient then is attached to the device and treated. Radiosurgery has high success rates in the treatment of arteriovenous malformations,28 meningiomas, acoustic schwannomas, and metastatic brain tumors.29
Stereotaxis: Past and Future Stereotaxis has come a long way since its origin as an elaborate, tedious method to locate a point within the brain. Present day technology enables real-time definition and visualization of the whole intracranial volume in stereotactic space. Still incorporating the three basic principles: geometry, reference points, and surgical instruments, stereotaxis today guides, with equal precision, a probe, the surgeon’s microscope, or a radiation beam, pushing the boundaries of what is considered surgically treatable intracranial disease. What was yesterday the realm of exclusive, high-tech, highbudget medical facilities, is now part of general neurosurgical practice.
W9418-10.qxd 10/31/03 3:42 PM Page 282
282
Section I
Introduction
Figure 10-14. In LINAC-based radiosurgery, a beam of x-rays is focused on the target point as the LINAC rotates around the patient. The patient is moved to another table position and another arc of focused radiation is executed. This results in multiple, noncoplanar arcs of radiation that intersect only at the target point.
Figure 10-15. This computer screen shot shows the treatment isodose line, one half of the treatment dose, and one fourth of the treatment dose to a left cavernous sinus meningioma. Axial, sagittal, and coronal views are displayed.
W9418-10.qxd 10/31/03 3:42 PM Page 283
Chapter 10
Stereotactic and Functional Neurosurgery
283
P earls 1. Stereotaxis was born as a technical aid in experimental neurology. 2. As previously noted, coordinates for visible targets are obtained directly from the scan, and referred exclusively to the Cartesian system represented by the stereotactic frame. 3. Once the probe has been advanced to the tentative target as obtained by map coordinates, physiologic confirmation of its positioning is mandatory. This may be accomplished by two methods, both of which are performed under local anesthesia: recording of spontaneous or evoked electrical activity; and/or electrical stimulation with either microelectrodes or macroelectrodes. The final position of the probe depends on the results of this physiologic testing.25
4. The most feared complication of stereotactic biopsy is hemorrhage. Fortunately symptomatic hemorrhage occurs in less than 2% of cases.15-17 5. Bilateral DBS of the globus pallidus and subthalamic nuclei has proven successful for Parkinson’s disease. If preliminary results are borne out by further studies, DBS may replace lesioning procedures in the treatment of Parkinson’s disease and other movement disorders.22 6. Radiosurgery has high success rates in the treatment of arteriovenous malformations,28 meningiomas, acoustic schwannomas, and metastatic brain tumors.29
References 16. 1. Spiegelmann R, Friedman WA: Principles of stereotaxis. In Crockard A, Hayward R, Hoff JT (eds): Neurosurgery: The Scientific Basis of Clinical Practice, 3rd ed. Oxford. Blackwell Scientific, 2000, pp 877–898. 2. Horsley V, Clarke RH: The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45–124. 3. Spiegel EA, Wycis HT: Stereoencephalotomy. Thalamotomy and related procedures. JAMA 1952;148:446–451. 4. Spiegel EA, Wycis HT: Stereoencephalotomy. New York, Grune & Stratton, 1952. 5. Spiegel EA, Wycis HT, Marks M, Lee AJ: Stereotaxic apparatus for operations on the human brain. Science 1947;106:349–350. 6. Talairach J, David M, Tournoux P, Corredor H, Kasina T: Atlas D’anatomie Stereotaxique. Paris, Masson et Cie, 1957. 7. Schaltenbrand G, Bailey P: Introduction to stereotaxis with an atlas of the human brain. Stuttgart, George Thieme Verlag, 1959. 8. Friedman WA, Coffey RJ: Stereotaxic surgical instrumentation. In Heilbrun MP (ed): Concepts in Neurological Surgery, vol 2. Philadelphia, Williams and Wilkins, 1988, pp 55–72. 9. Leksell L: A stereotaxic apparatus for intracerebral surgery. Acta Chir Scandinava 1949;99:229–233. 10. Brown RA: A computerized tomography-computer graphics approach to stereotaxic localization. J Neurosurg 1979;50:715–720. 11. Spiegelmann R, Friedman WA: Rapid determination of thalamic CTstereotactic coordinates: A method. Acta Neurochir 1991;110:77–81. 12. Tasker RR, Organ LW, Hawrylyshyn PA: The Thalamus and Midbrain of Man: A Physiological Atlas Using Electrical Stimulation. Springfield, Ill, Charles Thomas, 1982. 13. Kelly PJ, Kall B, Goerss S: Transposition of volumetric information derived from computed tomography scanning into stereotactic space. Surg Neurol 1984;21:465–471. 14. Friedman WA, Sceats DJ, Nestok BR, Ballinger WE: A comparison of preoperative and pathologic diagnosis of intracranial lesions. A review of 100 consecutive stereotactic biopsies. Neurosurgery 1989;25:180– 184. 15. Apuzzo MLJ, Chandrasoma PT, Cohen D, Zee CS, Zelman V: Computed imaging stereotaxy: experience and perspective related to
17. 18.
19.
20.
21. 22.
23. 24.
25. 26. 27. 28.
29.
500 procedures applied to brain masses. Neurosurgery 1987;20:930– 937. Kaye AH, Laws ER: Brain Tumors: An Encyclopedic Approach. Edinburgh, Churchill Livingstone, 1995. Ostertag CB, Mennel HD, Kiessling M: Stereotactic biopsy of brain tumors. Surg Neurol 1980;14:275–283. Laitinen LV, Bergenheim AT, Hariz MI: Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53–61. Lang AE, Lozano AM, Montgomery E, Duff J, Tasker R, Hutchinson W: Posteroventral medial pallidotomy in advanced Parkinson’s disease. N Engl J Med 1997;337:1036–1042. Vitek JL, Baksy RAE, Hashimoto T, et al: Microelectrode—guided pallidotomy: Technical approach and its application in medically intractable Parkinson’s disease. J Neurosurg 1998;88:1027–1043. Alkhani A, Lozano AM: Pallidotomy for Parkinson disease: A review of contemporary literature. J Neurosurg 2001;94:43–49. Ghika J, Villemure JG, Fankhauser H, Favre J, Assal G, Ghika-Schmid F: Efficiency and safety of bilateral contemporaneous pallidal stimulation (deep brain stimulation) in levodopa-responsive patients with Parkinson’s disease with severe motor fluctuations: A 2-year follow-up review. J Neurosurg 1998;89:713–718. Leksell L: The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316–319. Kjellberg RN, Koehler AM, Preston WM, Sweet WH: Intracranial lesions made by the Bragg peak of a proton beam. In Haley TJ, Snider RS (eds): Response of the Nervous System to Ionizing Radiation (Second International Symposium). Boston, Little, Brown, 1964, pp 36–53. Betti OO, Derechinsky VE: Hyperselective encephalic irradiation with linear accelerator. Acta Neurochir Suppl 1984;33:385–390. Colombo F, Benedetti A, Pozza F, et al: stereotactic irradiation by linear accelerator. Neurosurgery 1985;16:154–160. Friedman WA, Bova FJ: The University of Florida radiosurgery system. Surg Neurol 1989;32:334–342. Friedman WA, Bova FJ, Mendenhall WM: LINAC radiosurgery for arteriovenous malformations: Outcome versus size. J Neurosurg 1995;82:180–189. Friedman WA, Buatti JM, Bova FJ, Mendenhall WM: LINAC Radiosurgery—A Practical Guide. Berlin, Springer-Verlag, 1998.
W9418-11.qxd 10/31/03 3:43 PM Page 285
Chapter 11 Pediatric Neurosurgery David H. Shafron, MD
Introduction Pediatric neurosurgery involves the diagnosis and treatment of disorders of the developing and mature nervous system. With improvements in diagnostic and multimodality therapeutic techniques, patients treated in infancy and early childhood may have normal life expectancies, although they often require long-term, if not indefinite, follow-up care. As our understanding of the pathogenesis and treatment of nervous system disorders grows at an exceptional rate, so does our appreciation for the many unresolved controversies inherent in the field. For most of the disorders seen on a daily basis, the optimal method of management remains unresolved. Our success with, and enthusiasm for, a particular intervention often overshadow the lack of stringent scientific proof of its superiority over other techniques. The literature is rife with controversy, dogma, and bias. As more evidence-based data accumulate, the answers to our most basic questions may emerge, either validating or dispelling long-held beliefs. Many treatment strategies may not be amenable to prospective, randomized study, due to ethical issues or other constraints. For certain disease processes, intervention must be individualized, taking into account not only a sick child, but often a frightened and confused family. The goal of this chapter is to provide a broad overview of some of the more common disorders encountered by the pediatric neurosurgeon; special emphasis will be given to hydrocephalus, by far the most prevalent pediatric neurosurgical disorder. For the most part, the remainder of this chapter presents issues relevant to the treatment of adults with brain and spinal cord pathologies; although there may
be some overlap, an attempt is made to highlight, when appropriate, the differences in presentation and treatment for these disorders in children.
Neurologic Evaluation The cornerstone to a successful neurologic evaluation in a child is observation. Watching the spontaneous movements of an infant, the play habits of a toddler, and the gait of a child provides valuable information before any formal testing. The first step of any assessment begins with a detailed history, searching for the earliest onset of symptoms, and the progression of complaints. As the patient matures through and beyond infancy, developmental assessment plays a paramount role in the evaluation of neurologic well-being. The Denver Developmental Standard Test measures social, motor, and language development as a function of age1 and can indicate the presence of subtle or gross neurologic insults. There is tremendous variability in skill acquisition for an individual child; of greatest importance are the continued gain of sequential milestones, or the loss of a previously acquired skill. Likewise, changes in scholastic performance may be an early sign of neurologic dysfunction. As will be discussed, the hereditary nature of certain disorders mandates a detailed family history, including both immediate and extended family members. Finally, the patient’s social situation must be explored because issues of parental understanding, compliance, and (sadly) resources may play a role in determining an optimal management plan. 285
W9418-11.qxd 10/31/03 3:43 PM Page 286
286
Section I
Introduction
A focused general physical examination must also be conducted to seek manifestations of nervous system disease. The evaluation of an infant or child must include the rate of head growth, presence and character of the fontanelle and cranial sutures, and a careful search for anomalies involving the appendicular and axial spine, such as syndactyly and scoliosis. The integument must be examined for the cutaneous stigmata of occult spinal dysraphism or the phakomatoses. The neurologic assessment of a preterm infant may be confounded by the need for mechanical ventilation and sedation, as well as their enormous sleep requirement. Primitive reflexes, including the grasp, suck, rooting, and Moro responses should be present in some form by 28 weeks’ gestation, and complete by full term.2 Healthy premature and term infants arouse with eye openings. After approximately 30 weeks’ gestation, infants will display pupillary constriction to light, and blink in response to a bright light. Extraocular movement in the neonate may be gauged by the vestibulo-ocular reflex, while older infants will track objects in space. Disorders of oculomotility, such as nystagmus, may be present in primary position but often are evident only with gaze, and may indicate chiasmal, diencephalic, brainstem, cerebellar, or craniovertebral junction pathology. Facial sensation can be evaluated with the corneal reflex, as well as the expected grimace or cry in response to a noxious stimulus. Facial movements are observed both while the infant is quiet, as well as during crying spells. Asymmetry can result from facial nerve injury, but may also be secondary to congenital hypoplasia of facial muscualture. Absence of the depressor anguli oris, causing a characteristic “asymmetric crying facies,” can be associated with major congenital anomalies.3 Hearing may be evaluated by observing an infant’s response to a loud stimulus, although more sophisticated electrophysiologic analyses can be performed if indicated. The gag response is present after 30 weeks’ gestation.2 Infancy precludes formal manual muscle testing, so the examiner must rely on observing the symmetry, velocity, and fluidity of spontaneous and induced movements to assess motor function. Hypotonia is an important sign of brain dysfunction, but is a nonspecific finding. The hypotonic infant displays little spontaneous movement, and occipital flattening (plagiocephaly) may develop early. The more common indicator of motor-tract dysfunction, spasticity, may not manifest until after the first year of life because of continued postnatal myelination. The dystonic posturing indicative of basal ganglia pathology may be seen early in life, although it often worsens later in childhood. Persistence of posturing or abnormal reflexes, such as the asymmetric tonic reflex, may indicate severe cortical or subcortical insult. Strength testing becomes more reliable as a child begins to follow commands. Changes in handedness, which normally develops after 2 years of age, can signify brain, spinal cord, or peripheral nerve pathology. Deep tendon reflexes can be
elicited at 33 weeks’ gestation and thereafter; the presence of an extensor4 response to plantar stimulation may be normal during the first 2 years of life.2,4 The sensory examination may be confined to the reaction to noxious stimuli in an infant or young toddler. Indirect cues to sensory disturbance must be sought, such as skin ulcerations or burns in asensate areas, or gait dysfunction secondary to proprioceptive problems. As the child matures and is able to provide verbal feedback, direct dermatomal and proprioceptive testing can be carried out. For the older child and adolescent with a nervous system complaint, an evaluation similar to that of an adult may be appropriate. These patients have a mature (or nearly so) nervous system, and a neurologic examination that takes into account mental status, and cranial nerve, motor, sensory, cerebellar, and reflex function is often sufficient. The examination must be tailored to a patient’s age and level of sophistication and development.
Hydrocephalus No treatment in the history of neurosurgery has had as important an impact in reducing morbidity and mortality than the use of cerebrospinal fluid (CSF) shunts. The treatment of hydrocephalus with CSF diversion was revolutionized in the 1950s, with the advent of the first valve-regulated shunting system by Nulsen and Spitz, at Case Western Reserve University in Cleveland, Ohio.5 Before this landmark development, there was no consistently safe or effective technique for the management of hydrocephalus. Despite nearly a half-century of innovation and modification, neurosurgery remains vexed by problems in attempting to redirect CSF flow, with the imperfections of modern shunting systems brought to light throughout the interim, most recently in a multicenter, randomized trial.6 The scope of hydrocephalus is enormous. The prevalence of CSF shunts in the United States was conservatively estimated to be greater than 127,000 based on data for 19887; this figure has no doubt increased substantially in the past 15 years. This study estimated nearly 70,000 admissions for hydrocephalus, and more than 36,000 shunting procedures annually, with approximately 40% of these for shunt revision. The evaluation and treatment of hydrocephalus dominates the pediatric neurosurgeon’s practice, but despite 50 years of progress, decisions regarding whom to shunt, when to shunt, and how to shunt are often cloudy. Hydrocephalus occurs whenever there is a disparity between CSF production and absorption; with the rare exception of CSF-producing tumors, hydrocephalus is due to an obstruction, either anatomic or functional, between the ventricular system and the arachnoid villi, the primary site of absorption. Hydrocephalus usually presents during the first decade, although it may arise later in life, when it occurs secondary to tumor or vascular disorders. Overwhelmingly,
W9418-11.qxd 10/31/03 3:43 PM Page 287
Chapter 11
Table 11-1 Etiology of Pediatric Hydrocephalus
From Drake JM, Kestle JR, Milner R, et al: Randomized trial of cerebrospinal fluid shunt valve design in pediatric hydrocephalus. Neurosurg 1998;43: 294, 303.
hydrocephalus in association with myelomeningocele and hemorrhage of prematurity are the most common etiologies in pediatric neurosurgical practice (Table 11-1), although in many cases the etiology cannot be determined. Clinical and Radiographic Features The clinical features of hydrocephalus in children depend on the age of presentation, the rapidity with which the CSF obstruction occurs, and the underlying cause. The compliant skull of an infant often allows mechanical distension to occur before the onset of overt signs of increased intracranial pressure (ICP), whereas the older child typically becomes symptomatic as ventricular distension and increased ICP develop. It is not uncommon to encounter hydrocephalus on prenatal ultrasound examination; in this situation, head size and ventricular distension may be monitored on serial imaging studies before confinement. Decisions regarding timing and method of delivery are then made with regard to ventricular and head size, respectively (see following discussion). The child diagnosed prenatally may be born with a head circumference greater than the 95th percentile, a full or bulging fontanelle, and spilt sutures; in this situation, the diagnosis is straightforward. However, quite often ventricular enlargement noted on prenatal studies is not clinically evident at delivery, and treatment decisions depend on continued observation of the infant. Ventriculomegaly in the preterm infant secondary to periventricular-intraventricular hemorrhage (PIVH) is often an “incidental” finding on surveillance ultrasound evaluation, and may be asymptomatic. In this population, the impairment in CSF absorption may be mild at first, and can increase, stabilize, or normalize in the ensuing weeks. Daily occipital-frontal circumference is monitored, in conjunction with serial ultrasound studies. Signs of hydrocephalus in this group include accelerated head growth, progressive widening of sutures, and increased fullness of the fontanelle. Lethargy, apneic and bradycardic spells, and poor feeding may ensue, indicating a process of decompensation. These patients often present a considerable challenge in management decisions, as will be discussed.
Pediatric Neurosurgery
287
Infantile hydrocephalus from causes other than PIVH typically presents with accelerated head growth, full fontanelle, and sutural widening. Percussion of the head can result in a hollow timbre if the ventricles are sufficiently dilated, and transillumination examination may be revealing. As venous outflow is impeded by an elevated ICP, scalp veins become dilated and more prominent. Paresis of upward gaze, the presence of ocular “sun-setting,” feeding intolerance, vomiting, and lethargy may occur in advancing cases. As the process continues, other ocular disturbances such as sixth nerve paresis, nystagmus, and optic atrophy may ensue, although true papilledema in the presence of open sutures is uncommon. In less acute cases, however, these signs may be absent, because the accommodating nature of the infant skull can permit significant ventricular distension. Other than an enlarged head, the only clues may be the failure to develop age-expected milestones, or the loss of those previously attained. Extremity tone may be decreased, normal, or increased. In the older child, features of hydrocephalus again depend on the etiology and development over time. If due to a slowly progressive congenital cause, head size may be at the upper limits of normal or slightly enlarged. Young children will often not voice complaints, but may exhibit behavioral changes, such as increased napping, diminished play activity, or enuresis (after toilet-training). The older child may readily complain of headache, usually worse with recumbancy. Vomiting may accompany, and briefly relieve the headache, possibly due to the hyperventilation or positional changes accompanying bouts of emesis. Parents or teachers may note a change in school performance. In the absence of obvious associated signs, patients may be misdiagnosed as harboring migraine headache or gastrointestinal motility disorders. In late childhood and adolescence, central nervous system neoplasms are the primary cause of hydrocephalus, and symptoms of increased ICP are often the initial, or only, manifestations of tumor progression. Lethargy, papilledema, oculomotor dysfunction, and hyperreflexia may be noted before the development of focal deficits. The use of radiographic imaging complements the clinical evaluation in the diagnosis of hydrocephalus. In the neonatal period, ultrasonography is often the initial study obtained; magnetic resonance imaging (MRI) or computed tomography (CT) may provide the necessary detail to determine a specific etiology, and to facilitate decision-making. In the older child or adolescent, CT or MRI studies are obtained, often by primary care personnel. Management decisions for children with hydrocephalus may be straightforward when they present with obviously symptomatic disease, and studies reveal significant ventriculomegaly. Often, however, neurosurgical consultation is obtained when clinical manifestations are mild or absent, and radiographic imaging results equivocal. Decisions regarding treatment options may be made over days to months.
W9418-11.qxd 10/31/03 3:43 PM Page 288
288
Section I
Introduction
The prenatal diagnosis of hydrocephalus presents important questions to parents, obstetricians, and neurosurgeons: “Can the child attain normal intelligence?” “Should pregnancy continue?” “When and how should the child be delivered?” The answers are often elusive. If massive and progressive ventriculomegaly is detected within the first trimester, intellect is usually significantly impaired.8 When ventricular size is normal or mildly enlarged on initial studies, and progresses during the latter stages of gestation, intervention may succeed in preserving intelligence. Progressive ventricular enlargement in utero is variable—absent in one study of 40 patients,9 present in 4% of 47 patients,10 and present in 45% of 20 patients11 in yet another series. A general guideline, although by no means absolute, suggests that a cortical mantle less than 1-cm thick is indicative of significant cognitive impairment, and delivery should be entertained after 32 to 34 weeks’ gestation as the mantle approaches this cut-off measurement, as detected on serial ultrasound studies. Amniocentesis can be performed at approximately 32 weeks, and delivery is carried out if the biophysical profile indicates lung maturity. In the absence of lung maturity, maternal steroid therapy is initiated, with subsequent delivery. Route of delivery depends on the fetal head size, and other associated anomalies; ventriculomegaly alone does not preclude vaginal delivery, although in most series cesarean sections predominate. The most common cause of ventriculomegaly diagnosed in utero is myelomeningocele.9,10,12 In a minority of cases, increased ventricular size is found as an isolated phenomenon, without an apparent underlying cause, and in these children the condition may stabilize or resolve.9,13 Among patients diagnosed with hydrocephalus in utero who survive, approximately 50% to 70% ultimately require CSF diversion.9,10 Associated brain or other systemic anomalies are common, and are associated with deficits in cognitive development.10,13–15 As an isolated finding, however, approximately 50% to 60% of patients with fetal ventriculomegaly can attain normal intelligence.13,15 Most patients diagnosed in utero with isolated ventriculomegaly that ultimately stabilized or resolved without requiring shunting demonstrated satisfactory cognitive outcome.9,13 Post-hemorrhagic Hydrocephalus As noted previously, hydrocephalus secondary to germinal matrix-intraventricular hemorrhage is among the most common indications for CSF diversion in most modern series. In the past several decades, the incidence of PIVH of prematurity has decreased,16 while the survival of very low birth-weight infants has improved dramatically as a result of advances in critical care medicine. The greater survival rates have resulted in a possible increase in the prevalence of post-hemorrhagic hydrocephalus (PHHC),17 estimated at approximately 1.7% overall for infants less than 32 weeks’ gestation.18 The true incidence of PIVH and resultant
Table 11-2 Grading Scale for PeriventricularIntraventricular Hemorrhage • Grade I • Grade II • Grade III • Grade IV
germinal matrix hemorrhage germinal matrix and intraventricular hemorrhage; ventricles normal in size germinal matrix and intraventricular hemorrhage; ventricles dilated germinal matrix and intraventricular hemorrhage, with parenchymal extension
Papile LA, Burstein J, Burstein R, et al: Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birth weights less than 1500 gm. J Pediatr 1978;92:529–534.
PHHC, however, is difficult to elucidate, as reports have varied in inclusion criteria (such as birth weight), and grade of bleeding severity. The most commonly accepted classification of PIVH is shown in Table 11-2.19 Grades I and II hemorrhage pose little risk of subsequent PHHC or neurologic injury, while grades III and IV bleeding harbor an increased risk of PHHC, as well as subsequent cognitive impairment.20–22 Risk of bleeding severity and PHHC is strongly correlated with lower birth weight and degree of prematurity20,23–26; grade II to IV PIVH is uncommon in infants older than 29 weeks’ gestation.27 Other risk factors for PIVH in the premature population, apparently independent of birth weight and age, include the need for intubation, transport to another facility,24 and respiratory distress syndrome.26 Fluctuations in blood pressure, low cerebral blood flow,28 and amniotic sac inflammatory changes29,30 have also been associated with development of PIVH. In a large-scale recent series, grade II to IV PIVH developed in 22% of infants less than 1000 g surviving more than 12 hours 26; another modern series demonstrated severe PIVH in 11.4% of infants less than 1000 g, and in 5% of those between 1000 and 1250 g.25 The etiology of PIVH is uncertain and probably multifactorial.31 Advances in perinatal care and neonatal cardiorespiratory support have undoubtedly contributed to the overall decline in the incidence in the past 20 years. Multiple class I (prospective, randomized) studies32 have shown that the use of indomethacin in low birth weight infants reduced the risk and severity of PIVH.33–36 Most infants with PIVH do not develop PHHC; of those that do, issues regarding the necessity, timing, and type of neurosurgical intervention are not well established, because most patients with post-hemorrhagic ventriculomegaly stabilize and do not require shunting.17,37–39 Even those who demonstrate progressive ventricular enlargement over the first month of life may not ultimately require treatment.40 The subset of infants with PIVH who do finally require CSF diversion manifests with progressive ventriculomegaly, enlarging head size, and perhaps signs of increased ICP. Given the possibility of stabilization and resolution of symptoms, however, the optimal time for surgical intervention is unknown. Rekate8 suggested a goal of achieving a cortical mantle size of 3.5 cm by
W9418-11.qxd 10/31/03 3:43 PM Page 289
Chapter 11
age of 5 months to optimize intellectual outcome. The neurosurgeon is faced with a dilemma: balancing the possibility of injury to the developing brain from progressive ventricular expansion against the risks of surgery for a potentially reversible situation.In this regard,several forms of intervention are advocated to allow the decision regarding shunt placement to evolve over weeks or months. Removal of (bloody) CSF in an intermittent or continuous fashion is a logical “middleground”in this controversy, and studies have evaluated the efficacy of intermittent lumbar puncture,41,42 ventricular-access reservoirs or subgaleal shunts,39,43–48 and external ventricular drainage.49–51 These strategies may in theory protect the cortical mantle during periods of ventricular enlargement, and reduce signs of elevated ICP, but there is no clear evidence for their efficacy in reducing the likelihood of shunt placement, or future cognitive disability.41,52 Based on experimental models developed in canines by Pang and colleagues,53 the use of intraventricular fibrinolytic therapy as a means of preventing PHHC and shunt requirement has been explored in several studies. Despite the apparent benefit seen in several pilot studies and uncontrolled trials,54–57 others failed to show a benefit.58,59 There is currently no class 1 or 2 evidence supporting the use of fibrinolytic therapy for the treatment of PIVH,60,61 and larger, randomized, and well-controlled protocols are needed to establish its safety and efficacy. Similarly, the use of diuretic therapy to prevent PHHC in patients with PIVH62,63 was recently examined in a large, randomized, multicenter study. The use of diuretic agents plus standard therapy, such as CSF removal, was found not only to be less effective, but more harmful than standard therapy alone.64 At this time, CSF removal may be the only acceptable option for the symptomatic child too small to undergo shunt placement. Myelomeningocele and Hydrocephalus Myelomeningocele (MMC), the most common associated anomaly noted in patients with fetal ventriculomegaly, will be discussed in greater detail later in this chapter. Approximately 80% to 90% of children born with MMC have hydrocephalus, and most require shunt placement. The cause of hydrocephalus in these patients is unknown and is probably multifactorial. Hindbrain anomalies predispose to fourth ventricular outlet obstruction, and there may be obstruction of the aqueduct from cephalad herniation of the cerebellum. For unknown reasons, the level of the MMC is associated with the risk of hydrocephalus, being greatest for those with higher lesions, and lowest for sacral MMC.65 For infants with overt symptoms of increased ICP, clearly enlarged ventricles, or those with symptoms from hindbrain herniation, simultaneous repair and shunting can be performed safely,66,67 and may reduce wound complications or CSF leak. In patients without symptoms or radiographic evidence of hydrocephalus, an expectant approach can be used after primary repair, observing the patient with serial neurologic exami-
Pediatric Neurosurgery
289
nation, head circumference measurement, and ultrasound imaging. The need for shunt placement typically manifests within days to weeks of the repair; if there is no indication for shunting before hospital discharge, careful follow-up evaluation will determine those requiring diversion. The use of endoscopic third ventriculostomy (see following discussion) for the initial treatment of hydrocephalus in the MMC population is controversial. The enlarged massa intermedia, thickened ventricular floor, and other associated anomalies often preclude safe passage of the scope,68 and the low success rates in neonates69 do not warrant its use. Treatment and Complications Ventriculoperitoneal shunt placement is the standard treatment for patients with hydrocephalus. The principal shunt complications, malfunction and infection, have been the objects of considerable study, and attempts to examine the multiple factors predisposing to shunt complications have largely been single-institutional reviews.70 Variables such as type of shunt, site of entry, cause of hydrocephalus, age of patient, length of operation, antibiotic regimen, and others have all been scrutinized, often with variable conclusions. There is significant disparity in shunt insertion practices among different institutions,71 and no technique has proven superior to others. Although systematic retrospective reviews can provide valuable information, the most effective method to examine factors predisposing to shunt failure is with a prospective, randomized study. Proximal ventricular catheter obstruction is usually, although not always,72 the most common cause of shunt failure, and the theory of reducing this risk with the use of anti-siphoning technology spurred the development of several modern shunt systems. The recent Shunt Design Trial6 found no differences in complication or shunt survival rates among standard differential-pressure, anti-siphon, or flow-limiting valve systems in 344 children randomized at first shunt insertion. The overall rates of shunt obstruction (31.4%) and infection (8.1%) were not significantly different between the systems. Kaplan-Meyer analysis showed shunt survival rates of 61% and 47%, at 1 and 2 years, respectively. Extended follow-up evaluation from this study continued to show no difference between the systems, with a 41% shunt survival at 4 years,73 and revealed that the rate of decrease in ventricular size after shunt placement was no different among the different designs.74 Another multi-institutional, randomized, prospective series involving 377 patients found no overall difference in failure rates between nonprogrammable and programmable valves (although it did not attempt to identify subsets of patients who may benefit from the programmable feature).75 For each group, the 2-year actuarial survival was identical, 48%, mirroring the rates seen in the 1998 Shunt Design Trial. Obstruction was the leading cause of failure, and the infection rate was approximately 10%. A prospective, nonran-
W9418-11.qxd 10/31/03 3:43 PM Page 290
290
Section I
Introduction
domized, single-institution series of 50 patients, comparing differential pressure and anti-siphon valves, showed no difference in shunt survival between the groups at 5 years (59%).76 Infections were seen in 11%, all in the siphoncontrol group. These recent prospective series highlight the continued high rates of shunt malfunction and infection in the modern era. The lack of apparent differences between the shunt systems underscores the need for continued innovations in shunt technology. In addition to choosing from among the scores of commercially available shunt systems, surgeons are faced with technical issues regarding shunt placement, often with little consensus. Previous series had shown efficacy for both the frontal77 and posterior78 approaches. A prospective, randomized study of 121 patients requiring a first shunt examined anterior versus posterior burr-hole placement. The posteriorly placed systems had longer survival, prompting the author79 to conclude that anteriorly placed shunts conferred no advantage. Follow-up analysis of the shunt-design trial revealed that catheter tips completely surrounded by CSF on follow-up imaging, in either the frontal or occipital horn, had the longest survival.80 These studies were not designed to identify reasons for shunt failure, so few conclusions regarding risk factors could be drawn from them. A single-institution retrospective analysis of 727 shunt procedures identified several variables associated with failure, including patient age less than 2 years, previous failure within 6 months, and placement for perinatal hemorrhage.81 Neither shunt system nor location had an effect on survival. Tuli and colleagues80 examined risk factors for recurrent shunt failure and also found that variables such as shunt location, type of shunt, and length of surgery, all factors within the surgeons’ control, were not significant. Patients younger than 1 year, and especially those who underwent a shunt procedure before 40 weeks’ gestation, had significantly higher failure rates than older children, and shunt revision in the previous 6 months was also found to confer a significant risk of future failure. Additionally, IVH, meningitis, and tumor-related hydrocephalus were associated with increased risk of failure in this series. The signs and symptoms of shunt malfunction are similar to those of hydrocephalus and increased ICP described previously, and are a reflection of the age, acuity of dysfunction, and degree of shunt dependency. Infants may display asymptomatic fluid accumulation in the subgaleal plane and along the tubing. Although uncommon, seizures may also occur de novo, or increase in frequency.6,72 Patients with Chiari II malformations can present with neck pain, lower cranial neuropathy, and symptoms related to syrinx.72 The diagnosis of shunt malfunction typically begins with concern on the part of a parent, with the ensuing workup involving clinical and radiographic examination. Overwhelmingly, diagnosis of shunt malfunction can be made on
clinical grounds, based on complaints and findings on examination. The time-honored technique of pumping a shunt to evaluate proximal or distal obstruction has no clear predictive value.81 Radiographic demonstration of ventricular enlargement is seen in the majority of documented malfunctions, but may be absent or equivocal in up to one third of cases82; normal or small ventricles cannot provide assurance in the face of clinical evidence to the contrary. Overall, the finding of discontinuity on shunt series radiographs is not common, but its use in planning a revision warrant its use as part of the evaluation. Tapping a shunt can provide supplemental information when clinical and radiographic evaluation is equivocal,83,84 and may be particularly valuable for the infant, in whom a viral syndrome can mimic symptoms often indistinguishable from those of shunt dysfunction or infection. Other ancillary tests, such as radionuclide studies, may be helpful in certain cases.85 After mechanical obstruction, shunt infection is the most common complication associated with CSF diversion. The causes of shunt infections are multifactorial, and the sources of contamination are often unknown.86 In addition to necessitating shunt replacement, which confers morbidity, central nervous system infections may have a profound effect on future cognitive development. Most shunt infections occur within several months of insertion, with Staphylococcus epidermidis and Staphylococcus aureus the most common organisms isolated.87 Fever, irritability, signs of shunt malfunction, and erythema around the incision site are the most common findings. Uncommonly, infection can present in a delayed fashion88; these cases are notable for the frequent association with abdominal pseudocyst, and the variety of organisms isolated, including enteric flora. The discovery of a remote infection should always prompt a workup for pseudocyst.88–90 The treatment of shunt infection usually mandates removal of all hardware, external ventricular drainage, intravenous (and rarely, intrathecal) antibiotics, and shunt replacement after approximately 1 week of sterile CSF culture results. Pseudocysts generally resolve upon removal of the peritoneal catheter, although aspiration may be required in select cases. The infection rates noted in the prospective randomized trials described previously, roughly 8% to 10%, are not unusual for large centers, although single institutional reviews have described significantly lower rates,91 even less than 1%.92 Retrospective analyses of risk factors for infection have yielded widely disparate findings. A prospective evaluation of risk factors for 299 patients undergoing shunt placement was recently reported.93 The rate of infection was 10.1%, with Staphylococcus species the most common isolated pathogens. Univariate analysis revealed that only postoperative CSF leak, contamination of implant by a breached glove, and patient age (less than 40 weeks’ gestation) were significant factors for infection. Other factors, such as cause
W9418-11.qxd 10/31/03 3:43 PM Page 291
Chapter 11
of hydrocephalus, duration of surgery, or previous shunt operation, were not associated with increased risk. An earlier study from the same institution94 described a significant decrease in shunt infection (from 12.9% to 3.8%) after instituting a protocol for shunt surgery that included early start times, minimal operating room personnel, and perioperative antibiotics. Further analysis in this report showed that antibiotic use and shorter duration of surgery were associated with lower infection rates, findings not noted in the recent, prospective analysis.93 Antibiotic prophylaxis is used in most, if not all, centers regularly performing shunt surgery, although there is no agreement regarding the most efficacious agent, nor the timing of perioperative antibiotic administration. A metaanalysis of randomized, controlled trials95 did confer a benefit to the use antibiotic prophylaxis in decreasing infection rates, with the strongest effect noted when baseline (control) rates were high. Endoscopic procedures are gaining popularity for the management of hydrocephalus in select pediatric patients. In patients with multiloculated hydrocephalus, the use of endoscopic cyst fenestration can be an effective means for reducing the number of shunt revisions, and simplifying existing shunt systems.96 Endoscopic third ventriculostomy (E3V) can obviate shunt placement, most commonly in patients with noncommunicating hydrocephalus. The best candidates are patients with aqueductal stenosis, or mass lesions obstructing flow into or through the fourth ventricle (triventriculomegaly). As an initial procedure for the treatment of hydrocephalus secondary to aqueductal stenosis in children, Tuli et al.97 reported nearly identical failure rates between E3V and shunts (44% and 45%, respectively) using survival analyses. In general, success rates are 50% to 90% for children with triventricular hydrocephalus.98–102 The use of E3V for hydrocephalus secondary to myelomeningocele and hemorrhage is not as well established,68,69,98,101 with variable success rates reported. Its use in infancy is also unclear; efficacy is generally, although not universally,98 lower in this group, presumably due to immature subarachoid pathways and inefficient absorption at the arachnoid villi.69,101,103 In experienced hands, complications of E3V are not common, usually less than 10%; reported complications include perforation of the basilar artery or its branches, injury to hypothalamic structures, and infection.69,99,101,102,104 Uncommonly, the ostomy can become occluded months or years after the procedure, and patients may present with chronic or acute symptoms of increased ICP, similar to shunt failure. As with every technical development, there is a steep learning curve, not only for the technical aspects involved in neuroendoscopy, but perhaps more importantly, for proper patient selection.98,101 As more long-term results are reported, greater insight into the best indications for its use should be clarified.
Pediatric Neurosurgery
291
Myelomeningocele Epidemiology and Etiology Myelomeningocele is the most common congenital defect of the nervous system compatible with life. It occurs as a result of nonclosure of the posterior neuropore, at approximately 26 days’ gestation (failure of primary neurulation), although the mechanism of this failure is unknown. The incidence is changing, with a distinct decrease over the past 50 years or more.105 Currently, the incidence in the United States is one per 2000 live births,106,107 although it may be higher in some U.S. populations.108 Rates in the United Kingdom are also falling, but are still more than double the U.S. incidence.109 The reasons for the worldwide decline are multifactorial, but are primarily due to an increase in elective abortion rates,107,110 and the use of periconceptual folic acid. The link between folate metabolism and neural tube defects was suspected more than 30 years ago. Despite earlier studies that cast doubt on a protective effect of maternal folate intake,111 such an effect is now firmly established109,112–117; in 1996, the U.S. Food and Drug Administration mandated folate fortification of all enriched grain products. Folate supplementation may reduce the risk of concurrent and recurrent neural tube defects by up to 60% and 72%, respectively.115 The current recommendation for women of childbearing potential is 0.4 mg/day, and 4 mg/day for those with a previously affected offspring.109 Increased maternal homocysteine levels have been linked to the development of neural tube defects, and the protection afforded by folate is likely due to its effects in homocysteine metabolism. Mutations in the gene for methylene tetrahydrofolate reductase (MTHFR), a folate-dependent enzyme in the homosysteine remethylation cycle, are associated with increased risk of neural tube defects.109,118 Deficient MTHFR activity results in increased serum homocysteine levels, which can be corrected by exogenous folate supplementation. Folatedependent enzymes, including MTHFR, are involved in amino acid methylation, as well as nucleotide (DNA and RNA) synthesis; the association of these processes and neural tube formation during embryogenesis remains to be fully elucidated. In additional to folic acid deficiency, other environmental factors have been linked to neural tube defects, such as maternal use of carbemazapine and valproic acid,119 and pre-pregnancy maternal obesity.120 Diagnosis and Management The diagnosis of myelomeningocele is often made during prenatal ultrasound screening.14 Prenatal management of MMC is usually dictated by the degree of hydrocephalus noted on sonography. Although studies regarding vaginal versus cesarean delivery (as well as timing of delivery) have
W9418-11.qxd 10/31/03 3:43 PM Page 292
292
Section I
Introduction
led to conflicting conclusions,121,122 most obstetricians favor the latter, soon after lung maturity has been ensured. There is continued debate regarding the potential for injury that labor and vaginal delivery may impart upon spinal cord function. Similarly, there is controversy as to the possible damaging effect of the intrauterine environment on the dysraphic and exposed spinal cord.123 MMC is immediately apparent at birth as an open spinal defect, usually at the lumbar or sacral level; thoracic lesions account for approximately 20% of lesions, and cervical MMC is rare. A fluid-filled, epithelialized sac is typically present, although it is often flattened or ruptured by intrauterine or perinatal trauma. The dorsal spinal elements are absent or atretic. When the roof of the sac is opened, the neural placode is seen, often herniated ventrally out of the confines of the spinal canal. The placode is composed of a dysraphic spinal cord, flayed open in the dorsal midline so that the normally dorsal aspect of the cord faces ventrally. The dura is also deficient over the lesion, similarly open in such a matter that its lateral margins end in a juctional zone at the margins of the defect. Immediately after birth, the defect should be covered in saline-moistened gauze, and prophylactic antibiotics initiated. Neurosurgical assessment includes evaluation of the head circumference, and character of the fontanelle and sutures; most infants will not exhibit gross manifestations of hydrocephalus at birth, particularly if there is CSF leakage from the sac. The location and size of the defect is noted. Spinal cord function is diminished, usually absent, below the level of the dysraphic cord. Motor and sensory levels are tested by observing both spontaneous movement, as well as response to noxious stimuli. The presence of joint deformity or contracture can assist in the assessment of motor level. The anus is typically patulous due to sacral involvement. Chiari II malformation (caudal herniation of vermis and brainstem) can lead to quadriparesis, often in association with lower cranial neuropathy, although these lesions are uncommonly symptomatic at birth. Repair of the MMC is usually carried out within 48 hours of birth, and shunting can be performed concurrent with closure if signs or symptoms of hydrocephalus, or ventricular enlargement, are present.124,125 The repair consists of a multilayer closure, beginning with the restoration of the placode to a cylindrical configuration by suturing together the pial margins. Dural edges are freed from the junctional zone, undermined, and approximated around the placode. Fascial flaps are then elevated and closed, followed by subcutaneous layer and skin closure. Some authors doubt that placode closure is beneficial, although it may reduce the risk of subsequent spinal cord tethering.126 Nearly all patients with MMC have anatomic evidence of Chiari II malformation, consisting of herniation of the cerebellar vermis (and often the tonsils), caudal brainstem, and fourth ventricle through an enlarged foramen magnum, often with kinking of the medulla. The calvaria, brains,
meninges, and spines of these patients display additional abnormalities, including a small posterior fossa, dysplasia of the falx and associated interdigitation of the septum pellucidum, hypoplasia of the tentorium (often in association with cephalad displacement of the cerebellum through a widened incisura), lacunar skull (lukenschadel), and enlarged massa intermedia; the ventricles may display a colpocephalic appearance, with disproportionately enlarged occipital horns. Cervical spinal fusion or segmentation anomalies may be present. The causes of the Chiari II and other MMCassociated abnormalities are unknown, but McLone and Knepper127 theorized that CSF leakage from the MMC leads to decompression of the developing ventricular system. This disturbance removes the distension forces that normally provide impetus for brain and calvarial development. Despite its nearly universal presence in MMC, Chiari II malformations become symptomatic in only a minority of patients, approximately 10% to 20%,128–130 resulting in potentially life-threatening lower cranial nerve dysfunction (stridor, nasopharyngeal regurgitation, aspiration, apnea), quadriparesis, and cerebellar signs. In selected patients, such symptoms and signs may in part be due to dysplasia or absence of brainstem nuclei,131 but because symptoms are not typically present at birth, in most cases they are thought to be due to acquired neural compression at the foramen magnum. Before ascribing such symptoms to the Chiari malformation, it is imperative to ensure a functional shunt, because even modest ventricular or brain distension may lead to identical symptoms. The majority of malformations become symptomatic in childhood,128,130 although they can manifest later.132 Abnormal CSF flow at the foramen magnum contributes to the formation of syringomyelia and syringobulbia (discussed later in this chapter). Once symptoms of hindbrain herniation develop (and shunt function is ensured), decompression should be carried out on an urgent basis, especially in infants. This involves bony removal of the rim of the foramen magnum, and cervical laminectomies along the entire descent of the herniated hindbrain. The dura is then opened, and dissection of cerebellar tissues is carried out to identify outflow from the fourth ventricle. Due to the low position of the transverse sinuses, as well as the presence of large venous lacunes, even the dural opening can be daunting. The abnormal anatomy of the herniated structures, as well as the presence of dense adhesions between cerebellum and brainstem, can complicate the procedure. If the decompression is performed in a timely fashion after the onset of symptoms, they can be reversed, although persistent brainstem dysfunction may occur, necessitating gastrostomy and tracheostomy.128,129 The outlook for patients with MMC has improved dramatically in the past several decades. Overall long-term survival data are limited,133 although at least 75% of patients survive to adulthood.134 As a group, children with MMC are below average, but within normal limits, on many facets of intelligence testing; however, cognitive problems (especially
W9418-11.qxd 10/31/03 3:43 PM Page 293
Chapter 11
performance intelligence quotient, visual-spatial function, and mathematical ability)135–137 are not uncommon. Sixty percent to 80% of children can attend normal classes, at a grade appropriate for age.135,137 Children without shunts may have superior intellectual development compared to those with shunts,137–139 although others have noted no difference between the groups. Previous shunt infections have been reported to negatively impact intelligence,138 although others have not seen such an effect.135 Patients with MMC are usually followed in a multidisciplinary fashion, with neurosurgical, urologic, orthopedic, and rehabilitation specialists. During childhood, scarring and tethering of the placode can lead to progressive pain and functional deterioration. Increased weakness, gait changes, and scoliosis, alone or in combination, may be seen in approximately half of children with symptomatic tethering, and back or lower extremity pain occurs in approximately one third of patients.140 Progressive foot deformity and worsening bowel or bladder function may accompany these symptoms. MRI typically confirms tethering of the placode to the dorsal aspect of the thecal sac. If operative untethering is accomplished during the early onset of symptoms, improvement can be seen in most cases, especially with respect to pain. Scoliosis tends to stabilize if untethering is performed before the curvature has progressed past approximately 40 to 50 degrees.140,141 In patients with MMC and progressive scoliosis, a syrinx related to Chiari II must be ruled out as a cause; tethering can also lead to syrinx formation, although these are usually low-level cysts. Many children will require a subsequent untethering procedure due to recurrent symptoms.142 Between 1997 and 1999, two centers in the United States began reporting initial experience with in utero closure of MMC defects.143–147 The rationale for this bold undertaking was based on experimental evidence that amniotic fluid and other aspects of the intrauterine milleau exert a toxic effect on the already injured spinal cord.123,148,149 These early series suggest that in utero closure may decrease the extent of
Pediatric Neurosurgery
293
subsequent hindbrain herniation,146 as well as reduce the risk of shunt dependence150; the evidence for improvement in motor function has been extremely limited.146,150,151 The ultimate effect of this procedure on hydrocephalus, symptomatic Chiari II, and neurologic function will require careful analysis in the coming decade.
Occult Spinal Dysraphism Occult spinal dysraphism (OSD), also known as spina bifida occulta, refers to a group of lesions characterized by abnormal embryogenesis of the distal spinal cord and associated midline structures. The spectrum of OSD is vast (Table 11-3), and includes disorders of primary neurulation (dermal sinus tract and associated inclusion cyst), secondary neurulation (terminal myelocystocele, tight filum terminale), and notochord development (split cord malformations, neurenteric cyst). Primary neurulation, discussed previously in relation to MMC, is the process through which the neural tube is formed by closure of the anterior and posterior neuropore, giving rise to the upper lumbar cord and segments cephalad to this. Secondary neurulation refers to the formation of the caudal-most segments of the spinal cord and filum terminale, and occurs after 26 days’ gestation through processes of canalization and retrogressive differentiation.152 The stage at which lipomyelomeningocele, the most common occult defect, develops is unclear.153 Despite their heterogenous etiology, they are grouped together because of their skin-covered (hence occult) appearance. The presence of anorectal abnormalities is not uncommon in these patients.154 Many of these lesions may be suspected in a neonate harboring cutaneous stigmata on the back, including hemangiomas, dermal appendage (“tail”), dimple or sinus tract above the gluteal fold; or a hairy patch. Lipomatous lesions often manifest with a visible subcutaneous lump above the gluteal fold, often asymmetric in location. The presence of
Table 11-3 Occult Spinal Dysraphism Lesion
Features
• Lipomyelomeningocele
• Subcutaneous fat, extending through bifid posterior elements and dura, attaching to low-lying distal cord • Midline, epithelialized tract extending intradurally from a lumbosacral dimple; may be associated with intradural (epi)dermoid cyst • Cystic dilatation of distal spinal cord, usually with intradural lipoma; often associated with cloacal exstrophy • Thickened filum terminale, often with fatty infiltration of filum; conus usually below inferior portion of L2 • Spinal cord or conus split along the parasagittal plane; may be in a single or duplicated dural canal; usually associated with cutaneous hairy patch • Extension of endodermally derived tissue, lined with alimentary tract mucosa, into spinal canal or cord; often at cervico-thoracic junction
• Dermal sinus tract • Terminal myelocystocele • Tight filum terminale • Split cord malformation • Neurenteric cyst
W9418-11.qxd 10/31/03 3:43 PM Page 294
294
Section I
Introduction
any of these markers, or the discovery of anorectal malformation, mandates further imaging; sonography can depict the level of the conus and overt anomalies, while MRI depicts detailed anatomy of the distal spinal cord and surrounding structures. These lesions may be otherwise asymptomatic, especially in the neonatal period, with no abnormalities of neurologic function. If not detected on the basis of cutaneous findings, they usually come to attention as the child grows, producing symptoms and signs of spinal cord tethering, including back and leg pain, weakness, gait changes, bladder and bowel dysfunction, orthopedic foot changes, and progressive scoliosis.155–159 Meningitis can occur when dermal sinus tracts communicate with the intradural compartment. Lipomatous lesions of the distal spinal cord are the most common of the occult defects, and the one for which the natural history is best established. The incidence is approximately one tenth of that for MMC, and there appears to be a female preponderance in most series. Terminology for these lesions remains confusing and inconsistent in the literature. Lipomyelomeningocele is characterized by a lipomatous mass that extends from the subcutaneous plane through dysraphic posterior elements and dura, and inserts into the caudally displaced distal spinal cord. The diffentiation between lipomyelomeningocele and lipomeningocele may be a matter of semantics; the former describes such a lesion in which neural elements herniate outside the confines of the canal, and is the term most commonly used, although some think that the elements are almost never found dorsal to the canal.160 Subclassifications of lipomyelomeningocele depend on where the lipoma enters the spinal cord, and include dorsal, caudal, and transitional variants. Additionally, many lipomas are confined to the filum terminale, and may appear as a mass, or simply a thickened filum. These classification schemes, however, are inadequate to explain many of the diverse lesions described in large series.161 Lipomyelomeningoceles are often associated with normal neurologic function at birth, but because of the tethering of the spinal cord, almost invariably become symptomatic with time and growth. In a recent series, 41% of infants less than 1 year old were symptomatic at presentation; neurologic signs, urologic dysfunction on formal assessment, and orthopedic deformity were each present in approximately 60% of these symptomatic patients.157 In a series of 177 patients from two institutions, two thirds were asymptomatic when evaluated at younger than 6 months, whereas most had incurred deficits by age 4.162 Because of the near universal development and progression of deficits with time, most authors advocate surgery at diagnosis to prevent or stabilize neurologic dysfunction. Bladder dysfunction is unlikely to recover completely after it is lost, and motor signs may not improve significantly once progressive weakness occurs. Surgery involves debulking the lipomatous mass to its interface with the neural elements, thereby untethering the spinal cord, followed by a multilayered closure. In the
case of a fatty filum, the thickened filum is simply divided to release the tethering. The results of untethering procedures vary among series, and are often difficult to compare due to differences in methods and terminology. Lesions of the filum result in almost no surgical morbidity, and delayed deterioration is extremely rare; surgery for both asymptomatic and symptomatic lesions is warranted.161,163,164 For symptomatic conus lesions, surgical untethering is accepted by almost all authors to prevent progressive deficit.161,163,164 Controversy remains in regard to asymptomatic conus lesions, which can be a more formidable surgical undertaking. LaMarca and colleagues163 reported postoperative neurologic morbidity in none of 71 asymptomatic patients and in two of 87 symptomatic patients following untethering of conus lesions. Delayed worsening occurred in 12.7% of asymptomatic patients and in 41.4% of symptomatic patients following initial repair. Actuarial analysis showed that delayed worsening occurs in less than 20% of asymptomatic patients following initial repair, while 60% of those initially with symptoms deteriorate after repair; they concluded that prophylactic untethering was warranted, as have numerous other authors.164–166 In a review of 94 patients with lipomyelomeningocele, delayed retethering occurred in 20% of patients after mean followup of nearly 6 years, and was most common in transitional lesions; actuarial analysis revealed approximately 40% delayed worsening due to retethering at 8 years.167 Prophylactic surgery for conus lipomas in asymptomatic patients has recently been questioned in a detailed review and personal series of 253 conus lesions, many with extensive longterm follow-up.161 In this series, long-term deficits developed in almost half of asymptomatic patients after primary repair, not all of which could be explained by recurrent tethering. These authors and others168 could not conclude that prophylactic surgery provided an outcome superior to the natural history of the disease.
Chiari I Malformations and Syringomyelia The Chiari I malformation (CM1) is a poorly understood disorder, characterized by cerebellar tonsillar descent below the foramen magnum, in the setting of an underdeveloped posterior cranial fossa.169 The disorder is distinct from the Chiari II malformation associated with myelomeningocele, and the Chiari III malformation, which is characterized by cerebellar herniation into a cervical encephalocele. The Chiari IV malformation refers to congenital absence of the cerebellum, and is not a type of hindbrain herniation. Although the disorder usually presents in young adults, MRI has facilitated the understanding of the CM1, and the disorder is diagnosed earlier. It is now clear that CM1 is not a disease of adults, but rather a congenital disorder that may present at any time of life, with an enormous variety of
W9418-11.qxd 10/31/03 3:43 PM Page 295
Chapter 11
symptoms and signs, and an array of associated radiographic findings, including syringomyelia, scoliosis, and craniocervical anomalies. Although the incidence of CM1 in children is unknown, an MRI database search found the abnormality in 0.9% of more than 8000 pediatric patients, 43% of whom were symptomatic170; similar searches in the general population (adults and children) found rates of 0.56%171 to 0.77%,172 with 69% and 86% of patients symptomatic, respectively. Most series report a female predominance among symptomatic adults with CM1, although pediatric reports do not demonstrate this proclivity.170,173 A minority of cases are associated with familial inheritance.169 Symptoms and signs of CM1 are caused by a variety of mechanisms, including direct compression of the cervicomedullary junction and cerebellum, and abnormal CSF dynamics, the latter responsible for syrinx development and many nonspecific symptoms. Although CM1 is the result of an anatomic anomaly, the pathophysiologic mechanisms responsible for the development of neurologic dysfunction may take years or decades to manifest clinically; this may explain the large percentage of asymptomatic pediatric patients. Headache, typically suboccipital and increased with Valsalva maneuver, and often with associated neck pain, is present in the majority of adults, although it may be less frequent in younger children.174,175 Varying degrees of weakness, numbness, coordination deficits, ocular, and otoneurologic disturbances are common, although the frequency varies significantly among both adult169,176,177 and pediatric series.170,173–175,178 Presentation is uncommon within the first few years of life, although motor delay178 and disturbances of lower cranial nerve function, usually in the form of breathing disturbances, may occur.175 Unexplained crying179 and paroxysmal rage behavior180 are uncommon as the sole manifestation of CM1 in young children. Chiari malformation has been discovered in patients with SCIWORA (spinal cord injury without radiographic abnormality) injuries of the cervical spine181,182 and in cases of sudden, unexplained death. In older children and teenagers, CM1 is often discovered during workup for unexplained scoliosis175,183 and is the only reason for referral in as many as 73% of pediatric patients184; scoliosis is usually, although not always, associated with syrinx in these patients. The degree to which CM1 and scoliosis occur in patients with “idiopathic scoliosis” is unknown, and there is mixed consensus as to whether all children with this condition should undergo MRI evaluation.185–188 The traditional hallmark of CM1 is tonsillar descent below the plane of the foramen magnum. Although acquired tonsillar herniation may occur after placement of lumboperitoneal shunts189–191 or in the setting of hydrocephalus or mass lesions, these comprise only a small minority of cases. Several reports have demonstrated small posterior cranial fossa volumes in patients with the condition,169,192–194 suggesting a primary mesodermal disorder resulting in underdevelopment of occipital somites194 and subsequent
Pediatric Neurosurgery
295
overcrowding of posterior fossa contents. The pathogenesis of syrinx formation is also unknown; recent reports analyzing CSF flow and subarachnoid pressures in both CM1 patients and unaffected controls revealed that tonsillar impaction at the foramen magnum results in elevation of the cervical subarachnoid pressure, forcing CSF transmurally into the spinal cord from the subarachnoid space.195,196 Milhorat and colleagues169 theorize that abnormal CSF dynamics due to crowding at the foramen magnum not only cause syrinx formation, but may lead to the headaches and otoneurologic disturbances seen in many patients. The radiographic findings in CM1 reflect the proposed pathophysiologic mechanisms described previously. Although the classic definition of the syndrome includes tonsillar descent below the foramen magnum, usually 5 mm or more, or 3 to 5 mm in association with syrinx (present in ≥50% pediatric cases) or brainstem kinking,197 this may be too restrictive, because 9% of the 364 patients evaluated by Milhorat et al.169 with MRI did not have significant tonsillar herniation. All patients in this series, however, had an MRI appearance of hindbrain overcrowding, and all patients evaluated with cine-MRI showed decreased CSF flow at the foramen magnum. Similarly, a small series of children with syringomyelia in the absence of tonsillar herniation was recently described, in whom the posterior fossa contents were compromised.198 In these patients with the so-called Chiari 0 malformation, syrinx resolution was seen following posterior fossa decompression.199 Other cranio-cervical anomalies may be present in CM1, including Klippel-Feil anomaly and basilar invagination.169,200 The treatment options for Chiari malformation include observation, posterior fossa decompression, and shunting of syringomyelia, if present. If hydrocephalus is present, placement of a ventriculoperitoneal shunt should be performed before considering these other options. Surgery is generally recommended for children with clear and persistent symptoms related to hindbrain compression and those with neurologic deficits. Asymptomatic patients may be observed conservatively, although they (or their parents) must be counseled about symptom development, and the potential risks of contact sports. The definitive treatment for symptomatic CM1 is suboccipital craniectomy, often combined with upper cervical laminectomy, to enlarge the crowded posterior fossa. The goal of surgery is re-establishment of normal CSF flow at the foramen magnum, although the mechanisms by which this is best accomplished are controversial.201 Most surgeons open the dura, often inspecting the intradural contents to ensure CSF flow from the fourth ventricle, and close the dura with a patch graft. Alternative techniques include bony decompression, opening only the outer dural layer,202 coagulation or resection of the cerebellar tonsils,183,203 and leaving the durotomy open.184 In patients with significant ventral compression at the foramen magnum, transoral decompression may be indicated.200,204 In patients with large syrinxes, some physicians favor primary shunting of the
W9418-11.qxd 10/31/03 3:43 PM Page 296
296
Section I
Introduction
syrinx over posterior fossa decompression.205 In general, all of these procedures are successful in reducing the signs and symptoms associated with hindbrain compression and syringomyelia in most children, although no technique has proven superior to the others. Patients with the least favorable results after surgery are typically those with advanced disease at presentation. Outcomes in children may be more favorable than those in adults, possibly due to the shorter disease duration and resiliency of the pediatric nervous system.183 Progressive scoliosis, seen in a large percentage of affected children, tends to stabilize or improve after treatment.
Craniosynostosis and Abnormal Head Shape Craniosynostosis is a term used to describe a vast spectrum of malformations resulting from premature closure of a cranial suture or sutures, and the resultant abnormalities of head growth and shape. Growth of the skull is a complex phenomenon, and though our understanding of this process at the genetic and molecular level has grown appreciably in the past few decades, the process is not fully understood. In simplest terms, calvarial growth is driven by development of the underlying brain. After approximately 2 months’ gestation, centers of ossification develop within the ectomenix, a rudimentary fibrous covering of the brain. Ultimately, these centers develop into the frontal, parietal, and occipital bones, and osteogenic fronts appear at the bony margins, the site of the calvarial sutures.206–208 At a suture, bone is deposited in a spicule configuration, and a subsequent microfracture and redeposition of bone continues during brain growth to allow calvarial expansion.209 With the exception of the metopic suture, the cranial sutures never fully fuse, even through adulthood, remaining separated by a microscopic layer of fibrous tissue.207 The reasons for premature suture fusion are unknown.210 Mutations in the various subtypes of the fibroblast growth factor receptors genes are well described in some of the familial (syndromic) and in isolated forms of synostosis.211–214 Other mutations, involving the TWIST, MSX2, and other genes have been described.213,215 The site of suture formation
is in large part influenced by dura; in tissue culture, sutures transplanted in the absence of their underlying dura undergo fusion, while transplantation with dura allows continued patency.216,217 These influences have also been demonstrated in animals.218 Mechanical influences can lead to suture closure in experimental animals. The ultimate causes are likely a multifactorial response to genetic predisposition, with a complex set of physical and biochemical interactions involving the dura, calvarium, and skull base. No matter the cause, the pathogenesis of the resultant abnormal skull shape is the result not only of the fused suture, but also the compensatory growth that occurs in patent sutures.219 When growth is restricted at a suture, the adjacent plates functionally become a single bone. Compensatory bone growth occurs in a symmetric fashion at sutures in line with the fused suture, and asymmetrically at sutures along the margins of the fused segment. The compensatory growth is greatest at sutures closest to the abnormally fused suture.207,219 This expansion of Virchow’s observations of 150 years ago can explain the head shape in the various forms of synostosis. More than 90 syndromes are associated with craniosynostosis and facial dysostosis,220 although the majority of cases occur in isolation. These nonsyndromic varieties result in characteristic head shape, reflecting the suture(s) that have fused prematurely (Table 11-4). Sagittal synostosis, the most common form, accounts for nearly half of all cases,221,222 and for unclear reasons, is at least three times more prevalent in males. Growth is resticted in the lateral plane, giving rise to a long, narrowed head (scaphocephaly or dolochocephaly), with prominent frontal and occipital bossing. Intelligence is usually normal, although increased intracranial pressure rarely occurs.223 Unilateral coronal synostosis is the second most common syndrome and results in frontal plagiocephaly. The ipsilateral forehead is flattened, with contralateral frontal bossing. Because of skull base shortening of the involved side, the sphenoid wing is displaced superiorly and anteriorly, and the brow is recessed; the resultant orbital involvement leads to the “harlequin eye” seen on skull radiographs. The ipsilateral ear is displaced anteriorly, and the nasal root deviates toward the affected suture. In bicoronal synostosis (turribrachycephaly), the head has a shortened anterior-posterior dimen-
Table 11-4 Craniosynostosis Affected Suture
Findings
• Sagittal • Unilateral coronal
• Long, narrow head; occipital and frontal bossing; midline “saddle” • Ipsilateral forehead flattened with recessed brow; harlequin eye; ipsilateral ear displaced anteriorly; nasal root deviated ipsilaterally; contralateral frontal bossing • Shortened anterior-posterior distance; widened and towered head; orbital rims recessed and hypoplastic • Triangulated appearance, with palpable or visible vertical ridge over forehead; bitemporal narrowing; hypotelorism common • Rare; ipsilateral occipital flattening; ipsilateral ear displaced posteriorly; contralateral parietal bossing
• Bicoronal • Metopic • Lambdoid
W9418-11.qxd 10/31/03 3:43 PM Page 297
Chapter 11
sion, a widened mediolateral distance, and is elongated vertically, with recession and hypoplasia of the superior orbital rims. Although it commonly occurs as an isolated phenomenon, bicoronal synostosis is the most typical pattern seen in the syndromic craniofacial dysostoses, in which there is significant deformity of the midface. Metopic synostosis leads to a characteristic triangular shape to the head when viewed from above or on axial CT images (trigononcephaly). The bitemporal width is narrowed, and there is a visible or palpable ridge along the suture on the pointed forehead. Hypotelorism is seen in most cases. The most common abnormality of head shape seen in pediatric neurosurgical practice is occipital plagiocephaly, nearly always due to positional molding. The incidence has dramatically increased recently, due to recommendations of the American Academy of Pediatrics that a prone sleeping position for infants be avoided to reduce sudden infant death syndrome.224 The condition can also result from neonatal torticollis. In occipital plagiocephaly, the ipsilateral occiput is flattened, with foreward displacement of the ear and ipsilateral frontal bossing, resulting in a parallelogram head configuration. More severe cases result in anterior midface displacement on the ipsilateral side. For unknown reasons, most affected infants prefer to lie on the right occiput. Benign extra-axial fluid collections are quite common in patients with positional molding, and this may render the skull more vulnerable to compressive forces.225 The deformation usually peaks by 5 to 6 months, when most infants are rolling over, and in general spending less time in the supine position. Unilateral lambdoid synostosis is quite rare, accounting for approximately 3% of craniosynostosis; in contrast to positional molding, the ipsilateal ear in true synostosis is displaced posteriorly and somewhat inferiorly, and bossing occurs most prominently in the contralateral parietal region, although ipsilateral frontal bossing can be seen. Crouzon’s syndrome is the most common craniofacial syndrome, occurring in approximately one in 25,000 births.226 It is familial in one half to two thirds of patients, with autosomal dominant inheritance.227 The disorder is characterized by bicoronal synostosis and resultant turribrachycephaly, proptosis and orbital hypoplasia, and midface anomalies characterized by hypoplasia and retrusion of the maxilla. Progressive sutural fusion at the skull base can occur during the first several years of life.228 Apert’s syndrome is less common, with a prevalence of one in 55,000 to 65,000 births,227,229 and is usually sporadic, although an autosomal dominant inheritance has been noted. The craniofacial features of Apert’s syndrome are quite complex, and involve bicoronal synostosis, as well as aberrant metopic, sagittal, anterior skull base sutures, and mid-face hypoplasia. In infancy, there is a large defect of the midline calvarium, extending from the nasion to the posterior fontanelle, which gradually fills with bone during the first several
Pediatric Neurosurgery
297
years.230 Syndactyly of the hands and feet are part of the syndrome, and multiple apendicular joints can be affected.227 The majority of these children are of subnormal intelligence.231 The relationship between craniosynostosis and hydrocephalus is not well understood. In nonsyndromic, singlesuture involvement, increased ICP is seen in a minority of patients, and is not clinically evident in most cases.223 In syndromic patients, abnormal CSF dynamics can result from calvarial vault restriction, veno-occlusive skull base involvement, or a combination of these. In a retrospective review of 1727 cases of craniosynostosis, nearly 30% of patients with Crouzan’s syndrome and 12% of patients with Apert’s syndrome required CSF diversion; in the latter group, asymptomatic ventricular dilation was seen in half of the cases.228 In a series of 22 syndromic patients observed prospectively from diagnosis, five patients required diversion before or just after initial reconstruction for papilledema or macrocephaly associated with ventriculomegaly; in all cases, papilledema resolved after reconstruction and shunting.232 Increased ICP developed in eight patients 12 to 37 months after primary repair, requiring either shunt, or if head size was less than the 25th percentile, re-expansion. In this report, asymptomatic papilledema was the only sign of delayed ICP elevation in four cases. The treatment of craniosynotosis depends on the involved suture(s) and the severity of the deformity. Patients with metopic or coronal synostosis typically undergo calvarial and orbital reconstruction within the first 6 to 12 months of life; those with syndromic involvement also require cranio-orbital reconstruction, and may require subsequent re-expansions, as well as delayed midface reconstructive procedures. Significant controversy remains concerning the optimal treatment for the most common entity, isolated sagittal synostosis. The dozens of published procedures can be categorized into three basic groups: strip craniectomy (without or with interposition material), extended strip craniectomy, and calvarial vault remodeling. The extended strip craniectomy techniques usually involve removal of the sagittal suture, combined with wedge, barrelstave, or other types of parietal and occipital osteotomy. These procedures typically minimize blood loss, are relatively fast and technically simple, minimize hospital stay, and provide a satisfactory outcome in the vast majority of cases.233,234 Proponents of primary calvarial reconstruction argue that remodeling procedures optimize rapid correction of head shape, do not depend on continued skull growth to achieve adequate cosmesis, minimize risk of premature restenosis and the need for subsequent procedures, and provide outcomes superior to those with strip or extended strip procedures based on anthropomorphic measurements.235–237 Recently, a technique of endoscopic-assisted strip craniectomy, combined with subsequent cranial remodeling orthosis, was described for sagittal synostosis.238,239 The authors demonstrated satisfactory early results
W9418-11.qxd 10/31/03 3:43 PM Page 298
298
Section I
Introduction
for head shape, less blood loss, shorter operating time, and greatly reduced length of hospitalization compared to more invasive procedures. Until prospective studies are performed comparing the various techniques, controlling for patient age and preoperative cranial shape, the procedure of choice is unknown.
Encephalocele Encephalocele (cephalocele) refers to a skin-covered defect in which there is herniation of brain and meninges through a defect in the calvarium or skull base. The term meningocele is used when there is no evidence of parenchyma within the defect. These lesions are classified according to the site of the defect and are broadly categorized into basal, anterior (sincipital), posterior, and convexity encephaloceles,240 although basal herniations may be categorized as a subtype of anterior lesions.241 Basal lesions primarily arise through defects in the sphenoid or ethmoid bones at the skull base.240 Sincipital encephaloceles, which arise in the region of the foramen cecum, can be further subdivided into frontoethmoidal (nasofrontal, naso-ethmoidal, and naso-orbital subtypes), and the uncommon interfrontal and craniofacial cleft subtypes.242 Posterior encephaloceles result from herniation through an occipital defect, and may be above or below the torcula.243 Convexity encephaloceles arise from herniations along the vertex.240 The overall incidence of these lesions is approximately 1 in 5000 live births,240,244 but this may be an underestimate due to fetal loss.245 In North America and Western cultures, occipital encephaloceles are by far the most common entity, accounting for approximately 90% of cases,243 with a female predominance.246 Anterior encephaloceles are the predominant lesions in Southeast Asia, with an incidence of approximately one in 5000 live births.247 Encephaloceles in other locations are less common. The embryologic bases of these lesions is unknown, and given the diversity of locations in which they occur, is likely multifactorial. A disorder of primary neurulation (akin to myelomeningocele defects) is unlikely, given the skin-covered nature of these lesions. It is more tenable that they are due to a disruption in the mesenchymal structures overlying the brain, with resultant herniation through the affected area.244,245 Encephaloceles typically occur in isolation, but may be associated with a variety of nervous system and systemic abnormalities.240 If large enough, encephaloceles can be documented on prenatal sonography, and maternal alpha-fetoprotein levels may be increased. The diagnosis is usually straightforward upon delivery, however, with an outpouching noted over the occipital or suboccipital region for posterior lesions. Anterior encephaloceles are characterized by a midline or paramedian mass near the nasion and can cause proptosis and hypertelorism. Naso-ethmoidal lesions may not be readily apparent at birth, but may present later with symptoms
of nasopharyngeal obstruction or CSF rhinorrhea. Basal encephaloceles may present with a midline mass, or in a delayed fashion, with obstructive symptoms or CSF leak. MRI is the most accurate means of delineating the contents of the sac and assessing the presence of associated anomalies, while CT is useful for characterizing the relationship of the herniation to skull base structures in anterior encephaloceles. Occipital lesions are often associated with anomalous venous sinus anatomy, and rarely these veins may be part of the sac contents; magnetic resonance venography can be used to examine the relationship of the herniation with respect to the venous system.248,249 The principles of treatment for encephaloceles include exploration of sac contents, with preservation of the normal brain present within the sac, if technically feasible. Anterior and basilar lesions may involve craniofacial reconstruction concurrent with or after the primary repair; the sac is typically explored both intradurally and extradurally to delineate clearly its anatomy and contents. Anterior defects usually contain atretic and nonfunctional frontal lobe tissue, which can be resected, while hypothalamic or chiasmal elements may be present in basal lesions. The goal for anterior and basal repairs is resection of the sac, and a watertight dural closure. For occipital and other convexity defects, the sac is explored and amputated, followed by dural closure. Cranioplasty is usually not needed for lesions with small necks. Because of the venting of CSF into the sac during development, hydrocephalus develops in a significant number of patients after closure of large occipital encephaloceles, which ultimately may require shunt placement. The long-term outcome is more favorable for patients with lesions containing very little or no brain tissue, and for anterior versus occipital defects. Those with underlying associated brain anomalies, significant amounts of brain herniation, or critical structures within the sac, generally have significant deficits. Most patients with anterior encephaloceles have normal or near-normal intelligence,241 while the majority of those with occipital lesions have persistent deficits, often secondary to seizures, hydrocephalus, or visual impairment.240,250
Intracranial Arachnoid Cysts Intracranial arachnoid cysts are thought to be congenital lesions that arise from anomalous splitting of the arachnoid membrane during embryogenesis.251 They typically present during early childhood, with signs and symptoms depending on their location, size, and age of the patient. In modern series, the most common location is the middle fossa (sylvian fissure), representing nearly one half 252 to two thirds253 of cases, followed by posterior fossa and supracellar locations, approximately 25% and 10%, respectively. The prevalence of these lesions is unknown because they often remain asymptomatic, but there is no question
W9418-11.qxd 10/31/03 3:43 PM Page 299
Chapter 11
Pediatric Neurosurgery
299
Figure 11-1. T2-weighted axial MR images of type I and type II arachnoid cysts. A, Type I cysts are small lesions located behind the sphenoid wing. B, Type II cysts are larger, splaying apart the sylvian fissure.
that the reported prevalence has increased with use of CT and MRI. Arachnoid cysts may be incidental findings, especially in adults, but are usually discovered during childhood for specific complaints. In 95 children presenting in the era of modern imaging, the average age at the time of surgery was 4.9 years254; the median age of 2.2 years noted in another series of 40 patients suggest that most come to attention at an earlier age.255 More than one half of patients presented during their first year of life in a report of 67 pediatric cases.256 A male predominance of approximately 65% has been noted in large series.254,257 Overall, the most common symptoms and signs are related to increased ICP or abnormal head growth, seen in more than one half of all cases.254,256,257 In series of middle fossa cysts, seizures were present in approximately 30% to 40% of patients.253,258 Posterior fossa lesions most commonly present with headache and vomiting; cerebellar and cranial nerve dysfunction are less common.259 Suprasellar lesions can produce ventricular obstruction, visual dysfunction, and endocrine disturbances, including growth disturbance and precocious puberty.255,260 Based upon CT appearance, Galassi261 divided middle fossa cysts into types I through III, depending on size and morphology. Type I cysts are small and spindle-shaped, located just behind the sphenoid at the temporal pole, and without mass effect on the ventricles or midline. Type II cysts
are moderate in size and have a triangular or quadrangular appearance, occupying the anterior and middle part of the fossa. These cysts splay apart the sylvian fissure, and may cause shift. Type III cysts are large, round or oval lesions occupying the entire fossa, causing significant compression of frontal and parietal lobes and pronounced midline shift. MRI is now the diagnostic study of choice for characterizing cyst location and size, and for preoperative planning (Fig. 11-1). The management options for arachnoid cysts in children include observation (no treatment), open resection and fenestration, cystoperitoneal shunt, ventriculoperitoneal shunt, and endoscopic fenestration. The optimal strategy is not known, because there have been no randomized analyses comparing the two most common techniques (fenestration and cystoperitoneal shunt), and most series have been weighted heavily toward one procedure. Treatment decisions are based on the presentation, cyst location, the presence of hydrocephalus, and the surgeon’s preference or experience. The natural history of arachnoid cysts is not well understood; the mechanisms by which they can grow include a “slit-valve” mechanism, in which CSF can enter but not readily escape,262 cyst wall production of fluid, and osmotic gradients favoring fluid uptake.240,263 There is no question that cysts can enlarge with time, although they often remain quiescent throughout life, and can, uncommonly, resolve without treatment.263–265 Observation and nonoperative
W9418-11.qxd 10/31/03 3:43 PM Page 300
300
Section I
Introduction
treatment are the initial management for small, incidentally discovered cysts. To ensure stability, repeat imaging should be undertaken within the first 6 to 12 months following diagnosis, and then on a regular schedule for at least several years, while monitoring for the development of symptoms or signs. For any lesion other than a very small type 1 cyst, contact sports should be avoided, due to the known risk of subdural hemorrhage or bleeding into the cyst.266–269 Patients presenting with atypical headaches, in whom a small arachnoid cyst is discovered during a workup, may be treated conservatively initially; if symptoms progress, or if there is any change in cyst size, treatment is probably warranted. Surgery is recommended for patients harboring clearly symptomatic cysts, especially related to mass effect or increased ICP. In young children with asymptomatic cysts causing pronounced mass effect, treatment is indicated to reduce the possibility of deleterious effects with development. Proponents of fenestration into the basal cisterns as a primary treatment option note not only the high success rates reported in the literature, but also the avoidance of shunt complication and dependence. The success rate for alleviating symptoms and obviating future procedures was 79% in a recent series of 44 patients with middle fossa cysts treated by fenestration.253 Fewel and colleagues254 noted a 73% success rate with fenestration alone among patients without hydrocephalus, although one half of all patients in the series treated with fenestration required subsequent shunt placement. Conversely, in a series of 77 middle fossa cysts, all were effectively treated with cystoperitoneal shunts, although shunt malfunctions developed in eight patients during the 7.7-year mean follow-up period.258 In a report of 40 pediatric patients, fenestration was efficacious in only 20% of 15 patients so treated, while cystoperitoneal shunting was successful in all 20 patients in whom it was attempted, with six patients (none with middle fossa cysts) requiring subsequent shunt revision.255 For cysts located in the cerebellopontine angle, open excision and fenestration is probably the most consistently effective treatment,259 and obviates shunt placement near critical neural and vascular structures. The use of neuroendoscopy for cyst exploration and fenestration has been described in a limited number of cases. This technique offers the advantages described for open craniotomy, and avoids the potential morbidity associated with shunt placement. Choi and colleagues270 reported successful outcomes in all 36 patients (31 children) undergoing endoscopic fenestration, with minimal morbidity. A smaller series, comprised mostly of adults, showed favorable outcomes in six of seven cases.271
Dandy-Walker Complex The Dandy-Walker complex refers to a spectrum of anomalies typified by varying degrees of cerebellar hypoplasia and cystic dilatation of the fourth ventricle.272 Limitations in
imaging in the pre-MRI era led to difficulty in making distinctions among patients with posterior fossa cystic fluid collections, and a variety of terms were used to categorize the variations seen, often creating confusion. Barkovitch and colleagues272 studied patients with multiplanar MRI, and concluded that what had historically been referred to as mega cisterna magna, Dandy-Walker malformation, and Dandy-Walker variant were actually similar disorders along a continuum, manifested by posterior fossa fluid collections directly communicating with the fourth ventricle, and no evidence of cerebellar atrophy. They suggested the term Dandy-Walker complex (DWC), and divided patients into DWC type A (vermis absent on axial views at the level of the fourth ventricle, and varying degrees of vermian aplasia or hypoplasia or vermian rotation seen on sagittal views), and DWC type B (vermis present on axial views, with little or no hypoplasia on sagittal views, with clear communication between cyst and fourth ventricle)272; the term “prominent cisterna magna” was used to describe patients with enlagement of the cistern due to cerebellar atrophy. In their analysis, patients with DWC type A were those who would have classically been referred to as having DandyWalker malformation or variant, and nearly all patients (11/12) had hydrocephalus. Enlargement of the posterior fossa and elevation of the tentorium and transverse sinuses was seen in most patients with DWC type A, and other brain anomalies, such as dysgenesis of the corpus callosum, were seen in more than one half (Fig. 11-2). Patients with DWC type B, who likely would have been categorized in the past as having mega-cisterna magna, also had hydrocephalus and associated anomalies, but less frequently. This classification scheme is useful in that it unifies radiologic criteria for the evaluation of treatment of these patients,273 although others have separated cystic malformations of the posterior fossa based on variations in cerebellar development.274 The cause of DWC is unknown; theories had focused on atresia of the fourth ventricular outflow foramina, although a subsequent study has shown these foramina to be patent in some patients with DWC. The etiology is likely more complex and multifactorial, and due to development anomalies of the anterior membranous area occurring during formation of the cerebellum and fourth ventricle.272,274 In addition to brain anomalies, such as callosal dysgenesis and occipital meningocele, DWC has also been associated with widespread malformations, especially cardiac defects.275–277 The true incidence of DWC is unknown.278 Patients with DWC typically present in childhood because of signs and symptoms of increased ICP, either due to mass effect from the cyst itself, or from the frequent associated hydrocephalus (pan-ventricular); it is now being diagnosed earlier, and at greater frequency, because of prenatal imaging.276,279 Patients with DWC type A usually present in infancy (mean, 0.4 years), while DWC type B manifests later (mean, 4.2 years).273 In cases in which hydrocephalus is absent, the
W9418-11.qxd 10/31/03 3:43 PM Page 301
Chapter 11
Pediatric Neurosurgery
301
Figure 11-2. T2-weighted MR images of Dandy-Walker complex, type A. A, Sagittal. There is complete absence of the cerebellar vermis, and hydrocephalus is present. B, Axial. Dysgenesis of the corpus callosum is depicted.
malformation may be discovered during workup for motor delay, intellectual impairment, or seizures. Surgery, in the form of CSF diversion, is indicated for patients who present with hydrocephalus, increased ICP, or symptoms or signs related to the mass effect from the posterior fossa cyst. Ventriculoperitoneal, cystoperitoneal, and ventriculocystoperitoneal shunts can all be used, and there is no concensus as to the preferred method; patency of the aqueduct should be assessed on the initial MRI studies to help in preoperative planning. Domingo and Peter273 found that cystoperitoneal shunts had lower rates of obstruction, and both ventricular and cyst size were reduced more effectively compared to ventriculoperitoneal shunts, although cystoperitoneal shunts had a high likelihood of poor shunt position; overall complication rates were high for both groups, with no statistical difference. Asai and colleagues280 noted that an isolated fourth ventricle developed in more than 40% of patients treated with a primary ventriculoperitoneal shunt; this was thought to be secondary to acquired aqueductal stenosis, necessitating additional cyst shunting, compared to only 10% of those treated initially with a cystoperitoneal shunt; cases of upward herniation of the cyst after ventriculoperitoneal shunting alone have been
reported.281 Bindal and colleagues277 noted conversion rates from single to double shunt systems in 42% of primary cystoperitoneal shunts and 30% of ventriculoperitoneal shunts, implying that each method is fraught with the problem of malcommunication between supratentorial and infratentorial compartments; primary shunting of both compartments is advocated in an effort to minimize this risk, and to avoid pressure differentials across the tentorium.282,283 The outlook for DWC is now improving, with advances in imaging and better understanding of the shunt requirements necessary for these patients. Approximately one half of patients are normal or mildly delayed intellectually,284,285 and one half have normal cerebellar function. There is no association between cerebellar appearance on imaging studies and ultimate functional outcome.284
Head and Spinal Injuries Head Injury The physiologic and clinical aspects of severe traumatic brain injury (TBI) are covered elsewhere in this text, and will
W9418-11.qxd 10/31/03 3:43 PM Page 302
302
Section I
Introduction
not be detailed in this section. Nonetheless, some important differences in etiology, pathophysiology, management, and outcome have been noted for the pediatric age group, and will be highlighted in the present section. In the United States, TBI is the leading cause of death in children, and is the major factor impacting outcome in the injured child.286 Each year, there are at approximately 100,000 to 150,000 hospital admissions related to head injury in the United States for children younger than 15,287,288 with the vast majority of these admissions being mild in nature. After the first year of life, injury kills more children than all diseases combined.289 At least 7000 children die each year as a result of head injury, representing 30% of all accidental deaths in children.288 For all ages, with the possible exception of infancy, motor vehicle accidents are the most common cause of serious brain injury and mortality.288 In children younger than 5, falls are the most prevalent overall cause of head injury, while recreation-type injuries (including bicycle accidents) are the most common cause in the 10- to 14-year-old age group; after age 15, motor vehicle accidents are the primary source of injury.288 After age 5, males are far more likely than females to suffer TBI.288 Nonaccidental trauma (NAT) is becoming recognized with greater frequency as a major cause of brain injury in children, and NAT-related brain injury may be the most common cause of trauma-related death in infancy.290,291 The evaluation of older children and adolescents with head injury is relatively straightforward, encompassing a routine general physical and neurologic examination, and assessment of level of consciousness; the most widely used index for this population is the Glascow Coma Scale (GCS), which evaluates eye opening, verbal response, and motor response.292 For the young child, this scale is not reproducible and often impractical, given their limited communications skills, and a variety of pediatric coma scales have been developed to circumvent shortcomings inherent in the GCS.293–295 CT should be routine in any child with alteration of consciousness, significant mechanism (e.g., patient ejected from vehicle, crash at speed >40 miles per hour, death in the passenger compartment), external signs of head trauma, or skull fracture noted on plain radiographs. The incidence of specific clinical and radiographic findings after head injury is difficult to ascertain because reported series vary widely with respect to definition of head or brain injury, inclusion criteria, and population base. In a series of more than 9000 consecutive children evaluated in an emergency department after any type of head injury, most (86%) did not suffer loss of consciousness, although symptoms and signs of brain injury (such as vomiting or alteration of consciousness) were common, seen on greater than one third of patients requiring admission after injury.296 In children younger than 2 years, skull fractures were present in 45% of those admitted following head injury, while only 11% of those 2 years and older had such fractures.296 Depressed skull fractures account for up to one fourth of
pediatric skull fractures,297 and are far more prevalent in children than adults after head injury. The presence of significant mass lesions such as epidural hematoma, subdural hematoma, and intracranial clot is less common in children compared to adults. In a series of 1906 consecutive children with injury resulting in skull fracture, loss of consciousness, or amnesia, the incidence of subdural and epidural hematoma was 2% and 1%, respectively,298 although other series have shown rates of pediatric subdural and epidural hematomata as high as 18% each after mild head injury (GCS 13–15).299 Severe head injury—defined at GCS £ 8—represents a minority of cases, but is the greatest source of mortality and acquired long-term morbidity in children. The brain sustains insult not only at the moment of impact (primary injury), but also as the result of secondary regenerative cascades mediated by ischemia, cellular energy failure, excitatory amino acid release, calcium fluxes, free radical formation, and resultant cellular death, a process termed excitotoxicity.300 Programmed cell death (apoptosis), also induced by trauma, is characterized by condensation of nuclear and cytoplasmic components, cell shrinkage, fragmentation, and degradation301; this process is ultrastructurally distinct from excitotoxic death. Secondary injury may progress in the hours and days following trauma, and contribute to brain swelling and delayed deterioration.302–304 Systemic insults after the initial trauma, such as shock and hypoxia, deprive the brain of substrates necessary for normal oxidative metabolism, further aggravating these cyclic cascades, and contributing to neuronal death. Injury to the developing brain differs from that of adults, although the specific reasons are unknown.305 In immature rat models of brain injury, apoptotic cell death occurs in a delayed fashion, with the most significant damage occurring in the youngest pups; mature rats did not demonstrate apoptotic death after identical injury, suggesting a particular age-dependent vulnerability to injury.301 An understanding of the effects of brain injury on cerebral perfusion and ICP, autoregulation, compliance, and metabolism is important when devising effective treatment strategies; these parameters are discussed elsewhere in this text and in several excellent recent reviews.306,307 In landmark reports on pediatric brain injury by Bruce and colleagues, in which the GCS was first used to assess children,308,309 several points were demonstrated, including the overall low incidence of mortality in children following traumatic brain injury, and the relatively high incidence of diffuse brain edema. In the relatively short time since that publication, there have been numerous series in the literature highlighting the pathophysiologic differences between children and adults after severe brain injury, and the implication of these differences with respect to management and outcome; these reports have resulted in significant controversy, and many questions remain unanswered.310–314 Head-injured children do appear to be more prone to diffuse cerebral swelling than
W9418-11.qxd 10/31/03 3:43 PM Page 303
Chapter 11
adults,315–317 although the mechanisms by which this occurs and the ultimate consequences are unknown. The assumption that hyperemic cerebral blood flow was responsible for this swelling, with hyperventilation advocated for its treatment315 is being reconsidered. Recent studies of cerebral perfusion in normal children313,318 and in children with brain injury319,320 have shown a tremendous variation in cerebral blood flow in both normal and injured children, and further work is necessary to clarify the issue. In light of our lack of understanding of the age-related differences in pathophysiologic response to TBI, “there is no reason to treat children differently from adults.”313 Treatment of severe brain injury is in large part that of ICP control, to preserve cerebral perfusion and the metabolic demands of the injured brain. The tenets of aggressive resuscitation, avoidance of systemic hypotension and hypoxia, and control of ICP and maintenance of adequate cerebral perfusion are applicable to all age groups and are discussed in another chapter of this text. For children with refractory intracranial hypertension, several “nonstandard” treatment methods have been advocated. Lumbar cistern CSF drainage has been advocated as a means of ICP control in the refractory patient,321,322 although its efficacy has not been demonstrated in large retrospective or prospective studies. Decompressive craniectomy as a means of controlling refractory ICP remains controversial, and is not routinely performed; several retrospective reports323–326 as well as small prospective trials327,328 have suggested its efficacy in children and younger patients, if performed before the onset of irreversible cerebral ischemia. The largest series on severe pediatric brain injury have differed with respect to patient age, timing of assessments, and definitions regarding severity of injury.312 In a prospective study involving analysis of over 8000 head-injured patients of all severity, the mortality rate for children with GCS 3 to 8 was 28.4%, significantly better than the adult rate of 47.7%.298 Within the pediatric population, mortality was highest in infants and toddlers, with declining rates throughout the remainder of childhood until mid-adolescence.298 Most large series have shown similar mortality rates in severely injured children, approximately 25% to 35%.312,329–332 Not surprisingly, multiple risk factors such as low initial GCS (particularly motor score)289; diffuse cerebral swelling, especially when combined with hemorrhage333,334; and low mean cerebral perfusion pressure332 have all been associated with poor outcome. The reasons for the improved survival in children compared to adults, and the seemingly worse outlook for critically injured infants, remain unknown. In addition to mortality, sustained neurologic deficits, especially cognitive impairment, are exceedingly common in children after severe brain injury.333–338 Intentionally inflicted trauma is a common cause of pediatric brain injury. In one series, at least 24% of such injuries in 100 consecutive patients younger than 2 years old were the result of child abuse.339 In some centers, nonaccidental brain
Pediatric Neurosurgery
303
injury is the most common cause of pediatric traumatic death,290 and an estimated one in 4000 children may suffer nonaccidental head injury during the first year of life.340 Such injuries are the result of forceful shaking, often, but not invariable, in conjunction with blunt trauma, in a patient whose body habitus renders him or her particularly vulnerable.341–345 The average victim age in a series of 48 patients was 7.85 months, with none older than 3 years of age.343 The term “shaken impact syndrome” was coined by Bruce and Zimmerman to describe the constellation of clinical and radiographic findings in abused infants.345 Because of vague circumstances surrounding shaken impact syndrome, abuse is typically inferred from examination and findings on cranial and skeletal imaging. The diagnosis has no strict criteria, but is usually made in the presence of extra-axial hemorrhage, retinal hemorrhage, external signs of trauma, and a vague or incompatible mechanism of injury. Typical scenarios provided by parents usually involve a fall or other minor trauma that is not compatible with the patient’s injury. Although level of consciousness may be normal in the most mild cases, the patient is often significantly depressed, with coma or posturing present in approximately one half of cases.346 Apneic events, inconsolability and irritability, and seizures are frequent, with a tense fontanelle present in severe cases. External evidence of head trauma, such as bruising or soft tissue swelling, is present in less than one half of patients.343,344 Retinal hemorrhages, often with preretinal and subretinal components, are present in the majority of patients,347 and are extremely rare after accidental trauma339 or cardiopulmonary rescuscitation348; retinal folds may be found in more severe cases.349 Radiographic findings, typically on CT, may rarely be the sole manifestation of intentional trauma. Acute subdural hemorrhage is the most common finding, seen in more than 80% of cases,350 often interhemispheric in location.351 Chronic subdural collections, with or without an acute component, as well as subarachnoid blood may be present. Diffuse edema on initial imaging is present in a significant percentage of patients, and may indicate prolonged apnea,352 or a prolonged delay in presentation after injury. Skull fractures may be present in one fourth of cases,343 while long bone and rib fractures are less prevalent.344 Treatment of inflicted brain injury is no different than that of accidental trauma. Outcome after nontraumatic brain injury depends on severity, but is associated with mortality rates as high as 27% to 37%.343,344 At autopsy, diffuse brain injury is universal, and there is often concomitant damage to the cervical spinal cord,341,353 which may contribute to hypoxic brain injury. Long-term dysfunction is common in survivors.354 Injuries of the Spine Spinal column and cord trauma in children differs from that sustained by adults with respect to frequency, mechanism,
W9418-11.qxd 10/31/03 3:43 PM Page 304
304
Section I
Introduction
levels involved, radiographic findings, and outcome. Pediatric spine and spinal cord injuries are approximately 20 times less common than severe head trauma.355 In large referral centers, pediatric spinal cord and spinal column injuries, including fracture, fracture and subluxation, subluxation only, and spinal cord injury without radiographic abnormalities (SCIWORA)356 are approximately 20 times less frequent than such injuries in adults,357,358 although the incidence rises throughout childhood and abruptly beyond the teenage years. The reasons for the rarity of spine trauma in children relate both to causative and anatomic factors. Young children are less likely to participate in many of the activities associated with spinal injury, and are more likely to be restrained and in the rear seat of vehicles, rendering them less vulnerable to severe impact injury than adults. The ligamentous structures of children are more elastic than those of adults, and their facets are oriented in a more horizontal configuration, allowing greater deformity without resultant fracture or dislocation.359 Their disproportionately larger head results in different inertial forces dissipated to the neck during deceleration.360 Such anatomic differences may also explain the predisposition for relatively unique pediatric injuries, such as atlantoaxial rotatory subluxation361 and occipito-atlantal dislocation. In young children, the most common causes of spinal injury are falls and pedestrian versus car accidents,358,360 although motor vehicle accidents are the primary cause in some series.362,363 The injuries sustained by older children and teenagers are due to motor vehicle accidents, followed by sports-related mishaps, with the former assuming greater importance with increasing age.357,358,362 Except in the youngest age groups, there is a strong male predominance. Half of pediatric patients with spinal column injuries are neurologically normal (by definition, none with SCIWORA are normal); among patients with deficit, incomplete injury is slightly more common than complete impairment.357,358,362 The cervical region is by far the most frequently affected level in children, with no significant difference between the thoracic, thoracolumbar, and lumbar regions.358,364 Associated head injury is present in approximately one third to one half of cervical spine injuries, especially those involving the upper cervical region.365 Noncontiguous injuries are present in approximately 10% to 15% of children. In young children, the upper cervical spine (occiput to C3) is disproportionately affected,357,360,362,363,365 with lower cervical levels increasingly affected throughout later childhood and beyond. SCIWORA is most common in younger patients (birth to 9 years), representing more than one third of spinal injuries in this group, and becomes less prevalent with advancing age.356–358,362 SCIWORA is associated with greater neurologic deficit in young patients, while older children and teenagers with SCIWORA suffer less severe impairment.366–368 Fractures (with or without subluxation) are common at all ages, but assume greater predominance in older children and
teenagers. Because of the frequency of SCIWORA in young patients, this population is more prone to neurologic deficit, while patients with fracture only, who tend to be older, are at lowest risk for neurologic injury. Interestingly, delayed neurologic deficit may develop in approximately one fourth358,360,367 of patients with SCIWORA, with onset of signs and symptoms hours to several days after injury, although others have not seen this phenomenon.357 The signs and symptoms of spine and spinal cord injuries are variable, and primarily depend on the age and level of consciousness of the patient. In an awake, cooperative patient, pain and spasm at the level of injury is nearly universal; in cervical injury, neck movement is voluntarily restricted, and the head may assume an abnormal posture in cases of subluxation. Neurologic findings range from no deficit to complete loss of sensory, motor, and autonomic function, and may change early in the course, as noted previously. Severe spinal cord injury, especially in the cervical region, may be associated with unexplained hypotension without compensatory tachycardia. Standard radiography remains invaluable in the evaluation of pediatric spine injury, and should be performed in any child with neck or back pain following injury, and in all patients with neurologic deficit or depressed mental status. In the cervical region, anatomic variants in children include pseudosubluxation of the upper cervical vertebrae, increased atlantodental interval, bifid arch of C1, and the anteriorly “wedged” vertebral bodies.369 Normal synchondroses, especially at C1 and C2, appear lucent and may be misdiagnosed as fractures. Flexion and extension cervical views may be obtained under supervision if pain does not prohibit active range of motion, and can indicate areas of hypermobility. CT imaging, often with coronal and sagittal reconstruction, is useful for evaluating areas not well visualized on plain radiographs, and is essential for further depicting osseous injury and subluxation. MRI is necessary in all children with suspected spinal cord or root injury, and in patients with any unexplained neurologic deficit. SCIWORA was defined before the advent of MRI; such imaging may reveal areas of ligamentous disruption, spinal cord edema, hemorrhage, or transection.369 The principles of treatment for children with spinal cord or column injuries do not differ in theory from adults: prevention of further injury by immediate immobilization of unstable segments, decompression of neural elements if necessary, and maintenance of long-term stability. All children with suspected spinal injuries must be immobilized with a hard collar during transport and evaluation. Because of their relatively large head, young children placed directly on a rigid transport board will be forced into a relatively flexed position; their back and shoulders must be “built up” to avoid this posture.370,371 Obviously, attention must be given to airway management and hemodynamic stability in the acute setting, especially in cases of cervical cord injury. Highdose methylprednisolone is usually initiated when neuro-
W9418-11.qxd 10/31/03 3:43 PM Page 305
Chapter 11
logic deficit is present, although its efficacy in the pediatric population is not known; the recent National Spinal Cord Injury Studies372,373 did not include children younger than 13 years, and results for older children were not separately analyzed. Problems inherent in the management of pediatric spine injuries, including skeletal immaturity, future growth potential, and relative difficulty with both internal and external fixation techniques, complicate treatment decisions in children. In several ways, however, young children with spinal column injuries are at a relative advantage compared to adults. Both ligamentous and bony injuries may experience degrees of healing not seen in adults and the presence of such injuries does not always mandate surgical fusion; less than one third of spinal column injuries in children require surgical stabilization,357,358,365 with the majority managed successfully with prolonged external immobilization. Indications for operative intervention include significant instability, irreducible subluxation, spinal cord compression with persistent or worsening deficit, and failure of external immobilization. The overall rarity of such cases mandates an individualized approach in the formulation of a surgical plan, taking into account the patient’s age, level and type of injury, degree of deficit, and other injuries. The outcome of children with spinal column and cord injuries is variable. Hamilton and Myles358 reported a mortality rate of 28% among children with spinal injury, more than double that of adults; 45% of cord-injured children in their report died, usually at the scene. Although death could not be definitively attributed to spinal column or cord injuries in many of these patients, the incidence of cervical spine injury, particularly occipital to C2, was significantly overrepresented in the group of children who died. The prognosis for children surviving spinal injury is generally favorable, and is directly related to the severity of the initial injury. Because of the tremendous healing capacity of the pediatric spine, long-term failure of either operative or nonoperative spinal stabilization is the exception. In all large series, the vast majority of patients with complete neurological deficit remain significantly impaired at long-term follow-up, while most patients with partial impairment recover useful function.357,358,362 In a recent report of 102 pediatric cervical injuries, 83% of the 46 patients with incomplete injury returned to normal.365
Neoplasms Brain tumors are the most common solid neoplasms affecting children, second only to leukemia when all tumor types are considered. The annual incidence of pediatric brain tumors is approximately 32.5 cases per million children.374 A recent survey revealed that despite the declining mortality rates for pediatric brain tumors, they now represent the most common cause of tumor-related death in children.375 The incidence of brain tumors in children appears to have
Pediatric Neurosurgery
305
increased over the past several decades, a jump which cannot be fully explained by advances in imaging.376 Overall, the difference in incidence between supratentorial and infratentorial lesions in children is not profound377; when children younger than 15 years of age are considered, there is a slight predominance of infratentorial (53.5%) versus supratentorial (46.5%) lesions,374 although in infancy the latter are far more common.378,379 Tumors of glial origin are the most common histologic type overall, and although several subtypes exist, the majority are benign lesions, unlike those of the adult population. Medulloblastoma is the most common malignant pediatric brain tumor,374,380 and is exceedingly rare in adults. Brain tumors are discussed elsewhere in this text, and a detailed survey of pediatric neoplasms is beyond the scope of this chapter. Rather, several of the more common brain tumors in children will be described. Medulloblastoma Medulloblastoma, or posterior fossa primitive neuroectodermal tumor, accounts for approximately 20% to 25% of pediatric brain tumors,374,377 and represents slightly less than one third of posterior fossa tumors. There is a distinct male predominance,381 and the median age at diagnosis is approximately 5 years old; there is a second peak during adulthood. Although the vast majority of these lesions occur sporadically, several syndromes are associated with medulloblstoma, including Turcot’s syndrome,382 Gorlin’s syndrome,383 and Li-Fraumeni syndrome.380 The clinical presentation of medulloblastoma relates to its location, typically arising from the medullary velum, and growing within the fourth ventricle. Most patients present with symptoms of hydrocephalus, including headache (usually early morning), vomiting, and progressive lethargy.381 In infancy, excessive fussiness and vomiting are found in conjunction with accelerated head growth. Because of its relatively fast growth, symptoms are usually present for less than 3 months in more than three fourths of patients.381 Lesions with more lateral growth may present with cerebellar symptoms, rather than those due to ventricular obstruction. On examination, papilledema is present in most patients, and truncal more so than limb ataxia may be seen. Sixth nerve paresis occurs due to hydrocephalus, and the presence of other cranial neuropathies is typically associated with brainstem involvement. MRI is now the standard imaging procedure for medulloblastoma (Fig. 11-3). They usually appear as midline posterior fossa lesions filling the fourth ventricle, and are typically hypointense on T1-weighted images and hyperintense on T2-weighted images. There is usually moderate to robust enhancement with gadolinium, although this may be heterogeneous. Even in experienced hands, the radiographic distinction between medulloblastoma and other posterior fossa tumors may be difficult.384 Because there is a 31% likelihood of visible metastatic disease at diagnosis, nearly
W9418-11.qxd 10/31/03 3:43 PM Page 306
306
Section I
Introduction
Figure 11-3. Axial (A) and sagittal (B) T1-weighted, postcontrast MR image of a medulloblastoma. The lesion fills the enlarged fourth ventricle, and there is moderate, heterogeneous contrast enhancement. The temporal horns are minimally enlarged, representing early hydrocephalus.
always within the spinal or intracranial subarachnoid space,385 both brain and spinal MRI should be obtained. Staging is based on tumor size and growth pattern, and the presence of metastatic disease.386 Neither the cause nor cell of origin of medulloblastoma is known, although the tumors are believed to represent malignant transformation of primitive cells in the external granule cell layer of the cerebellum. Deletions or alterations of multiple chromosomes, especially chromosome 17, have been noted, and amplification of the oncogene n-myc has been observed.387,388 Alterations in proto-oncogenes, such as Smoothened, and the Patched tumor-suppressor gene, have been linked to medulloblastoma.389 The initial treatment of medulloblastoma involves control of hydrocephalus and surgical resection. Because there is no difference in outcome between total and neartotal resection, critical structures (brainstem) should not be violated in an attempt to remove all visible tumor.390 Hydrocephalus is managed initially with external drainage, because only one third of patients or fewer require long-term diversion.391 Following resection, chemotherapy and (if age permits) radiation therapy are usually used, although optimal regimens are still unknown.392 Prognosis, as well as treatment strategies, are based on risk stratification; patients are considered high risk of they are less than three years of age, have greater than 1.5 cm2 residual tumor after resection, or have cytologic or radiographic evidence of tumoral dis-
semination. Survival may be as high as 85% at 5 years with aggressive adjunctive treatment,393,394 although typical overall survival rates are approximately 40% to 70%, for high- and low-risk patients, respectively.380,395 Cerebellar Astrocytoma Cerebellar astrocytoma represents approximately 25% to 35% of all posterior fossa tumors in the pediatric population,380,396 but occurs only rarely in adults. The average age at presentation is 7 years,397,398 and a slight male predominance may exist,397,399 although this is not universally accepted.398,400 Because of the benign nature of most of these lesions, symptoms are more insidious, with an average duration of 5 months before diagnosis.399,401 Headache and vomiting, secondary to hydrocephalus, and ataxia are present in the majority of patients.397,399–401 Papilledema is present in 55% to 85%,397,399–401 and cerebellar signs are more pronounced with this lesion as compared to medulloblastoma, with truncal ataxia more common than appendicular disturbance (74% to 88% vs. 39% to 58%).399,400 Head tilt and nystagmus are present in fewer than 20% of patients. Radiographically, these lesions may be cystic, often with a densely enhancing mural nodule, although this “classic” appearance is seen in a minority of cases400; when present in
W9418-11.qxd 10/31/03 3:43 PM Page 307
Chapter 11
Pediatric Neurosurgery
307
Figure 11-4. A and B, Axial T1-weighted postcontrast MR images of cerebellar astrocytoma. Unlike the classic cystic lesions, this lesion is solid, with irregular contrast uptake. The lesion arose from the superior vermis, and caused significant hydrocephalus.
children this appearance is almost pathognomonic for astrocytoma.396 Solid tumors, often with a microcystic component, are probably more common, and a “false” cystic pattern, in which the cyst wall shows enhancement, may be found occasionally.396,399 Astrocytomas are located more commonly within the cerebellar hemisphere in which case the cystic pattern is more often seen than in the vermis, where the solid lesions predominate.380,401 Hydrocephalus is present in most cases (Fig. 11-4). Histologically, most cerebellar astrocytomas are benign; the pilocytic variant is seen in approximately 70% to 80%400,401 of cases, while the fibrillary (diffuse) pattern is present in most of the remainder; mixed and anaplastic or malignant subtypes, similar to supratentorial astrocytomas of adults, are rare. The treatment of these lesions involves alleviation of the hydrocephalus, usually by temporary external drainage, and gross surgical resection. In the classical cystic tumor with mural nodule, only the solid portion requires resection; in cases in which there is diffuse enhancement of the wall (false cyst), the cyst wall may be resected. Brainstem invasion, which may occur in 8% to 40% of cases,399–401 usually precludes total removal. Overall, the survival at 10 years approaches 90%.377 Most recurrences occur in those with known postoperative residual disease, and are quite rare when postoperative imaging reveals total resection.400 Small residual tumor may involute spontaneously, but this is the
exception; the residual may either be re-resected or followed with serial imaging. Recurrent disease may be treated with resection if technically feasible. The role for radiation therapy for the treatment of these lesions is not clear; some reports have described a protective effect,402–404 while others have shown no benefit.400 Ependymoma Ependymomas account for 5% to 10% of pediatric brain tumors overall, and in young children, the most common group affected, the majority of these occur in the posterior fossa, arising from the floor of the fourth ventricle. Overall, approximately 60% to 70% of ependymomas in children arise in this location; when they occur supratentorially, half occur in a periventricular location.380 Approximately one third of cases occur in children younger than 2 years of age, and two thirds occur before age 7405; there is a male predominance.406 Symptoms and signs relate to tumor location. Because most infratentorial lesions arise from the floor of the fourth ventricle, most patients usually present with headache and vomiting, secondary to ventricular obstruction. As ependymomas may arise from or extend through the fourth ventricular apertures, symptoms may include those related to cranial nerve deficit, torticollis, or cerebellar distur-
W9418-11.qxd 10/31/03 3:43 PM Page 308
308
Section I
Introduction
bance.406,407 Because of their site of origin near the area postrema, unexplained vomiting may predate the development of other symptoms, and can prompt workup for gastrointestinal disease. Younger children and infants typically have a slightly more protracted duration of symptoms before diagnosis than older children, due to the often nonspecific nature of complaints.408 Supratentorial lesions may present with symptoms of hydrocephalus, focal deficits, or seizures. On imaging studies, fourth ventricular ependymomas may be difficult to distinguish from medulloblastoma, but several characteristics are more common in the former, including the presence of calcification (best seen on CT), and extension through the fourth ventricular foramina, into the cerebellopontine angle or cisterna magna. The signal characteristics are often variable, and most enhance in a heterogeneous manner409 (Fig. 11-5). Complete brain and spinal MRI should be obtained perioperatively, because dissemination may occur in approximately 20% of patients,377 although usually the spread is detected only on CSF analysis and not on imaging studies. Histologically, intracranial ependymomas may be subdivided broadly into benign and anaplastic variants, the former occurring in about 62% to 75% of cases,380,410,411 although the criteria regarding classification has been subject to debate.412 Chromosomal analysis has revealed multiple abnormalities in tumor karyotype, most commonly involving chromosome 22.413
The primary treatment of ependymoma remains surgical resection and control of hydrocephalus, as described previously. Because of their common origin from the fourth ventricular floor, and growth through the posterior fossa subarachnoid space, gross total resections are often impossible,411,414 and are impractical in the event of known dissemination, because the risk of deficit related to brainstem or cranial nerve injury is high. Supratentorial lesions are usually more amenable to total removal. Tumor progression is inevitable when residual tumor remains, however, and the benefits of aggressive resection must be balanced against quality-of-life issues.407 After complete resection, the value of adjuvant therapy is unknown,415 but radiation may be advised, given the risk of recurrence even after a documented gross removal.414 After incomplete resection, radiation therapy has been shown to prolong survival and delay disease progression.411,416–418 Prolonged chemotherapy may benefit infants in whom radiation is delayed,411 but its effectiveness in older children is not known, nor is the most efficacious regimen.419 There is little question that young age at diagnosis and incomplete resection are each associated with shorter survival,414,420 although intensive adjuvant therapy appears to provide some benefit for young children.411 Overall, the 5-year survival rate for patients following incomplete resection is approximately 22%, while total resection improves the rate to 80%.414
Figure 11-5. A and B, Axial T1-weighted postcontrast MR images of an anaplastic ependymoma. There is invasion of the brainstem and right cerebellar peduncles, and extension toward the right cerebellopontine angle.
W9418-11.qxd 10/31/03 3:43 PM Page 309
Chapter 11
Cerebellar Mutism Neurologic deficits related to resection of the posterior fossa tumors described previously typically depend on the degree to which the lesions infiltrate or involve critical neural or vascular structures, and the aggressiveness of the resection; such deficits are often manifest immediately after surgery. A syndrome nearly exclusive to the pediatric population, characterized by the delayed onset of transient mutism and oropharyngeal dysfunction after surgery, is seen in approximately 8% of patients undergoing such resections,421–423 with a higher incidence noted for vermian lesions. The onset of mutism usually occurs after a brief period of normal speech (1 to 3 days), and can last for weeks to months.421,422,424,425 Most affected patients have some degree of swallowing difficulty, emotional lability, or decreased volitional limb movement, which tend to improve prior to the mutism.421,423 Gradually, the mutism is replaced by dysarthric speech, which continues to improve with time. The cause of this syndrome is not known, but damage incurred during resection cannot explain the latent nature of onset. Most cases are associated with splitting of portions of the inferior vermis,421,423 but can occur after opening the superior vermis.426 The anatomic substrate for the mutism and pseudobulbar signs, and the role of cerebellar structures in the initiation and modulation of speech, attention, movement, and cognition in general, are unknown.427,428 Pollack and colleagues421 blindly evaluated postoperative MRI studies of affected and non-affected patients undergoing similar resections, and found that bilateral edema within the middle cerebellar peduncles was strongly correlated with development of the syndrome, although this finding was not absolute; such edema could explain the delayed onset and ultimate recovery of function. They hypothesized that interruption of dentate nuclei efferent and afferent tracts could compromise the volitional input necessary for initiating speech, swallowing, and kinesis.
Brainstem Gliomas Primary brainstem gliomas comprise approximately 25% to 30% of posterior fossa tumors in children.380,429 Before the routine use of MRI, the classification of these lesions, and therefore their treatment, was a problem because the extent of involvement was difficult to ascertain.430,431 Approximately 60% of gliomas are diffuse astrocytomas, and nearly 80% of these diffuse lesions primarily involve the pons, although most extend above or below the pons.431 Diffuse pontine tumors are nearly exclusive to the pediatric population, with median age at diagnosis approximately 7 to 8 years.432,433 Symptoms and signs are usually of short duration, with a median onset less than 1 month before diagnosis. Cranial neuropathies, often bilateral, and cerebellar dysfunction are present in most patients at initial presentation, and long
Pediatric Neurosurgery
309
tract signs are seen in about half 433; hydrocephalus is generally a late finding, and is rare with diffuse pontine tumors.431 Focal brainstem tumors are less common than diffuse lesions, and symptoms are usually more insidious. Symptoms and signs depend on the level affected, and the presence or absence of hydrocephalus. Lesions within the midbrain commonly present with oculomotor abnormality or pyramidal involvement429,434; tectal tumors often cause aqueductal obstruction, leading to symptoms of hydrocephalus.419, 435 Lesions confined to the lower brainstem or cervico-medullary junction lead to a combination of long tract signs and lower cranial neuropathy. A subclass of tumors that primarily grow into the fourth ventricle in an exophytic manner cause symptoms secondary to ventricular obstruction.436 MRI has revolutionized the understanding of brainstem tumors. Diffuse pontine tumors appear as an enlarged pons, typically with extension into the midbrain or medulla; approximately one third to one half of tumors show areas of enhancement with gadolinium.431,433 The infiltrative aspects are best appreciated on T2 imaging. Focal lesions show variable patterns of enhancement on MRI.431,434 For diffuse pontine tumors, biopsy is usually unnecessary, and may be hampered by sampling error; the majority are malignant astrocytomas on pathologic examination.437,438 In a series of 12 patients, all focal midbrain lesions, both tectal and tegmental, were nonpilocytic, low-grade astrocytomas.434 Focal lesions at the cervico-medullary junction are most often low-grade gliomas, most commonly astrocytoma.439 Treatment and prognosis depend on the type of brainstem tumor. For diffuse pontine gliomas, steroids and radiation therapy may temporarily improve symptoms, but progression is inevitable. Even with aggressive hyperfractionated radiation, approximately one third survive for 1 year, and 10% are alive at 3 years.432 Outlook for patients with tectal tumors is far better; when patients present with aqueductal obstruction, CSF diversion alone is the primary treatment, using shunting, and more recently, third ventriculostomy. Of 16 such patients treated with shunts,419 progressive symptoms developed in only four at a mean of nearly 8 years after diversion. Biopsy specimens of these lesions revealed three benign gliomas, and one anaplastic astrocytoma; all patients were treated with radiation and remained stable more than 4 years after treatment.419 For tumors at the cervico-medullary junction, resection is usually indicated, and can result in long-term survival for nearly all patients, although cranial nerve dysfunction and long tract deficits may remain after resection.440,441 Dorsally exophytic tumors are also treated with resection of the ventricular component, with no attempt to excise tumor infiltrating the brainstem. Of 18 patients treated in this manner, only four showed progression at a mean follow-up of 9 years.436 Radiation may be beneficial for patients with residual tumors after resection, although this component often
W9418-11.qxd 10/31/03 3:43 PM Page 310
310
Section I
Introduction
remains quiescent for long periods. Chemotherapy has not shown effectiveness for any type of brainstem glioma.406,430 Craniopharyngioma Craniopharyngiomas are the most common non-glial brain tumor in children, and account for approximately 5% to 9% of pediatric brain tumors.374,442 Although they can arise at any age, there are two fairly distinct peaks of occurrence, the earlier typically occurring in the first or second decades, and a later onset after age 50.443 In children, the mean age at diagnosis is approximately 7 to 9 years.444–446 There do not appear to be racial or gender differences,447 although several series have demonstrated a small male predominance.444,448 Because of their slow growth, and proclivity for the sellar region, presentation is usually due to hydrocephalus, endocrine dysfunction, or visual impairment, with symptoms present an average of 10 months before diagnosis.445 Headache, often with vomiting, is present in more than 75%,446 and visual disturbance due to optic nerve or chiasm compression is present in approximately 60%.444,446 Papilledema is seen at presentation in approximately 30% to 40%.445,446 Although endocrine dysfunction is rarely the sole manifestation of craniopharyngioma, more than 80% of children have abnormal hormonal function at presentation. Growth hormone deficiency can be detected in 75% of affected children; small stature is seen in at least half of patients, and decreased rate of growth occurs in more than 75%.449 Decreased levels of leutinizing or follicle-stimulating hormone occurs in approximately 40%, and disturbances of cortisol and thyroid function each occur in one-fourth of patients.449 Abnormal posterior pituitary function is less common at presentation.449,450 Radiographically, craniopharygiomas can be identified on plain radiographs due to calcification and expansion of the sella, although these studies are now uncommonly performed. These lesions are usually intrasellar and suprasellar in location, although they rarely may be entirely within the sella.451 CT imaging typically reveals a combination of cystic and solid components, with the latter usually showing contrast enhancement; the majority of tumors are at least 50% cystic.452 Calcifications are seen on CT in more than 90% of pediatric craniopharyngiomas.380 MRI is valuable in revealing the tumoral relation to vital structures, such as the anterior cerebral arterial branches, the visual apparatus, and the pituitary stalk and tuber cinerum. In the suprasellar region, tumor growth may be broadly classified as prechiasmatic, subchiasmatic, or retrochiasmatic.380 Prechiasmatic tumors cause posterior displacement of the chiasm and elevation of the anterior cerebral artery branches, while subchiasmatic lesions elevate the optic apparatus. Retrochiasmatic tumors displace the chiasm anteriorly, and involute or grow within the third ventricle. Prechiasmatic tumors have a greater tendency to cause early visual dysfunction, while retrochias-
matic growth may result in signs and symptoms of ventricular obstruction.442,453 Histologically, craniopharyngiomas are benign cystic epithelial lesions, of unknown embryologic origin.454 They are thought to arise from cell rests along Rathke’s pouch, the precursor of the infidibulum and anterior lobe of the pituitary gland. The cystic component typically contains a dark, “crank-case” fluid, often harboring cholesterol crystals. In children, the epithelial component is usually of the adamantinomatous variety, similar in appearance to primitive tooth-forming tissues, while tumors in adults typically harbor a papillary pathology. Keratin nodules are found within the epithelial component of most lesions; this “wet” keratin appearance is nearly unique to craniopharyngiomas. Although the tumors are largely extra-axial in location, adamantinomatous tumors often have foci of brain invasion, usually into adjacent hypothalamic tissue.455,456 The management of craniopharyngiomas includes a thorough preoperative endocrine workup, and correction of hormonal (usually cortisol and thyroid) deficiencies that could impact anesthesia and surgery. Ophthalmologic evaluation is important in identifying evidence of optic nerve or chiasmatic impairment, and for obtaining baseline acuity and visual fields. If time permits, baseline neuropsychological testing may be obtained preoperatively. Hydrocephalus, if present, is usually managed via temporary external ventricular drainage at the time of surgery. The optimal treatment for craniopharyngiomas in children is unknown because the excellent long-term survival seen with most forms of therapy allows more intense focus on quality-of-life issues. These tumors may be managed with surgery alone, limited resection followed by radiation therapy, or intracavitary approaches. Several management algorithms have been advocated, based on factors including patient age, size and location of tumor, presence of hydrocephalus, and hypothalamic function,442,457 suggesting the importance of multiple patient- and tumor-related variables in formulating a treatment plan. Surgery is usually carried out via subfrontal or pterional approach, although subdiaphragmatic lesions can be removed from a transphenoidal approach.443,458 Surgery may be curative, but aggressive resection may be at the expense of permanent endocrine or neurologic deficit. Gross total resection can be obtained in most cases,444,453,459 although adherence to optic or vascular structures, poor visualization, hypothalamic invasion, and dense calcification may preclude total removal.443 Most patients will have permanent pituitary dysfunction, primarily diabetes insipidus, after gross resection,453,459 and depend on lifelong replacement therapy, although sparing the infundibulum reduces the risk of anterior pituitary dysfunction.450 Although visual function often improves after resection, worsened acuity and new field deficits are not uncommon.453,459 Major neurologic deficits, usually related to vascular injury or vasospasm, are relatively uncommon in experienced centers. After documented total
W9418-11.qxd 10/31/03 3:43 PM Page 311
Chapter 11
resection, recurrence is uncommon,442,459 usually less than 20% of cases, and long-term survival is seen in greater than 90%. Limited resection alone usually results in tumor regrowth; the addition of adjuvant radiation offers similar control rates compared to radical resection, and carries a significantly lower risk of major morbidity, especially diabetes insipidus.445,446,460–465 Long-term sequellae after radiation therapy, including endocrine dysfunction, visual impairment, and cognitive difficulty, are not uncommon, however, and usually occur in a delayed fashion.466 Stereotactic radiosurgery may provide tumor control without the morbidity of conventional external beam therapy, although proximity to visual structures may limit its use for tumors in this region467,468; stereotactic radiation therapy may allow equivalent dose delivery without risking long-term visual injury.469 For lesions with a proportionally large cystic component, intracavitary therapy, using bleomycin470,471 or radioactive agents (usually phosphorus or yttrium)472,473 may provide excellent long-term tumor control; worsening of visual function after such treatment is common, occurring in as many as half of patients, while new onset endocrine dysfunction is relatively rare.472 Neurofibromatosis Type I A small subset of pediatric brain tumors occurs in the setting of the phakomatoses. The most common of these neurocutaneous disorders is neurofibromatosis type 1 (NF1), which affects approximately one in 3000 people.474 It is transmitted in an autosomal dominant pattern and is eventually fully penetrant, although with a wide variety of clinical features that tend to become more frequent and severe with age; half of all cases are sporadic mutations.475 The cause of NF1 is due to a mutation on chromosome 17q11.2. The NF1 gene is a tumor-suppressor gene whose protein product, neurofibromin, is a GTPase activator protein (GAP) involved in cell growth regulation.476–478 The mechanism through which loss of neurofibromin activity results in tumorigenesis is unknown. With the advent of modern imaging, greater insight into the neoplastic and non-neoplastic processes affecting the brain of NF1 patients is being uncovered, although it is diffucult to correlate clinical, radiographic, and histologic findings. The most common findings on MRI are areas of increased T2 signal in the white matter of the supratentorial or infratentorial compartment, nearly always without mass effect or contrast enhancement. These unidentified bright objects, present in approximately 60% to 80% of patients, may increase in size and frequency throughout childhood.479–481 The etiology of these foci is uncertain, but given their tendency toward resolution by adulthood,482 and the lack of associated focal deficits, they are thought to be non-neoplastic in nature,480,483 and do not require biopsy or treatment. Gliomas involving the optic nerves, chiasm, or hypothalamus (optic pathway gliomas) are found in approx-
Pediatric Neurosurgery
311
imately 15% of children with NF1,479,484 although approximately half of these patients have visual or endocrinologic disturbance at the time of diagnosis. Optic pathway gliomas in patients with NF1 may involve only the optic nerve (unilaterally or bilaterally), a pattern that is distinctively rare in non-NF1 patients.485,486 Involvement of the chiasm, seen in nearly all non-NF1 optic pathway gliomas,486 is seen in 70% of NF1 gliomas.485 The clinical manifestations of symptomatic optic pathway gliomas differ significantly between children with and without NF1. Though decreased visual acuity is common in both groups of patients, sporadic cases typically present with symptoms and signs related to increased intracranial pressure or nystagmus, which is extremely rare in NF1. Conversely, symptomatic children with NF1 commonly have precocious puberty or proptosis, which is quite uncommon in sporadic cases.485 The natural history of optic pathway gliomas in these two groups is also vastly different. Whereas most non-NF1 patients experience tumor growth and progression of symptoms if not treated, this occurs in less than 20% of symptomatic NF1 cases followed for more than 3 years.485 Progression is even less common among asymptomatic or minimally symptomatic patients.487 The reasons for the relatively indolent nature of optic pathway tumors in NF1 patients are unknown, because low grade astrocytoma is seen pathologically in all NF1 and nearly all non-NF1 tumors.485 Because of the quiescent nature of these lesions in NF1, they do not require therapy unless significant symptoms develop, and in this event, options for treatment are controversial. Tumors anterior to the chiasm associated with significant proptosis or visual loss may be safely resected. For lesions involving the chiasm, chemotherapy may be preferred, because radiation therapy, the traditional therapy for these unresectable lesions, can lead to cognitive disturbance, endocrine dysfunction, or moyamoya disease.488,489 Gliomas, usually low-grade astrocytomas, occur less commonly in other locations in NF1 patients, such as the cerebellum or cerebral hemispheres. In contrast to the unidentified bright objects described previously, these tumors are associated with mass effect, changes on T1weighted images, and often enhance with contrast material.489 Aggressive resection is usually recommended for these tumors, with adjuvant treatment reserved for unresectable or recurrent lesions, or those with aggressive histologic findings.489 Fewer than 10% of NF1 patients have brainstem tumors associated with mass effect or changes on both T1and T2-weighted images (distinguishing them from unidentified bright objects), and only half of these patients are symptomatic. In a series of 21 such patients, Pollack and colleagues490 subdivided these lesions into diffuse, infiltrating tumors (causing brainstem enlargement), focally enhancing lesions, tectal tumors, and focal, non-enhancing masses. In general, the behavior of most of these lesions was indolent, and therapy was reserved for those tumors exhibiting progressive growth or neurologic deficit, although hydro-
W9418-11.qxd 10/31/03 3:43 PM Page 312
312
Section I
Introduction
cephalus was treated if present. Although they were the most common type of tumor in this series, none of the diffuse, infiltrating tumors, so aggressive in patients without NF1, progressed during the course of follow-up, raising the possibility that these lesions may represent a hamartomatous process. The focally enhancing tumors were the most likely to exhibit growth; histologic studies of these lesions showed low-grade, non-pilocytic astrocytoma.490 Neurofibromas are seen in most patients with NF1; they usually come to the attention of neurosurgeons when they are associated with deficit or pain due to spinal or major peripheral nerve involvement. In most instances, these are histologically benign, although malignant degeneration may occur in up to 15% of cases.491 Neurofibromas tend to involve a significant portion of fascicles of the affected nerve, making functional resection difficult, although not unfeasible, for fusiform tumors.491 The less common plexiform neurofibromas, however, involve nearly all fascicles along a significant length of the affected nerve or plexus, and complete resection is always associated with sacrifice of the nerve.491,492 Their behavior may be more aggressive in younger NF1 patients. Neurofibromas causing spinal cord compression are relatively infrequent; when necessary, canal decompression can be performed safely to preserve neurologic function.493 Spinal Cord Tumors Intramedullary spinal cord tumors are uncommon, compromising between 2% to 4% of central nervous glial neoplasms; among intraspinal tumors, intramedullary tumors account for over 35% of pediatric cases, a significantly higher percentage than is seen in adults.494,495 Pediatric spinal cord tumors tend to be evenly distributed throughout childhood, and there is no gender predominance. As opposed to adults, in whom ependymomas are the most common lesion, lowgrade astrocytomas are the most common pediatric tumor, accounting for nearly half of cases.494 In many reports, ependymomas are second in frequency,496,497 and are more common in older children, although in the largest pediatric series to date,494 gangliogliomas were the second most common tumor, accounting for more than one fourth of 164 cases, followed by ependymoma. The reasons for this discrepancy are unknown, and may be related to historical inaccuracy in diagnosing ganglioglioma. Malignant lesions are uncommon in all series. Tumors occur in a relatively even distribution along the spinal cord, and span an average of more than five bony levels.494 Symptoms of benign intramedullary tumors in children are usually insidious, with an average duration of approximately 1 year before diagnosis.494 Loss of function or delay in attaining motor milestones is the most common complaint of parents, seen in approximately two thirds of patients, followed by axial pain, often worse at night. Slowly progressive scoliosis and urinary complaints are each present
in approximately one third of children.494 Cervical tumors commonly present with head tilt, and less commonly, with arm weakness. Hydrocephalus is seen in more than half of those with malignant spinal cord tumors, usually due to meningeal infiltration, and in a small percentage of benign tumors, secondary to rostral extension of tumor or associated cyst toward the obex.498 Because of the slow growth of these tumors, most children have only mild deficits at presentation, with overt plegia being uncommon. Histologically aggressive tumors have a shorter duration of symptoms, and ependymomas, more so than other tumors, often present with bilaterally symmetric dysesthesia.495 While MRI is the procedure of choice for intramedullary tumors, children often undergo plain radiography as an initial procedure, particularly if pain or scoliosis are present before the development of overt neurologic signs. In addition to demonstrating mild curvature, these studies often show a widened canal and increased interpedicular distance, which, if subtle, may be overlooked initially. MRI is of enormous value in operative planning, as the rostral-caudal extent of the tumor may not be apparent on CT scan or myelography. Intrinsic tumors are typically hypointense on T-1 weighted images, and bright on T-2 sequences. Cysts may be present either at the poles or within the tumor itself. Ependymomas tend to be centrally located and usually enhance intensely and homogeneously with gadolinium, while astrocytomas may be eccentric, and show patchy enhancement.499 The optimal management of intramedullary spinal cord tumors in children is controversial.500,501 Although historical as well as recent series496 have noted long-term relapse-free survival in the majority of children treated with radiation therapy without resection, aggressive resection is considered the most effective treatment for low-grade pediatric spinal cord tumors,500,501 and carries a relatively low risk of serious neurologic deficit in experienced centers.494 Gross total resection of ependymomas, which often have a distinct cleavage plane, is associated with 5- and 10-year, progression-free survival rates of 100% and 86%, respectively; corresponding rates for ganglioglioma are 67% and 47%.494 Low-grade astrocytomas, typically more infiltrative into the surrounding cord, are nonetheless associated with long-term, progression-free survival after total or subtotal resection alone.494,500 High-grade neoplasms are associated with the highest rates of recurrance, even after aggressive resection and adjuvant therapy, although long-term survival may be seen in a minority of patients with anaplastic astrocytoma; cord glioblastomas are uniformly fatal within several years. For recurrent, low-grade tumors, resection may again result in long survival, and may be preferred over radiation therapy.494,501 The role of radiation therapy in the treatment of low-grade spinal cord tumors is unclear, because no series has carefully compared this modality against surgery for primary or recurrent tumors. After gross resection, radiation should not be given, as the potential deleterious effects on
W9418-11.qxd 10/31/03 3:43 PM Page 313
Chapter 11
the cord as well as the pediatric spine may outweigh its potential benefit.501 Chemotherapy has no role in the management of low-grade tumors, but may play a palliative role in the treatment of malignant cord tumors in children.502
Vascular Malformations and Disease Cerebrovascular diseases are uncommon in pediatric patients because atherosclerosis, and most vascular malformations and aneurysms, usually become symptomatic during adulthood. Nonetheless, diseases of the cerebral circulation in children remain an important cause of long-term morbidity, and their potential for cumulative damage mandates careful scrutiny and understanding of both their natural history and management. Arteriovenous Malformations Pial arteriovenous malformations (AVMs) usually present during young adulthood, with intraparenchymal, subarachnoid, or intraventricular hemorrhage, new onset seizure, progressive deficit, or headache as the most common symptoms. Most pediatric AVMs manifest with hemorrhage, seen in 80% of cases.503 Sudden headache, neurologic deficit, seizures, or lethargy may occur, depending on the location and extent of hemorrhage. Only 15% to 20% of childhood AVMs present with seizures.503,504 In infancy, congestive heart failure leads to diagnosis in more than half of patients, while approximately one third present with seizures; only 15% of infants present with acute hemorrhage.505 As in adults, most AVMs in children are hemispheric in location (two thirds among a series of 160 patients), while cerebellar and brainstem lesions occur in something less than one fourth of cases.503 The diagnosis of AVM can be suggested on CT scan or MRI, in which acute hemorrhage may be detected. Nidal calcifications are seen in a minority of cases in children; contrasted images may reveal dilated vessels feeding and draining the malformation. While MR angiography can be useful in displaying the nidus, angiography remains the gold standard for depicting the anatomy of the AVM, and should be performed expeditiously in a child presenting with hemorrhage, if the patient is clinically stable. The “classic” appearance is that of an inverted wedge-shaped nidus, with the apex approaching the ventricle, although small, globular niduses or indistinct fistulae can occur. In children, the appearance of an AVM may reflect an immature or developing arterialvenous communication; a “fluffy,” or diffuse-appearing nidus may be indicative of an evolving process.506,507 In a small subset of patients, especially those with large, compressive clots, the presence of a small nidus may be obscured due to mass effect; rarely, an AVM may spontaneously thrombose before hemorrhage. In these situations, consid-
Pediatric Neurosurgery
313
Table 11-5 Spetzler-Martin Grading Scale for Arteriovenous Malformations Feature • AVM (nidus) size 6 cm • Eloquence Noneloquent Eloquent • Venous Drainage Superficial Deep
Points 1 2 3 0 1 0 1
Grade = total points (1–5); Grade 6 AVM implies large, diffuse, inoperable lesion. Spetzler RF, Martin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–483.
eration must also be given to the presence of a cavernous malformation as the cause of an occult bleed. Upon the completion of imaging evaluations, AVMs in children can be assigned a Spetzler-Martin grade508 (Table 11-5) based on size, pattern of venous drainage, and involvement of eloquent cortex; smaller AVMs are thought to be more likely than large lesions to present with hemorrhage in children,503 similar to those seen in adults. Historically, AVMs have been presumed to be congenital lesions arising from anomalous formation of capillary and arteriolar channels, although the causes remain elusive. Mullan and colleagues509 believe that AVMs develop due to errors in the formation and subsequent absorption of cortical surface veins, usually after 9 to 10 weeks’ gestation. AVMs most commonly present during adulthood, and growth or enlargement of these lesions is well known to occur after birth, suggesting that other mechanisms, such as hemodynamic vascular recruitment or humoral factors, may be involved in the development of clinically relevant AVMs. Recent work has shown increased expression of vascular endothelial growth factors in recurrent AVMs, which lends credence to postnatal factors in their genesis.510 The majority of AVMs occur as isolated phenomena, although multifocal lesions have been reported511; AVMs may also be associated with familial disorders, such as hereditary hemorrhagic telangiectasia.512 The natural history of AVMs in children is unknown, because most children present with hemorrhage and are therefore not managed expectantly. AVMs presenting in childhood may have a higher frequency of eloquent brain involvement, especially thalamic, basal ganglia, or posterior fossa structures,503,504 compared to adult patients. The annual risk of bleeding of a newly discovered AVM is probably not different between children and adults, but the lifetime risk in children is obviously much higher; if one assumes an annual bleed risk of 3%, the chance of eventual hemorrhage equals 1-0.97x, where x equals the remaining years of expected life.513
W9418-11.qxd 10/31/03 3:43 PM Page 314
314
Section I
Introduction
The optimal treatment strategies for children harboring AVMs is unknown, because the largest reported series have been biased toward a single mode of treatment, either surgical503 or endovascular,514 and the options available for treatment are constantly improving. The goals of intervention are elimination of lifetime bleeding risk and preservation of function. For small and accessible lesions (Spezler-Martin I and II), surgery alone can achieve cure in the majority of patients.503,515 Endovascular procedures alone historically result in cures in approximately 5% of patients,516 and are usually a prelude to other forms of treatment. Multimodality therapy, including surgery, embolization techniques, and radiation surgery, is becoming increasingly common for the management of large or poorly accessible lesions. A recent series of 40 pediatric patients treated with multimodality therapy demonstrated anatomic cure in more than 90% of patients completing treatment.504 Radiosurgery is particularly successful for smaller pediatric AVMs, with obliteration rates of 80% for lesions less than 3 mL in volume, although the reported overall cure rate of less than 60% at a mean of 3 years follow-up implies a continued risk of hemorrhage in a significant number of patients.517 The ultimate outcome of children with AVMs is strongly related to condition at time of presentation, as well as the resultant morbidity associated with future hemorrhages in incompletely obliterated cases.503,515 Most patients presenting in good condition after hemorrhage remain so after treatment,503,504 with new deficits relatively uncommon in current practice. In patients with unresected or incompletely obliterated AVMs, nearly 40% suffered recurrent hemorrhage.503 Seizures are typically reduced after treatment; 11 of 15 patients were seizure-free after radiosurgery,518 and nearly 60% remained so after resection.503 Even after complete obliteration, children are at risk, albeit low, for delayed recurrence or regrowth. Kader and associates519 reported five such recurrences, from 1 to 9 years after angiographically confirmed resection, and none in more than 650 adults from the same institutions. Although they postulated that spasm or mass effect could have led to a false-negative postoperative angiography result, the absence of recurrent lesions in adults implied an inherent difference in pediatric AVMs, suggesting that AVMs in children may be in a dynamic state of development. Subsequent analysis by this group confirmed higher rates of astrocytic endothelial growth factor in the original specimens of the recurrent patients compared to controls.510 The authors indicated that children may benefit from delayed angiography to assure continued obliteration.
tures. In general, the manner of presentation depends on age. Neonates usually have high-output cardiac failure; hydrocephalus or seizures develop in infants; and older children and adults present with hydrocephalus, hemorrhage, seizures, or neurologic deterioration; most cases present within the first several years.520–522 In neonates and infants, cranial bruits are often present, and the diagnosis must be suspected in patients with unexplained cardiac failure that is usually unresponsive to medical management. Imaging studies reveal a midline mass in the area of the vein of Galen, often in association with hydrocephalus; CT scan may reveal diffuse calcifications secondary to prolonged ischemia. Angiography remains essential in evaluating the architecture of these malformations. Yasargil523 described four types of malformations: type I malformations have one or several feeders, usually from the pericollosal or choriodal arteries; type II malformations are fed primarily from posterior cerebral or thalamoperforator branches; type III lesions are a combination of types I and II. Types I to III are true subarachnoid fistulas, while type IV malformations are parenchymal AVMs with secondary dilatation of the vein of Galen. The causes of anomalous communication between primitive deep arterial feeders and the vein of Galen, or its precursor, the median prosencephalic vein of Markowski, are unknown.522,524 The ultimate endpoint of therapy for true vein of Galen malformations is elimination of the anomalous fistula, although the immediate goal of treatment is often a reduction of abnormal flow to a point where patients are rendered clinically stable. Before the advent of modern endovascular techniques, the treatment for vein of Galen malformations was surgical interruption of the fistulae,520,521,525 often complicated by high rates of permanent morbidity or death; most untreated patients, especially neonates, succumbed to complications secondary to cardiac failure. Surgery is now generally reserved for treatment of hydrocephalus and to provide access for transtorcular embolization.522,526 Most modern series describe transarterial or transvenous endovascular approaches, alone or in combination, occasionally using transtorcular embolization. While less than half of malformations are completely obliterated in this manner, symptoms secondary to shunting are reduced in the majority of patients.522,527–530 The majority of patients now survive, usually with no or minor neurologic deficits, with the highest mortality rates in neonates with cardiac failure.522,527–529 Patients with residual lesions can undergo staged endovascular procedures if symptoms persist, whereas asymptomatic patients can be observed expectantly.
Vein of Galen Malformations
Cavernous Malformations
Vein of Galen malformations are a rare subclass of pediatric vascular malformations, characterized by high flow arterial shunting through the galenic system, usually in association with aneurysmal dilation of the midline deep venous struc-
Cavernous malformations are angiographically occult vascular lesions that can present at any age, though they are relatively rare in the elderly. In large series, most patients become symptomatic by the fourth decade of life, with pedi-
W9418-11.qxd 10/31/03 3:43 PM Page 315
Chapter 11
atric patients comprising approximately one fourth of patients531,532; the average age in a series of 40 pediatric patients was approximately 10 years.533 A bimodal distribution within children may exist, with one peak during the first several years, and a later peak in the early teens.534,535 As with pial AVMs, children most commonly present with focal deficits related to hemorrhage; seizures, headache, and irritability are less prevalent,533,536 although in early reports, seizures were the most frequent symptom.534 Increasingly, incidentally discovered lesions have been noted, especially with MRI screening of family members of affected patients. The deficits incurred by hemorrhage are generally less severe than those associated with AVM. Cavernous malformations are multiple in up to 25% of children.533,537 Familial clustering of cavernous malformations are noted in up to 30% of patients, most notably in the Hispanic population, and has usually been linked to chromosome 7q,538–541 although other foci exist.542 Imaging studies are usually successful in revealing areas of hemorrhage, although the malformation may be obscured by blood products. CT is excellent for depicting acute blood and calcifications, although MRI is the more sensitive study, and often reveals a lesion of mixed intensity, indicating areas of both recent hemorrhage as well as hemosiderin deposition. MRI is often superior to CT in depicting multiple lesions, especially those that have not bled.537 Cavernous malformations may occur anywhere along the central nervous system, including the spinal cord; they are most common within the hemispheres, but brainstem lesions are present in approximately one fourth to one third of patients.533,543 Pathologically, cavernous malformations are composed of thin-walled, dilated vascular channels, often filled with clot, and with little or no intervening brain; on an ultrastructural level, these channels lack a collagen support matrix, and the endothelial cells lining the malformation often lack tight junctions.544 Associated venous angiomas are often seen in proximity to cavernous malformations, especially in the posterior fossa.545,546 The natural history of cavernous malformations in adults is becoming better characterized, with annual risk of hemorrhage less than 1% in patients presenting without a bleed, and 4.5% for those presenting with hemorrhage.543 In children, however, the natural history is not well known; as with AVMs, children presumably have a greater overall risk of future bleeds. These malformations can arise de novo after exposure to ionizing radiation,547,548 and new lesions may occur in more than one fourth of patients with familial malformations,549,550 and less commonly in non-familial cases. Lesions can also exhibit significant growth over time in children, likely secondary to microhemorrhage and subsequent angiogenesis.533 Surgery remains the mainstay of treatment for symptomatic malformations in children. It is recommended for all symptomatic patients with surgically accessible lesions, because complete resection eliminates the potential for hem-
Pediatric Neurosurgery
315
orrhage, and eliminates or reduces seizure frequency in most patients533,551; incomplete removal is not protective. For children with incidentally discovered lesions, treatment options include observation, with resection only after documented radiographic change or onset of symptoms, or prophylactic resection. For patients with multiple lesions, only the symptomatic lesion should be addressed surgically, although resection of clinically silent lesions may be considered if accessible through the same approach. Brainstem malformations are usually resected only when the lesion or clot extends to a pial surface, or in cases of progressive neurologic deterioration533,552; these lesions are the most technically challenging, but can be managed safely in experienced centers.553,554 Radiosurgery has been described for patients with inaccessible lesions, and may reduce the annual hemorrhage rate in those with frequent bleeds,555,556 but the results have been largely disappointing557,558; its effect in pediatric cases is unknown. Moyamoya Disease Moyamoya disease is a disease of unknown etiology characterized by cerebral ischemia or hemorrhage, secondary to progressive occlusion of the distal carotid arteries and anterior circle of Willis, and the development of compensatory collateral circulation. The disease is relatively rare in the Western hemisphere, and is most prevalent in Japan, and less so in other Asian cultures.559,560 The disease is most prevalent in children, with a second peak in the third and fourth decades.561 Nearly all affected children present with transient or permanent deficits secondary to ischemia, often precipitated by bouts of hyperventilation, and often associated with headache; seizures or hemorrhage may uncommonly be the initial manifestation.559,560,562 The long-term consequences depend on the ishemic damage incurred prior to the development of sufficient collateralization, but a gradual worsening of cognitive function has been noted in untreated patients.563 Approximately 10% of Japanese cases are familial,561 although this tendency is not present in Western cultures. CT or MRI may show areas of ischemia or infarction, but angiography remains central to the diagnosis. Angiographic changes include progressive stenosis or occlusion of the anterior circle of Willis, and the subsequent development of enlarging collateral vessels, initially from the basal perforators off the circle of Willis (which appear similar to a puff of smoke, “moyamoya” in Japanese), and later from the extracranial circulation, which gradually fill the middle cerebral territory.564,565 The process is nearly always bilateral. Nuclear blood flow studies usually show reduced perfusion at baseline, and more commonly, diminished reserve.560 The cause of moyamoya disease is unknown, and by strict definition is an idiopathic process, without associated syndrome or illness; similar vascular changes can be seen in children after whole brain radiation, and in children with neurofi-
W9418-11.qxd 10/31/03 3:43 PM Page 316
316
Section I
Introduction
bromatosis type 1 (NF 1); interestingly, linkage analysis in Japanese families has shown a focus on chromosome 17q, the site of the gene responsible for NF 1.566 The treatment for moyamoya involves surgical augmentation of collateral flow to areas at risk for permanent ischemia. More than 40 procedures have been described, and can be broadly categorized into direct and indirect anastomoses559; it is uncertain whether any technique is superior. The direct procedures involve a surgical bypass, usually from superficial temporal artery to middle cerebral artery, while indirect procedures usually involve placement of inverted dura, scalp vessels (usually superior temporal artery), temporalis muscle, or a combination of these directly on the brain surface. Some series combine direct and indirect anastomoses, and procedures are commonly performed on both hemispheres. In general, most current procedures result in long-term clinical stabilization and elimination of ischemic symptoms in the majority of patients, and delayed angiography usually reveals robust extracranial-to-intracranial collateral vessel development. Indirect procedures are increasing in popularity,559 and reports have shown favorable clinical and radiographic outcome in most patients.562,567 Three fourths of patients have no further ischemic symptoms after the first year, which is far faster than the natural course of the disease,568 although postoperative infarction may occur in up to 10% of patients.569 Indirect procedures may be most suitable for children, in whom the small vessels make direct anastomosis difficult. Because most children stabilize after surgery, the long-term outcome is usually related to the presence of a fixed deficit before the development of collateral circulation. Young children (younger than 3 years) are more likely to suffer a significant preoperative stroke than older children, and thus have a poorer prognosis.570
Infectious Disorders Bacterial infection may occur within any intracranial compartment, including the subdural and epidural spaces, or within the brain parenchyma or cerebrospinal fluid pathways. With the exception of shunt infection, the incidence of these processes in children is low in Western societies, where better access to health care and the availability of antibiotics have resulted in effective treatment for many of the conditions predisposing to such infections. The potential for significant morbidity, if improperly diagnosed or treated, and the common occurrence among certain populations, mandate a thorough understanding of the causes and management of intracranial bacterial infections. Subdural Infection (Empyema) Causes of subdural suppuration are typically a reflection of patient age, and may also vary with respect to geographical
location or patient population. Subdural empyema is primarily a disease of children and young adults, and for unknown reasons, there is a strong male predominance.571,572 More than 85% of 699 recent (CT era) cases from a native South African population occurred in patients less than 20 years old, with 32% of the total occurring in children 10 years or younger, and 54% occurring in the 11- to 20-yearold age group.571 More than 70% of 102 cases in children ages 5 years and younger were associated with meningitis, while 86% of infections in older children and teenagers were the result of paranasal sinus disease; only 8% of pediatric cases were secondary to an otogenic source, although chronic middle ear infection was the most common etiology in an Indian patient population.573 In the United States, sinusitis remains the most common condition predisposing to subdural infection in children, and is seen primarily in the second decade,572,574,575 after development of the paranasal sinuses; otogenic sources are a less common cause.576 A recent study of intracranial infection highlighted the relative rarity of pediatric cases, 7%, in a United States tertiary care setting; two thirds of the 41 cases were of postoperative origin, and the mean age was 50 years.577 Infection usually spreads from thrombophlebitis of valveless extracranial and intracranial veins, although direct extension may uncommonly occur. In infantile meningitis, a significant percentage of patients develop subdural effusions in response to the infection; although these collections are usually sterile, empyema may result from inoculation of the fluid.578 Less common causes of subdural infection in children include trauma, scalp infection, or iatrogenic inoculation, in which staphylococcal species predominate. Overall, streptococcal species are the most frequently isolated organisms, although mixed flora and negative culture results are also common.571,579 In infants, Hemophilus has been the most common pathogen, followed by aerobic Streptococcus,578,580 although immunization against the type B may lower the prevalence of Hemophilus. The symptoms and signs associated with subdural empyema depend on the age of the patient, the underlying cause, and the site and severity of infection. Infantile infections, usually due to associated meningitis, typically present with fever (refractory to or recrudescent after antibiotics), bulging fontanelle, and lethargy; seizures are not uncommon.581 Older children and teenagers usually have a history of recent or chronic otorhinologic symptoms, although purulent discharge may be absent571; fever, headache, lethargy, and seizures are the most common complaints. A significant number of patients have a depressed level of consciousness and focal deficits, including hemiparesis and cranial neuropathy, and facial swelling may be present.571,572,579 Infratentorial empyema, usually the result of mastoid or middle ear disease, typically presents with depressed level of consciousness, fever, and ear drainage. Meningismus and signs of increased ICP are common, while focal signs are less prevalent.581
W9418-11.qxd 10/31/03 3:43 PM Page 317
Chapter 11
The diagnosis of subdural empyema may be suspected on clinical grounds. Ancillary test results such as from white blood cell count and sedimentation rate are usually abnormal, but are nonspecific, and blood culture specimens are frequently sterile. CT reveals the presence of a hypodense or isodense subdural fluid collection, usually with rim enhancement and associated mass effect, although MRI is more sensitive582; associated paranasal or mastoid disease is detected on most studies. Associated osteomyelitis, epidural infection, and parenchymal abscess may be present, and venous thrombosis is an unusual but potentially devastating occurrence. Empyemas are usually seen overlying the convexity, with frontal and panhemispheric collections most common, although interhemispheric collections are frequently seen.571,579 Treatment involves the use of prolonged intravenous antibiotics and surgical drainage; an attempt should be made to obtain culture specimens before antibiotic administration, and associated sinus or middle ear collections should be evacuated. Anticonvulsants are initiated if seizures have occurred, and are often given prophylactically. Subdural pus may be drained via burr-hole or open craniotomy; although the former approach can be adequate in select cases,579 an open procedure may allow a more aggressive and thorough drainage,580,581,583 and may reduce the need for subsequent procedures. Infants can be treated successfully with percutaneous subdural taps or catheter drainage.578,584 Serial imaging is required during treatment to ensure that no reaccumulation has occurred. In the modern era, mortality is usually less than 15%, and is usually seen only in those with advanced cases at the time of presentation.579,585 Detailed large-scale outcome data specific to children are lacking; in the South African study of 699 patients, most of whom were children or teenagers, seizures, focal deficit, or other permanent morbidity were noted in one fourth of survivors.571 Bok and Peter579 reported excellent outcome for 70% of 70 patients 20 years or younger. Mortality in this group was 9%, and 6%, all age 5 or younger, were left with severe handicaps. Epidural Infection Isolated epidural abscesses are significantly less frequent than subdural infections; in areas where intracranial suppuration remains common, subdural empyema is nine times more frequent,585 and epidural abscess associated with subdural empyema is more frequent than isolated epidural infection. Like subdural infection, epidural abscess usually affects older children and teenagers, with more than 80% of cases reported during the first two decades.585 Paranasal sinus disease remains the culprit in three fourths of young patients, with middle ear infection responsible for most of the remaining cases. Because meningitis does not predispose to epidural infection, the process is rare in the first 5 years.585,586 In the United States, epidural abscess may be a
Pediatric Neurosurgery
317
more frequent complication of otorhinologic disease than subdural empyema.587,588 The bacterial profile and pathophysiology is similar to that seen in subdural empyema, although the reasons for confinement within the epidural compartment are not clear. Fever, headache, and facial swelling are common, although symptoms are more insidious than in subdural infection, and focal signs and lethargy are exceedingly rare; seizures occur in approximately 10%.585 Subgaleal abscess (Pott’s puffy tumor) is seen in slightly less than half of patients (Fig. 11-6). CT imaging, and especially MRI, reveal the presence of a rim-enhancing extradural process, and the distinction between intradural and extradural infection is usually apparent.582 Treatment principles are the same as those for subdural empyema, although burr-hole drainage may be more successful in the setting of epidural infection. Nearly all patients have excellent outcomes. Brain Abscess Parenchymal brain abscess is the most serious of the intracranial suppurative processes; despite advances in diagnostic and treatment modalities over the past 30 years, and improved survival, there remains a significant potential for long-term sequelae in affected children. Like other forms of intracranial infection, brain abscess is strikingly common in the pediatric population,589 although the conditions with which it is associated are more diverse than those in extra-axial infection, and again depend on the population examined. Cyanotic congenital heart disease, especially tetralogy of Fallot, and other conditions with right-to-left cardiac shunting are among the most common associated conditions in children with brain abscesses in many recent series.589–593 Of 149 patients with brain abscesses, 103 were associated with cardiac shunting; more than half of these were seen in children 10 and younger, and three fourths occurred in the first two decades of life.592 Meningitis is the most common antecedent in neonates and young infants, but can lead to abscess formation at any age.589,590,593–595 Otorhinologic sources are responsible for a significant percentage of brain abscess, although they were found to be the cause in less than 15% of recent cases in the United States,587 and in only 6% in a large Chinese report593; otitis and mastoid infection appear to be far more important than paranasal sinusitis.591,595 Trauma is an important cause in most pediatric series, usually seen in teenagers, although it remains more common among adults. Interestingly, no known antecedent is noted in up to 40% of cases,591 although this figure is significantly lower, approximately 5% to 15%, in most large pediatric series.589,590,593 These cases are usually presumed to be the result of bacteremic spread from an occult source, such as lung or soft tissue. The presentation of intraparenchymal abscesses reflects the infectious and space-occupying nature of these lesions. Combining four pediatric series totaling 346 patients,589–591,593
W9418-11.qxd 10/31/03 3:43 PM Page 318
318
Section I
Introduction
Figure 11-6. Noncontrast CT scan of an epidural abscess, presumably from sinus disease. A, The extensive subgaleal involvement over the forehead (Pott’s puffy tumor) is demonstrated. B, The epidural abscess causes mild mass effect on the frontal lobes.
headache was the most common sign or symptom (62%), followed by vomiting (53%), papilledema (53%), fever (46%), seizures (31%), and hemiparesis (32%). Patients who were alert and neurologically normal were exceedingly rare, as were deeply comatose patients (each seen in less than 10%); approximately one third were alert and with minor or moderate deficts.589 Another one third were obtunded, with moderate deficit, and approximately 15% were stuporous, with severe deficit.589 The pathogenesis of abscess formation is complex, and varies according to the primary cause, as does the responsible pathogen. Patients with cyanotic cardiac disease are at risk for several reasons, including the lack of normal pulmonary filtration, low oxygen tension, and polycythemia and elevated blood viscosity; this scenario sets the stage for bacterial deposition from septic emboli within abnormally perfused and oxygenated brain. Abscesses are multiple in approximately one third of these patients, and are classically found in the middle cerebral distribution at the grey-white interface; a wide variety of pathogens is seen, most commonly Peptostreptococcus, Staphylococcus, and Streptococcus species, although negative culture results are quite common.591,592 Infection from the middle ear, mastoid, and paranasal sinuses gains access to the brain via a similar
thrombophlebitic process seen in other forms of intracranial suppuration, and are usually solitary. Temporal or cerebellar abscesses generally result from an otogenic or mastoid focus, whereas paranasal sinus disease, more common in older children and teens, usually results in frontal abscess formation. The organisms cover a wide spectrum, and are often mixed. Abscesses due to neonatal sepsis and meningitis occur in any location, and are often multiple; their occurrence may result from an infectious vasculitis.596 The most frequent organisms are those seen in neonatal sepsis and meningitis, including group B Streptococcus, Escherichia coli, and Proteus species. Citrobacter diversus abscesses may complicate meningitis in more than 75% of affected patients.597 Direct focal bacterial deposition may occur as a result of open fracture, surgical procedure, or other penetrating injury, and are due to Staphylococcus or mixed flora. After bacterial inoculation, the infected brain responds with a characteristic pattern of inflammatory changes, initially with a focal area of cerebritis, progressing over several weeks into an encapsulated mass containing necrotic, purulent brain.598 These changes are reflected on CT and MRI, which may not show the classical rim-enhancing necrotic mass early in the course of abscess development. MRI is more sensitive in demonstrating small areas of infection,
W9418-11.qxd 10/31/03 3:43 PM Page 319
Chapter 11
particularly in the case of multiple abscesses, and more clearly demonstrates the relationship of the abscess wall to the ventricular system, which has important implications for management and outcome. The treatment strategy for intracranial abscess in children is no different than that of adult infection; goals are eradication of any primary source, bacterial identification, longterm antibiotic therapy, and, in most cases, drainage or excision. Very small lesions, or those in the earliest stages of cerebritis may be cured with antibiotic therapy alone,599 but in the absence of organism identification (blood culture, sinus aspiration), some form of drainage is necessary. Modern imaging has revolutionized the treatment of bacterial abscesses, and increasingly, aspiration techniques are being used for management,589,591,600,601 often coupled with image guidance. Such techniques are often more practical and less morbid than open excision, especially in cases of multiple abscesses, and may be preferable in children.591,602 In the pre-CT era, mortality rates for brain abscess ranged from 25% to 50%, while recent pediatric series generally report rates below 15%589,591; severely affected neonates remain at higher risk than older children. Intraventricular rupture is associated with a significant risk of death and disability,592 as is an initial moribund presentation. Long-term morbidity, especially intellectual impairment, remains high, especially among infants and young children.594,603,604 Pediatric Movement Disorders In the past 15 years, the pediatric neurosurgeon has assumed a greater role in the treatment of pediatric movement disorders. The two most common disorders, spasticity and dystonia, are usually seen in the setting of cerebral palsy, a condition occurring in approximately two per 1000 live births,605 although any cause of brain injury, including trauma, tumor, cerebral ischemia, or infection, as well as hereditary and metabolic disorders, may result in spasticity, dystonia, or a mixed disorder. The treatment of children with movement disorders has traditionally involved a multidisciplinary approach, using oral medications, rehabilitative therapy, orthopedic intervention, and neurosurgical procedures, including intramuscular botulinum injection, dorsal rhizotomy, and intrathecal baclofen delivery. Spasticity is defined as a velocity-dependent increase in muscle tone, often associated with hyperactive deep-tendon reflexes, weakness, and a breakdown in isolated movements and coordination. Affected patients may also have pain associated with tonic muscle contraction, and long-standing spasticity can result in contracture development and joint deformity. Any insult to the descending inhibitory tracts within the brain or spinal cord may interfere with segmental reflex circuitry, and result in a disorder characterized by spasticity, but the exact pathophysiologic mechanisms responsible for increased tone are unknown.606 Spasticity may be manifest in the lower extremities (spastic diplegia),
Pediatric Neurosurgery
319
all extremities (spastic quadriplegia), or may be unilateral. Dystonia, which often co-exists with spasticity in braininjured patients, is characterized by involuntary and sustained muscular contraction, resulting in abnormal twisting and repetitive movements, or abnormal postures607; such abnormal movements may increase in severity with attempted voluntary activity, or during periods of excitement or startle. Primary dystonias are hereditary disorders, many which have known genetic loci and patterns of inheritance,608 while secondary dystonias are due to an acquired insult to the developing or mature brain, usually within the basal ganglia. Heredo-degenerative dystonias are seen in the context of progressive neurologic or systemic conditions, such as Huntington’s disease or HallervordenSpatz disease.608,609 Dystonia can be focal or segmental in nature, or can involve only one side of the body (hemidystonia). Generalized dystonia, involving both axial and appendicular muscle groups, is the pattern most commonly seen in the context of cerebral palsy and the degenerative conditions. The specific pathways involved in the genesis of dystonia are unknown. Lesions affecting the feedback loop between the motor cortices and basal ganglia, causing impaired inhibition of the motor cortex, may lead to dystonic movement,610–612 although the exact pathophysiologic disturbances are an area of controversy. The evaluation of a child with a movement disorder requires a detailed perinatal history and evolution of the disability over time. If the cause of the disorder is in doubt (absence of prematurity, hypoxia, or other known insult), a careful family history must be obtained, and genetic or metabolic testing may be required if such a disorder is suspected. Brain MRI may not be necessary in cases which are clearly due to perinatal insult, but is needed to search for structural lesions that can impact motor function in questionable cases; imaging may be helpful to assess ventricular size, as well as basal ganglia or white matter lesions in patients with cerebral palsy. Input from a multidisciplinary team requires a thorough assessment of the type and severity of the movement disorder, addressing not only the abnormal motor pattern, but also the degree to which this affects strength, coordination, and function. Spasticity is usually assessed utilizing the Ashworth scale, a five-point grading system based upon resistance to passive range of motion, in which a score of 1 is normal tone, and 5 is a rigid, immobile limb.613 Dystonia is assessed by observing the patient’s pattern of involuntary movement at rest, and often during volitional activity. A variety of scales have been used to grade dystonia; the recently described Barry-Albright dystonia scale assesses the severity and distribution of dystonia, as well as its impact on function.614 Input from patients and parents regarding goals of treatment is a vital part of the evaluation process, and has important implications for decisions regarding therapeutic intervention. The optimal treatment of a patient with spasticity depends on the age of the patient and the severity of the dis-
W9418-11.qxd 10/31/03 3:43 PM Page 320
320
Section I
Introduction
ability. Mild spasticity may not require any treatment, particularly in cases of spastic diplegia associated with cerebral palsy, as some children are only minimally incapacitated, and may actually improve with time.615 For most children with spasticity, oral agents, alone or in combination, are the initial form of therapy. Baclofen, a GABAB agonist, reduces spasticity by causing hyperpolarization of neurons in the spinal cord, resulting in enhanced presynaptic inhibition. In a double-blind study, baclofen was found to be effective in reducing tone in children with spasticity due to cerebral palsy.616 Benzodiazepines also increase postsynaptic inhibition by enhancing the affinity of GABA to the GABAA receptor, and are effective at reducing tone in children with cerebral palsy.617 The beneficial effects of both baclofen and the benzodiazepines may be limited by sedation, and cessation of either of these drugs may be associated with withdrawal. Alpha-2 adrenergic agonists, such as tizanadine, decrease spasticity by interfering with excitatory amino acid release618 and inhibition of spinal cord interneuronal activity.619 Multiple double-blind, randomized studies have demonstrated efficacy in tone reduction similar to baclofen and benzodiazepines in adults with spasticity from various causes,620,621 although tizanidine’s effect in children has not been similarly evaluated. Dantrolene inhibits calcium release from the sarcoplasmic reticulum, inhibiting excitationcontraction coupling within skeletal muscle. Although it has been shown effective in treating spasticity in children with cerebral palsy,617 its use has been supplanted by newer, centrally acting agents. Botulinum toxin type A (Botox) is a purified form of one of the seven serotypes of toxins produced by Clostridia botulinum, and causes muscle weakness by binding to presynaptic cholinergic nerve terminals at the neuromuscular junction, inhibiting release of acetylcholine into the synaptic cleft.622 In patients with spasticity, botulinum toxin type A is injected directly into spastic muscles, resulting in temporary weakness, to equilibrate agonist-antagonist forces across the affected joints. The clinical response begins within several days, peaks within a few weeks, and lasts for several months, as vesicle turnover at the motor end-plate restores cholinergic transmission. Botulinum toxin type A therapy for children with spasticity is usually combined with oral agents and aggressive physiotherapy; because it only works on a temporary basis, serial injections are usually used in younger children until more definitive forms of therapy, if necessary, are used. Multiple double-blind, randomized studies have demonstrated functional improvement in children with cerebral palsy and other forms of spasticity,623–626 with a paucity of side effects. By directly causing muscular relaxation, botulinum toxin type A therapy may reduce the development of contracture formation, delaying or potentially obviating orthopedic intervention.627 The sectioning of dorsal rootlets involved in the spastic reflex response was described in the early 1900s, although the popularity of selective dorsal rhizotomy for the treatment of
spasticity escalated after Peacock’s reports in the 1980s demonstrated its efficacy for children with spastic cerebral palsy.628,629 The ideal candidates for the procedure are children with spastic diplegia, approximately 4 to 8 years old, who are independent or partially dependent ambulators; relative contraindications include the presence of significant dystonia or upper extremity involvement. The procedure performed in most centers involves a multilevel osteoplastic laminotomy, intradural stimulation of individual dorsal rootlets between L2 and S2, and the sectioning of those with abnormal electrophysiologic or clinical responses, although the validity of the electrophysiological parameters used to “selectively” cut individual rootlets has been questioned.630 Outcomes after selective dorsal rhizotomy have been the subject of multiple reports that were recently analyzed by Steinbok.631 In this evidence-based review of 63 articles, the procedure was deemed effective in reducing lower extremity tone, increasing range of motion, and improving lower extremity function. Additionally, there was weaker evidence that rhizotomy may improve upper extremity function via suprasegmental interneuronal effects in some patients, and may reduce the need for subsequent orthopedic procedures.631 Intrathecal baclofen (ITB) has become an increasingly popular treatment for children with spasticity due to its ability to concentrate the drug at its site of action within the spinal cord, while minimizing the side effects associated with systemic delivery. Candidates for ITB include patients with spasticity or a mixed movement disorder, if they are of sufficient size to accommodate a pump. Response to ITB is usually gauged via instillation of a test dose via lumbar puncture, with subsequent implantation of an externally programmable pump and intrathecal catheter in those with clear improvement in lower extremity tone. The use of ITB has been shown to be beneficial in children with spasticity due to CP and other insults.632–634 Unlike rhizotomy, ITB may be used in patients with spastic quadriplegia or dystonia. Additionally, ITB is not a destructive procedure, and the dose can be titrated to effect. Its main disadvantages include the need for serial pump refills and eventual replacement, and the risks of pump or catheter dysfunction, and less commonly, infection.633 ITB has not been subject to the same level of scrutiny as rhizotomy, and there have been no reports prospectively comparing the efficacy of these procedures, although the cost of ITB therapy is significantly higher than that of rhizotomy.635 The treatment of dystonia in children is often more challenging than that for spasticity. Generalized dystonia, the pattern most prevalent in cerebral palsy, may be modestly improved with oral baclofen, but the doses necessary to impart significant benefit are often associated with unacceptable sedation. The anticholinergic agent trihexyphenidyl may be used alone or in combination with baclofen in cases of generalized dystonia, and appears to be more effective in younger patients, with a lower side-effect profile.636,637 One
W9418-11.qxd 10/31/03 3:43 PM Page 321
Chapter 11
Pediatric Neurosurgery
321
P earls 1. The recent Shunt Design Trial6 found no differences in complication or shunt survival rates among standard differential-pressure, anti-siphon (Delta), or flow-limiting (Orbis-Sigma) valve systems in 344 children randomized at first shunt insertion. The overall rates of shunt obstruction (31.4%) and infection (8.1%) were not significantly different between the systems. Kaplan-Meyer analysis showed shunt survival rates of 61% and 47%, at 1 and 2 years, respectively. 2. Most shunt infections occur within several months of insertion, with Staphylococcus epidermidis and Staphylococcus aureus the most common organisms isolated.87 Fever, irritability, signs of shunt malfunction, and erythema around the incision site are the most common findings. Uncommonly, infection can present in a delayed fashion88; these cases are notable for the frequent association with abdominal pseudocyst, and the variety of organisms isolated, including enteric flora. 3. A prospective evaluation of risk factors for infection in 299 patients undergoing shunt placement was recently reported.93 The rate of infection was 10.1%, with Staphylococcus species the most common isolated pathogens. Postoperative CSF leak, contamination of implant by a breached glove, and patient age less than 40 weeks’ gestation were found to be significant risk factors. 4. MMC is immediately apparent at birth, as an open spinal defect, usually at the lumbar or sacral level; thoracic lesions account for approximately 20% of lesions, and cervical MMC is rare. 5. Despite its nearly universal presence in MMC, Chiari II malformations become symptomatic in only a minority of patients, approximately 10% to 20%,128–130 resulting in lower cranial nerve dysfunction (stridor, nasopharyngeal regurgitation, aspiration, apnea), quadriparesis, and cerebellar signs. Symptoms secondary to lower cranial neuropathy may be life-threatening.
of the primary dystonias is characterized by dramatic response to l-dopa therapy,638 although this agent is not effective for the secondary dystonias. Botulinum toxin type A can reduce dystonia by its effect on the neuromuscular junction, and is most practical for patients with nongeneralized involvement639,640; intramuscular injection may secondarily affect cortical excitatory and inhibitory circuitry, as measured by response to transcranial magnetic stimulation.641 The use of intrathecal baclofen for generalized dystonia in children has recently been shown to be effective
6. In spina bifida patients with symptomatic spinal cord tethering, stabilization or improvement can be seen in many cases, especially with respect to pain, if detethering is performed during the early onset of symptoms. Scoliosis tends to stabilize if untethering is performed before the curvature has progressed past approximately 40 to 50 degrees.140,141 7. Because of the near universal development and progression of deficits in patients with tethered cord due to lipomyelomeningocele or fatty filum, most authors advocate surgery at the time of diagnosis, to prevent or stabilize neurological dysfunction. Bladder dysfunction is unlikely to recover completely after it is lost, and motor signs may not improve significantly once progressive weakness occurs. 8. The pathogenesis of syrinx formation in Chiari malformation (CM1) is unknown; recent reports analyzing CSF flow and subarachnoid pressures in both CM1 patients and unaffected controls revealed that tonsillar impaction at the foramen magnum results in elevation of the cervical subarachnoid pressure, forcing CSF transmurally into the cord from the subarachnoid space.195,196 9. Because of vague circumstances surrounding shaken baby (impact) syndrome, abuse is typically inferred from examination and findings on cranial and skeletal imaging. The diagnosis has no strict criteria, but is usually made in the presence of extra-axial hemorrhage, retinal hemorrhage, external signs of trauma, and a vague or incompatible mechanism of injury. 10. Severe spinal cord injury, especially in the cervical region, may be associated with unexplained hypotension without compensatory tachycardia. 11. Because of their relatively large head size, young children placed directly upon a rigid transport board will be forced into a relatively flexed position; their back and shoulders must be “built up” to avoid this posture.370,371
not only in reducing dystonia scores, but in improving quality of life.642,643 In these reports, the doses necessary to control dystonia were significantly higher than those used for pure spasticity. The use of thalamotomy and pallidotomy for dystonia in children is limited, and has shown modest success.644 As experience with deep brain stimulation expands, as it has for the treatment of parkinsonian symptoms, this nondestructive modality may find utility in the treatment of children with refractory hyperkinetic movement disorders.
W9418-11.qxd 10/31/03 3:43 PM Page 322
322
Section I
Introduction
References 1. Frankenburg WK, Dodds J, Archer P, Shapiro H, Bresnick B: The Denver II: A major revision and restandardization of the Denver Developmental Screening Test. Pediatrics 1992;89:91–97. 2. Painter MJ: Neurological evaluation of newborns, infants, and older children. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:3–19. 3. Lahat E, Heyman E, Barkay A, Goldberg M: Asymmetric crying facies and associated congenital anomalies: Prospective study and review of the literature. J Child Neurol 2000;15:808–810. 4. van Gijn J: The Babinski sign: The first hundred years. J Neurol 1996;243:675–683. 5. Nulsen FE, Spitz EB: Treatment of hydrocephalus by direct shunt from ventricle to jugular vein. Surg Forum 1952;2:399–403. 6. Drake JM, Kestle JR, Milner R, et al: Randomized trial of cerebrospinal fluid shunt valve design in pediatric hydrocephalus. Neurosurgery 1998;43:294–303. 7. Bondurant CP, Jimenez DF: Epidemiology of cerebrospinal fluid shunting. Pediatr Neurosurg 1995;23:254–258. 8. Rekate HL: Treatment of hydrocephalus. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:47–73. 9. Rosseau GL, McCullough DC, Joseph AL: Current prognosis in fetal ventriculomegaly. J Neurosurg 1992;77:551–555. 10. Hudgins RJ, Edwards MS, Goldstein R, et al: Natural history of fetal ventriculomegaly. Pediatrics 1988;82:692–697. 11. Vintzileos AM, Campbell WA Weinbaum PJ, Nochimson DJ: Perinatal management and outcome of fetal ventriculomegaly. Obstet Gynecol 1987;69:5–11. 12. Wilhelm C, Keck C, Hess S, Korinthenberg R, Breckwoldt M: Ventriculomegaly diagnosed by prenatal ultrasound and mental development of the children. Fetal Diagn Ther 1998;13:162–166. 13. Gupta JK, Bryce FC, Lilford RJ: Management of apparently isolated fetal ventriculomegaly. Obstet Gynecol Surv 1994;49:716–721. 14. Drugan A, Krause B, Canady A, Zador IE, Sacks AJ, Evans MI: The natural history of prenatally diagnosed cerebral ventriculomegaly. JAMA 1989;261:1785–1788. 15. Twining P, Jaspan T, Zuccollo J: The outcome of fetal ventriculomegaly. Br J Radiol 1994;67:26–31. 16. Gurtner P, Bass T, Gudeman SK, Penix JO, Philput CB, Schinco FP: Surgical management of posthemorrhagic hydrocephalus in 22 low-birth-weight infants. Childs Nerv Syst 1992;8:198–202. 17. Du Plessis AJ: Posthemorrhagic hydrocephalus and brain injury in the preterm infant: Dilemmas in diagnosis and management. Semin Pediatr Neurol 1998;5:161–179. 18. Fernell E, Hagberg G, Hagberg B: Infantile hydrocephalus epidemiology: An indicator of enhanced survival. Arch Dis Child Fetal Neonatal Ed 1994;70:F123–128. 19. Papile LA, Burstein J, Burstein R, Koffler H: Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birth weights less than 1,500 gm. J Pediatr 1978;92:529–534. 20. Vohr B, Allan WC, Scott DT, et al: Early-onset intraventricular hemorrhage in preterm neonates: Incidence of neurodevelopmental handicap. Semin Perinatol 1999;23:212–217. 21. Resch B, Gedermann A, Maurer U, Ritschl E, Muller W: Neurodevelopmental outcome of hydrocephalus following intra/periventricular hemorrhage in preterm infants: Short- and long-term results. Childs Nerv Syst 1996;12:27–33. 22. Pikus HJ, Levy ML, Gans W, Mendel E, McComb JG: Outcome, cost analysis, and long-term follow-up in preterm infants with massive grade IV germinal matrix hemorrhage and progressive hydrocephalus. Neurosurgery 1997;40:983–988. 23. Perlman JM, Volpe JJ: Intraventricular hemorrhage in extremely small premature infants. Am J Dis Child 1986;140:1122–1124.
24. Gleissner M, Jorch G, Avenarius S: Risk factors for intraventricular hemorrhage in a birth cohort of 3721 premature infants. J Perinat Med 2000;28:104–110. 25. Perlman JM, Rollins N: Surveillance protocol for the detection of intracranial abnormalities in premature neonates. Arch Pediatr Adolesc Med 2000;154:822–826. 26. Tommiska V, Heinonen K, Ikonen S, et al: A national short-term follow-up study of extremely low birth weight infants born in Finland in 1996–1997. Pediatrics 2001;107:E2. 27. Harding D, Kuschel C, Evans N: Should preterm infants born after 29 weeks’ gestation be screened for intraventricular haemorrhage? J Paediatr Child Health 1998;34:57–59. 28. Meek JH, Tyszczuk L, Elwell CE, Wyatt JS: Low cerebral blood flow is a risk factor for severe intraventricular haemorrhage. Arch Dis Child Fetal Neonatal Ed 1999;81:F15–18. 29. DiSalvo D: The correlation between placental pathology and intraventricular hemorrhage in the preterm infant. The Developmental Epidemiology Network Investigators. Pediatr Res 1998;43:15–19. 30. Vergani P, Patane L, Doria P, Borroni C, Cappellini A, Pezzullo JC, Ghidini A: Risk factors for neonatal intraventricular haemorrhage in spontaneous prematurity at 32 weeks gestation or less. Placenta 2000;21:402–407. 31. Duncan CC, Chiang VL: Intraventricular hemorrhage and posthemorrhagic hydrocephalus. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:107–124. 32. Guidelines for the Management of Severe Head Injury. Park Ridge, IL, American Association of Neurological Surgeons, 1995. 33. Hanigan WC, Kennedy G, Roemisch F, Anderson R, Cusack T, Powers W: Administration of indomethacin for the prevention of periventricular–intraventricular hemorrhage in high–risk neonates. J Pediatr 1988;112:941–947. 34. Bada HS, Green RS, Pourcyrous M, et al: Indomethacin reduces the risks of severe intraventricular hemorrhage. J Pediatr 1989;115: 631–637. 35. Ment LR, Oh W, Ehrenkranz RA, et al. Low–dose indomethacin and prevention of intraventricular hemorrhage: a multicenter randomized trial. Pediatrics 1994;93:543–550. 36. Fowlie PW: Intravenous indomethacin for preventing mortality and morbidity in very low birth weight infants. Cochrane Database Syst Rev 2000;(2):CD000174. 37. Ment LR, Duncan CC, Scott DT, Ehrenkranz RA: Posthemorrhagic hydrocephalus. Low incidence in very low birth weight neonates with intraventricular hemorrhage. J Neurosurg 1984;60:343–347. 38. Allan WC, Dransfield DA, Tito AM: Ventricular dilation following periventricular–intraventricular hemorrhage: outcome at age 1 year. Pediatrics 1984;73:58–162. 39. Gurtner P, Bass T, Gudeman SK, Penix JO, Philput CB, Schinco FP: Surgical management of posthemorrhagic hydrocephalus in 22 low-birth-weight infants. Childs Nerv Syst 1992;8:198–202. 40. Volpe JJ. Neonatal intracranial hemorrhage. Pathophysiology, neuropathology, and clinical features. Clin Perinatol 1977;4:77–102. 41. Ventriculomegaly Trial Group: Randomised trial of early tapping in neonatal posthaemorrhagic ventricular dilatation. Arch Dis Child 1990;65:3–10. 42. Anwar M, Kadam S, Hiat IM, Hegyi T: Serial lumbar punctures in prevention of post–hemorrhagic hydrocephalus in preterm infants. J Pediatr 1985;107:446–450. 43. Marlin AE: Protection of the cortical mantle in premature infants with posthemorrhagic hydrocephalus. Neurosurgery 1980;7:464–468. 44. Benzel EC, Reeves JP, Nguyen PK, Hadden TA: The treatment of hydrocephalus in preterm infants with intraventricular haemorrhage. Acta Neurochir (Wien) 1993;122:200–203. 45. Gaskill SJ, Marlin AE, Rivera S: The subcutaneous ventricular reservoir: An effective treatment for posthemorrhagic hydrocephalus. Childs Nerv Syst 1988;4:291–295.
W9418-11.qxd 10/31/03 3:43 PM Page 323
Chapter 11 46. Anwar M, Doyle AJ, Kadam S, Hiatt IM, Hegyi T: Management of posthemorrhagic hydrocephalus in the preterm infant. J Pediatr Surg 1986;21:334–347. 47. McComb JG, Ramos AD, Platzker AC, Henderson DJ, Segall HD: Management of hydrocephalus secondary to intraventricular hemorrhage in the preterm infant with a subcutaneous ventricular catheter reservoir. Neurosurgery 1983;13:295–300. 48. Fulmer BB, Grabb PA, Oakes WJ, Mapstone TB: Neonatal ventriculosubgaleal shunts. Neurosurgery 2000;47:80–83. 49. Heep A, Engelskirchen R, Holschneider A, Groneck P: Primary intervention for posthemorrhagic hydrocephalus in very low birthweight infants by ventriculostomy. Childs Nerv Syst 2001;17:47–51. 50. Cornips E, Van Calenbergh F, Plets C, devlieger H, Casaer P: Use of external drainage for posthemorrhagic hydrocephalus in very low birth weight premature infants. Childs Nerv Syst 1997;13:369–374. 51. Weninger M, Salzer HR, Pollak A, et al: External ventricular drainage for treatment of rapidly progressive posthemorrhagic hydrocephalus. Neurosurgery 1992;31:52–57. 52. Whitelaw A: Repeated lumbar or ventricular punctures in newborns with intraventricular hemorrhage. Cochrane Database Syst Rev 2001;(1):CD000216. 53. Pang D, Sclabassi RJ, Horton JA: Lysis of intraventricular blood clot with urokinase in a canine model: Part 3. Effects of intraventricular urokinase on clot lysis and posthemorrhagic hydrocephalus. Neurosurgery 1986;19:553–572. 54. Whitelaw A, Rivers RP, Creighton L, Gaffney P: Low dose intraventricular fibrinolytic treatment to prevent posthaemorrhagic hydrocephalus. Arch Dis Child 1992;67:12–14. 55. Whitelaw A, Saliba E, Fellman V, Mowinckel MC, Acolet D, Marlow N: Phase I study of intraventricular recombinant tissue plasminogen activator for treatment of posthaemorrhagic hydrocephalus. Arch Dis Child Fetal Neonatal Ed 1996;75:F20–26. 56. Hudgins RJ, Boydston WR, Hudgins PA, Adler SR: Treatment of intraventricular hemorrhage in the premature infant with urokinase. A preliminary report. Pediatr Neurosurg 1994;20:190–197. 57. Hudgins RJ, Boydston WR, Hudgins PA, Morris R, Adler SM, Gilreath CL: Intrathecal urokinase as a treatment for intraventricular hemorrhage in the preterm infant. Pediatr Neurosurg 1997;26:281–287. 58. Luciano R, Velardi F, Romagnoli C, Papacci P, De Stefano V, Tortorolo G: Failure of fibrinolytic endoventricular treatment to prevent neonatal post-haemorrhagic hydrocephalus. A case–control trial. Childs Nerv Syst 1997;13:73–76. 59. Hansen AR, Volpe JJ, Goumnerova LC, Madsen JR: Intraventricular urokinase for the treatment of posthemorrhagic hydrocephalus. Pediatr Neurol 1997;17:213–217. 60. Whitelaw A: Intraventricular streptokinase after intraventricular hemorrhage in newborn infants. Cochrane Database Syst Rev 2001;(1):CD000498. 61. Haines SJ, Lapointe M: Fibrinolytic agents in the management of posthemorrhagic hydrocephalus in preterm infants: The evidence. Childs Nerv Syst 1999;15:226–234. 62. Libenson MH, Kaye EM, Rosman NP, Gilmore HE: Acetazolamide and furosemide for posthemorrhagic hydrocephalus of the newborn. Pediatr Neurol 1999;20:185–191. 63. Shinnar S, Gammon K, Bergman EW, Jr, Epstein M, Freeman JM: Management of hydrocephalus in infancy: Use of acetazolamide and furosemide to avoid cerebrospinal fluid shunts. J Pediatr 1985;107:31– 37. 64. International PHVD Drug Trial Group: International randomised controlled trial of acetazolamide and furosemide in posthaemorrhagic ventricular dilatation in infancy. Lancet 1998;352:433–440. 65. Rintoul NE, Sutton LN, Hubbard AM, Cohen B, Melchionni J, Pasquariello PS, Adzick NS: A new look at myelomeningoceles: Functional level, vertebral level, shunting, and the implications for fetal intervention. Pediatrics 2002;109:409–413.
Pediatric Neurosurgery
323
66. Parent AD, McMillan T: Contemporaneous shunting with repair of myelomeningocele. Pediatr Neurosurg 1995;22:132–135. 67. Miller PD, Pollack IF, Pang D, Albright AL: Comparison of simultaneous versus delayed ventriculoperitoneal shunt insertion in children undergoing myelomeningocele repair. J Child Neurol 1996;11:370– 372. 68. Brockmeyer D, Abtin K, Carey L, Walker ML: Endoscopic third ventriculostomy: An outcome analysis. Pediatr Neurosurg 1998;28:236– 240. 69. Teo C, Jones R: Management of hydrocephalus by endoscopic third ventriculostomy in patients with myelomeningocele. Pediatr Neurosurg 1996;25:57–63. 70. Piatt JH, Jr, Carlson CV: A search for determinants of cerebrospinal fluid shunt survival: Retrospective analysis of a 14-year institutional experience. Pediatr Neurosurg 1993;19:233–241. 71. James HE, Bruce DA: Ventriculoperitoneal shunts for hydrocephalus: A focus group discussion on the selection of shunt systems in pediatrics. A report of the Pediatric Neurosurgery Research Group meeting, December 1992. Childs Nerv Syst 1995;11:449–451. 72. Lee TT, Uribe J, Ragheb J, Morrison G, Jagid JR: Unique clinical presentation of pediatric shunt malfunction. Pediatr Neurosurg 1999;30:122–126. 73. Kestle J, Drake J, Milner R, Sainte-Rose C, et al: Long-term follow-up data from the Shunt Design Trial. Pediatr Neurosurg 2000;33:230– 236. 74. Tuli S, O’Hayon B, Drake J, Clarke M, Kestle J: Change in ventricular size and effect of ventricular catheter placement in pediatric patients with shunted hydrocephalus. Neurosurgery 1999;45:1329–1333. 75. Pollack IF, Albright AL, Adelson PD: A randomized, controlled study of a programmable shunt valve versus a conventional valve for patients with hydrocephalus. Hakim-Medos Investigator Group. Neurosurgery 1999;45:1399–1408. 76. Jain H, Sqouros S, Walsh AR, Hockley AD: The treatment of infantile hydrocephalus: “Differential-pressure” or “flow-control” valves. A pilot study. Childs Nerv Syst 2000;16:242–246. 77. Albright AL, Haines SJ, Taylor FH: Function of parietal and frontal shunts in childhood hydrocephalus. J Neurosurg 1988;69:883–886. 78. Sainte-Rose C, Piatt JH, Renier D, et al: Mechanical complications in shunts. Pediatr Neurosurg 1991;17:2–9. 79. Bierbrauer KS, Storrs BB, McLone DG, Tomita T, Dauser R: A prospective, randomized study of shunt function and infections as a function of shunt placement. Pediatr Neurosurg 1990;16:287–291. 80. Tuli S, Drake J, Lawless J, Wigg M, Lamberti-Pasculli M: Risk factors for repeated cerebrospinal shunt failures in pediatric patients with hydrocephalus. J Neurosurg 2000;92:31–38. 81. Piatt JH Jr: Pumping the shunt revisited. A longitudinal study. Pediatr Neurosurg 1996;25:73–76. 82. Iskandar BJ, McLaughlin C, Mapstone TB, Grabb PA, Oakes WJ: Pitfalls in the diagnosis of ventricular shunt dysfunction: Radiology reports and ventricular size. Pediatrics 1998;101:1031–1036. 83. Noetzel MJ, Baker RP: Shunt fluid examination: Risks and benefits in the evaluation of shunt malfunction and infection. J Neurosurg 1984;61:328–332. 84. Sood S, Kim S, Ham SD, Canady AI, Greninger N: Useful components of the shunt tap test for evaluation of shunt malfunction. Childs Nerv Syst 1993;9:157–161. 85. May CH, Aurisch R, Kornrumpf D, Vogel S: Evaluation of shunt function in hydrocephalic patients with the radionuclide 99mTcpertechnetate. Childs Nerv Syst 1999;15:239–244. 86. Duhaime AC, Bonner K, McGowan KL, Schut L, Sutton LN, Plotkin S: Distribution of bacteria in the operating room environment and its relation to ventricular shunt infections: A prospective study. Childs Nerv Syst 1991;7:211–214. 87. Davis SE, Levy ML, McComb JG, Masri-Lavine L: Does age or other factors influence the incidence of ventriculoperitoneal shunt infections? Pediatr Neurosurg 1999;30:253–257.
W9418-11.qxd 10/31/03 3:43 PM Page 324
324
Section I
Introduction
88. Baird C, O’Connor D, Pittman T: Late shunt infections. Pediatr Neurosurg 1999;31:269–273. 89. Salomao JF, Leibinger RD: Abdominal pseudocysts complicating CSF shunting in infants and children. Report of 18 cases. Pediatr Neurosurg 1999;31:274–278. 90. Roitberg BZ, Tomita T, McLone DG: Abdominal cerebrospinal fluid pseudocyst: A complication of ventriculoperitoneal shunt in children. Pediatr Neurosurg 1998;29:267–273. 91. Morissett I, Gourdeau M, Francoeur J: CSF shunt infections: A fifteenyear experience with emphasis on management and outcome. Can J Neurol Sci 1993;20:118–122. 92. Choux M, Genitori L, Lang D, Lena G: Shunt implantation: Reducing the incidence of shunt infection. J Neurosurg 1992;77:875– 880. 93. Kulkarni AV, Drake JM, Lamberti-Pasculli M: Cerebrospinal fluid shunt infection: A prospective study of risk factors. J Neurosurg 2001;94:195–201. 94. Kestle JR, Hoffman HJ, Soloniuk D, Humphreys RP, Drake JM, Hendrick EB: A concerted effort to prevent shunt infection. Childs Nerv Syst 1993;9:163–165. 95. Haines SJ, Walters BC: Antibiotic prophylaxis for cerebrospinal fluid shunts: A metanalysis. Neurosurg 1994;34:87–92. 96. Lewis AI, Keiper GL Jr, Crone KR: Endoscopic treatment of loculated hydrocephalus. J Neurosurg 1995;82:780–785. 97. Tuli S, Alshail E, Drake J: Third ventriculostomy versus cerebrospinal fluid shunt as a first procedure in pediatric hydrocephalus. Pediatr Neurosurg 1999;30:11–15. 98. Cinalli G, Sainte-Rose C, Chumas P, et al: Failure of third ventriculostomy in the treatment of aqueductal stenosis in children. J Neurosurg 1999;90:448–454. 99. Goumnerova LC, Frim DM: Treatment of hydrocephalus with third ventriculocisternostomy: Outcome and CSF flow patterns. Pediatr Neurosurg 1997;27:149–152. 100. Valenzuela S, Trellez A: Pediatric neuroendoscopy in Chile. Analysis of the first 100 cases. Childs Nerv Syst 1999;15:457–460. 101. Scarrow AM, Levy EI, Pascucci L, Albright AL: Outcome analysis of endoscopic III ventriculostomy. Childs Nerv Syst 2000;16:442– 444. 102. Brockmeyer D, Abtin K, Carey L, Walker ML: Endoscopic third ventriculostomy: An outcome analysis. Pediatr Neurosurg 1998;28: 236–240. 103. Jones RF, Kwok BC, Stening WA, Vonau M: Neuroendoscopic third ventriculostomy. A practical alternative to extracranial shunts in noncommunicating hydrocephalus. Acta Neurochir Suppl 1994;61:79–83. 104. Abtin K, Thompson BG, Walker ML: Basilar artery perforation as a complication of endoscopic third ventriculostomy. Pediatr Neurosurg 1998;28:35–41. 105. Stein SC, Feldman JG, Friedlander M, Klein RJ: Is myelomeningocele a disappearing disease? Pediatrics 1982;69:511–514. 106. Northrup H, Volcik KA: Spina bifida and other neural tube defects. Curr Probl Pediatr 2000;30:313–332. 107. Roberts HE, Moore CA, Cragan JD, Fernhoff PM, Khoury MJ: Impact of prenatal diagnosis on the birth prevalence of neural tube defects, Atlanta, 1990–1991. Pediatrics 1995;96:880–883. 108. Hendricks KA, Simpson JS, Larsen RD: Neural tube defects along the Texas-Mexico border, 1993–1995. Am J Epidemiol 1999;149:1119– 1127. 109. Locksmith GJ, Duff P: Preventing neural tube defects: The importance of periconceptional folic acid supplements. Obstet Gynecol 1998;91: 1027–1034. 110. Cuckle H, Wald N: The impact of screening for open neural tube defects in England and Wales. Prenat Diagn 1987;7:91–99. 111. Mills JL, Rhoads GG, Simpson JL, et al: The absence of a relation between the periconceptional use of vitamins and neural-tube defects. National Institute of Child Health and Human Development Neural Tube Defects Study Group. N Engl J Med 1989;321:430–435.
112. Lumley J, Watson L, Watson M, Bower C: Periconceptional supplementation with folate and/or multivitamins for preventing neural tube defects. Cochrane Database Syst Rev 2000;(2):CD001056. 113. Murphy M, Whiteman D, Stone D, Botting B, Schorah C, Wild J: Dietary folate and the prevalence of neural tube defects in the British Isles: The past two decades. BJOG 2000;107:885–889. 114. Laurence KM, James N, Miller MH, Tennant GB, Campbell H: Double-blind randomised controlled trial of folate treatment before conception to prevent recurrence of neural-tube defects. Br Med J (Clin Res Ed) 1981;282:1509–1511. 115. Werler MM, Shapiro S, Mitchell AA: Periconceptional folic acid exposure and risk of occurrent neural tube defects. JAMA 1993;269:1257–1261. 116. Shaw GM, Schaffer D, Velie EM, Morland K, Harris JA: Periconceptional vitamin use, dietary folate, and the occurrence of neural tube defects. Epidemiology 1995;6:219–226. 117. Thompson S, Torres M, Stevenson R, Dean J, Best R: Periconceptional vitamin use, dietary folate and occurrent neural tube defected pregnancies in a high risk population. Ann Epidemiol 2000;10:476. 118. van der put NM, Blom HJ: Neural tube defects and a disturbed folate dependent homocysteine metabolism. Eur J Obstet Gynecol Reprod Biol 2000;92:57–61. 119. Lindhout D, Omtzigt JG: Teratogenic effects of antiepileptic drugs: Implications for the management of epilepsy in women of childbearing age. Epilepsia 1994;35:S19–28. 120. Shaw GM, Todoroff K, Schaffer DM, Selvin S: Maternal height and prepregnancy body mass index as risk factors for selected congenital anomalies. Paediatr Perinat Epidemiol 2000;14:234–239. 121. Merrill DC, Goodwin P, Burson JM, Sato Y, Williamson R, Weiner CP: The optimal route of delivery for fetal meningomyelocele. Am J Obstet Gynecol 1998;179:235–240. 122. Shurtleff DB, Luthy DA, Nyberg DA, Mack LA: The outcome of fetal myelomeningocele brought to term. Eur J Pediatr Surg 1994;4:25–28. 123. Heffez DS, Aryanpur J, Hutchins GM, Freeman JM: The paralysis associated with myelomeningocele: clinical and experimental data implicating a preventable spinal cord injury. Neurosurgery 1990;26:987–992. 124. Chadduck WM, Reding DL: Experience with simultaneous ventriculo-peritoneal shunt placement and myelomeningocele repair. J Pediatr Surg 1988;23:913–916. 125. Miller PD, Pollack IF, Pang D, Albright AL: Comparison of simultaneous versus delayed ventriculoperitoneal shunt insertion in children undergoing myelomeningocele repair. J Child Neurol 1996;11:370– 372. 126. McLone DG, Dias MS: Complications of myelomeningocele closure. Pediatr Neurosurg 1991;17:267–273. 127. McLone DG, Knepper PA: The cause of Chiari II malformation: A unified theory. Pediatr Neurosci 1989;15:1–12. 128. Pollack IF, Pang D, Albright AL, Krieger D: Outcome following hindbrain decompression of symptomatic Chiari malformations in children previously treated with myelomeningocele closure and shunts. J Neurosurg 1992;77:881–888. 129. Pollack IF, Kinnunen D, Albright AL: The effect of early craniocervical decompression on functional outcome in neonates and young infants with myelodysplasia and symptomatic Chiari II malformations: Results from a prospective series. Neurosurgery 1996;38:703– 710. 130. Vandertop WP, Asai A, Hoffman HJ, et al: Surgical decompression for symptomatic Chiari II malformation in neonates with myelomeningocele. J Neurosurg 1992;77:541–544. 131. Gilbert JN, Jones KL, Rorke LB, Chernoff GF, James HE: Central nervous system anomalies associated with meningomyelocele, hydrocephalus, and the Arnold–Chiari malformation: Reappraisal of theories regarding the pathogenesis of posterior neural tube closure defects. Neurosurgery 1986;18:559–564.
W9418-11.qxd 10/31/03 3:43 PM Page 325
Chapter 11 132. Teo C, Parker EC, Aureli S, Boop FA: The Chiari II malformation: A surgical series. Pediatr Neurosurg 1997;27:223–229. 133. Dillon C, Davis BE, Duguay S, Seidel KD, Shurtleff DB: Longevity of patients born with myelomeningocele. Eur J Pediatr Surg 2000;10:33– 34. 134. Bowman RM, McLone DG, Grant JA, Tomita T, Ito JA: Spina bifida outcome: A 25-year prospective. Pediatr Neurosurg 2001;34:114–120. 135. Steinbok P, Irvine B, Cochrane DD, Irwin BJ: Long-term outcome and complications of children born with meningomyelocele. Childs Nerv Syst 1992;8:92–96. 136. Wills KE, Holmbeck GN, Dillon K, McLone DG: Intelligence and achievement in children with myelomeningocele. J Pediatr Psychol 1990;15:161–176. 137. Casari EF, Fantino AG: A longitudinal study of cognitive abilities and achievement status of children with myelomeningocele and their relationship with clinical types. Eur J Pediatr Surg 1998;8:52–54. 138. Mapstone TB, Rekate HL, Nulsen FE, Dixon MS, Jr, Glaser N, Jaffe M: Relationship of CSF shunting and IQ in children with myelomeningocele: A retrospective analysis. Childs Brain 1984;11:112–118. 139. Hunt GM: Open spina bifida: Outcome for a complete cohort treated unselectively and followed into adulthood. Dev Med Child Neurol 1990;32:108–118. 140. Herman JM, McLone DG, Storrs BB, Dauser RC: Analysis of 153 patients with myelomeningocele or spinal lipoma reoperated upon for a tethered cord. Presentation, management and outcome. Pediatr Neurosurg 1993;19:243–249. 141. Pierz K, Banta J, Thomson J, Gahm N, Hartford J: The effect of tethered cord release on scoliosis in myelomeningocele. J Pediatr Orthop 2000;20:362–365. 142. Cochrane DD, Rassekh SR, Thiessen PN: Functional deterioration following placode untethering in myelomeningocele. Pediatr Neurosurg 1998;28:57–62. 143. Tulipan N, Bruner JP: Myelomeningocele repair in utero: A report of three cases. Pediatr Neurosurg 1998;28:177–180. 144. Bruner JP, Tulipan NE, Richards WO: Endoscopic coverage of fetal open myelomeningocele in utero. Am J Obstet Gynecol 1997;176:256– 257. 145. Bruner JP, Richards WO, Tulipan NB, Arney TL: Endoscopic coverage of fetal myelomeningocele in utero. Am J Obstet Gynecol 1999; 180:153–158. 146. Sutton LN, Adzick NS, Bilaniuk LT, Johnson MP, Combleholme TM, Flake AW: Improvement in hindbrain herniation demonstrated by serial fetal magnetic resonance imaging following fetal surgery for myelomeningocele. JAMA 1999;282:1826–1831. 147. Adzick NS, Sutton LN, Crombleholme TM, Flake AW: Successful fetal surgery for spina bifida. Lancet 1998;352:1675–1676. 148. Meuli M, Meuli-Simmen C, Yinling CD, et al: In utero repair of experimental myelomeningocele saves neurological function at birth. J Pediatr Surg 1996;31:397–402. 149. Olutoye OO, Adzick NS: Fetal surgery for myelomeningocele. Semin Perinatol 1999;23:462–473. 150. Bruner JP, Tulipan N, Paschall RL, et al: Fetal surgery for myelomeningocele and the incidence of shunt-dependent hydrocephalus. JAMA 1999;282:1819–1825. 151. Sutton LN, Sun P, Adzick NS: Fetal neurosurgery. Neurosurgery 2001;48:124–142; discussion 142–144. 152. McLone DG, Dias MS: Normal and abnormal early development of the nervous system. In Cheek WR, Marlin AE, Mclone DG, Reigel DH, Walker ML (eds): Pediatric Neurosurgery. Surgery of the Developing Nervous System. Philadelphia, WB Saunders, 1994:1–39. 153. Catala M: Embryogenesis. Why do we need a new explanation for the emergence of spina bifida with lipoma? Childs Nerv Syst 1997;13:336– 340. 154. Warf BC, Scott, RM: Anorectal and urogenital anomalies associated with spinal cord tethering. In Albright AL, Pollack IF, Adelson PD
155.
156.
157.
158.
159. 160. 161. 162. 163. 164. 165.
166. 167.
168.
169.
170.
171. 172.
173.
174.
175. 176.
177. 178. 179.
Pediatric Neurosurgery
325
(eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:353–362. Koyanagi I, Iwasaki Y, Hida K, Abe H, Isu T, Akino M: Surgical treatment supposed natural history of the tethered cord with occult spinal dysraphism. Childs Nerv Syst 1997;13:268–274. Kanev PM, Lemire RJ, Loeser JD, Berger MS: Management and longterm follow-up review of children with lipomyelomeningocele, 1952–1987. J Neurosurg 1990;73:48–52. Byrne RW, Hayes EA, George TM, McLone DG: Operative resection of 100 spinal lipomas in infants less than 1 year of age. Pediatr Neurosurg 1995;23:182–186. Koyanagi I, Iwasaki Y, Hida K, Abe H, Isu T, Akino M, Aida T: Factors in neurological deterioration and role of surgical treatment in lumbosacral spinal lipoma. Childs Nerv Syst 2000;16:143–149. Jamil M, Bannister CM: A report of children with spinal dysraphism managed conservatively. Eur J Pediatr Surg 1992;2:26–28. Sutton LN: Lipomyelomeningocele. Neurosurg Clin North Am 1995;6:325–338. Pierre-Kahn A, Zerah M, Renier D, et al: Congenital lumbosacral lipomas. Childs Nerv Syst 1997;13:298–334. Kanev PM, Bierbrauer KS: Reflections on the natural history of lipomyelomeningocele. Pediatr Neurosurg 1995;22:137–140. La Marca F, Grant JA, Tomiat T, McLone DG: Spinal lipomas in children: outcome of 270 procedures. Pediatr Neurosurg 1997;26:8–16. Xenos C, Sgouros S, Walsh R, Hockley A: Spinal lipomas in children. Pediatr Neurosurg 2000;32:295–307. Hoffman HJ, Taecholarn C, Hendrick EB, Humphreys RP: Management of lipomyelomeningoceles. Experience at the Hospital for Sick Children, Toronto. J Neurosurg 1985;62:1–8. Chapman PH: Congenital intraspinal lipomas: Anatomic considerations and surgical treatment. Childs Brain 1982;9:37–47. Colak A, Pollack IF, Albright AL: Recurrent tethering: A common long-term problem after lipomyelomeningocele repair. Pediatr Neurosurg 1998;29:184–190. Van Calenbergh F, Vanvolsem S, Verpoorten C, Lagae L, Casaer P, Pletas C: Results after surgery for lumbosacral lipoma: The significance of early and late worsening. Childs Nerv Syst 1999;15:439–442. Milhorat TH, Chou MW, Trinidad EM, et al: Chiari I malformation redefined: Clinical and radiographic findings for 364 symptomatic patients. Neurosurgery 1999;44:1005–1017. Wu YW, Chin CT, Chan KM, Barkovich AJ, Ferriero DM: Pediatric Chiari I malformations: Do clinical and radiologic features correlate? Neurology 1999;53:1271–1276. Elster AD, Chen MY: Chiari I malformations: Clinical and radiologic reappraisal. Radiology 1992;183:347–353. Meadows J, Kraut M, Guarnieri M, Haroun RI, Carson BS: Asymptomatic Chiari Type I malformations identified on magnetic resonance imaging. J Neurosurg 2000;92:920–926. Park JK, Gleason PL, Madsen JR, Goumnerova LC, Scott RM: Presentation and management of Chiari I malformation in children. Pediatr Neurosurg 1997;26:190–196. Dyste GN, Menezes AH: Presentation and management of pediatric Chiari malformations without myelodysplasia. Neurosurgery 1988;23:589–597. Nagib MG: An approach to symptomatic children (ages 4-14 years) with Chiari type I malformation. Pediatr Neurosurg 1994;21:31–35. Pillay PK, Awad IA, Little JR, Hahn JF: Symptomatic Chiari malformation in adults: A new classification based on magnetic resonance imaging with clinical and prognostic significance. Neurosurgery 1991;28:639–645. Bindal AK, Dunsker SB, Tew JM Jr: Chiari I malformation: Classification and management. Neurosurgery 1995;37:1069–1074. Dure LS, Percy AK, Cheek WR, Laurent JP: Chiari type I malformation in children. J Pediatr 1989;115:573–576. Listernick R, Tomita T: Persistent crying in infancy as a presentation of Chiari type I malformation. J Pediatr 1991;118:567–569.
W9418-11.qxd 10/31/03 3:43 PM Page 326
326
Section I
Introduction
180. Hudgins RJ: Paroxysmal rage as a presenting symptom of the Chiari I malformation. Report of two cases. J Neurosurg 1999;91:328–329. 181. Bondurant CP, Oro JJ: Spinal cord injury without radiographic abnormality and Chiari malformation. J Neurosurg 1993;79:833–838. 182. Callaway GH, O’Brien SJ, Tehrany AM: Chiari I malformation and spinal cord injury: Cause for concern in contact athletes? Med Sci Sports Exerc 1996;28:1218–1220. 183. Park JK, Gleason PL, Madsen JR, Goumnerova LC, Scott RM: Presentation and management of Chiari I malformation in children. Pediatr Neurosurg 1997;26:190–196. 184. Krieger MD, McComb JG, Levy ML: Toward a simpler surgical management of Chiari I malformation in a pediatric population. Pediatr Neurosurg 1999;30:113–121. 185. Shen WJ, McDowell GS, Burke SW, Levine DB, Chutorian AM: Routine preoperative MRI and SEP studies in adolescent idiopathic scoliosis. J Pediatr Orthop 1996;16:350–353. 186. Evans SC, Edgar MA, Hall-Craggs MA, Powell MP, Taylor BA, Noordeen HH: MRI of “idiopathic” juvenile scoliosis. A prospective study. J Bone Joint Surg Br 1996;78:314–317. 187. Maiocco B, Deeney VF, Coulon R, Parks PF Jr: Adolescent idiopathic scoliosis and the presence of spinal cord abnormalities. Preoperative magnetic resonance imaging analysis. Spine 1997;22: 2537–2541. 188. Winter RB, Lonstein JE, Heithoff KB, Kirkham JA: Magnetic resonance imaging evaluation of the adolescent patient with idiopathic scoliosis before spinal instrumentation and fusion. A prospective, double-blinded study of 140 patients. Spine 1997;22:855–858. 189. Johnston I, Jacobson E, Besser M: The acquired Chiari malformation and syringomyelia following spinal CSF drainage: A study of incidence and management. Acta Neurochir (Wien) 1998;140:417–427. 190. Payner TD, Prenger E, Berger TS, Crone KR: Acquired Chiari malformations: Incidence, diagnosis, and management. Neurosurgery 1994;34:429–434. 191. Chumas PD, Armstrong DC, Drake JM, et al: Tonsillar herniation: The rule rather than the exception after lumboperitoneal shunting in the pediatric population. J Neurosurg 1993;78:568–573. 192. Badie B, Mendoza D, Batzdorf U: Posterior fossa volume and response to suboccipital decompression in patients with Chiari I malformation. Neurosurgery 1995;37:214–218. 193. Stovener LJ, Bergan U, Nilsen G, Sjaastad O: Posterior cranial fossa dimensions in the Chiari I malformation: Relation to pathogenesis and clinical presentation. Neuroradiology 1993;35:13–118. 194. Nishikawa M, Sakamoto H, Hakuba A, Nakanishi N, Inoue Y: Pathogenesis of Chiari malformation: A morphometric study of the posterior cranial fossa. J Neurosurg 1997;86:40–47. 195. Oldfield EH, Maraszko K, Shawker TH, Patronas NJ: Pathophysiology of syringomyelia associated with Chiari I malformation of the cerebellar tonsils. Implications for diagnosis and treatment. J Neurosurg 1994;80:3–15. 196. Heiss JD, Patronas N, DeVroom HL, et al: Elucidating the pathophysiology of syringomyelia. J Neurosurg 1999;91:553–562. 197. Elster AD, Chen MY: Chiari I malformations: Clinical and radiologic reappraisal. Radiology 1992;183:347–353. 198. Tubbs RS, Elton S, Grabb P, Dockery SE, Bartolucci AA, Oakes WJ: Analysis of the posterior fossa in children with the Chiari 0 malformation. Neurosurgery 2001;48:1050–1054. 199. Iskandar BJ, Hedlund GL, Grabb PA, Oakes WJ: The resolution of syringohydromyelia without hindbrain herniation after posterior fossa decompression. J Neurosurg 1998;89:212–216. 200. Dyste GN, Menezes AH, VanGilder JC: Symptomatic Chiari malformations. An analysis of presentation, management, and long-term outcome. J Neurosurg 1989;71:159–168. 201. Haroun RI, Guarnieri M, Meadow JJ, Kraut M, Carsons BS: Current opinions for the treatment of syringomyelia and chiari malformations: Survey of the Pediatric Section of the American Association of Neurological Surgeons. Pediatr Neurosurg 2000;33:311–317.
202. Munshi I, Frim D, Stine-Reyes R, Weir BK, Hekmatpanah J, Brown F. Effects of posterior fossa decompression with and without duraplasty on Chiari malformation-associated hydromyelia. Neurosurgery 2000;46:1384–1389. 203. Fischer EG: Posterior fossa decompression for Chiari I deformity, including resection of the cerebellar tonsils. Childs Nerv Syst 1995;11:625–629. 204. Grabb PA, Mapstone TB, Oakes WJ: Ventral brain stem compression in pediatric and young adult patients with Chiari I malformations. Neurosurgery 1999;44:520–827. 205. Hida K, Iwasaki Y, Koyanagi I, Sawamura Y, Abe H: Surgical indication and results of foramen magnum decompression versus syringosubarachnoid shunting for syringomyelia associated with Chiari I malformation. Neurosurgery 1995;37:673–678. 206. Delashaw JB, Persing JA, Jane JA: Cranial deformation in craniosynostosis. A new explanation. Neurosurg Clin North Am 1991;2:611– 620. 207. Sun PS, Persing JA: Craniosynostosis. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:219–242. 208. Heutink P, Vermeij-Keers C, Oostra BA: The genetic background of craniosynostosis syndromes. Eur J Hum Genet 1995;3:312–323. 209. Burke MJ, Winston KR, Williams S: Normal sutural fusion and the etiology of single sutural craniosynostosis: The microspicule hypothesis. Pediatr Neurosurg 1995;22:241–246. 210. Alden TD, Lin KY, Jane JA: Mechanisms of premature closure of cranial sutures. Childs Nerv Syst 1999;15:670–675. 211. Mulliken JB, Steinberger D, Kunze S, Muller U: Molecular diagnosis of bilateral coronal synostosis. Plast Reconstr Surg 1999;104:1603– 1615. 212. Tsai FJ, Wu JY, Lee CC, Tsa CH: A unique point mutation in the fibroblast growth factor receptor 3 gene (FGFR3) causes non-syndromic craniosynostosis. Acta Paediatr 2000;89:672–674. 213. Lajeunie E, Catala M, Renier D: Craniosynostosis: From a clinical description to an understanding of bone formation of the skull. Childs Nerv Syst 1999;15:676–680. 214. Reardon W, Winter RM, Rutland P, Pulleyn LJ, Jones BM, Malcolm S: Mutations in the fibroblast growth factor receptor 2 gene cause Crouzon syndrome. Nat Genet 1994;8:98–103. 215. Wilkie AO. Craniosynostosis: Genes and mechanisms. Hum Mol Genet 1997;6:1647–1656. 216. Opperman LA, Passarelli RW, Morgan EP, Reintjes M, Ogle RC: Cranial sutures require tissue interactions with dura mater to resist osseous obliteration in vitro. J Bone Miner Res 1995;10:1978–1987. 217. Opperman LA, Chhabra A, Nolen AA, Bao Y, Ogle RC: Dura mater maintains rat cranial sutures in vitro by regulating suture cell proliferation and collagen production. J Craniofac Genet Dev Biol 1998;18:150–158. 218. Levine JP, Bradley JP, Roth DA, McCarthy JG, Longaker MT: Studies in cranial suture biology: Regional dura mater determines overlying suture biology. Plast Reconstr Surg 1998;101:1441–1447. 219. Delashaw JB, Persing JA, Broaddus WC, Jane JA: Cranial vault growth in craniosynostosis. J Neurosurg 1989;70:159–165. 220. Cohen MM Jr: Sutural biology and the correlates of craniosynostosis. Am J Med Genet 1993;47:581–616. 221. Sloan GM, Wells KC, Raffel C, McComb JG: Surgical treatment of craniosynostosis: Outcome analysis of 250 consecutive patients. Pediatrics 1997;100(1):E2. 222. Lajeunie E, Le Merrer M, Bonaiti-Pellie C, Marchac D, Renier D: Genetic study of scaphocephaly. Am J Med Genet 1996;62:282–285. 223. Thompson DN, Malcolm GP, Jones BM, Harkness WJ, Hayward RD: Intracranial pressure in single-suture craniosynostosis. Pediatr Neurosurg 1995;22:235–240. 224. Kane AA, Mitchell LE, Craven KP, Marsh JL: Observations on a recent increase in plagiocephaly without synostosis. Pediatrics 1996;97:877– 885.
W9418-11.qxd 10/31/03 3:43 PM Page 327
Chapter 11 225. Sawin PD, Muhonen MG, Menezes AH: Quantitative analysis of cerebrospinal fluid spaces in children with occipital plagiocephaly. J Neurosurg 1996;85:428–434. 226. Cohen MM Jr, Kreiborg S: Birth prevalence studies of the Crouzon syndrome: Comparison of direct and indirect methods. Clin Genet 1992;41:12–15. 227. Goodrich JT: Craniofacial syndromes. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:243–259. 228. Cinalli G, Sainte-Rose C, Kollar EM, et al: Hydrocephalus and craniosynostosis. J Neurosurg 1998;88:209–214. 229. Cohen MM Jr, Kreiborg S, Lammer EJ, et al: Birth prevalence study of the Apert syndrome. Am J Med Genet 1992;42:655–659. 230. Kreiborg S, Cohen MM Jr: The infant Apert skull. Neurosurg Clin North Am 1991;2:551–554. 231. Renier D, Arnaud E, Cinalli G, Sebag G, Zerah M, Marchac D: Prognosis for mental function in Apert’s syndrome. J Neurosurg 1996;85:66–72. 232. Pollack IF, Losken HW, Biglan AW: Incidence of increased intracranial pressure after early surgical treatment of syndromic craniosynostosis. Pediatr Neurosurg 1996;24:202–209. 233. Albright AL: Operative normalization of skull shape in sagittal synostosis. Neurosurg 1985;17:329–331. 234. Albright AL, Pollack IF, Adelson PD, Solot JJ: Outcome data and analysis in pediatric neurosurgery. Neurosurgery 1999;45:101–106. 235. Burstein FD, Hudgins RJ, Cohen SR, Boydston WR: Surgical correction of severe scaphocephalic deformities. J Craniofac Surg 1994;5:228–235. 236. Boop FA, Shewmake K, Chadduck WM: Synostectomy versus complex cranioplasty for the treatment of sagittal synostosis. Childs Nerv Syst 1996;12:371–375. 237. Panchal J, Marsh JL, Park TS, Kaufman B, Pilgram T, Huang SH: Sagittal craniosynostosis outcome assessment for two methods and timings of intervention. Plast Reconstr Surg 1999;103:1574–1584. 238. Jimenez DF, Barone CM: Endoscopic craniectomy for early surgical correction of sagittal craniosynostosis. J Neurosurg 1998;88:77–81. 239. Barone CM, Jimenez DF: Endoscopic craniectomy for early correction of craniosynostosis. Plast Reconstr Surg 1999;104:1965–1973. 240. Peter JC, Fieggen G: Congenital malformations of the brain—a neurosurgical perspective at the close of the twentieth century. Childs Nerv Syst 1999;15:635–645. 241. Macfarlane R, Rutka JT, Armstrong D, et al: Encephaloceles of the anterior cranial fossa. Pediatr Neurosurg 1995;23:148–158. 242. Suwanwela C, Suwanwela N: A morphological classification of sincipital encephalomeningoceles. J Neurosurg 1972;36:201–211. 243. Chapman PH, Swearingen B, Caviness VS: Subtorcular occipital encephaloceles. Anatomical considerations relevant to operative management. J Neurosurg 1989;71:375–381. 244. Pollack IF: Management of encephaloceles and craniofacial problems in the neonatal period. Neurosurg Clin North Am 1998;9:121–139. 245. McLone DG: Congenital malformations of the central nervous system. Clin Neurosurg 2000;47:346–377. 246. Jimenez, DF, Barone CM: Encephaloceles, meningoceles, and dermal sinuses. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:189–208. 247. Hoving EW: Nasal encephaloceles. Childs Nerv Syst 2000;16:702–706. 248. Bartels RH, Merx JL, van Overbeeke JJ: Falcine sinus and occipital encephalocele: A magnetic resonance venography study. J Neurosurg 1998;89:738–741. 249. Otsubo Y, Sato H, Sato N, Ito H: Cephaloceles and abnormal venous drainage. Childs Nerv Syst 1999;15:329–332. 250. Brown MS, Sheridan-Pereira M: Outlook for the child with a cephalocele. Pediatrics 1992;90:914–919. 251. Rengachary SS, Watanabe I: Ultrastructure and pathogenesis of intracranial arachnoid cysts. J Neuropathol Exp Neurol 1981;40: 61–83.
Pediatric Neurosurgery
327
252. Abtin K, Walker ML: Congenital arachnoid cysts and the DandyWalker complex. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:125–141. 253. Kang JK, Lee KS, Lee IW, et al: Shunt-independent surgical treatment of middle cranial fossa arachnoid cysts in children. Childs Nerv Syst 2000;16:111–116. 254. Fewel ME, Levy ML, McComb JG: Surgical treatment of 95 children with 102 intracranial arachnoid cysts. Pediatr Neurosurg 1996;25:165– 173. 255. Ciricillo SF, Cogen PH, Harsh GR, Edwards MS: Intracranial arachnoid cysts in children. A comparison of the effects of fenestration and shunting. J Neurosurg 1991;74:230–235. 256. Pascual-Castroviejo I, Roche MC, et al: Primary intracranial arachnoidal cysts. A study of 67 childhood cases. Childs Nerv Syst 1991;7:257–263. 257. Oberbauer RW, Haase J, Pucher R: Arachnoid cysts in children: A European co-operative study. Childs Nerv Syst 1992;8:281–286. 258. Arai H, Sato K, Wachi A, Okuda O, Takeda N: Arachnoid cysts of the middle cranial fossa: Experience with 77 patients who were treated with cystoperitoneal shunting. Neurosurg 1996;39:1108–1112. 259. Jallo GI, Woo HH, Meshki C, Epstein FJ, Wisoff JH: Arachnoid cysts of the cerebellopontine angle: Diagnosis and surgery. Neurosurgery 1997;40:31–37. 260. Adan L, Bussieres L, Dinand V, Zerah M, Pierre-Kahn A, Brauner R: Growth, puberty and hypothalamic-pituitary function in children with suprasellar arachnoid cyst. Eur J Pediatr 2000;159:348–355. 261. Galassi E, Tognetti F, Gaist G, Fagioli L, Frank F, Frank G: CT scan and metrizamide CT cisternography in arachnoid cysts of the middle cranial fossa: Classification and pathophysiological aspects. Surg Neurol 1982;17:363–369. 262. Santamarta D, Aguas J, Ferrer E: The natural history of arachnoid cysts: Endoscopic and cine-mode MRI evidence of a slit-valve mechanism. Minim Invasive Neurosurg 1995;38:133–137. 263. McDonald PJ, Rutka JT: Middle cranial fossa arachnoid cysts that come and go. Report of two cases and review of the literature. Pediatr Neurosurg 1997;26:48–52. 264. Przybylo HJ, Radkowski MA, Przybylo J, McLone D: Spontaneous resolution of an asymptomatic arachnoid cyst. Pediatr Neurosurg 1997;26:312–314. 265. Yamauchi T, Saeki N, Yamaura A: Spontaneous disappearance of temporo-frontal arachnoid cyst in a child. Acta Neurochir (Wien) 1999;141:537–540. 266. Sener RN: Arachnoid cysts associated with post-traumatic and spontaneous rupture into the subdural space. Comput Med Imaging Graph 1997;21:341–344. 267. Albuquerque FC, Giannotta SL: Arachnoid cyst rupture producing subdural hygroma and intracranial hypertension: Case reports. Neurosurgery 1997;41:951–955. 268. Cappelen J, Unsgaard G: Arachnoid cysts of the middle cranial fossa and traumatic complications. Childs Nerv Syst 1986;2:225–227. 269. Auer LM, Gallhofer B, Ladurner G, Sager WD, Heppner F, Lechner H: Diagnosis and treatment of middle fossa arachnoid cysts and subdural hematomas. J Neurosurg 1981;54:366–369. 270. Choi JU, Kim DS, Huh R: Endoscopic approach to arachnoid cyst. Childs Nerv Syst 1999;15:285–291. 271. Schroeder HW, Gaab MR, Niendorf WR: Neuroendoscopic approach to arachnoid cysts. J Neurosurg 1996;85:293–298. 272. Barkovich AJ, Kjos BO, Norman D, Edwards MS: Revised classification of posterior fossa cysts and cystlike malformations based on the results of multiplanar MR imaging. AJR 1989;153:1289–1300. 273. Domingo Z, Peter J: Midline developmental abnormalities of the posterior fossa: Correlation of classification with outcome. Pediatr Neurosurg 1996;24:111–118. 274. Tortori-Donati P, Fondelli MP, Rossi A, Carini S: Cystic malformations of the posterior cranial fossa originating from a defect of the poste-
W9418-11.qxd 10/31/03 3:43 PM Page 328
328
275.
276.
277. 278.
279.
280.
281.
282.
283.
284.
285. 286. 287.
288. 289.
290.
291. 292. 293. 294. 295.
296.
297. 298.
Section I
Introduction
rior membranous area. Mega cisterna magna and persisting Blake’s pouch: Two separate entities. Childs Nerv Syst 1996;12:303–308. Pascual-Castroviejo I, Velez A, Pascual-Pascual SI, Roche MC, Villarejo F: Dandy-Walker malformation: Analysis of 38 cases. Childs Nerv Syst 1991;7:88–97. Ecker JL, Shipp TD, Bromley B, Benacerraf B: The sonographic diagnosis of Dandy-Walker and Dandy-Walker variant: Associated findings and outcomes. Prenat Diagn 2000;20:328–332. Bindal AK, Storrs BB, McLone DG: Management of the Dandy-Walker syndrome. Pediatr Neurosurg 1990;16:163–169. Hirsch JF, Pierre-Kahn A, Renier D, Sainte-Rose C, Hoppe-Hirsch E: The Dandy-Walker malformation. A review of 40 cases. J Neurosurg 1984;61:515–522. Ulm B, Ulm MR, Deutinger J, Bernaschek G: Dandy-Walker malformation diagnosed before 21 weeks of gestation: Associated malformations and chromosomal abnormalities. Ultrasound Obstet Gynecol 1997;10:167–170. Asai A, Hoffman HJ, Hendrick EB, Humphreys RP: Dandy-Walker syndrome: Experience at the Hospital for Sick Children, Toronto. Pediatr Neurosci 1989;15:66–73. Pillay P, Barnett GH, Lanzeiri C, Cruse R: Dandy-Walker cyst upward herniation: The role of magnetic resonance imaging and double shunts. Pediatr Neurosci 1989;15:74–78. Raimondi AJ, Samuelson G, Yarzagaray L, Norton T: Atresia of the foramina of Luschka and Magendie: the Dandy-Walker cyst. J Neurosurg 1969;31:202–216. Rekate HL: Treatment of hydrocephalus. In McLaurin RL, Schut L, Venes JL, Epstein F (eds): Pediatric Neurosurgery: Surgery of the Developing Nervous System. Philadelphia, WB Saunders, 1989:200– 218. Gerszten PC, Albright AL: Relationship between cerebellar appearance and function in children with Dandy-Walker syndrome. Pediatr Neurosurg 1995;23:86–92. Maria BL, Zinreich SJ, Carson BC, Rosenbaum AE, Freeman JM: Dandy-Walker syndrome revisited. Pediatr Neurosci 1987;13:45–51. Tepas JJ 3rd, DiScala C, Ramenofsky ML, Barlow B: Mortality and head injury: The pediatric perspective. J Pediatr Surg 1990;25:92–95. Kraus JF, Fife D, Conroy C: Pediatric brain injuries: The nature, clinical course, and early outcomes in a defined United States’ population. Pediatrics 1987;79:501–507. Kraus JF, Rock A, Hemyari P: Brain injuries among infants, children, adolescents, and young adults. Am J Dis Child 1990;144:684–691. Childhood injuries in the United States. Division of Injury Control, Center for Environmental Health and Injury Control, Centers for Disease Control. Am J Dis Child 1990;144:627–646. Duhaime AC, Christian CW, Rorke LB, Zimmerman RA: Nonaccidental head injury in infants—the “shaken-baby syndrome.” N Engl J Med 1998;338:1822–1829. Christoffel KK: Violent death and injury in US children and adolescents. Am J Dis Child 1990;144:697–706. Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 1974;2:81–84. Hahn YS, McLone DG: Risk factors in the outcome of children with minor head injury. Pediatr Neurosurg 1993;19:135–142. Yager JY, Johnston B, Seshia SS: Coma scales in pediatric practice. Am J Dis Child 1990;144:1088–1091. Durham SR, Clancy RR, Leuthardt E, et al: CHOP Infant Coma Scale (“Infant Face Scale”): A novel coma scale for children less than two years of age. J Neurotrauma 2000;17:729–737. Lloyd DA, Carty H, Patterson M, Butcher CK, Roe D: Predictive value of skull radiography for intracranial injury in children with blunt head injury. Lancet 1997;349:821–824. Ersahin Y, Mutluer S, Mirzai H, Palali I: Pediatric depressed skull fractures: Analysis of 530 cases. Childs Nerv Syst 1996;12:323–331. Luerssen TG, Klauber MR, Marshall LF: Outcome from head injury related to patient’s age. A longitudinal prospective study
299. 300. 301.
302. 303.
304.
305.
306. 307. 308. 309.
310.
311.
312.
313. 314. 315.
316.
317.
318.
319.
320.
321.
of adult and pediatric head injury. J Neurosurg 1988;68:409– 416. Mandera M, Wencel T, Bazowski P, Krauze J: How should we manage children after mild head injury? Childs Nerv Syst 2000;16:156–160. Faden AI: Pharmacological treatment of central nervous system trauma. Pharmacol Toxicol 1996;78:12–17. Bittigau P, Sifringer M, Pohl D, et al: Apoptotic neurodegeneration following trauma is markedly enhanced in the immature brain. Ann Neurol 1999;45:724–735. Adelson PD: Animal models of traumatic brain injury in the immature: A review. Exp Toxicol Pathol 1999;51:130–136. Stein SC, Spettell CM: Delayed and progressive brain injury in children and adolescents with head trauma. Pediatr Neurosurg 1995;23:299–304. McIntosh TK, Juhler M, Wieloch T: Novel pharmacologic strategies in the treatment of experimental traumatic brain injury: 1998. J Neurotrauma 1998;15:731–769. Adelson PD, Robichaud P, Hamilton RL, Kochanek PM: A model of diffuse traumatic brain injury in the immature rat. J Neurosurg 1996;85:877–884. Zwienenberg M, Muizelaar JP: Cerebral perfusion and blood flow in neurotrauma. Neurol Res 2001;23:167–174. Adelson PD, Kochanek PM: Head injury in children. J Child Neurol 1998;13:2–15. Bruce DA, Schut L, Bruno LA, Wood JH, Sutton LN: Outcome following severe head injuries in children. J Neurosurg 1978;48:679–688. Bruce DA, Raphaely RC, Goldberg AI, Zimmerman RA, Bilaniuk LT, Schut L, Kuhl DE: Pathophysiology, treatment and outcome following severe head injury in children. Childs Brain 1979;5:174–191. Muizelaar JP, Marmarou A, DeSalles AA, et al: Cerebral blood flow and metabolism in severely head-injured children. Part 1: Relationship with GCS score, outcome, ICP, and PVI. J Neurosurg 1989;71:63–71. Muizelaar JP, Ward JD, Marmarou A, Newlon PG, Wachi A: Cerebral blood flow and metabolism in severely head-injured children. Part 2: Autoregulation. J Neurosurg 1989;71:72–76. Luerssen TG, Klauber MR: Outcome from pediatric head injury: On the nature of prospective and retrospective studies. Pediatr Neurosurg 1995;23:34–40. Zwienenberg M, Muizelaar JP: Severe pediatric head injury: The role of hyperemia revisited. J Neurotrauma 1999;16:937–943. Johnson DL, Krishnamurthy S: Severe pediatric head injury: myth, magic, and actual fact. Pediatr Neurosurg 1998;28:167–172. Bruce DA, Alavi A, Bilaniuk L, Dolinskas C, Obrist W, Uzzell B: Diffuse cerebral swelling following head injuries in children: The syndrome of “malignant brain edema”. J Neurosurg 1981;54:170–178. Aldrich EF, Eisenberg HM, Saydjari C, et al: Diffuse brain swelling in severely head-injured children. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 1992;76:450–454. Lang DA, Teasdale GM, Macpherson P, Lawrence A: Diffuse brain swelling after head injury: More often malignant in adults than children? J Neurosurg 1994;80:675–680. Chiron C, Raynaud C, Maziere B, et al: Changes in regional cerebral blood flow during brain maturation in children and adolescents. J Nucl Med 1992;33:696–703. Adelson PD, Clyde B, Kochanek PM, Wisniewski SR, Marion DW, Yonas H: Cerebrovascular response in infants and young children following severe traumatic brain injury: A preliminary report. Pediatr Neurosurg 1997;26:200–207. Sharples PM, Stuart AG, Matthews DS, Aynsley-Green A, Eyre JA: Cerebral blood flow and metabolism in children with severe head injury. Part 1: Relation to age, Glasgow coma score, outcome, intracranial pressure, and time after injury. J Neurol Neurosurg Psychiatry 1995;58:145–152. Baldwin HZ, Rekate HL: Preliminary experience with controlled external lumbar drainage in diffuse pediatric head injury. Pediatr Neurosurg 1991;17:115–120.
W9418-11.qxd 10/31/03 3:43 PM Page 329
Chapter 11 322. Levy DI, Rekate HL, Cherny WB, Manwaring K, Moss SD, Baldwin HZ: Controlled lumbar drainage in pediatric head injury. J Neurosurg 1995;83:453–460. 323. Kunze E, Meixensberger J, Janka M, Sorensen N, Roosen K: Decompressive craniectomy in patients with uncontrollable intracranial hypertension. Acta Neurochir Suppl 1998;71:16–18. 324. Guerra WK, Gaab MR, Dietz H, Mueller JU, Piek J, Fritsch MJ: Surgical decompression for traumatic brain swelling: Indications and results. J Neurosurg 1999;90:187–196. 325. Litofsky NS, Chin LS, Tang G, Baker S, Giannotta SL, Apuzzo ML: The use of lobectomy in the management of severe closed-head trauma. Neurosurgery 1994;34:628–632. 326. Polin RS, Shaffrey ME, Bogaev CA, et al: Decompressive bifrontal craniectomy in the treatment of severe refractory posttraumatic cerebral edema. Neurosurgery 1997;41:84–92. 327. Taylor A, Butt W, Rosenfeld J, et al: A randomized trial of very early decompressive craniectomy in children with traumatic brain injury and sustained intracranial hypertension. Childs Nerv Syst 2001;17:154–162. 328. Gaab MR, Rittierodt M, Lorenz M, Heissler HE: Traumatic brain swelling and operative decompression: A prospective investigation. Acta Neurochir Suppl (Wien) 1990;51:326–328. 329. Esparza J, M-Portillo J, Sarabia M, Yuste JA, Roger R, Lamas E: Outcome in children with severe head injuries. Childs Nerv Syst 1985;1:109–114. 330. Zuccarello M, Facco E, Zampieri P, Zanardi L, Andrioli GC: Severe head injury in children: Early prognosis and outcome. Childs Nerv Syst 1985;1:158–162. 332. Walker ML, Storrs BB, Mayer T: Factors affecting outcome in the pediatric patient with multiple trauma. Further experience with the modified injury severity scale. Childs Brain 1984;11:387–397. 332. Downard C, Hulka F, Mullins RJ, Piatt J, Chesnut R, Quint P, Mann NC: Relationship of cerebral perfusion pressure and survival in pediatric brain-injured patients. J Trauma 2000;49:654–658. 333. Feickert HJ, Drommer S, Heyer R: Severe head injury in children: Impact of risk factors on outcome. J Trauma 1999;47:33–38. 334. Aldrich EF, Eisenberg HM, Saydjari C, et al: Diffuse brain swelling in severely head-injured children. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 1992;76:450–454. 335. Jaffe KM, Fay GC, Polissar NL, et al: Severity of pediatric traumatic brain injury and neurobehavioral recovery at one year—A cohort study. Arch Phys Med Rehabil 1993;74:587–595. 336. McDonald CM, Jaffe KM, Fay GC, et al: Comparison of indices of traumatic brain injury severity as predictors of neurobehavioral outcome in children. Arch Phys Med Rehabil 1994;75: 328–337. 337. Wesson DE, Williams JI, Spence LJ, Filler RM, Armstrong PF, Pearl RH: Functional outcome in pediatric trauma. J Trauma 1989;29:589– 592. 338. Koskiniemi M, Kyykka T, Nybo T, Jarho L: Long-term outcome after severe brain injury in preschoolers is worse than expected. Arch Pediatr Adolesc Med 1995;149:249–254. 339. Duhaime AC, Alario AJ, Lewander WJ, et al: Head injury in very young children: Mechanisms, injury types, and ophthalmologic findings in 100 hospitalized patients younger than two years of age. Pediatrics 1992;90:179–185. 340. Barlow KM, Minns RA: Annual incidence of shaken impact syndrome in young children. Lancet 2000;356:1571–1572. 341. Hadley MN, Sonntag VK, Rekate HL, Murphy A: The infant whiplashshake injury syndrome: A clinical and pathological study. Neurosurgery 1989;24:536–540. 342. Gilliland MG, Folberg R: Shaken babies—some have no impact injuries. J Forensic Sci 1996;41:114–116. 343. Duhaime AC, Gennarelli TA, Thibault LE, Bruce DA, Margulies SS, Wiser R: The shaken baby syndrome. A clinical, pathological, and biomechanical study. J Neurosurg 1987;66:409–415.
Pediatric Neurosurgery
329
344. Alexander R, Sato Y, Smith W, Bennett T: Incidence of impact trauma with cranial injuries ascribed to shaking. Am J Dis Child 1990;144:724–726. 345. Bruce DA, Zimmerman RA: Shaken impact syndrome. Pediatr Ann 1989;18:482–484, 486–489, 492–494. 346. Hahn YS, Raimondi AJ, McLone DG, Yamanouchi Y: Traumatic mechanisms of head injury in child abuse. Childs Brain 1983;10:229– 241. 347. Kivlin JD: A 12-year ophthalmologic experience with the shaken baby syndrome at a regional children’s hospital. Trans Am Ophthalmol Soc 1999;97:545–581. 348. Odom A, Christ E, Kerr N, et al: Prevalence of retinal hemorrhages in pediatric patients after in-hospital cardiopulmonary resuscitation: A prospective study. Pediatrics 1997;99:E3. 349. Gaynon MW, Koh K, Marmor MF, Frankel LR: Retinal folds in the shaken baby syndrome. Am J Ophthalmol 1988;106:423–425. 350. Dias MS, Backstrom J, Falk M, Li V: Serial radiography in the infant shaken impact syndrome. Pediatr Neurosurg 1998;29:77–85. 351. Hymel KP, Rumack CM, Hay TC, Strain JD, Jenny C: Comparison of intracranial computed tomographic (CT) findings in pediatric abusive and accidental head trauma. Pediatr Radiol 1997;27:743–747. 352. Johnson DL, Boal D, Baule R: Role of apnea in nonaccidental head injury. Pediatr Neurosurg 1995;23:305–310. 353. Shannon P, Smith CR, Deck J, Ang LC, Ho M, Becker L: Axonal injury and the neuropathology of shaken baby syndrome. Acta Neuropathol (Berl) 1998;95:625–631. 354. Ewing-Cobbs L, Prasad M, Kramer L, Landry S: Inflicted traumatic brain injury: Relationship of developmental outcome to severity of injury. Pediatr Neurosurg 1999;31:251–258. 355. Durkin MS, Olsen S, Barlow B, Virella A, Connolly ES Jr: The epidemiology of urban pediatric neurological trauma: Evaluation of, and implications for, injury prevention programs. Neurosurg 1998;42: 300–310. 356. Pang D, Wilberger JE Jr: Spinal cord injury without radiographic abnormalities in children. J Neurosurg 1982;57:114–129. 357. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag VK: Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg 1988;68:18–24. 358. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174 hospital admissions. J Neurosurg 1992;77:700–704. 359. Fesmire FM, Luten RC: The pediatric cervical spine: developmental anatomy and clinical aspects. J Emerg Med 1989;7:133–142. 360. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinal injury: The very young. J Neurosurg 1988;68:25–30. 361. Subach BR, McLaughlin MR, Albright AL, Pollack IF: Current management of pediatric atlantoaxial rotatory subluxation. Spine 1998;23:2174–2179. 362. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebral column injury. Neurosurg 1992;30:385–390. 363. Kokoska ER, Keller MS, Rallo MC, Weber TR: Characteristics of pediatric cervical spine injuries. J Pediatr Surg 2001;36:100–105. 364. Grabb PA, Hadley MN: Spinal column trauma in children. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:935–953. 365. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VK: Pediatric cervical spine injuries: Report of 102 cases and review of the literature. J Neurosurg 2000;92:12–17. 366. Dickman CA, Zabramski JM, Hadley MN, Rekate HL, Sonntag VK: Pediatric spinal cord injury without radiographic abnormalities: Report of 26 cases and review of the literature. J Spinal Disord 1991;4:296–305. 367. Pang D, Pollack IF: Spinal cord injury without radiographic abnormality in children—the SCIWORA syndrome. J Trauma 1989;29: 654–664. 368. Osenbach RK, Menezes AH: Spinal cord injury without radiographic abnormality in children. Pediatr Neurosci 1989;15:168–174.
W9418-11.qxd 10/31/03 3:43 PM Page 330
330
Section I
Introduction
369. Grabb PA, Pang D: Magnetic resonance imaging in the evaluation of spinal cord injury without radiographic abnormality in children. Neurosurgery 1994;35:406–414. 370. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emergency transport and positioning of young children who have an injury of the cervical spine. The standard backboard may be hazardous. J Bone Joint Surg Am 1989;71:15–22. 371. Nypaver M, Treloar D: Neutral cervical spine positioning in children. Ann Emerg Med 1994;23:208–211. 372. Bracken MB, Shepard MJ, Collins WF, et al: A randomized, controlled trial of methylprednisolone or naloxone in the treatment of acute spinal-cord injury. Results of the Second National Acute Spinal Cord Injury Study. N Engl J Med 1990;322:1405–1411. 373. Bracken MB, Shepard MJ, Holford TR, et al: Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the Third National Acute Spinal Cord Injury Randomized Controlled Trial. National Acute Spinal Cord Injury Study. JAMA 1997;277:1597– 1604. 374. Gjerris F, Agerlin N, Borgesen SE, et al: Epidemiology and prognosis in children treated for intracranial tumours in Denmark 1960–1984. Childs Nerv Syst 1998;14:302–311. 375. Linet MS, Ries LA, Smith MA, Tarone RE, Devesa SS: Cancer surveillance series: Recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst 1999;91:1051– 1058. 376. Bleyer WA: Epidemiologic impact of children with brain tumors. Childs Nerv Syst 1999;15:758–763. 377. Pollack IF: Brain tumors in children. N Engl J Med 1994;331:1500– 1507. 378. Asai A, Hoffman HJ, Hendrick EB, Humphreys RP, Becker LE: Primary intracranial neoplasms in the first year of life. Childs Nerv Syst 1989;5:230–233. 379. Lieberman DM, Russo CL, Berger MS: Brain tumors during the first two years of life. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:463–490. 380. Mainprize TG, Taylor MD, Rutka JT: Pediatric brain tumors: A contemporary prospectus. Clin Neurosurg 2000;47:259–302. 381. Park TS, Hoffman HJ, Hendrick EB, Humphreys RP, Becker LE: Medulloblastoma: Clinical presentation and management. Experience at the hospital for sick children, toronto, 1950–1980. J Neurosurg 1983;58:543–552. 382. Hamilton SR, Liu B, Parsons RE, et al: The molecular basis of Turcot’s syndrome. N Engl J Med 1995;332:839–847. 383. Evans DG, Farndon PA, Burnell LD, Gattamaneni HR, Birch JM: The incidence of Gorlin syndrome in 173 consecutive cases of medulloblastoma. Br J Cancer 1991;64:959–961. 384. Albright AL: Medulloblastomas. In: Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:591–608. 385. Albright AL, Wisoff JH, Zeltzer PM, Boyett JM, Rorke LB, Stanley P: Effects of medulloblastoma resections on outcome in children: A report from the Children’s Cancer Group. Neurosurgery 1996;38:265– 271. 386. Chang CH, Housepian EM, Herbert C Jr: An operative staging system and a megavoltage radiotherapeutic technic for cerebellar medulloblastomas. Radiology 1969;93:1351–1359. 387. Bigner SH, Mark J, Friedman HS, Biegel JA, Bigner DD: Structural chromosomal abnormalities in human medulloblastoma. Cancer Genet Cytogenet 1988;30:91–101. 388. Bayani J, Zielenska M, Marrano P, et al: Molecular cytogenetic analysis of medulloblastomas and supratentorial primitive neuroectodermal tumors by using conventional banding, comparative genomic hybridization, and spectral karyotyping. J Neurosurg 2000;93:437– 448.
389. Taipale J, Chen JK, Cooper MK, et al: Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 2000;406:1005–1009. 390. Gajjar A, Sanford RA, Bhargava R, et al: Medulloblastoma with brain stem involvement: The impact of gross total resection on outcome. Pediatr Neurosurg 1996;25:182–187. 391. Dias MS, Albright AL: Management of hydrocephalus complicating childhood posterior fossa tumors. Pediatr Neurosci 1989;15:283– 289. 392. Chintagumpala M, Berg S, Blaney SM: Treatment controversies in medulloblastoma. Curr Opin Oncol 2001;13:154–159. 393. Packer RJ, Sutton LN, Elterman R, et al: Outcome for children with medulloblastoma treated with radiation and cisplatin, CCNU, and vincristine chemotherapy. J Neurosurg 1994;81:690–698. 394. Thomas PR, Deutsch M, Kepner JL, et al: Low-stage medulloblastoma: Final analysis of trial comparing standard-dose with reduced-dose neuraxis irradiation. J Clin Oncol 2000;18:3004–3011. 395. Zeltzer PM, Boyett JM, Finlay JL, et al: Metastasis stage, adjuvant treatment, and residual tumor are prognostic factors for medulloblastoma in children: Conclusions from the Children’s Cancer Group 921 randomized phase III study. J Clin Oncol 1999;17:832–845. 396. Steinbok P, Mutat A: Cerebellar astrocytomas. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:641–662. 397. Pencalet P, Maixner W, Sainte-Rose C, et al: Benign cerebellar astrocytomas in children. J Neurosurg 1999;90:265–273. 398. Morreale VM, Ebersold MJ, Quast LM, Parisi JE: Cerebellar astrocytoma: Experience with 54 cases surgically treated at the Mayo Clinic, Rochester, Minnesota, from 1978 to 1990. J Neurosurg 1997;87:257– 261. 399. Abdollahzadeh M, Hoffman HJ, Blazer SI, et al: Benign cerebellar astrocytoma in childhood: Experience at the Hospital for Sick Children 1980–1992. Childs Nerv Syst 1994;10:380–383. 400. Smoots DW, Geyer JR, Lieberman DM, Berger MS: Predicting disease progression in childhood cerebellar astrocytoma. Childs Nerv Syst 1998;14:636–648. 401. Sgouros S, Fineron PW, Hockley AD: Cerebellar astrocytoma of childhood: Long-term follow-up. Childs Nerv Syst 1995;11:89–96. 402. Ilgren EB, Stiller CA: Cerebellar astrocytomas: therapeutic management. Acta Neurochir (Wien) 1986;81:11–26. 403. Akyol FH, Atahan IL, Zorlu F, Gurkaynak M, Alanyali H, Ozyar E: Results of post-operative or exclusive radiotherapy in grade I and grade II cerebellar astrocytoma patients. Radiother Oncol 1992; 23:245–248. 404. Lapras C, Patet JD, Lapras C Jr, Mottolese C: Cerebellar astrocytomas in childhood. Childs Nerv Syst 1986;2:55–59. 405. Nazar GB, Hoffman HJ, Becker LE, Jenkin D, Humphreys RP, Hendrick EB: Infratentorial ependymomas in childhood: Prognostic factors and treatment. J Neurosurg 1990;72:408–417. 406. Allen JC, Siffert J, Hukin J: Clinical manifestations of childhood ependymoma: A multitude of syndromes. Pediatr Neurosurg 1998; 28:49–55. 407. Sutton LN, Goldwein JW, Schartz D: Ependymomas. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:609–628. 408. Comi AM, Backstrom JW, Burger PC, Duffner PK: Clinical and neuroradiologic findings in infants with intracranial ependymomas. Pediatric Oncology Group. Pediatr Neurol 1998;18:23–29. 409. Osborne A: Diagnostic Radiology. St. Louis, Mosby, 1994:936. 410. Robertson PL, Zeltzer PM, Boyett JM, et al: Survival and prognostic factors following radiation therapy and chemotherapy for ependymomas in children: A report of the Children’s Cancer Group. J Neurosurg 1998;88:695–703. 411. Duffner PK, Krischer JP, Sanford RA, et al: Prognostic factors in infants and very young children with intracranial ependymomas. Pediatr Neurosurg 1998;28:215–222.
W9418-11.qxd 10/31/03 3:43 PM Page 331
Chapter 11 412. Merchant TE, Haida T, Wang MH, Finlay JL, Leibel SA: Anaplastic ependymoma: Treatment of pediatric patients with or without craniospinal radiation therapy. J Neurosurg 1997;86:943–949. 413. Mazewski C, Soukup S, Ballard E, Gotwals B, Lampkin B: Karyotype studies in 18 ependymomas with literature review of 107 cases. Cancer Genet Cytogenet 1999;113:1–8. 414. Pollack IF, Gerszten PC, Martinez AJ, et al: Intracranial ependymomas of childhood: Long-term outcome and prognostic factors. Neurosurgery 1995;37:655–666. 415. Awaad YM, Allen JC, Miller DC, Schneider SJ, Wisoff J, Epstein FJ: Deferring adjuvant therapy for totally resected intracranial ependymoma. Pediatr Neurol 1996;14:216–219. 416. Salazar OM, Castro-Vita H, VanHoutte P, Rubin P, Aygun C: Improved survival in cases of intracranial ependymoma after radiation therapy. Late report and recommendations. J Neurosurg 1983;59:65265– 65269. 417. Vanuytsel LJ, Bessell EM, Ashley SE, Bloom HJ, Brada M: Intracranial ependymoma: Long-term results of a policy of surgery and radiotherapy. Int J Radiat Oncol Biol Phys 1992;23:313–319. 418. Garrett PG, Simpson WJ: Ependymomas: Results of radiation treatment. Int J Radiat Oncol Biol Phys 1983;9:1121–1124. 419. Pollack IF, Pang D, Albright AL: The long-term outcome in children with late-onset aqueductal stenosis resulting from benign intrinsic tectal tumors. J Neurosurg 1994;80:681–688. 420. Nazar GB, Hoffman HJ, Becker LE, Jenkin D, Humphreys RP, Hendrick EB: Infratentorial ependymomas in childhood: Prognostic factors and treatment. J Neurosurg 1990;72:408–417. 421. Pollack IF, Polinko P, Albright AL, Towbin R, Fitz C: Mutism and pseudobulbar symptoms after resection of posterior fossa tumors in children: Incidence and pathophysiology. Neurosurgery 1995;37:885– 893. 422. Van Calenbergh F, Van de Laar A, Plets C, Goffin J, Casaer P: Transient cerebellar mutism after posterior fossa surgery in children. Neurosurgery 1995;37:894–898. 423. Dailey AT, McKhann GM 2nd, Berger MS: The pathophysiology of oral pharyngeal apraxia and mutism following posterior fossa tumor resection in children. J Neurosurg 1995;83:467–475. 424. Aguiar PH, Plese JP, Ciquini O, Marino R: Transient mutism following a posterior fossa approach to cerebellar tumors in children: A critical review of the literature. Childs Nerv Syst 1995;11:306–310. 425. Ersahin Y, Mutluer S, Cagli S, Duman Y: Cerebellar mutism: Report of seven cases and review of the literature. Neurosurgery 1996;38:60–65. 426. Dietze DD Jr, Mickle JP: Cerebellar mutism after posterior fossa surgery. Pediatr Neurosurg 1990;16:25–31. 427. Vandeinse D, Hornyak JE: Linguistic and cognitive deficits associated with cerebellar mutism. Pediatr Rehabil 1997;1:41–44. 428. Siffert J, Poussaint TY, Goumnerova LC, et al: Neurological dysfunction associated with postoperative cerebellar mutism. J Neurooncol 2000;48:75–81. 429. Abbott R, Goh KYC: Brainstem gliomas. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:629–640. 430. Packer RJ, Nicholson HS, Johnson DL, Vezina LG: Dilemmas in the management of childhood brain tumors: Brainstem gliomas. Pediatr Neurosurg 1991;17:37–43. 431. Fischbein NJ, Prados MD, Wara W, Russo C, Edwards MS, Barkovich AJ: Radiologic classification of brain stem tumors: Correlation of magnetic resonance imaging appearance with clinical outcome. Pediatr Neurosurg 1996;24:9–23. 432. Packer RJ, Boyett JM, Zimmerman RA, et al: Outcome of children with brain stem gliomas after treatment with 7800 cGy of hyperfractionated radiotherapy. A Childrens Cancer Group Phase I/II Trial. Cancer 1994;74:1827–1834. 433. Moghrabi A, Kerby T, Tien RD, Friedman HS: Prognostic value of contrast-enhanced magnetic resonance imaging in brainstem gliomas. Pediatr Neurosurg 1995;23:293–298.
Pediatric Neurosurgery
331
434. Vandertop WP, Hoffman HJ, Drake JM, et al: Focal midbrain tumors in children. Neurosurgery 1992;31:186–194. 435. Squires LA, Allen JC, Abbott R, Epstein FJ: Focal tectal tumors: Management and prognosis. Neurology 1994;44:953–956. 436. Pollack IF, Hoffman HJ, Humphreys RP, Becker L: The long-term outcome after surgical treatment of dorsally exophytic brain-stem gliomas. J Neurosurg 1993;78:859–863. 437. Epstein F, Wisoff JH: Intrinsic brainstem tumors in childhood: Surgical indications. J Neurooncol 1988;6:309–317. 438. Mantravadi RV, Phatak R, Bellur S, Liebner EJ, Haas R: Brain stem gliomas: An autopsy study of 25 cases. Cancer 1982;49:1294– 1296. 439. Epstein F, Wisoff J: Intra-axial tumors of the cervicomedullary junction. J Neurosurg 1987;67:483–487. 440. Robertson PL, Allen JC, Abbott IR, Miller DC, Fidel J, Epstein FJ: Cervicomedullary tumors in children: A distinct subset of brainstem gliomas. Neurology 1994;44:1798–1803. 441. Epstein F, Constantini S: Practical decisions in the treatment of pediatric brain stem tumors. Pediatr Neurosurg 1996;24:24–34. 442. Einhaus SL, Sanford RA: Craniopharyngiomas. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:545–562. 443. Fahlbusch R, Honegger J, Paulus W, Huk W, Buchfelder M: Surgical treatment of craniopharyngiomas: Experience with 168 patients. J Neurosurg 1999;90:237–250. 444. Pierre-Kahn A, Sainte-Rose C, Renier D: Surgical approach to children with craniopharyngiomas and severely impaired vision: Special considerations. Pediatr Neurosurg 1994;21(suppl 1):50–56. 445. Sanford RA: Craniopharyngioma: results of survey of the American Society of Pediatric Neurosurgery. Pediatr Neurosurg 1994;21(suppl 1):39–43. 446. Scott RM, Hetelekidis S, Barnes PD, Goumnerova L, Tarbell NJ: Surgery, radiation, and combination therapy in the treatment of childhood craniopharyngioma—a 20-year experience. Pediatr Neurosurg 1994;21(suppl 1):75–81. 447. Bunin GR, Surawicz TS, Witman PA, Preston-Martin S, Davis F, Bruner JM: The descriptive epidemiology of craniopharyngioma. J Neurosurg 1998;89:547–551. 448. Rajan B, Ashley S, Gorman C, et al: Craniopharyngioma—a long–term results following limited surgery and radiotherapy. Radiother Oncol 1993;26:1–10. 449. Sklar CA: Craniopharyngioma: endocrine abnormalities at presentation. Pediatr Neurosurgery 1994;21(suppl 1):18–20. 450. Honegger J, Buchfelder M, Fahlbusch R: Surgical treatment of craniopharyngiomas: Endocrinological results. J Neurosurg 1999;90:251– 257. 451. Harwood–Nash DC: Neuroimaging of childhood craniopharyngioma. Pediatr Neurosurg 1994;21(suppl 1):2–10. 452. De Vile CJ, Grant DB, Kendall BE, et al: Management of childhood craniopharyngioma: Can the morbidity of radical surgery be predicted? J Neurosurg 1996;85:73–81. 453. Hoffman HJ, De Silva M, Humphreys RP, Drake JM, Smith ML, Blaser SI: Aggressive surgical management of craniopharyngiomas in children. J Neurosurg 1992;76:47–52. 454. Miller DC: Pathology of craniopharyngiomas: Clinical import of pathological findings. Pediatr Neurosurg 1994;21(suppl 1):11–17. 455. Adamson TE, Wiestler OD, Kleihues P, Yasargil MG: Correlation of clinical and pathological features in surgically treated craniopharyngiomas. J Neurosurg 1990;73:12–17. 456. Weiner HL, Wisoff JH, Rosenberg ME, et al: Craniopharyngiomas: A clinicopathological analysis of factors predictive of recurrence and functional outcome. Neurosurgery 1994;35:1001–1010. 457. Hayward R: The present and future management of childhood craniopharyngioma. Childs Nerv Syst 1999;15:764–769. 458. Laws ER Jr: Transsphenoidal removal of craniopharyngioma. Pediatr Neurosurg 1994;21(suppl 1):57–63.
W9418-11.qxd 10/31/03 3:43 PM Page 332
332
Section I
Introduction
459. Tomita T, McLone DG: Radical resections of childhood craniopharyngiomas. Pediatr Neurosurg 1993;19:6–14. 460. Yasargil MG, Curcic M, Kis M, Siegenthaler G, Teddy PJ, Roth P: Total removal of craniopharyngiomas. Approaches and long–term results in 144 patients. J Neurosurg 1990;73:3–11. 461. Hoffman HJ: Surgical management of craniopharyngioma. Pediatr Neurosurg 1994;21(suppl 1):44–49. 462. Hetelekidis S, Barnes PD, Tao ML, et al: 20-year experience in childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1993;27:189– 195. 463. Mark RJ, Lutge WR, Shimizu KT, Tran LM, Selch MT, Parker RG: Craniopharyngioma: Treatment in the CT and MR imaging era. Radiology 1995;197:195–198. 464. Vernet O, Montes JL, Farmer JP, Blundell JE, Bertrand G, Freeman CR: Long term results of multimodality treatment of craniopharyngioma in children. J Clin Neurosci 1999;6:199–203. 465. Fischer EG, Welch K, Shillito J Jr, Winston KR, Tarbell NJ: Craniopharyngiomas in children. Long–term effects of conservative surgical procedures combined with radiation therapy. J Neurosurg 1990;73:534–540. 466. Habrand JL, Ganry O, Couanet D, et al: The role of radiation therapy in the management of craniopharyngioma: A 25-year experience and review of the literature. Int J Radiat Oncol Biol Phys 1999;44:255–263. 467. Kobayashi T, Tanaka T, Kida Y: Stereotactic gamma radiosurgery of craniopharyngiomas. Pediatr Neurosurg 1994;21(suppl 1):69–74. 468. Mokry M: Craniopharyngiomas: A six year experience with Gamma Knife radiosurgery. Stereotact Funct Neurosurg 1999;72(suppl 1): 140–149. 469. Tarbell NJ, Barnes P, Scott RM, et al. Advances in radiation therapy for craniopharyngiomas. Pediatr Neurosurg 1994;21(suppl 1):101– 107. 470. Hader WJ, Steinbok P, Hukin J, Fryer C: Intratumoral therapy with bleomycin for cystic craniopharyngiomas in children. Pediatr Neurosurg 2000;33:211–218. 471. Takahashi H, Nakazawa S, Shimura T: Evaluation of postoperative intratumoral injection of bleomycin for craniopharyngioma in children. J Neurosurg 1985;62:120–127. 472. Van den Berge JH, Blaauw G, Breeman WA, Rahmy A, Wijngaarde R: Intracavitary brachytherapy of cystic craniopharyngiomas. J Neurosurg 1992;77:545–550. 473. Pollock BE, Lunsford LD, Kondziolka D, Levine G, Flickinger JC: Phosphorus–32 intracavitary irradiation of cystic craniopharyngiomas: Current technique and long–term results. Int J Radiat Oncol Biol Phys 1995;33:437–446. 474. Friedman JM: Epidemiology of neurofibromatosis type 1. Am J Med Genet 1999;89:1–6. 475. Rasmussen SA, Friedman JM: NF1 gene and neurofibromatosis 1. Am J Epidemiol 2000;151:33–40. 476. Xu GF, O’Connell P, Viskochil D, et al: The neurofibromatosis type 1 gene encodes a protein related to GAP. Cell 1990;62:599–608. 477. Shen MH, Harper PS, Upadhyaya M: Molecular genetics of neurofibromatosis type 1 (NF1). J Med Genet 1996;33:2–17. 478. Xu H, Gutmann DH: Mutations in the GAP–related domain impair the ability of neurofibromin to associate with microtubules. Brain Res 1997;759:149–152. 479. Es SV, North KN, McHugh K, Silva MD: MRI findings in children with neurofibromatosis type 1: A prospective study. Pediatr Radiol 1996;26:478–487. 480. Sevick RJ, Barkovich AJ, Edwards MS, Koch T, Berg B, Lempert T: Evolution of white matter lesions in neurofibromatosis type 1: MR findings. AJR 1992;159:171–175. 481. Faro SH, Higginson SN, Koenigsberg R, et al: Phakomatoses. Part I: Neurofibromatosis type 1: Common and uncommon neuroimaging findings. J Neuroimaging 2000;10:138–146. 482. Aoki S, Barkovich AJ, Nishimura K, et al: Neurofibromatosis types 1 and 2: Cranial MR findings. Radiology 1989;172:527–534.
483. Pollack IF, Mulvihill JJ: Neurofibromatosis. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:719–740. 484. Listernick R, Charrow J, Greenwald MJ, Esterly NB: Optic gliomas in children with neurofibromatosis type 1. J Pediatr 1989;114:788–792. 485. Listernick R, Darling C, Greenwald M, Strauss L, Charrow J: Optic pathway tumors in children: The effect of neurofibromatosis type 1 on clinical manifestations and natural history. J Pediatr 1995;127:718– 722. 486. Chateil JF, Soussotte C, Pedespan JM, Brun M, Le Manh C, Diard F: MRI and clinical differences between optic pathway tumours in children with and without neurofibromatosis. Br J Radiol 2001;74:24–31. 487. Listernick R, Charrow J, Greenwald M, Mets M: Natural history of optic pathway tumors in children with neurofibromatosis type 1: A longitudinal study. J Pediatr 1994;125:63–66. 488. Hoffman HJ, Soloniuk DS, Humphreys RP, et al: Management and outcome of low–grade astrocytomas of the midline in children: A retrospective review. Neurosurgery 1993;33:964–971. 489. Pollack IF, Mulvihill JJ: Special issues in the management of gliomas in children with neurofibromatosis 1. J Neurooncol 1996;28:257–268. 490. Pollack IF, Shultz B, Mulvihill JJ: The management of brainstem gliomas in patients with neurofibromatosis 1. Neurology 1996;46: 1652–1660. 491. Donner TR, Voorhies RM, Kline DG: Neural sheath tumors of major nerves. J Neurosurg 1994;81:362–373. 492. Needle MN, Cnaan A, Dattilo J, et al: Prognostic signs in the surgical management of plexiform neurofibroma: The Children’s Hospital of Philadelphia experience, 1974–1994. J Pediatr 1997;131:678–682. 493. Pollack IF, Colak A, Fitz C, Wiener E, Moreland M, Mulvihill JJ: Surgical management of spinal cord compression from plexiform neurofibromas in patients with neurofibromatosis 1. Neurosurgery 1998;43:248–255. 494. Constantini S, Miller DC, Allen JC, Rorke LB, Freed D, Epstein FJ: Radical excision of intramedullary spinal cord tumors: Surgical morbidity and long–term follow–up evaluation in 164 children and young adults. J Neurosurg 2000;93(suppl 2):183–193. 495. Muszynski CA, Constantiti S, Epstein FJ: Intraspinal intramedullary neoplasms. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:697– 709. 496. O’Sullivan C, Jenkin RD, Doherty MA, Hoffman HJ, Greenberg ML: Spinal cord tumors in children: Long–term results of combined surgical and radiation treatment. J Neurosurg 1994;81:507–512. 497. Steinbok P, Cochrane DD, Poskitt K: Intramedullary spinal cord tumors in children. Neurosurg Clin North Am 1992;3:931–945. 498. Rifkinson–Mann S, Wisoff JH, Epstein F: The association of hydrocephalus with intramedullary spinal cord tumors: A series of 25 patients. Neurosurgery 1990;27:749–754. 499. Koeller KK, Rosenblum RS, Morrison AL: Neoplasms of the spinal cord and filum terminale: Radiologic–pathologic correlation. Radiographics 2000;20:1721–1749. 500. Przybylski GJ, Albright AL, Martinez AJ: Spinal cord astrocytomas: long–term results comparing treatments in children. Childs Nerv Syst 1997;13:375–382. 501. Nadkarni TD, Rekate HL: Pediatric intramedullary spinal cord tumors. Critical review of the literature. Childs Nerv Syst 1999;15:17– 28. 502. Hardison HH, Packer RJ, Rorke LB, Schut L, Sutton LN, Bruce DA: Outcome of children with primary intramedullary spinal cord tumors. Childs Nerv Syst 1987;3:89–92. 503. Humphreys RP, Hoffman HJ, Drake JM, Rutka JT: Choices in the 1990s for the management of pediatric cerebral arteriovenous malformations. Pediatr Neurosurg 1996;25:277–285. 504. Hoh BL, Ogilvy CS, Butler WE, Loeffler JS, Putman CM, Chapman PH: Multimodality treatment of nongalenic arteriovenous malformations in pediatric patients. Neurosurg 2000;47:346–357.
W9418-11.qxd 10/31/03 3:43 PM Page 333
Chapter 11 505. Rodesch G, Malherbe V, Alvarez H, Zerah M, Devictor D, Lasjaunias P: Nongalenic cerebral arteriovenous malformations in neonates and infants. Review of 26 consecutive cases (1982–1992). Childs Nerv Syst 1995;11:231–241. 506. Mickle JP, Glasser RS: Cerebral vascular diseases in children. In Youmans JR (ed): Neurological Surgery, 4th Ed. Philadelphia, WB Saunders, 1996:1243–1263. 507. Luessenhop AJ: Natural history of cerebral arteriovenus malformations. In Wilson CB, Stein BM (eds): Intracranial Arteriovenous Malformations. Baltimore, Williams and Wilkins, 1984:12–23. 508. Spetzler RF, Martin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–483. 509. Mullan S, Mojtahedi S, Johnson DL, Macdonald RL: Embryological basis of some aspects of cerebral vascular fistulas and malformations. J Neurosurg 1996;85:1–8. 510. Sonstein WJ, Kader A, Michelsen WJ, Llena JF, Hirano A, Casper D: Expression of vascular endothelial growth factor in pediatric and adult cerebral arteriovenous malformations: An immunocytochemical study. J Neurosurg 1996;85:838–845. 511. Iizuka Y, Rodesch G, Garcia-Monaco R, et al: Multiple cerebral arteriovenous shunts in children: Report of 13 cases. Childs Nerv Syst 1992;8:437–444. 512. Willemse RB, Mager JJ, Westermann CJ, Overtoom TT, Mauser H, Wolbers JG: Bleeding risk of cerebrovascular malformations in hereditary hemorrhagic telangiectasia. J Neurosurg 2000;92:779–784. 513. Kondziolka D, McLaughlin MR, Kestle JR: Simple risk predictions for arteriovenous malformation hemorrhage. Neurosurgery 1995;37:851– 855. 514. Lasjaunias P, Hui F, Zerah M, et al: Cerebral arteriovenous malformations in children. Management of 179 consecutive cases and review of the literature. Childs Nerv Syst 1995;11:66–79. 515. Di Rocco C, Tamburrini G, Rollo M: Cerebral arteriovenous malformations in children. Acta Neurochir (Wien) 2000;142:145–156. 516. Frizzel RT, Fisher WS 3rd: Cure, morbidity, and mortality associated with embolization of brain arteriovenous malformations: A review of 1246 patients in 32 series over a 35-year period. Neurosurgery 1995;37:1031–1039. 517. Levy EI, Niranjan A, Thompson TP, et al: Radiosurgery for childhood intracranial arteriovenous malformations. Neurosurgery 2000;47: 834–841. 518. Gerszten PC, Adelson PD, Kondziolka D, Flickinger JC, Lunsford LD: Seizure outcome in children treated for arteriovenous malformations using gamma knife radiosurgery. Pediatr Neurosurg 1996;24:139–144. 519. Kader A, Goodrich JT, Sonstein WJ, Stein BM, Carmel PW, Michelsen WJ: Recurrent cerebral arteriovenous malformations after negative postoperative angiograms. J Neurosurg 1996;85:14–18. 520. Hoffman HJ, Chuang S, Hendrick EB, Humphreys RP: Aneurysms of the vein of Galen. Experience at The Hospital for Sick Children, Toronto. J Neurosurg 1982;57:316–322. 521. Johnston IH, Whittle IR, Besser M, Morgan MK: Vein of Galen malformation: Diagnosis and management. Neurosurgery 1987;20:747–758. 522. Mickle JP, Herbert A: Vein of Galen and dural malformations in childhood. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:1033– 1051. 523. Yasargil MG: Microneurosurgery: Volume IIIB, AVM of the Brain, Clinical Considerations, General and Special Operative Techniques, Surgical Results, Nonoperated Cases, Cavernous and Venous Angiomas, Neuroanesthesia. New York, Thieme; 1988:323–357. 524. Raybaud CA, Strother CM, Hald JK: Aneurysms of the vein of Galen: Embryonic considerations and anatomical features relating to the pathogenesis of the malformation. Neuroradiology 1989;31:109–128. 525. Yasargil MG, Antic J, Laciga R, Jain KK, Boone SC: Arteriovenous malformations of vein of Galen: Microsurgical treatment. Surg Neurol 1976;195–200.
Pediatric Neurosurgery
333
526. Mickle JP, Quisling RG: The transtorcular embolization of vein of Galen aneurysms. J Neurosurg 1986;64:731–735. 527. Lasjaunias P, Garcia-Monaco R, Rodesch G, et al: Vein of Galen malformation. Endovascular management of 43 cases. Childs Nerv Syst 1991;7:360–367. 528. Lasjaunias P, Hui F, Zerah M, et al: Cerebral arteriovenous malformations in children. Management of 179 consecutive cases and review of the literature. Childs Nerv Syst 1995;11:66–79. 529. Lylyk P, Vinuela F, Dion JE, et al: Therapeutic alternatives for vein of Galen vascular malformations. J Neurosurg 1993;78:438–445. 530. Horowitz MB, Jungreis CA, Quisling RG, Pollack I: Vein of Galen aneurysms: A review and current perspective. AJNR 1994;15:1486– 1496. 531. Maraire JN, Awad IA: Intracranial cavernous malformations: Lesion behavior and management strategies. Neurosurgery 1995;37:591–605. 532. Herter T, Brandt M, Szuwart U: Cavernous hemangiomas in children. Childs Nerv Syst 1988;4:123–127. 533. Frim DM, Scott RM: Management of cavernous malformations in the pediatric population. Neurosurg Clin North Am 1999;10: 513–518. 534. Fortuna A, Ferrante L, Mastronardi L, Acqui M, d’Addetta R: Cerebral cavernous angioma in children. Childs Nerv Syst 1989;5:201–207. 535. Edwards M, Baumgartner J, Wilson C: Cavernous and other cryptic vascular malformations of the pediatric age group. In Awad IA, Barrow DL (eds): Cavernous Malformations. Park Ridge, IL, American Association of Neurological Surgeons, 1993;163. 536. Mazza C, Scienza R, Beltramello A, Da Pian R: Cerebral cavernous malformations (cavernomas) in the pediatric age-group. Childs Nerv Syst 1991;7:139–146. 537. Scott RM, Barnes P, Kupsky W, Adelman LS: Cavernous angiomas of the central nervous system in children. J Neurosurg 1992;76:38–46. 538. Gunel M, Awad IA, Anson J, Lifton RP: Mapping a gene causing cerebral cavernous malformation to 7q11.2-q21. Proc Natl Acad Sci USA 1995;92:6620–6624. 539. Gunel M, Awad IA, Finberg K, et al: A founder mutation as a cause of cerebral cavernous malformation in Hispanic Americans. N Engl J Med 1996;334:946–951. 540. Notelet L, Chapon F, Khoury S, et al: Familial cavernous malformations in a large French kindred: Mapping of the gene to the CCM1 locus on chromosome 7q. J Neurol Neurosurg Psychiatry 1997;63:40– 45. 541. Dubovsky J, Zabramski JM, Kurth J, et al: A gene responsible for cavernous malformations of the brain maps to chromosome 7q. Hum Mol Genet 1995;4:453–458. 542. Gunel M, Awad IA, Finberg K, et al: Genetic heterogeneity of inherited cerebral cavernous malformation. Neurosurgery 1996;38:1265– 1271. 543. Kondziolka D, Lunsford LD, Kestle JR: The natural history of cerebral cavernous malformations. J Neurosurg 1995;83:820–824. 544. Wong JH, Awad IA, Kim JH: Ultrastructural pathological features of cerebrovascular malformations: A preliminary report. Neurosurgery 2000;46:1454–1459. 545. Maraire JN, Awad IA: Intracranial cavernous malformations: Lesion behavior and management strategies. Neurosurgery 1995;37:591–605. 546. Abe T, Singer RJ, Marks MP, Norbash AM, Crowley RS, Steinberg GK: Coexistence of occult vascular malformations and developmental venous anomalies in the central nervous system: MR evaluation. AJNR 1998;19:51–57. 547. Chang SD, Vanefsky MA, Havton LA, Silverberg GD: Bilateral cavernous malformations resulting from cranial irradiation of a choroid plexus papilloma. Neurol Res 1998;20:529–532. 548. Maeder P, Gudinchet F, Meuli R, de Tribolet N: Development of a cavernous malformation of the brain. AJNR 1998;19:1141–1143. 549. Zabramski JM, Wascher TM, Spetzler RF, et al: The natural history of familial cavernous malformations: Results of an ongoing study. J Neurosurg 1994;80:422–432.
W9418-11.qxd 10/31/03 3:43 PM Page 334
334
Section I
Introduction
550. Labauge P, Brunereau L, Levy C, Laberge S, Houtteville JP: The natural history of familial cerebral cavernomas: A retrospective MRI study of 40 patients. Neuroradiology 2000;42:327–332. 551. Zevgaridis D, van Velthoven V, Ebeling U, Reulen HJ: Seizure control following surgery in supratentorial cavernous malformations: A retrospective study in 77 patients. Acta Neurochir (Wien) 1996;138:672– 677. 552. Scott RM: Brain stem cavernous angiomas in children. Pediatr Neurosurg 1990;16:281–286. 553. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM: Cavernous malformations of the brain stem. A review of 139 cases. Acta Neurochir (Wien) 1994;130:35–46. 554. Morcos JJ, Heros RC, Frank DE: Microsurgical treatment of infratentorial malformations. Neurosurg Clin North Am 1999;10:441–474. 555. Maesawa S, Kondziolka D, Lunsford LD: Stereotactic radiosurgery for management of deep brain cavernous malformations. Neurosurg Clin North Am 1999;10:503–511. 556. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR: Reduction of hemorrhage risk after stereotactic radiosurgery for cavernous malformations. J Neurosurg 1995;83:825–831. 557. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link MJ: Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000;93:987–991. 558. Karlsson B, Kihlstrom L, Lindquist C, Ericson K, Steiner L: Radiosurgery for cavernous malformations. J Neurosurg 1998;88: 293–297. 559. Matsushima Y: Moyamoya disease. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:1053–1069. 560. Kim SK, Wang KC, Kim DG, et al: Clinical feature and outcome of pediatric cerebrovascular disease: A neurosurgical series. Childs Nerv Syst 2000;16:421–428. 561. Fukui M: Current state of study on moyamoya disease in Japan. Surg Neurol 1997;47:138–143. 562. Scott RM: Moyamoya syndrome: A surgically treatable cause of stroke in the pediatric patient. Clin Neurosurg 2000;47:378–384. 563. Matsushima Y, Aoyagi M, Nariai T, Takada Y, Hirakawa K: Long–term intelligence outcome of post-encephalo-duro-arterio-synangiosis childhood moyamoya patients. Clin Neurol Neurosurg 1997;99(suppl 2):S147–S150. 564. Suzuki J, Kodama N: Moyamoya disease—a review. Stroke 1983;14:104–109. 565. Matsushima T, Inoue T, Suzuki SO, Fujii K, Fukui M, Hasuo K: Surgical treatment of moyamoya disease in pediatric patients— Comparison between the results of indirect and direct revascularization procedures. Neurosurgery 1992;31:401–405. 566. Yamauchi T, Tada M, Houkin K, et al: Linkage of familial moyamoya disease (spontaneous occlusion of the circle of Willis) to chromosome 17q25. Stroke 2000;31:930–935. 567. Adelson PD, Scott RM: Pial synangiosis for moyamoya syndrome in children. Pediatr Neurosurg 1995;23:26–33. 568. Matsushima Y, Aoyagi M, Suzuki R, Tabata H, Ohno K: Perioperative complications of encephalo-duro-arterio-synangiosis: Prevention and treatment. Surg Neurol 1991;36:343–353. 569. Kim SK, Wang KC, Kim DG, et al: Clinical feature and outcome of pediatric cerebrovascular disease: A neurosurgical series. Childs Nerv Syst 2000;16:421–428. 570. Karasawa J, Touho H, Ohnishi H, Miyamoto S, Kikuchi H: Long-term follow-up study after extracranial–intracranial bypass surgery for anterior circulation ischemia in childhood moyamoya disease. J Neurosurg 1992;77:84–89. 571. Nathoo N, Nadvi SS, van Dellen JR, Gouws E: Intracranial subdural empyemas in the era of computed tomography: A review of 699 cases. Neurosurgery 1999;44:529–535. 572. Dill SR, Cobbs CG, McDonald CK: Subdural empyema: Analysis of 32 cases and review. Clin Infect Dis 1995;20:372–386.
573. Pathak A, Sharma BS, Mathuriya SN, Khosla VK, Khandelwal N, Kaak VK: Controversies in the management of subdural empyema. A study of 41 cases with review of literature. Acta Neurochir (Wien) 1990;102:25–32. 574. Lerner DN, Choi SS, Zalzal GH, Johnson DL: Intracranial complications of sinusitis in childhood. Ann Otol Rhinol Laryngol 1995;104:288–293. 575. Maniglia AJ, Goodwin WJ, Arnold JE, Ganz E: Intracranial abscesses secondary to nasal, sinus, and orbital infections in adults and children. Arch Otolaryngol Head Neck Surg 1989;115:1424–1429. 576. Gower DJ, McGuirt WF, Kelly DL Jr: Intracranial complications of ear disease in a pediatric population with special emphasis on subdural effusion and empyema. South Med J 1985;78:429–434. 577. Hlavin ML, Kaminski HJ, Fenstermaker RA, White RJ: Intracranial suppuration: A modern decade of postoperative subdural empyema and epidural abscess. Neurosurgery 1994;34:974–980. 578. Curless RG: Subdural empyema in infant meningitis: Diagnosis, therapy, and prognosis. Childs Nerv Syst 1985;1:211–214. 579. Bok AP, Peter JC: Subdural empyema: Burr holes or craniotomy? A retrospective computerized tomography-era analysis of treatment in 90 cases. J Neurosurg 1993;78:574–578. 580. Feuerman T, Wackym PA, Gade GF, Dubrow T: Craniotomy improves outcome in subdural empyema. Surg Neurol 1989;32:105–110. 581. Nathoo N, Nadvi SS, van Dellen JR: Infratentorial empyema: Analysis of 22 cases. Neurosurgery 1997;41:1263–1268. 582. Weingarten K, Zimmerman RD, Becker RD, Heier LA, Haimes AB, Deck MD: Subdural and epidural empyemas: MR imaging. AJR 1989;152:615–621. 583. Nathoo N, Nadvi SS, Gouws E, van Dellen JR: Craniotomy improves outcomes for cranial subdural empyemas: Computed tomography–era experience with 699 patients. Neurosurgery 2001;49:872–877. 584. Pattisapu JV, Parent AD: Subdural empyemas in children. Pediatr Neurosci 1987;13:251–254. 585. Nathoo N, Nadvi SS, van Dellen JR: Cranial extradural empyema in the era of computed tomography: A review of 82 cases. Neurosurgery 1999;44:748–753. 586. Smith HP, Hendrick EB: Subdural empyema and epidural abscess in children. J Neurosurg 1983;58:392–397. 587. Giannoni C, Sulek M, Friedman EM: Intracranial complications of sinusitis: a pediatric series. Am J Rhinol 1998;12:173–178. 588. Go C, Bernstein JM, de Jong AL, Sulek M, Friedman EM: Intracranial complications of acute mastoiditis. Int J Pediatr Otorhinolaryngol 2000;52:143–148. 589. Ciurea AV, Stoica F, Vasilescu G, Nuteanu L: Neurosurgical management of brain abscesses in children. Childs Nerv Syst 1999;15:309–317. 590. Jadavji T, Humphreys RP, Prober CG: Brain abscesses in infants and children. Pediatr Infect Dis 1985;4:394–398. 591. Tekkok IH, Erbengi A: Management of brain abscess in children: Review of 130 cases over a period of 21 years. Childs Nerv Syst 1992;8:411–416. 592. Takeshita M, Kagawa M, Yato S, et al: Current treatment of brain abscess in patients with congenital cyanotic heart disease. Neurosurgery 1997;41:1270–1278. 593. Wong TT, Lee LS, Wang HS, et al: Brain abscesses in children—a cooperative study of 83 cases. Childs Nerv Syst 1989;5:19–24. 594. Renier D, Flandin C, Hirsch E, Hirsch JF: Brain abscesses in neonates. A study of 30 cases. J Neurosurg 1988;69:877–882. 595. Ersahin Y, Mutluer S, Guzelbag E: Brain abscess in infants and children. Childs Nerv Syst 1994;10:185–189. 596. Boop FA, Jacobs RF, Young RF: Brain abscesses and encephalitis in children. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:1203– 1226. 597. Graham DR, Band JD: Citrobacter diversus brain abscess and meningitis in neonates. JAMA 1981;245:1923–1925.
W9418-11.qxd 10/31/03 3:43 PM Page 335
Chapter 11 598. Britt RH, Enzmann DR, Placone RC Jr, Obana WG, Yeager AS: Experimental anaerobic brain abscess. Computerized tomographic and neuropathological correlations. J Neurosurg 1984;60:1148–1159. 599. Rosenblum ML, Mampalam TJ, Pons VG: Controversies in the management of brain abscesses. Clin Neurosurg 1986;33:603–632. 600. Barlas O, Sencer A, Erkan K, Eraksoy H, Sencer S, Bayindir C: Stereotactic surgery in the management of brain abscess. Surg Neurol 1999;52:404–410. 601. Mamelak AN, Mampalam TJ, Obana WG, Rosenblum ML: Improved management of multiple brain abscesses: A combined surgical and medical approach. Neurosurgery 1995;36:76–85. 602. Hirsch JF, Roux FX, Sainte-Rose C, Renier D, Pierre-Kahn A: Brain abscess in childhood. A study of 34 cases treated by puncture and antibiotics. Childs Brain 1983;10:251–265. 603. Buonaguro A, Colangelo M, Daniele B, Cantone G, Ambrosio A: Neurological and behavioral sequelae in children operated on for brain abscess. Childs Nerv Syst 1989;5:153–155. 604. Carey ME, Chou SN, French LA: Long–term neurological residua in patients surviving brain abscess with surgery. J Neurosurg 1971;34:652–656. 605. Albright AL: Spasticity and movement disorders. In Albright AL, Pollack IF, Adelson PD (eds): Principles and Practice of Pediatric Neurosurgery. New York, Thieme, 1999:1157–1173. 606. Thompson FJ, Parmer R, Reier PJ, Wang DC, Bose P: Scientific basis of spasticity: Insights from a laboratory model. J Child Neurol 2001;16:2–9. 607. Fahn S: Concept and classification of dystonia. Adv Neurol 1988;50: 1–8. 608. Jarman PR, Warner TT: The dystonias. J Med Genet 1998;35:314–318. 609. Fahn S, Bressman SB, Marsden CD: Classification of dystonia. Adv Neurol 1998;78:1–10. 610. Hallett M: The neurophysiology of dystonia. Arch Neurol 1998;55:601–603. 611. DeLong MR: Primate models of movement disorders of basal ganglia origin. Trends Neurosci 1990;13:281–285. 612. Parent A, Cicchetti F: The current model of basal ganglia organization under scrutiny. Mov Disord 1998;13:199–202. 613. Ashworth B: Preliminary trial of carisoprodol in multiple sclerosis. Practitioner 1964;192:540–542. 614. Barry MJ, VanSwearingen JM, Albright AL: Reliability and responsiveness of the Barry-Albright dystonia scale. Dev Med Child Neurol 1999;41:404–411. 615. Nelson KB, Ellenberg JH: Children who “outgrew’ cerebral palsy. Pediatrics 1982;69:529–536. 616. Milla PJ, Jackson AD: A controlled trial of baclofen in children with cerebral palsy. J Int Med Res 1977;5:398–404. 617. Krach LE: Pharmacotherapy of spasticity: Oral medications and intrathecal baclofen. J Child Neurol 2001;16:31–36. 618. Lapierre Y, Bouchard S, Tansey C, Gendron D, Barkas WJ, Francis GS: Treatment of spasticity with tizanidine in multiple sclerosis. Can J Neurol Sci 1987;14:513–517. 619. Coward DM: Tizanidine: Neuropharmacology and mechanism of action. Neurology 1994;44(suppl 9):S6–S10. 620. Groves L, Shellenberger MK, Davis CS: Tizanidine treatment of spasticity: A meta-analysis of controlled, double-blind, comparative studies with baclofen and diazepam. Adv Ther 1998;15:241–251. 621. Wallace JD: Summary of combined clinical analysis of controlled clinical trials with tizanidine. Neurology 1994;44:S60–68. 622. Edgar TS: Clinical utility of botulinum toxin in the treatment of cerebral palsy: Comprehensive review. J Child Neurol 2001;16:37–46. 623. Ubhi T, Bhakta BB, Ives HL, Allgar V, Roussounis SH: Randomised double blind placebo controlled trial of the effect of botulinum toxin on walking in cerebral palsy. Arch Dis Child 2000;83:481–487.
Pediatric Neurosurgery
335
624. Fehlings D, Rang M, Glazier J, Steele C: An evaluation of botulinumA toxin injections to improve upper extremity function in children with hemiplegic cerebral palsy. J Pediatr 2000;137:331–337. 625. Koman LA, Mooney JF 3rd, Smith BP, Walker F, Leon JM: Botulinum toxin type A neuromuscular blockade in the treatment of lower extremity spasticity in cerebral palsy: A randomized, double-blind, placebo-controlled trial. BOTOX Study Group. J Pediatr Orthop 2000;20:108–115. 626. Sutherland DH, Kaufman KR, Wyatt MP, Chambers HG, Mubarak SJ: Double-blind study of botulinum A toxin injections into the gastrocnemius muscle in patients with cerebral palsy. Gait Posture 1999;10:1– 9. 627. Eames NW, Baker R, Hill N, Graham K, Taylor T, Cosgrove A: The effect of botulinum toxin A on gastrocnemius length: Magnitude and duration of response. Dev Med Child Neuro 1999;41:226–232. 628. Peacock WJ, Arens LJ: Selective posterior rhizotomy for the relief of spasticity in cerebral palsy. S Afr Med J 1982;62:119–124. 629. Peacock WJ, Arens LJ, Berman B: Cerebral palsy spasticity. Selective posterior rhizotomy. Pediatr Neurosci 1987;13:61–66. 630. Steinbok P, Keyes R, Langill L, Cochrane DD: The validity of electrophysiological criteria used in selective functional posterior rhizotomy for treatment of spastic cerebral palsy. J Neurosurg 1994;81:354–361. 631. Steinbok P: Outcomes after selective dorsal rhizotomy for spastic cerebral palsy. Childs Nerv Syst 2001;17:1–18. 632. Albright AL, Barron WB, Fasick MP, Polinko P, Janosky J: Continuous intrathecal baclofen infusion for spasticity of cerebral origin. JAMA 1993;270:2475–2477. 633. Albright AL: Baclofen in the treatment of cerebral palsy. J Child Neurol 1996;11:77–83. 634. Armstrong RW, Steinbok P, Cochrane DD, Kube SD, Fife SE, Farrell K: Intrathecally administered baclofen for treatment of children with spasticity of cerebral origin. J Neurosurg 1997;87:409–414. 635. Steinbok P, Daneshvar H, Evans D, Kestle JR: Cost analysis of continuous intrathecal baclofen versus selective functional posterior rhizotomy in the treatment of spastic quadriplegia associated with cerebral palsy. Pediatr Neurosurg 1995;22:255–264. 636. Fahn S: High-dosage anticholinergic therapy in dystonia. Adv Neurol 1983;37:177–188. 637. Hoon AH Jr, Freese PO, Reinhardt EM, et al: Age-dependent effects of trihexyphenidyl in extrapyramidal cerebral palsy. Pediatr Neurol 2001;25:55–58. 638. Nygaard TG, Marsden CD, Fahn S: Dopa–responsive dystonia: Long–term treatment response and prognosis. Neurology 1991;41: 174–181. 639. Quirk JA, Sheehan GL, Marsden CD, Lees AJ: Treatment of nonoccupational limb and trunk dystonia with botulinum toxin. Mov Disord 1996;11:377–383. 640. Arens LJ, Leary PM, Goldschmidt RB: Experience with botulinum toxin in the treatment of cerebral palsy. S Afr Med J 1997;87:1001– 1003. 641. Gilio F, Curra A, Lorenzano C, Modugno N, Manfredi M, Berardelli A: Effects of botulinum toxin type A on intracortical inhibition in patients with dystonia. Ann Neurol 2000;48:20–26. 642. Albright AL, Barry MJ, Painter MJ, Shultz B: Infusion of intrathecal baclofen for generalized dystonia in cerebral palsy. J Neurosurg 1998;88:73–76. 643. Albright AL, Barry MJ, Shafron DH, Ferson SS: Intrathecal baclofen for generalized dystonia. Dev Med Child Neurol 2001;43:652–657. 644. Speelman D, van Manen J: Cerebral palsy and stereotactic neurosurgery: Long term results. J Neurol Neurosurg Psychiatry 1989;52:23–30.
W9418-12.qxd 11/3/03 2:58 PM Page 337
Chapter 12 Central Nervous System Infections Kenneth H. Rand, MD, with a contribution on Vertebral Osteomyelitis, Epidural Abscess, and External Ventricular Drain Infections, Arthur J. Ulm, MD and David W. Pincus, MD, PhD
Acute Bacterial Meningitis Definition Bacterial meningitis can be defined as a pyogenic inflammatory response to bacterial invasion of the pia arachnoid membranes surrounding the central nervous system. This infection typically involves the entire length of the neuraxis including the brain, spinal cord, and optic nerves because the subarachnoid space is continuous. Clinically, one sees the acute onset of headache, fever, and stiff neck with or without focal neurological signs over hours to a few days. Epidemiology The most common bacterial causes of bacterial meningitis in the United States include Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae, Listeria monocytogenes, and group B streptococcus. Collectively, these agents account for over 95% of cases. Table 12-1 summarizes the age distribution, predisposing conditions and fatality rates for the most common agents. Streptococcus pneumoniae S. pneumoniae is the most common cause of bacterial meningitis in the United States today in all age groups, except infants in the immediate neonatal period. The risk of pneumococcal meningitis varies with age, being significantly higher in infants than in young children and adults. When patients are older than 70 years of age, the incidence increases again to approximately double the average
for young and middle-aged adults. In children the most common predisposing conditions are sinus or middle ear infection, which lead to transient bacteremia and hematogenous seeding of the central nervous system.1 In adults, the major risk factors include alcoholism, splenectomy, human immunodeficiency virus (HIV), diabetes, sinusitis, spontaneous bacterial peritonitis, and acquired immunodeficiencies. Pneumococcal meningitis is the most common form of recurrent meningitis in patients who have cerebrospinal fluid (CSF) leaks. S. pneumonia is spread by respiratory transmission in the general population and results in colonization of the nasopharynx with rates commonly in the range of 5% to 10% of healthy adults. During the winter, carriage rates can increase to 20% to 30% in certain populations. Overcrowded environments such as day care centers, barracks, and prisons may serve as foci for spreading of these organisms. Neisseria meningitidis The incidence of meningitis due to N. meningitidis in the United States has been estimated at 0.6 per 100,000 population per year.2 The incidence of meningococcal meningitis is at least tenfold higher in third world countries. Strains of N. meningitidis are classified according to serologic recognition of epitopes on their capsule and outer membrane. N. meningitidis is classified into serogroups A, B, C, Y, W135 and other less common types based on polysaccharide capsule antigens. In the United States, strains from sera groups B and C cause the majority of infections, whereas in underdeveloped countries sera groups A and C predominate. Over the past 30 years in industrialized nations, clonal outbreaks of strains 337
W9418-12.qxd 11/3/03 2:58 PM Page 338
338
Section I
Introduction
Table 12-1 Age, Proportion of Cases, Predisposing Factors, and Approximate Fatality Rates for the Most Common Agents of Acute Bacterial Meningitis Microorganism
Age
Haemophilus influenzae
ª 4 mo–3 y
Streptococcus pneumoniae
All
Neisseria meningitidis
Percentage of Cases 50 yr
5–10
Newborn period, Immunocompromise, Age, Alcoholism/cirrhosis
ª 15
CSF, cerebrospinal fluid.
belonging to the sera group B have predominated. For example, in Northern Europe a pattern of hyperendemic infection with attack rates between 4 and 50 per 100,000 population has been observed since the mid 1970s. Most of these isolates have been identified as belonging to the ET5, lineage 3 serologic subgroup. Strains of this serotype have been observed to slowly circulate through the population and periodically result in isolated outbreaks. Such strains have been observed in the United States and in almost all parts of the world. Some clonal complexes such as ET-37 can nonetheless express the group C polysaccharide capsule type, but also sera group B, W135, and Y. Thus, epidemiologic studies using modern molecular typing methods show a low-level endemic background of cases punctuated by school dormitory–based or other epidemiologically linked local outbreaks of meningitis. Haemophilus influenzae Until the introduction of the H. influenzae type b, polysaccharide capsular vaccine, H. influenzae was the most common etiology of acute bacterial meningitis in children younger than 5 years of age. Before the introduction of the H. influenzae type b vaccination in 1986 it was estimated that there were approximately 12,920 cases of meningitis in the United States, compared with 5,755 in 1995, a reduction of 55%.2 Before the availability of this vaccine, it was estimated that invasive Haemophilus infection developed in as many as one in 200 children and included epiglottitis, septicemia, arthritis, and soft tissue infections in addition to meningitis. In a study in Washington state (King County), from 1977 to
1986, the incidence of H. influenzae meningitis in the age group younger than 5 months old was 63.5 per 100,000 population and increased to 128.2 per 100,000 in the 6- to 11month age group, decreasing to fewer than 8.5 per 100,000 after the age of 5 years.3 The occurrence of H. influenzae meningitis is directly related to the development of type specific anticapsular antibodies.4 The development of antibodies to polyribosylribitol phosphate, whether vaccine induced or occurring naturally, is directly related to protection from invasive Haemophilus infection. These antibodies have been shown to be opsonic and bactericidal against H. influenzae in vitro, and are protective in vivo as shown by numerous clinical studies. H. influenzae can be classified into six serologically distinct antigenic types based on capsular polysaccharides A-F. Of these, only type b is pathogenic. In the prevaccine era, colonization with nontypable strains of H. influenzae led to the development of cross-reacting antibodies, which were protective against infection due to type b. Following the introduction of the H. influenzae type b vaccine, there has been a profound reduction in the number of invasive infections due to H. influenzae in the United States. For example, Murphy and colleagues5 found a reduction of 85% to 92% in the incidence of invasive H. influenzae type b disease between 1983 to 1984 and 1991 after widespread use of the vaccine. Listeria monocytogenes L. monocytogenes is a significant cause of neonatal meningitis. This arises from a generally asymptomatic colonization of the genital or gastrointestinal tract in the mother before
W9418-12.qxd 11/3/03 2:58 PM Page 339
Chapter 12
delivery of an infant with transmission at the time of birth. L. monocytogenes causes meningitis most often in adults who are immunocompromised because of steroids, drugs for transplantation, diabetes, alcoholism, or who are extremely elderly. The organism is widespread in the animal population, both domestic and wild. Transmission to humans occurs through fecal-oral transmission from the animal reservoirs. Outbreaks in the United States and elsewhere have involved milk products such as Swiss cheese, undercooked chicken, hot dogs, seafood, vegetables, and a variety of other items. Stool carriage of this organism among asymptomatic adults has been documented in the range of 1%; thus the human reservoir of this agent is substantial. However, healthy adults exposed to the organism will generally not become ill unless exposed, in an outbreak type setting, to high levels of infectious organisms. Under normal circumstances, a significant degree of susceptibility of the host such as neonatal age group or the elderly, as indicated previously, is required for infection. Gram-Negative Bacilli Escherichia coli with the K1 capsular polysaccharide antigen accounts for a majority of the cases of gram-negative meningitis in the newborn.6 Carriage rates of the E. coli K1 serotypes vary in different populations, but range from 7% to 38% in women of child-bearing age and may be as high as 50% in nursing personnel.5,7–9 In a 15-year study of bacterial meningitis in a children’s hospital in Seattle, Washington, there was a total of 28 cases of E. coli meningitis comprising 3% of the total meningitis cases. However, 13 of these were in the neonatal group and accounted for 15% of neonatal meningitis.10 Other gram-negative organisms such as Klebsiella, Enterobacter, Pseudomonas, Citrobacter, and Salmonella may also cause meningitis in the neonatal period, with an epidemiology similar to that of the E. coli K1 sera type. Beyond the neonatal period, gram-negative meningitis is rare. However, gram-negative meningitis is highly significant in hospital acquired cases of meningitis. The vast majority of these cases are seen following neurocranial surgery, spinal surgery, and in patients who have suffered head trauma. Group B Streptococcus Streptococcus agalactiae (group B streptococcus) is the single most frequent cause of neonatal meningitis. This organism has been cultured from vaginal secretions in 30% to 35% of women before delivery, and transmission to the infant during delivery can result in neonatal meningitis within the first week of life. Alternatively, the organism may be acquired within the first few days after birth from adult contacts, either relatives or hospital personnel, and meningitis may develop during the first 1 to 2 months after birth. Group B streptococci are divided into six main serotypes: Ia, Ib/c, Ia/c, II, III, IV, based on capsular polysaccharide antigens. The vast majority of neonatal meningitis is caused by the type III
Central Nervous System Infections
339
group and virulence factors such as production of higher levels of neuraminidase have been described as an explanation for this. A recent analysis of 128 isolates suggested the clonality of certain invasive strain types.11 Other Etiologic Agents Acute meningitis can be produced by almost every organism known to medicine, including such organisms as group A streptococci, nonpneumococcal alpha hemolytic streptococci, Neisseria gonorrhea, Salmonella species, Flavobacterium meningosepticum, non–H. influenzae species, and even anthrax. Organisms such as Staphylococcus aureus and Staphylococcus epidermidis are extremely unusual as causes of primary bacterial meningitis. However, following trauma, the placement of CSF shunts, and neurosurgical procedures, these organisms are quite common causes of bacterial meningitis. Many other species such as mycobacteria, nocardia, fungi, spirochetes, brucella, and leptospira can also produce meningitis. However, with the exception of leptospira, the presentation of these illnesses tends to be more chronic and generally these would not be considered agents of acute bacterial meningitis. Pathogenesis of Meningeal Invasion Colonization of the respiratory tract is the critical first step, which precedes infection by the three major bacterial species causing meningitis. Biologically, colonization is mediated by attachment of these organisms to cell surface receptors and/or affinity for nasopharyngeal mucosa, which permits the organism to replicate in the upper airway for prolonged periods of time. All three major pathogens typically colonize the upper airway without producing symptoms. Both hostand pathogen-specific factors are critical in the development of invasive disease; many of these have been identified but are not fully understood. For example, splenectomy definitely predisposes to invasive disease by S. pneumoniae while having very little effect on the incidence of invasion by N. meningitidis, despite the fact that both are encapsulated organisms. The first step in colonization of the upper airway is attachment of the bacteria to the surface of the host mucosal epithelial cells. This adherence is mediated by fimbriae or pili in the case of gram-negative organisms. The pili of meningococcus are filamentous glycoproteins that are attached to the bacterial surface, traverse the polysaccharide capsule, and extend beyond the surface of the bacterium where they can bind to specific receptors on nasopharyngeal cells, in this instance the CD46 receptor.12,13 After receptor binding, further interaction with the host cell is established by certain outer membrane proteins on the meningococcus, designated Opa and Opc.14 Binding of the outer membrane proteins to specific receptors promotes engulfment of the meningococci by the epithelial cells and allows transportation of the meningococci across the cell. Meningococci also possess other outer membrane proteins that function as IgA pro-
W9418-12.qxd 11/3/03 2:58 PM Page 340
340
Section I
Introduction
teases, which can specifically degrade surface IgA, further enhancing the probability of invasive disease.15,16 Once the mucosal barrier has been penetrated, the development of meningococcal disease depends on the survival of the organism in the bloodstream. Here the most important virulence factor for survival of meningococci is the polysaccharide capsule, which protects against complement mediated phagocytosis by neutrophils in the reticular endothelial system.17 Host defense is clearly determined by the existing humoral antibody to specific polysaccharide capsular types and the cellular responses of the innate immune system. Protective antibody for meningococcal disease is acquired through maternal transmission and is protective for the first several months after birth. Colonization by nonpathogenic Neisseria, and a possibly cross-reacting gram-negative organism such as E. coli K1, induces protective antibodies. The protective effect of antibodies is to promote optimization of phagocytosis through opsonization and specific lysis via complement activation. For this reason, patients who are deficient in complement factor C5 are susceptible to repeated invasive infections by N. meningitidis. In fact, individuals with an inherited deficiency of any of the terminal components of complement C5, C6, C7, C8 have approximately 6000-fold greater risk of invasive disease.18–20 Interestingly, the overall mortality rate in these patients is only about 2%, less than that of patients whose complement system is normal. In two European studies of a total of 127 patients with fulminant meningococcal infection, there was only one case of complement deficiency.21,22 Colonization of the upper airway by H. influenzae is apparently also mediated by fimbrial attachment to epithelial cells. Alpha fimbriae enhance the binding to the anterior nasopharynx and b fimbriae facilitate binding to the posterior ciliated nasopharyngeal cells.23 H. influenzae type b lacking fimbriae appear not to colonize the nasopharynx and, interestingly, H. influenzae type b isolates from spinal fluid do not express fimbria, suggesting that although their presence is important in colonization and attachment, it is not necessary to cause meningitis.24,25 Virulence factors for the invasion of S. pneumoniae seem to be primarily a function of the capsular polysaccharide type. There are 90 known sera types, 18 of which are responsible for approximately 87% of bacteremic pneumococcal disease.26,27 The factors involved in invasion of the subarachnoid space have been extensively studied by Kim and colleagues28 using E. coli as a model system. However, while the specific details of the mechanisms described may or may not relate directly to invasion by meningococci and S. pneumoniae, the general principles probably do. Using human brain microvascular endothelial cells (BMEC), Prasadarao and co-workers29 were able to show that E. coli expressing the outer membrane protein A (OmpA) gene exhibited a 25- to 50-fold greater ability to attach and invade these vascular endothelial cells than E. coli not expressing the (OmpA) gene. The ability of E. coli to invade these cells in vitro was specific for the brain-derived endothelial cells and
was not found in endothelial cells obtained from human umbilical vein. On electron microscopy, invasive strains of E. coli were shown to migrate through the BMEC in enclosed vacuoles, which were dependent on recruitment of F-actin. Thus, transport through the cell appears to be dependent on cytoskeletal rearrangement involving both microfilaments and microtubules. Using a bacteremic neonatal rat model, these workers were also able to show that mutations of E. coli that affect expression of surface proteins, specifically OmpA as well as others, significantly affect the ability of these strains to actually invade the central nervous system in vivo.28 Thus the factors that appear to be important for the attachment and transport across brain endothelial cells in vitro affect the in vivo outcome of infection. On entering into the subarachnoid space, bacterial replication proceeds virtually unchecked by host defense mechanisms. By virtue of the blood-brain barrier, both immunoglobulin and complement levels are far lower in CSF than in serum and interstitial fluid. In addition, leukocyte proteases derived from an initial influx of leukocytes have actually been shown to degrade complement components in CSF from patients with meningitis.30 The major host response to the invasion of the subarachnoid space by pathogenic microorganisms is a rapid influx of polymorphonuclear leukocytes. The influx of neutrophils can be produced experimentally by the intracisternal injection of either encapsulated or nonencapsulated S. pneumoniae, heat-killed unencapsulated S. pneumoniae, and even pneumococcal cell walls.31 Purified lipopolysaccharide (LPS) from gram-negative bacteria is known to be extremely potent in the development of inflammation, and intracisternal injection of purified LPS from H. influenzae also elicits a strong inflammatory response.32,33 The mechanism by which LPS and presumably other bacterial cell wall components act to stimulate inflammation is probably through the induction of inflammatory cytokines such as interleukin 1 (IL-1) or tumor necrosis factor (TNF).34,35 In vitro studies with LPS, and with IL-1 and TNF show that incubation with endothelial cell monolayers leads to a rapid transient increased expression of the intercellular adhesion molecules (ICAM-1 and ICAM-2) as well as the selectin molecules such as ELAM-1. As a result, neutrophils are able to bind to central nervous system vascular endothelial cells at vastly increased rates and then subsequently migrate by diapedesis into the subarachnoid space. The pathophysiologic consequences of this intense neutrophil response in the subarachnoid space accounts for most, if not all, of the serious clinical and pathologic consequences of meningitis, such as a decrease in the blood-brain barrier and increased intracranial pressure (ICP). The increase in ICP occurs through several mechanisms. Vasogenic cerebral edema is caused by the increased permeability of the blood-brain barrier, which is a direct result of inflammatory bacterial products or the inflammatory cytokines released in response to these materials. The alterations in brain cellular membranes lead to cytotoxic cerebral
W9418-12.qxd 11/3/03 2:58 PM Page 341
Chapter 12
edema, that is, increased intracellular water content, potassium leakage, and a shift in brain metabolism to anaerobic glycolysis with increased lactate production. Also as a result of the inflammation in the subarachnoid space (SAS), there is decreased ability to reabsorb CSF, which leads to interstitial edema in brain parenchyma. All three of these mechanisms contribute to the risk of increased intracranial pressure and brain herniation. Clinical Manifestations The typical clinical presentation of meningitis in adults consists of fever, headache, and stiff neck, and varying degrees of altered consciousness. The majority of bacterial meningitis in adults is caused by S. pneumoniae, and the nature of the presentation of meningitis may depend on the underlying initial infection by the pneumococcus. For example, in pneumococcal pneumonia and sepsis, the onset of fever and hypotension may present with an acute shaking chill, which may evolve into a picture of meningitis in some patients. Alternatively, the disease may present with a more gradual onset consisting of symptoms of upper respiratory infection with or without symptoms of bronchitis and pneumonia for several days before the onset of meningitic symptoms. In the classic series by Carpenter and Petersdorf 36 approximately 27% of the patients had a sudden onset of headache, confusion, lethargy, and alteration of consciousness in the 24 hours before hospitalization. In contrast, 53% presented with a more slowly progressive course over 1 to 7 days. In a review of 493 episodes, Durand and associates37 found that 95% of the patients with bacterial meningitis had fever greater than 37.7°C on admission; neck stiffness was present in 88% of patients. Only 22% were alert while 51% were confused or lethargic and 22% were responsive only to pain. Within the first 24 to 48 hours of onset, 29% had focal seizures and/or focal neurologic findings. The most common predisposing factors were pneumonia, sinusitis, otitis media, alcoholism, diabetes, and some form of immunosuppression. A variety of other predisposing conditions have been identified such as malignancy, sickle cell disease, organ transplantation, splenectomy, dialysis, and steroid or other immunosuppressive therapy. In many of these settings, the presentation of meningitis may not be classic because of alteration of the immune response and the diagnosis only made upon investigation of altered sensorium, persistent headache, or newonset seizures. The presentation of meningococcal meningitis as such is quite similar to that of pneumococcal meningitis, and clinically one cannot distinguish between the two. However, meningococcal meningitis is part of the spectrum of meningococcal sepsis and the manifestations of meningococcal septicemia may precede the meningitis by 12 to 24 hours. Depending on their severity, signs of sepsis may dominate the clinical presentation. The initial presentation in meningococcemia may be completely nonspecific, with the patient simply complaining of not feeling well, but not
Central Nervous System Infections
341
having any clinical signs or symptoms of meningitis. Such patients may progress to irreversible shock and die before the development of meningitis. Early in meningococcemia a subtle petechial rash may develop that precedes the development into fulminant disseminated intravascular coagulation (DIC) and development of a massive purpuric rash. Indeed the purpura may be so severe that in some cases actual necrosis of the digits of the fingers and toes results. Purpura fulminans in the patient with meningococcemia is classically associated with hemorrhagic necrosis of the adrenal gland, commonly called Waterhouse-Friderichsen syndrome. Thus, the clinical manifestation of meningitis in patients with meningococcemia depends on the balance of factors between sepsis and shock and those of meningitis. It is important to recognize that in the elderly the presentation of meningitis may be subtler than that in young adults and children. For example, in a review of 54 cases, Gorse and colleagues found that confusion was a predominating symptom in presentation of disease in the elderly and was statistically more common than in that of the younger age group. In addition, pneumonia was much more often likely to be present in the older age group as well.38 There may also be alteration of physical signs in the elderly because there is very commonly cervical rigidity due to osteoarthritis and cervical spondylolysis in this age group and true nuchal rigidity has to be distinguished by careful physical examination. In addition, there may be hypertonicity of the neck muscles for diseaserelated reasons such as Parkinsonism or nonspecific conditions. The meningitis itself may progress more rapidly in the elderly, and elderly patients are more likely to present in coma than are younger patients. With the development of coma, nuchal rigidity may be markedly less pronounced. In children, the presentation of meningitis is fundamentally similar to that in young and middle-aged adults. There is a tendency for fever to be higher in children, and nonspecific symptoms such as irritability, nausea and vomiting, respiratory symptoms and photophobia are probably more common in children as well. In addition to nuchal rigidity, the classic physical signs of meningeal inflammation are Kernig’s and Brudzinski’s signs.39,40 Although Brudzinski originally described several signs of inflammation of the meninges, the best known of these is the “nape of the neck” sign which is now known as Brudzinski’s sign. This sign is elicited when flexion of the neck results in flexion of the hips and knees. Kernig’s sign is elicited with the patient in the supine position and the thigh flexed on the abdomen with the knee flexed. Upon passive extension of the leg in the presence of meningeal irritation the patient resists extension. Kernig’s and Brudzinski’s signs can only be elicited in approximately 50% of children with acute bacterial meningitis. Diagnosis Bacterial meningitis has to be differentiated from aseptic meningitis, encephalitis, brain abscess, subdural empyema,
W9418-12.qxd 11/3/03 2:58 PM Page 342
342
Section I
Introduction
and noninfectious conditions affecting the CNS. The differentiation from encephalitis can be difficult and initially is made on clinical grounds. In meningitis, meningeal signs such as stiff neck, photophobia, and so on tend to be more prominent, whereas in encephalitis the altered state of consciousness, confusion, and degree of obtundation are the predominant symptoms. As the level of consciousness declines, in both meningitis and encephalitis, differentiation between the two may only be possible with laboratory and radiologic findings. Because acute bacterial meningitis is a medical emergency, treatment should be implemented on clinical grounds without waiting for proof by laboratory or radiographic studies. In the case of brain abscess, the presentation generally is one of intracranial mass lesion with focal neurologic deficits, headache and a subacute onset. Fever may be present only in up to 50% of cases. In general, nuchal rigidity is not prominent in patients with brain abscess. Patients with subdural empyema similarly seem to present with a more localized headache and focal neurologic symptoms as well as an altered level of consciousness. Depending on the location of the lesion, fever and stiff neck may be present. However, the frequent occurrence of focal neurologic symptoms should suggest the possibility of a mass lesion. Fever and altered mental status with or without meningismus may occur in a variety of systemic infections as well as noninfectious conditions. For example, Rocky Mountain spotted fever can present with fever, shock, and a petechial rash, which must be differentiated from early meningococcemia. Meningococcal disease may initially present simply as meningococcemia with shock and skin rash; meningeal signs may not be prominent. Likewise, staphylococcal sepsis typically presents with high fever, with or without localizing signs, which may include encephalopathy. Noninfectious conditions, such as subarachnoid hemorrhage, can present precipitously with severe headache, loss of consciousness, and even fever and nuchal rigidity. The neuroleptic malignant syndrome, usually related to drugs, typically presents with very high fever, generalized rigidity, fluctuating levels of consciousness, and autonomic instability with blood pressures ranging from hypertensive to hypotensive levels along with arrhythmias and diaphoresis. Laboratory abnormalities in the neuroleptic malignant syndrome include increases of liver function enzyme concentrations and a striking increase of creatine kinase to levels exceeding 10,000 IU/L. Cerebrospinal Fluid Analysis Laboratory confirmation of the diagnosis of meningitis can almost always be made by the analysis of the spinal fluid. However, the decision to perform a lumbar puncture emergently upon presentation of the patient is not straightforward. Because of the presence of increased ICP, there is a significant increased risk of uncal or tonsillar herniation, which can lead to serious neurologic consequences and/or death.
For example, Renick and co-workers carried out a study of 445 children with meningitis. There were 19 episodes of herniation, which occurred in 17 children who had lumbar puncture. Twelve of these episodes occurred in the first 12 hours after the procedure and the remainder over the subsequent 12-hour period. Computed tomography (CT) scans were normal in five of the 14 episodes at the time of the procedure. Most of the patients with impending herniation are clinically identifiable based on coma, marked obtundation (Glasgow Coma Score 80 %) Usual range: 500–3000 typically >90% PMNs Can be normal in meningococcemia 2000 g Penicillin allergic: trimethoprim/sulfamethoxazole 15–20 mg/kg/d trimethoprim, 75–100 mg/kg/d sulfamethoxazole in 3 to 4 doses Ampicillin 2 g IV q 4 h plus ceftriaxone 2 g IV q 12 h or cefotaxime 2 g IV q 6 h, plus gentamicin 2 mg IV loading dose, then 1.7 mg/kg q 8 h Plus dexamethasone 0.4 mg/kg IV q 12 h ¥ 2 d For penicillin allergy: trimethoprim/sulfamethoxazole
W9418-12.qxd 11/3/03 2:58 PM Page 348
348
Section I
Introduction
Barbiturates decrease the CNS metabolic demand for oxygen and thus decrease cerebral blood flow with a resulting decrease in ICP. Phenobarbital is given at an initial dose of 5 to 10 mg/kg, at a rate of 1 mg/kg per minute followed by 1 to 3 mg/kg per hour.48 Such treatment requires an ICP monitoring device and/or an electroencephalogram (EEG) to monitor cerebral electrical activity. Phenobarbital is given until the ICP is reduced below 20 mm Hg or until approximately 90% burst suppression on the EEG (nine of the ten screens of the EEG are flat) has been achieved. It is recommended that serum phenobarbital concentrations be kept within the range of 20 to 40 mg/mL, although obtaining levels in “real-time” may be impossible. The clinical examination and monitoring may have to suffice. Pentobarbital is preferred because of its relatively short half-life of 24 hours compared with significantly longer half-life agents such as phenobarbital. It is important to recognize that there may be significant cardiac depression with arrhythmias and even hypotension from the use of high dose barbiturate treatment.
tis and septicemia) among college freshmen living in dormitories is 4.6 cases/105 population, compared with rates of 0.6 to 0.7/105 among college students as a whole, and 1.5/105 among noncollege students, aged 18 to 23.64 Similar figures were found in the United Kingdom. The vaccine contains immunogenic polysaccharide capsular material from serogroups A, C, Y, and W-135. The capsular material from serogroup B, which accounts for approximately 30% of cases in the United States, is unfortunately nonimmunogenic. The vaccine has not been tested in controlled clinical trials but produces protective antibody levels in over 90% of young adults, and has been used by the military since 1971.65 The vaccine has few side effects and is believed to be protective for at least 3 to 5 years. Vaccination for S. pneumoniae is recommended for immunocompromised and splenectomized patients to prevent fulminant sepsis and pneumonia, but is not thought to be cost effective for the general public.
Viral Meningitis and Encephalitis Seizures Seizures occur in approximately 30% to 40% of both children and adults with acute bacterial meningitis within the first few days of illness.62 If not treated, these seizures may progress to status epilepticus, which in turn can result in anoxic damage to the temporal lobe, cerebellum, and thalamus.61,63 The main principles of therapy are to control seizure activity quickly and definitively. Initially, short-acting anticonvulsants such as lorazepam or diazepam are given, followed by the long-acting agent, phenytoin. Lorazepam is given IV in doses of 1 to 4 mg in adults, and 0.05 mg/kg in children. Phenytoin is given IV at a dose of 18 to 20 mg/kg and at a rate of no more than 50 mg/min. The rate should be decreased if signs of toxicity such as hypotension or a prolonged QT interval develop. If phenytoin is not successful in controlling seizure activity, intubation and treatment with IV phenobarbital may be necessary. Patients must be watched carefully for signs of toxicity such as hypotension and respiratory depression. Phenobarbital should be given IV at a rate of 100 mg/min until seizure activity stops, up to an initial dose of 20 mg/kg. In children, the rate should be decreased to 30 mg/min. If these measures fail to control seizures, general anesthesia and additional phenobarbital may be necessary. Vaccination for Meningitis Following the widespread use of the conjugated H. influenzae capsular vaccines in the late 1980s and early 1990s, the incidence of Haemophilus meningitis has dramatically declined.10 Recently, the U.S. Public Health Service Advisory Committee on Immunization Practices has recommended that college students be targeted for meningococcal vaccination. The reasoning behind this approach is the observation that the incidence of meningococcal disease (both meningi-
Viral infections of the CNS occur as part of the spectrum of systemic viral infections. Viruses are obligate intracellular parasites that can only replicate within a cell. By definition they contain only one nucleic acid, either DNA or ribonucleic acid (RNA), which is surrounded by a protein coat called the capsid. This nucleocapsid is in turn surrounded by surface proteins called capsomers. In addition, some viruses, such as the herpes viruses, retroviruses, and most of the respiratory viruses, have lipid envelopes. Once inside the cell, viruses may replicate their nucleic acids within the cell’s nucleus or cytoplasm. Following replication of the nucleic acid, viral structural proteins are produced that either self-assemble or are actively assembled, generally at the cell membrane, where the completed viral particles are released. For every viral particle that infects a cell, several thousand viral particles are produced in a productive infection. The outcome of viral infection of a given cell may be lytic, as just described, or may lead to latent infection. In the latter case, the virus does not kill the host cell; rather, the viral DNA may be carried as an episome, as in the case of herpes viruses, or a DNA copy of its RNA genome may be integrated into the host chromosome, as in the case of the retroviruses. Virus infections of the CNS may be classified as exogenous due to infection with a viral agent acquired outside the host. Alternatively, CNS infections may be produced by reactivation of viruses that remain latent in the host. The majority of viral CNS infections are due to exogenously acquired enteroviruses, arborviruses, and less commonly, herpes viruses, respiratory viruses or lymphocytic choriomeningitis virus. Herpes simplex encephalitis is unique in that it may occur as part of the primary infection or be seen in patients in whom the infection has been latent for many years. CNS infections due to the other herpes viruses, such as Epstein-
W9418-12.qxd 11/3/03 2:58 PM Page 349
Chapter 12
Barr virus, varicella, or cytomegalovirus may occasionally be seen as part of the primary infection and may also occur as reactivated infections in patients also infected with human immunodeficiency virus (HIV). Epidemiology Aseptic meningitis and meningoencephalitis are the most common viral CNS infections encountered in the United States. The overwhelming majority of these cases are caused by enteroviruses, which produce disease in outbreaks occurring during the summer months, generally July and August, but occurring from May to October in warmer parts of the United States. While virtually all of the different serotypes of echovirus and coxsackievirus can produce meningitis and meningoencephalitis, in addition to other syndromes, the most prevalent serotype in the United States during the years 1997 and 1999 reported by the Centers for Disease Control (CDC) was echovirus 30, which accounted for 37.5% of all isolates, followed by echovirus 11 with 13.8%, and echovirus 9 with 8.7%. Of the more than 75 known enterovirus serotypes, the 15 most common of these account for 90% to 95% of all isolates each year. Thus, the pattern is one in which certain strains, such as echovirus 30, cause disease endemically while other strains occur in scattered outbreaks varying from year to year in different locations. Enteroviruses are transmitted from person to person by the fecal-oral route and their activity tends to be increased in areas of overcrowding, poverty, and poor hygienic conditions. Arboviruses account for the majority of epidemic cases of encephalitis. Their occurrence follows an identical seasonal distribution to that of aseptic meningitis and meningoencephalitis due to the enteroviruses. However, the mode of transmission is completely different. Arboviruses are spread by the bite of infected mosquitos, which are part of a complex cycle of enzootic transmission between birds, mosquitos and small mammals. Horses and human beings get infected incidentally to this natural cycle. From 1996 to 1997, a total of 252 cases of La Crosse encephalitis were reported to the CDC in patients ranging in age from 5 months to 70 years. Onset ranged between late June and early November. One hundred thirty-nine of these cases were reported from West Virginia and only one fatality was recorded.66 In contrast, during the same period there were only 15 cases of St. Louis encephalitis, and 19 cases of Eastern equine encephalitis five of which were fatal, illustrating the differences in frequency and severity between the different types of encephalitis viruses. The distribution of cases of equine encephalitis across the United States differs depending on the particular virus. For example, St. Louis encephalitis is found throughout the Midwest and South, as far north as New York and Michigan, with cases even reported on the West Coast. Eastern equine encephalitis, however, is essentially confined to the Southeast, and California encephalitis
Central Nervous System Infections
349
and La Crosse encephalitis tend to be seen primarily in the northern Midwest. The epidemiology of these diseases may be affected in part by prevention efforts from the public health authorities. For example, many states maintain surveillance systems, which include testing of mosquitoes for the presence of virus, as well as sentinel chicken flocks to determine arbovirus activity. Such efforts lead to early recognition of an outbreak and warnings by public health authorities for the population to take precautions such as insect repellants, wearing long sleeve shirts, and avoiding outdoor activity in the early evening hours when transmission is most likely to occur. In addition, mosquito control activities may be undertaken and diminish the infection as well. In August of 1999, an outbreak of encephalitis was detected in New York City, focused on the borough of Queens, where 62 patients were confirmed infected, seven of whom died with infection due to an agent identified as West Nile virus. This followed a massive die off among birds, particularly crows, that had been noticed in the month before the outbreak. Most of those affected with serious illness were elderly, although one patient was 29 years old.67,68 Initially, there was some question as to whether this virus would be able to survive winter conditions in the Northeast. However, surveillance data from the CDC found West Nile virus in 237 mosquito pools from 15 counties in the states of New York, New Jersey, Connecticut, and Massachusetts in the summer of 2000.69 Fourteen thousand seventy-one West Nile virus–infected bird carcasses from 79 counties in a total of six states were also noted. At least 12 persons were hospitalized with confirmed West Nile virus infection in the summer of 2000, most of them in New York City. In 2001, 42 human cases were reported to the CDC in Florida (ten cases), New York (ten cases), Connecticut (six cases), Maryland (six cases), New Jersey (six cases), Pennsylvania (three cases), and Georgia (one case). Overall, the median age was 70.5 with a range of 36 to 90 years, and two persons (4.8%) died. At least 26 states, mostly east of the Mississippi River, reported infections in birds and horses.70 A total of 4156 cases of West Nile virus infection, with 284 (approx. 7%) fatalities, were reported in 2002 from almost all states east of the Rocky Mountains. Thus, it seems likely that this virus will become endemic in the United States. It is interesting that this virus has been well known in Africa and the Middle East for many years. However, it was recently recognized as the cause of a large urban outbreak in Bucharest, Romania, and thus its epidemiology may be changing on a worldwide basis for unknown reasons. Although rabies is rare among humans in the United States, potential exposures to rabid animals lead to between 16,000 and 39,000 persons receiving post–rabies exposure prophylaxis each year. Since the 1950s, the incidence of rabies in domestic animals has declined dramatically because of immunization of dogs and other domestic animals. However, there has been a dramatic increase in the level of
W9418-12.qxd 11/3/03 2:58 PM Page 350
350
Section I
Introduction
endemic rabies in wild animals such as raccoons, skunks and bats, particularly on the East Coast.71 Pathogenesis Viral infection of the central nervous system occurs via two distinct routes: hematogenous and neuronal. In the case of the enteroviruses and arboviruses, CNS infection occurs as part of a systemic infection with the virus carried to the central nervous system via the bloodstream. In the case of herpes simplex and rabies encephalitis, the virus is carried to the CNS via nerve cells themselves. Because viruses must replicate intracellularly, the ability to cause disease is largely determined by whether viral surface proteins can attach to specific receptors on specific cells in affected tissues. One of the most well-documented examples of this phenomenon comes from animal model studies of reovirus type 1. In neonatal mice inoculated orally, type 1 reovirus grows to very high titers in intestinal tissue, whereas type 3 reovirus replicates minimally there. Genetic studies showed that ability to replicate was determined by the M2 gene segment, which codes for an outer capsid polypeptide of the virus. However, if reovirus type 3 is injected directed intravenously, it causes fatal encephalitis whereas reovirus virus type 1 does not. When an S-1 gene from type 3 reovirus is recombined with type 1 reovirus, neural entry is conferred and the ability to infect the central nervous system occurs. Another example of viral tropism being determined by the combination of viral surface proteins and specific tissue receptors is that of the binding of the HIV GP120 to the CD4+ receptor on T4 lymphocytes. The distribution of HIV also includes macrophages, dendritic cells and certain neural cells, all of which express the CD4+ receptor. Cells that do not express this receptor generally do not become infected with HIV.72 So important are these surface binding sites for their respective cellular receptors that several viruses such as rhinovirus, influenza virus, and poliovirus have evolved sophisticated molecular mechanisms to protect these sites from the host immune response. For example, receptors embedded in a molecular “canyon” may be so small that antibodies cannot bind to it, or the sequence of the attachment site is highly conserved while surrounded topologically by a hypervariable region so that the host’s humoral immune response cannot keep up with the number of variations.73,74 Like poliovirus, enteroviruses spread in the population through fecal-oral transmission. These viruses survive stomach acid, replicate in the intestine, and an initial viremia leads to infection of multiple organs within the body. A secondary viremia from these sources can infect the CNS. The prompt production of antibody disrupts this second viremia and prevents invasion of the CNS. In the case of the arboviruses the natural host defense mechanisms of the skin and mucous membrane are bypassed by the direct injection of the skin through the infected mosquito. Once again, local
replication is followed by viremia and infection of the brain is probably determined by viral tropism and the rapidity of the host immune response. In the case of viremic infections of the CNS, invasion of the brain involves attachment to the endothelial cells, presumably via specific receptors. Following invasion, an acute inflammatory reaction is generally seen with a perivascular distribution within the brain parenchyma and varying degrees of infection in the meninges depending on the particular agent involved. The perivascular inflammatory response is predominantly mononuclear although polymorphonuclear leukocytes may be seen. Infection of neural cells results in degenerative changes and phagocytosis by tissue macrophages or microglial cells. Some pathologic features are unique to certain viruses such as the production of multinucleated giant cells in the case of HIV infection of the brain, and the characteristic inclusions seen in herpes simplex infections, namely, the Cowdry type A intranuclear inclusion body, and the characteristic Negri bodies in the case of rabies.75,76 Some viral infections, most notably herpes simplex and rabies, spread to the CNS via a neuronal route. In the case of herpes simplex virus, the distribution involves the medial part of the temporal lobe bilaterally with one temporal lobe generally much more involved than the other. Autopsy studies on patients who died during active HSV encephalitis show the presence of virus in the olfactory bulbs, olfactory tracts, and the tracts of the limbic system that end in the hippocampus, amygdala, insula, cingulate gyrus, and olfactory cortex.77 Thus, the virus appears to gain access to the CNS from the nasal mucosa to the olfactory bulbs and olfactory tracts, although the mechanism by which the virus does this is unknown. About two thirds of cases of herpes simplex encephalitis in adults and older children occur in patients who have antibody to the virus at the time the infection begins. Many of these patients have a history of cold sores dating back 20 to 30 years. In the other one third of patients, antibody to herpes simplex is lacking at the time of onset of symptoms, indicating that the encephalitis is part of the primary infection. Approximately 90% to 95% of the cases of herpes simplex encephalitis in older children and adults are due to herpes simplex type I, with the remaining 5% to 10% due to herpes simplex type II. Neonatal herpes appears to be different, in that 90% to 95% of these cases are due to herpes simplex type II acquired from maternal or other sources at the time of birth. Infection of the CNS in neonates is part of a systemic viremic spread and there is no temporal lobe localization. Neuronal spread also accounts for the invasion of the central nervous system by rabies. Rabies infection may result from contact with saliva or other secretions from infected animals as well as the animal bite itself. Rabies replicates initially at the local site of inoculation and for this reason emergency preventive measures such as thorough cleansing of the wound and infiltration with human rabies immunoglobulin can be effective in preventing infection with this agent. In
W9418-12.qxd 11/3/03 2:58 PM Page 351
Chapter 12
the process of local replication, the virus gets into the nerve sheaths and is transported via infection of the nerve cells to the central nervous system. The rapidity of this process in reaching the CNS is a function of the distance of the nerve endings from the central nervous system. Thus bites on the lower extremities may take months to produce symptoms in the CNS, whereas bites in the face may reach the CNS in weeks. It is also important to recognize that the initial incident may be forgotten because of the time of the onset of the CNS symptoms and also because the inoculation may be inapparent as has been reported for bats.78 Therefore, one must always have a high index of suspicion for rabies for any unknown encephalitis, especially in patients who exhibit signs of hyperirritability. Clinical Manifestations The clinical presentation of viral meningitis includes fever, stiff neck, photophobia, and varying degrees of nonspecific symptoms such as malaise, myalgias, nausea, vomiting, abdominal pain, or diarrhea. The presence of impairment of consciousness, such as obtundation, disorientation, seizures, or localized neurologic signs or symptoms, should suggest brain parenchymal involvement and a diagnosis of encephalitis or meningoencephalitis. Stiff neck, while prominent, is generally less intense than that in bacterial meningitis. When the patient presents with meningitis, the most important consideration is to rule out bacterial meningitis as discussed previously. Because patients with viral meningitis tend to be less clinically ill, there is generally less need to obtain radiologic studies before performing lumbar puncture. In viral meningitis, the CSF typically shows an elevated white blood cell count, which may be predominantly polymorphonuclear leukocytes within the first 24 to 48 hours, although this number rarely exceeds 80% of total CSF white cells. The protein level is generally mildly elevated, and the glucose concentration is normal with the occasional exception of approximately 10% to 20% of patients with mumps and much less often in patients with enterovirus or herpes simplex virus. When the initial spinal fluid shows over 50% polymorphonuclear leukocytes, it is not uncommon for clinicians to repeat the lumbar puncture over the next 12 to 24 hours to determine whether there is a shift toward a lymphocytic predominance as one would expect in viral meningitis. In viral meningitis, the CSF Gram stain result will be negative and routine bacterial cultures will show no growth. For this reason, the term viral meningitis is often used interchangeably with the term aseptic meningitis. It is important to recognize that a wide range of nonviral illnesses can present similarly to meningitis and that, in addition to the enteroviruses, certain other viral agents can present with meningitis as well (Table 12-4). A parameningeal infectious focus will characteristically be associated with a CSF pleocytosis, somewhat elevated protein concentration, and a normal glucose concentration. Thus, a
Central Nervous System Infections
351
patient with an epidural or brain abscess could present with mild headache, fever, and a CSF picture identical to that of viral meningitis if radiologic studies were not done. Some cases of severe sinusitis, such as sphenoid sinusitis or frontal sinusitis may be associated with CSF pleocytosis. Patients with fungal and tuberculous meningitis will also present with headache, fever, and stiff neck, but in general the clinical course is much longer than in acute viral meningitis, usually at least 1 to 3 weeks before the patient seeks medical attention. Spinal fluid in these cases should show a low glucose, almost always less than 40 mg/dL. It is important to recognize that cryptococcal meningitis may present with completely normal spinal fluid analysis. Fever, headache, and nonspecific CSF findings can also be seen in such bacterial infections as syphilis, ehrlichiosis, and noninfectious conditions such as sarcoidosis, Behçet’s disease, and uveoparotitis. Viruses other than enteroviruses can also produce aseptic meningitis. The classic example is that of herpes simplex type II, which produces typical aseptic meningitis with low-grade fever, headache, stiff neck, and photophobia as part of primary genital herpes infection. It is, therefore, very important to question any potentially sexually active patient about the possibility of genital herpetic lesions, and in women perform a pelvic examination if indicated. Aseptic meningitis can also be seen as part of the syndrome of primary HIV infection. Certain strains of leptospirosis will typically present with aseptic meningitis. However, most cases present in conjunction with systemic disease and severe involvement of other organs such as lung, liver, and kidney. Lymphocytic choriomeningitis virus does occur with some frequency, particularly in patients with exposure to rodents such as house mice and pet hamsters. The laboratory diagnosis of viral meningitis is generally one of exclusion, as described previously. Viral cultures that grow enteroviruses from the spinal fluid are diagnostic. However, these results are positive in at most 30% to 50% of cases. Polymerase chain reaction (PCR) has become available for the diagnosis of enteroviral meningitis and this technique correlates very well with the results of viral culture. Unfortunately, this test is only available through reference laboratories or highly specialized research laboratories, and is costly. In the case of herpes simplex meningitis, the diagnosis is confirmed if herpes simplex is cultured from the spinal fluid or detected by PCR. In the case of aseptic meningitis associated with systemic infections such as leptospirosis, syphilis, or Ehrlichia, standard serologic tests generally available at state public health laboratories or reference laboratories would be diagnostic. Finally, certain drugs such as sulfa and nonsteroidal anti-inflammatory agents can produce acute syndromes of aseptic meningitis. Other agents that have been associated with aseptic meningitis include intravenous immunoglobulin, certain vaccines, and the intrathecal administration of drugs.79
W9418-12.qxd 11/3/03 2:58 PM Page 352
352
Section I
Introduction
Table 12-4 Laboratory Diagnosis of Selected Viral Diseases Virus Herpes viruses Herpes simplex
CNS Disease
Serologic Results
Viral Culture Results
Direct Antigen and PCR Results
Temporal lobe encephalitis
Almost always negative in CSF, throat, etc
Varicella
Encephalitis in HIV patients
About two thirds seropositive on admission, not helpful in diagnosis Not helpful diagnostically
Cytomegalovirus
Encephalitis in HIV patients
Almost always positive, not helpful diagnostically
Urine, throat, blood may be positive—consistent with, but not proof of, encephalitis
Epstein-Barr virus
Rare encephalitis in mononucleosis
Not available
Human herpes virus 6 and 7 (HHV 6 and 7)
Seizures, encephalitis in 1–3 year olds
MonoSpot test good presumptive test, may be negative in up to 20% in 1st wk VCA-IgG and IgM, positive IgM virtually diagnostic Reference laboratories only
CSF–PCR diagnostic FA can be done on brain biopsy No antigen test, PCR available (CSF) FA can be done on brain biopsy Antigenemia test on blood, positive test consistent with, but not proof of encephalitis Not available
Reference laboratories only
Not available
Respiratory viruses Influenza A and B Parainfluenza 1–3 Adenovirus Respiratory Syncytial virus (RSV)
Usually negative in CSF, throat, etc
Parainfluenza occasionally, others rarely cause encephalitis
Not useful
Nasopharyngeal swabs Throat washings, Cultures via bronchoscopy—excellent sensitivity, diagnostic if positive
Direct antigen ELISA available for RSV (excellent sensitivity), and Influenza (moderately sensitivity)
Summertime outbreaks of meningitis, meningoencephalitis
Not useful, too many serotypes, too much cross-reactivity
Send stool, throat, CSF If throat or CSF culture positive, diagnostically definitive If stool culture positive, presumptive (enteroviruses may be shed in stool for weeks)
None for direct antigen, CSF–PCR diagnostic
Summertime outbreaks of encephalitis
Diagnostic, if positive Send serum and CSF—done in reference laboratories and state public health laboratories
Generally not available
Direct antigen not available but PCR for West Nile virus diagnostic
HIV
Encephalopathy
ELISA, confirm with Western Blot
Research laboratories only
Use PCR in CSF or serum
Lymphocytic chorimeningitis virus
Meningitis
Reference laboratory
Not available
Not available
JC virus
Progressive Multifocal leukoencephalopathy, mostly in HIV, other immunocompromised patients
Reference laboratory
Not available
Direct antigen not available; PCR available in reference laboratory
Rabies
Encephalitis
Reference or state public health laboratories, antibody is generally undetectable before day 6, 50% by day 8, and 100% by day 15
Research laboratories only
Direct FA staining of hair follicles in skin biopsy from nape of neck above the hairline—50% positive in 1st wk, higher later
Enterovirus Coxsackie Echovirus
Arboviruses St. Louis encephalitis California encephalitis Western equine Eastern equine La Crosse West Nile virus
CSF, cerebrospinal fluid; ELISA, enzyme-linked immunosorbent assay; FA, fluorescent antibody; HIV, human immunodeficiency virus; PCR, polymerase chain reaction.
W9418-12.qxd 11/3/03 2:58 PM Page 353
Chapter 12
Serologic studies for the diagnosis of enterovirus infections are not recommended. Because there are over 75 different enterovirus serotypes, testing is only possible for a subset of these. In addition there is tremendous overlap in the serologic response between the different serotypes such that a diagnostic seroconversion to more than one enterovirus serotype may be found. Moreover, because there is no effective treatment, there is no need for expensive laboratory testing that cannot affect patient outcome. Another common diagnostic misconception is the usefulness of CSF antibodies. With the exception of the Venereal Diseases Research Laboratories (VDRL), which indicate active CNS syphilis, CSF IgM for West Nile virus and the ratio of measles antibodies in the spinal fluid to those in serum in extremely rare cases of subacute sclerosing panencephalitis, there is no useful diagnostic value in CSF antibody testing. In essence, all agents that are diagnosable can be diagnosed from serum studies rather than the use of spinal fluid. A list of common viral infections and preferred diagnostic methods is provided in Table 12-4. The hallmark of the presentation of viral encephalitis is the prominence of an altered level of consciousness with or without focal neurologic signs and symptoms in the setting of an acute febrile illness. The differential diagnosis is essentially the same as that for acute bacterial meningitis and acute viral meningitis as discussed previously. Varying degrees of nuchal rigidity can be present in patients with encephalitis due to the enteroviruses that in effect produce combined disease, for example, meningoencephalitis. Lumbar puncture reveals spinal fluid with a similar picture to that of viral meningitis. Radiologic studies, particularly MRI, are critical in arriving at a definitive or provisional diagnosis in cases of encephalitis. It is important to recognize that in some cases arbovirus encephalitis, particularly Eastern equine encephalitis, very commonly have focal lesions on MRI. However, their distribution is not consistent and differs from that of herpes simplex encephalitis (Figs. 12-3 and 12-4). Diagnosis of arbovirus encephalitis can almost always be made with serologic studies because antibody titers to all of the common arboviruses are generally present at the time the patient presents with the illness. Because there is only a very low background frequency of these antibodies in the general healthy population, a positive arbovirus serology can be accepted as clinically definitive. Arbovirus isolation is only possible in specialized research laboratories, as is PCR for West Nile virus. The diagnosis of herpes simplex encephalitis is of critical importance because of the ability of antiviral agents to treat this infection and improve the outcome. Unfortunately, almost nothing in the clinical presentation of patients with herpes encephalitis is of any value in diagnosing this condition. As part of an antiviral trial in the early 1980s, Whitley and colleagues obtained biopsy specimens from a total of 202 patients with signs and symptoms of encephalitis.80 Of these 202 patients, 113 had herpes simplex virus isolated
Central Nervous System Infections
353
Figure 12-3. Axial low convexity FLAIR image through the central nuclear structures demonstrates cytogenic edema within the basal ganglia and thalami bilaterally. In addition there is vasogenic edema in the subinsular white matter tracts, especially on the right, and in the posterior limb of the internal capsule on the right. Such findings would be consistent with either acute encephalitis versus acute disseminated postinfectious encephalomyelitis. In this instance, elevated titers for eastern equine encephalitis virus were observed. (Courtesy Kenneth H. Rand, MD, University of Florida.)
from brain tissue, only four of the remaining patients were thought to have herpes simplex encephalitis based on serology and other data. Subsequent studies confirmed these numbers using PCR. When the patients with positive brain biopsy results and those with negative brain biopsy results were compared statistically for the presence of findings such as alteration of consciousness, fever, headache, CSF pleocytosis, personality change, seizures, vomiting, hemiparesis, and memory loss, there were no differences between the two groups. In general, herpes encephalitis presents with a 3- to 5-day history of fever, and headache with or without systemic signs that progresses to obtundation and coma; the latter may appear abruptly with the onset of seizures. Herpes encephalitis can occur in any age group from childhood through old age, and occurs with equal frequency at all times of the year. The incidence is estimated at one in 250,000 to one in 500,000 people per year and is thought to account for approximately 10% to 20% of viral encephalitides in the United States. A diagnosis of herpes encephalitis is made highly likely from MRI findings that show bilateral temporal lobe involvement, which is generally asymmetrical (see Fig. 12-4). In untreated patients, this may progress to hemor-
W9418-12.qxd 11/3/03 2:58 PM Page 354
354
Section I
Introduction
Figure 12-4. Axial, low convexity, proton-density MR image demonstrating cytogenic edema (increased signal intensity represented as increased regional brightness) within the cortical and immediate subcortical portions of mainly the right temporal lobe. Similar, but less extensive, changes are seen in the lower portion of the left insula and in the right basifrontal cortex. Since this distribution of the edema does not conform to any specific vascular territory, an inflammatory process more likely. A frontotemporal distribution of mainly cortical edema is most consistent with herpes simplex encephalitis in an acute phase. (Courtesy of Ronald Quisling, MD, University of Florida.)
rhagic lesions with mass effect and herniation. PCR for herpes simplex virus in the spinal fluid is almost uniformly positive in patients with herpes encephalitis, despite the fact that one cannot grow the virus from spinal fluid or other peripheral sites in more than 1% to 4% of patients.81–83 Serologic diagnosis is not particularly helpful early on, although in general all patients with herpes encephalitis will show a significant increase in herpes simplex titer both in the spinal fluid and serum, and will show seroconversion in the case of those undergoing primary infection. Conditions that mimic herpes simplex encephalitis, and viral encephalitis in general, were well described by Whitley and associates84 and include brain abscess, subdural empyema, cerebritis due to Listeria, Mycoplasma, and infections with fungi, tuberculosis, cryptococcus, rickettsia, toxoplasmosis, mucor, and even routine bacterial meningitis caused by agents such as pneumococcus and meningococcus. Tumors, subdural hematomata, CNS lupus, and adrenal leukodystrophy may also mimic the signs and symptoms of encephalitis. It is important to recognize that strokes may present with fever as well, and toxic encephalopathy such as neuroleptic malignant syndrome or Reye’s syndrome can also mimic encephalitis.
Encephalitis can sometimes occur as part of systemic infection with common viruses that do not normally produce encephalitis. For example, Epstein-Barr virus can, on occasion, present with seizures and even coma that fortunately, in general, resolve with complete recovery from the disease. CNS involvement with toxoplasmosis, lymphoma, varicella, and cytomegalovirus in patients with HIV can mimic the presentation of the patients with encephalitis. Table 12-5 lists nonviral infections that may present as encephalitis. Although rare, it is important to be alert to the possibility that a patient with encephalitis, in fact, has rabies. Rabies generally presents with a prodrome of 1 to 4 days consisting of fever, headache, malaise, myalgias, fatigue, anorexia, nausea, vomiting, sore throat, and a cough. The patient may complain in particular of paresthesias or fasciculations at the site of the original animal bite due to viral replication at the site of inoculation and/or in the dorsal ganglia of the sensory nerve supplying that area. Between 50% and 80% of patients with rabies will have some manifestations of this sort. The disease quickly progresses to an encephalitic phase consisting of agitation, excitation, and excessive motor activity. Patients may experience hallucinations, become combative, and develop muscle spasms with opisthotonus. Seizures are not infrequent. Periods of hallucinations and aberrant men-
Table 12-5 Common Nonviral Causes of Encephalomyelitis Infectious Ehrlichia Rocky Mountain spotted fever Bacterial endocarditis Brain abscess/cerebritis Staphylococcus aureus sepsis Syphilis (primarily meningovascular) Lyme disease Leptospirosis Mycoplasma Listeria (rhomboid encephalitis) Typhus Legionella Cat-scratch disease Nocardia Tuberculosis Cryptococcus Histoplasmosis Coccidioidomycosis Amebae Malaria Trypanosomiasis Noninfectious Drugs Carcinoma Lymphoma Vasculitis Behçet’s disease
W9418-12.qxd 11/3/03 2:58 PM Page 355
Chapter 12
tation may alternate with lucid periods that get progressively shorter as the disease progresses. Hyperesthesias with excessive reactivity to normal stimulation of light, sound, and touch are very common; autonomic nervous system changes such as dilated pupils, increased salivation, lacrimation, perspiration, and postural hypotension occur. Ultimately brainstem function is affected with cranial nerve palsies, optic neuritis, and the characteristic hydrophobia due to the painful, violent involuntary contractions of the muscles of respiration, the pharynx, and the larynx initiated by attempts to swallow. Eventually the disease progresses to cardiorespiratory depression, coma, and death. Occasionally, rabies may present as an ascending paralysis clinically similar to the Guillain-Barré syndrome and in fact corneal transplants from two patients presumed to have died from GuillainBarré syndrome actually transmitted clinical rabies, resulting in the death of the recipients. The laboratory diagnosis of rabies requires viral isolation, positive serologic study results (assuming the patient has not been immunized) or demonstration of the characteristic Negri bodies in brain tissue. Viral antigen can also be demonstrated by immunofluorescent antibody in infected tissue including corneal scrapings, or skin biopsy or brain biopsy specimens. The yield from skin biopsy of the nape of the neck above the hairline is apparently significantly better than that of corneal scrapings.85 In patients who have traveled overseas, a large number of infectious diseases may present with encephalitis either as a primary entity or as part of a systemic disease, including agents such as Japanese B encephalitis, Murray Valley encephalitis, Omsk hemorrhagic fever, Kyasanur forest disease complex, Powassan virus, louping ill, Russian springsummer encephalitis, Rift Valley fever, yellow fever, dengue, chikungunya, Hantaan virus, Puumala virus, and the highly fatal hemorrhagic fevers, Marburg, Ebola, and Lassa. Following the bite of several species of monkeys, encephalitis may develop due to Monkey B virus, a herpes virus. This virus is related to human herpes simplex but humans have little native ability to contain it, in contrast to its natural host in whom it produces “cold sores.” It is transmitted to humans from saliva in monkeys and reaches the brain via nerves at the site of the monkey bite. Since 1932 about 40 cases have been described, with fatal outcome in approximately 70%. Treatment with ganciclovir or acyclovir may be useful. Consultations with the CDC or the Southwest Foundation for Biomedical Research in San Antonio TX (210-674-1410) is advised for the management of monkey bites or suspected cases due to Monkey B virus. Treatment No specific drug or serologic therapy is currently available for enterovirus or arbovirus infections. In general, viral meningitis due to enteroviruses is clinically mild and most patients can be treated without admission to the hospital
Central Nervous System Infections
355
unless there is a serious differential diagnostic question about bacterial meningitis. In patients with any suggestion of encephalitic symptoms or brain parenchymal involvement, assuming that appropriate radiologic studies have ruled out conditions such as brain abscess and subdural empyema, most physicians empirically treat the patient with acyclovir intravenously in doses appropriate for herpes encephalitis. Because of the rarity of complications from acyclovir it is difficult to argue with this practice and brain biopsy certainly is not justified to prove the presence of herpes encephalitis before treatment. Before the availability of antiviral agents for the treatment of herpes encephalitis, the disease was fatal in approximately 70% of patients, with an additional 20% to 25% surviving with severe disabilities. A National Institutes of Health– sponsored trial in the 1980s proved conclusively that acyclovir was superior to vidarabine, which had been shown to be superior to a placebo in earlier studies. Patients who received acyclovir had a 19% mortality at 6 months compared with a mortality of 55% for those who received the vidarabine. Furthermore, the outcome of surviving patients was significantly better for patients who received acyclovir, with 38% having only minor impairment or returning to baseline compared with only 13% among the vidarabine recipients. The most important host factors in determining the outcome of treatment are age and level of consciousness at the time that treatment is begun. Thus, those patients treated with acyclovir who had a Glasgow coma score greater than 10 had 100% survival, whereas those with a Glasgow coma score less than 6 had only a 50% survival; survival was significantly better for patients under the age of 30 than those above 30.84 The dosage of acyclovir is 10 mg/kg IV every 8 hours for 14 to 21 days.
Brain Abscess Brain abscesses have been recognized from the days of Hippocrates in 460 bc, but did not enter the realm of medical consciousness until William Macewen developed surgical procedures in the 1890s for the management of this entity. By definition, a brain abscess is a localized suppurative infection of the brain parenchyma.86 Brain abscess is fundamentally an uncommon disease. For example, during a 9-year period from 1990 to 1999, 57 intracranial complications developed among 2890 cases of chronic otitis media, an incidence of 1.97% in a highly selected susceptible population.87,88 The incidence in the general population has been estimated at 1.3 to 100,000 person years, with the rates slightly higher in children between 5 and 9 years of age and in adults older than 60 years of age. Most series document a male preponderance of between 2 : 1 and 3 : 1 and the age distribution is somewhat dependent on the associated underlying etiologies.89–91 Thus, some series have shown a bimodal age distribution with a peak in the pediatric age group and
W9418-12.qxd 11/3/03 2:58 PM Page 356
356
Section I
Introduction
after the age of 40.92,93 While the etiology and distribution of associated diseases has remained essentially unchanged over the years for pyogenic brain abscesses, the AIDS epidemic has led to the emergence of a large group of patients with brain abscess due to toxoplasmosis. Pathogenesis Brain abscess develops as a localized area of cerebritis initially consisting of bacteria in the brain parenchyma together with inflammation and edema. Over the next several days, this area of infection becomes more localized with the development of necrosis in the middle and a ring-enhancing capsule. Ultimately, host defenses lead to the development of a well-formed capsule. The most common predisposing conditions for the development of a brain abscess are infections in the middle ear, paranasal sinuses, mastoids, and teeth. It is believed that bacteria reach the brain through valveless emissary veins, which traverse the cranium into the venous drainage system of the brain. Alternatively, direct extension through an area of osteitis or osteomyelitis adjacent to the sinus or middle ear infection could provide access to the CNS. The other major mechanism for seeding the brain parenchyma is metastatic transmission from an extracranial focus of infection. Hematogenous brain abscesses tend to be multiple, to be located at the gray-to-white matter junction, and to follow a vascular distribution within the brain. Pyogenic lung abscess and bronchiectasis are frequently noted as underlying associated conditions.92,94,95 Hematogenous dissemination from a contiguous focus of infection has been described. Other distant foci that have been associated with brain abscess include wound infections, osteomyelitis, pelvic infection, cholecystitis and other intra-abdominal foci. Any procedure that results in a transient bacteremia can on occasion be associated with the subsequent development of a brain abscess. Despite its chronicity and high level of bacteremia endocarditis accounts for only 1% to 5% of cases of brain abscess.96 A significant number of brain abscesses are associated with penetrating trauma such as gunshot wounds, depressed skull fractures with retained bone fragments, cranial penetration from objects such as pencils, animal bites, or even as a complication of cervical traction associated with pin site infection. In approximately 25% of cases, no underlying etiology can be found. Microbiology The bacterial etiology of brain abscess is to a great extent dependent on the location of the abscess and the predisposing factors. Thus, aerobic, anaerobic, and microaerophilic streptococci are the most frequently isolated bacterial species. S. aureus makes up 10% to 25% of the isolates in most series. However, its occurrence is most likely secondary to trauma, infection following neurological surgery, and
endocarditis.96 In addition to streptococci, brain abscess associated with paranasal sinus or chronic otitis media infection may be caused by Haemophilus species, Bacteroides species, other anaerobes, and P. aeruginosa in the case of chronic otitis media. The bacteria found in intracranial abscesses from patients with hematogenous spread depend on the underlying source, for example, S. aureus and viridans streptococci associated with endocarditis. If the source of bacteremia is intraabdominal, Enterobacteriaceae, enterococci, and anaerobes may be found, while a urinary tract origin is likely to lead to infection with Pseudomonas and/or Enterobacteriaceae, but not anaerobes. Anaerobes, including actinomyces, may be associated with spread from a lung abscess. Although S. aureus is the most common organism complicating penetrating trauma, Clostridium species and Enterobacteriaceae must be considered. The nature of the trauma is important too, because if it occurs in water, associated organisms such as Pseudomonas and Aeromonas would also have to be considered. The organisms associated with postoperative infections include S. aureus, S. epidermidis, Enterobacteriaceae, and Pseudomonas. On occasion, unusual organisms such as nocardia can produce brain abscesses. In a series of 11 cases and review of 120 cases of nocardial brain abscess in the literature, concomitant pulmonary disease was present in 34%. Most of the brain abscesses were single but approximately one third were multiple and overall 38% of the cases occurred in patients who were immunocompromised by virtue of HIV or iatrogenic causes.97 Rarely, Mycobacterium tuberculosis may produce a space-occupying lesion (tuberculoma). While uncommon in the United States, tuberculoma is the most common cause of brain abscess in some developing countries. Yeast and fungal infections are quite rare as causes of brain abscess, but it is important to recognize that they do occur. Candida albicans almost never causes isolated brain abscesses, but may cause microabscesses in association with disseminated candidiasis. Cryptococcus usually produces meningitis but cryptococcomas are frequently seen if sensitive radiographic techniques are used. Agents of Phaeohyphomycosis, such as Cladosporium, Bipolaris, Curvularia, and Wangiella, as well as the agents of Chromoblastomycosis have all been reported to cause brain abscess. Aspergillosis is well known to cause brain abscess but is almost always limited to the immunocompromised population particularly the transplant patients. Zygomycoses such as Mucor rhizopus, and Rhizomucor produce brain infection by direct extension from the paranasal sinuses in poorly controlled diabetics. Even protozoa and other parasites may cause brain abscess. As mentioned previously, toxoplasmosis is probably the most frequent protozoal cause of brain abscess observed in the United States and is almost entirely associated with HIV infection. Strongyloides, Entamoeba, Echinococcus, Paragonimus, Trichinosis, Sparganosis, and Angiostrongylus have all been reported, particularly in developing nations.
W9418-12.qxd 11/3/03 2:58 PM Page 357
Chapter 12
Central Nervous System Infections
357
Rarely, brain abscess due to Naegleria and Acanthamoeba occurs in this country.
Clinical Manifestations Headache of varying degree is the most consistent symptom among patients with brain abscess. The headache is generally not well localized and may be mild and difficult to differentiate from ordinary headaches. In many series, fever is present in only 50% or less of cases, and focal neurologic signs and symptoms such as hemiparesis, aphasia, ataxia, and sensory deficits may only be present in one third to one half of cases. Papilledema as a reflection of increased ICP is present in a minority of cases.98,99 Likewise, seizures are observed in approximately 25% to 45% of patients by the time they present. The seizures are most often generalized and most commonly associated with frontal lobe lesions. To some extent, the presenting signs and symptoms are dependent on the location of the abscess. For example, cerebellar abscesses often present with nystagmus, ataxia, vomiting, and dysmetria.100 Frontal lobe abscesses generally present with headache, drowsiness, and deterioration of mental status together with hemiparesis and unilateral motor signs. Temporal lobe abscesses may present with or without asphasia or dysphasia, depending on whether the abscess is in the dominant hemisphere. Pituitary abscesses may simulate a tumor and can present with visual field defects and endocrine abnormalities. Brainstem abscesses typically exhibit facial weakness, fever, headache, hemiparesis, dysphasia, and vomiting. The differential diagnosis includes a wide range of other infections such as meningitis, subdural empyema, epidural abscess, and viral encephalitis; noninfectious causes include migraine, intracerebral and subarachnoid hemorrhage, venous sinus thrombosis, and malignancies.
Diagnosis The introduction of CT scanning in the 1970s revolutionized the diagnosis of mass lesions in the CNS. For brain abscess, CT scans are 95% to 99% sensitive, and provide information on the size, location, and stage of the abscess together with the extent of surrounding edema and the presence or absence of mass effect such as midline shift, hydrocephalus, and impending herniation. Characteristic findings are a ring-enhancing lesion in the contrast-enhanced CT scan or MR image with a hypodense center reflecting the necrotic center of the abscess surrounded by a variable zone of edema (Fig. 12-5). The major problem in radiographic diagnosis is the differentiation from tumors including neuroblastomas as well as metastatic lesions. In one study, eight of 26 patients with a brain abscess were initially diagnosed as having a tumor.101 Another study noted that in 18% of CT scans from 100 patients with confirmed brain abscess, the
Figure 12-5. Axial, mid-convexity, T1-weighted, post– gadolinium-enhanced, MR image demonstrates a cavitary lesion in the right frontal region. There is abnormal enhancement surrounding the margin of this centrally necrotic mass. The deepest, innermost portion of the enhancement appears thinner, typical of a brain abscess. There is extensive edema surrounding the abscess cavity producing a right-to-left subfalcine shift. In addition to the abnormal enhancement of the cavity, there is also enhancement of the adjacent pial surface of brain consistent with meningeal inflammation. These are features of a subacute organized brain abscess with central suppuration. (Courtesy of Ronald Quisling, MD, University of Florida.)
initial findings could not be radiographically distinguished from that of a malignancy.102 MRI provides soft tissue resolution and detail that is superior to that achieved with CT. In addition, there is no exposure to ionizing radiation; the cost, however, is substantially greater. On T1-weighted images, brain abscesses appear hypointense and show ring enhancement following administration of the contrast agent gadolinium. On T2-weighted sequences, the central area of necrosis appears hyperintense and is surrounded by a well-defined hypointense capsule and readily discernible surrounding edema. One of the advantages of MRIs is that they can detect the cerebritis stage before the formation of the abscess with a fully developed capsule and, thus, can diagnose pyogenic brain infection earlier and with greater accuracy than CT scan. Newer methods such as MR spectroscopy, which can detect products of bacterial metabolism such as lactate, acetate, or pyruvate, should improve our ability to differentiate brain abscess from malignancy as these methods become more widely available in clinical practice.103 Use of radionuclide brain scans with agents such as indium-
W9418-12.qxd 11/3/03 2:58 PM Page 358
358
Section I
Introduction
111–labeled leukocytes probably does not provide any advantage over conventional radiographic techniques. Thallium 201 single-photon emission computed tomography is a promising new technique, but its use in differentiating toxoplasma encephalitis from intracerebral lymphoma in patients with AIDS has to be regarded as preliminary at this time. Routine laboratory studies are not particularly useful in the diagnosis of brain abscess. Patients should have routine laboratory tests such as CBC, differential cell count, erythrocyte sedimentation rate, and standard chemistries performed. If the sedimentation rate is elevated, it may be useful to observe the rate to document a therapeutic response. Lumbar puncture in a patient with a space-occupying lesion is to some extent contra-indicated and probably should not be done unless there is a clear clinical suspicion of meningitis or meningeal carcinomatosis to justify the risk of herniation. The CSF in patients with brain abscesses generally shows findings similar to any other parameningeal focus of infection, in that the cell count and protein show mild elevations and the glucose is normal. Cultures are generally sterile, unless there is some anatomic connection between the abscess and the spinal fluid as may occur in cases in which the brain abscess is secondary to trauma or to a postoperative complication. While reasonable empiric treatment can be devised for most common brain abscesses, culture of the material and transport to the laboratory under strictly anaerobic conditions is essential for optimal identification of the causative agent(s). In addition, a biopsy can be obtained or material sent for pathologic examination to rule out malignancy as well as for stains for unusual microorganisms, if needed. The choice of antibiotics is to be determined both by the spectrum of microbiologic agents known to cause brain abscess and the degree to which individual antibiotics penetrate the blood-brain barrier and enter into the abscess cavity itself. Treatment Treatment for brain abscess represents expert consensus based on empiric treatment, rather than randomized controlled trials. Brain abscesses that develop in contiguity with frontal sinus infection may be assumed to contain mixed aerobic and anaerobic flora. In this situation, even if anaerobic bacteria are not recovered, treatment should be given with high-dose penicillin ranging from 10 to 20 million units per day together with metronidazole 7.5 mg/kg IV every 6 hours or 15 mg/kg IV every 12 hours. If there is any suspicion that the abscess may have arisen from a dental focus, anaerobic culture should be held for 7 to 14 days to detect the growth of Actinomycosis. However, Actinomycosis should respond to standard therapy with penicillin. Brain abscesses that are related to chronic otitis media and mastoiditis should be treated with a combination of antibiotics that will cover anaerobes as well as Enterobacteriaceae and
Pseudomonas. A combination of cefotaxime, ceftazidime, or ceftriaxone plus metronidazole would work well in this particular setting. Although culture specimens may not always grow anaerobes, particularly if they are fastidious, the absence of growth of Enterobacteriaceae or Pseudomonas from abscess material from a patient who has not received IV antibiotics can be relied on to exclude these particular organisms. Similarly, the absence of a culture showing S. aureus would also be very good evidence that this agent is not involved in the particular process. In general, S. aureus is much more likely to be a pathogen in the setting of endocarditis, metastatic infection to the CNS, and in the setting of trauma or postoperative infection. If a brain abscess is associated with a neurosurgical procedure, vancomycin should be included in the regimen to cover both methicillinresistant S. aureus and coagulase-negative staphylococci. The dosage of third-generation cephalosporins is 2 g IV every 4 hours for ceftazidime and 2 g IV every 12 hours for ceftriaxone. Patients with a Nocardia brain abscess should be treated with higher doses of trimethoprim-sulfamethoxazole, 15 mg/kg/day of the trimethoprim component in three to five divided doses until the disease is under control; thereafter, the dose can be lowered to one double-strength trimethoprim-sulfamethoxazole tablet orally twice daily for 3 to 6 months in nonimmunocompromised patients and up to 1 year in the immunocompromised. Patients with severe immunocompromise due to advanced AIDS probably need lifelong treatment for this disease.
Tuberculous Meningitis Tuberculosis has been known to humankind since antiquity, having been demonstrated recently by molecular methods in mummies from both the new and the old world dating to 1000 to 1500 years bc.104,105 Tuberculosis was recognized on clinical grounds in the 18th century and with the isolation of the organism by Robert Koch in 1882, its ability to produce CNS disease was quickly recognized. Epidemiology Tuberculosis is a truly worldwide disease, with between one and two billion people infected. It is estimated that approximately 8 to 10 million new cases occur each year and two to three million people die from tuberculosis worldwide every year. In the United States and Western Europe, the number of cases of tuberculosis has declined dramatically during the past 100 years. For example, in the early 1950s there were approximately 84,000 cases per year in the United States; this declined to 20,000 to 25,000 cases per year in the mid 1980s. This downward trend abruptly changed in 1984 and by the early 1990s there was actually an increase in the number of cases to approximately 27,000 new infections per year in the United States. Investigations by the CDC showed
W9418-12.qxd 11/3/03 2:58 PM Page 359
Chapter 12
that much of the increase could be explained by the occurrence of HIV infections leading to an increasing pool of highly susceptible individuals and a resultant increase in urban outbreaks.106 Fortunately, increased surveillance and treatment efforts nationwide resulted in a resumption of the downtrend in the number of tuberculosis cases. The development of effective antiretroviral therapy in the mid to late 1990s further hastened this decline; approximately 13,000 new cases were reported for the year 2000 in the United States. On a worldwide basis, however, the uncontrolled HIV epidemic, particularly in sub-Saharan Africa and parts of Asia, will unfortunately lead to an increasing disease burden from tuberculosis in those areas for the foreseeable future. Pathogenesis Transmission of tuberculosis occurs by airborne droplet nuclei, which directly reach the alveolar spaces. During replication in the aveoli and within alveolar macrophages, the tubercle bacilli are transported via lymphatics to the pleural surface of the lung and thence to the hilar and mediastinal lymph nodes. During the 3 to 4 weeks of this infection there is virtually no immune response and the tubercle bacilli readily disseminate via the bloodstream and seed all organs in the body including bone marrow, liver, spleen, kidney, meninges, genitourinary tract and brain parenchyma itself. The overwhelming majority of these primary infections occurs in children and are asymptomatic, or nearly so. With healing, the only residue are a few calcified lymph nodes in the hilar region, characteristically designated a “Ghon” complex, and a lifelong positive skin test result. Lifelong persistence of tuberculosis DNA has been demonstrated in macrophages and nonphagocyte cells in histologically normal lung tissue from individuals dying of unrelated causes (e.g., trauma) by in situ PCR.107 In children with normal immunity, the primary infection rarely progresses to symptomatic pulmonary disease. However, in immunosuppressed patients of any age and in patients with HIV infection depending on the degree of immunosuppression, progression to active cavitary tuberculosis is much more common. The immunologic processes involved in controlling tuberculosis are almost entirely dependent on intact cell-mediated immunity. T lymphocytes are stimulated to produce lymphokines, which in turn attract and activate mononuclear macrophages. Macrophages may successfully kill the tubercle bacilli. However, very commonly they are killed by the organisms in the process. A focus of macrophages and lymphocytes develop with central necrosis, which is termed caseation. The tubercle thus is the pathologic hallmark of the caseating granuloma seen microscopically in mycobacterial disease. Tuberculous meningitis occurs as a reactivation of metastatic foci in the meninges and brain parenchyma, which have been present asymptomatically for months to
Central Nervous System Infections
359
years following primary infection. Pathologically, tuberculous meningitis exhibits a thick exudate at the base of the brain, particularly involving the optic nerves at the optic chiasm, the pons, and cerebellum. The histologic appearance depends on the stage of the disease. Initially it consists of polymorphonuclear leukocytes, macrophages, and lymphocytes. But later, after a phase of lymphocytic proliferation, granulomas with caseous centers become a prominent feature. Another feature of tuberculous infection of the meninges is involvement of the blood vessels traversing the meninges. Small- and medium-sized arteries are most often involved, although capillaries and veins may be similarly affected. The changes involve granuloma formation and inflammation of the adventitia, which causes a reactive cellular proliferation of the intima, which in turn may lead to occlusion of the vessel and infarction of the areas supplied by the vessel. Clinically, this phenomenon is most often found in the distribution of the middle cerebral artery due to its location in relation to disease at the base of the brain. Hydrocephalus is one of the most frequent complications of tuberculous meningitis, commonly accompanying symptomatic primary infection in children. Hydrocephalus occurs either by mechanical blockage of the spinal aqueduct or the foramina of Luschka due to the exudate at the base of the brain, or to edema of the surrounding brain parenchyma with the same result. Hydrocephalus may also be caused by blockage of CSF reabsorption at the base of the brain due to the intense infiltrate. The former mechanism leads to noncommunicating hydroencephalus and the latter mechanism to communicating hydrocephalus. Clinical Presentation The classical clinical presentation of tuberculous meningitis in adults is fever and headache, together with meningismus that becomes progressively more severe over a period of 2 to 3 weeks. However, the duration of prodromal symptoms can be quite variable and some patients have been reported with symptoms for several months before seeking medical attention. None of these signs or symptoms is universally present in all patients with tuberculous meningitis. Several recent studies have documented that fever is present in approximately 60% to 80% of patients. Stiff neck and meningismus are reported approximately as frequently as is headache among adult patients.108–110 Other signs that are commonly observed in these patients include lethargy and other behavioral changes in 30% to 70%, seizures in 10% to 15%, and cranial nerve palsies in up to 20% to 30% of adults. Occasionally, abnormal movements such as chorea, hemiballismus, athetosis, myoclonus, and cerebellar signs and symptoms are observed. Localizing neurologic symptoms due to tuberculomas depend on the size and location of the mass lesion. Strokes due to tuberculous vasculitis usually involve the distribution of the middle cerebral artery and produce symptoms related to that distribution. The most
W9418-12.qxd 11/3/03 2:58 PM Page 360
360
Section I
Introduction
common cranial nerve abnormalities involve the sixth cranial nerve followed by the third, fourth, and seventh cranial nerves, but may even involve the second, eighth, tenth, eleventh, and twelfth cranial nerves.111 Higher degrees of mental obtundation, such as coma, are present or develop in approximately 30% of adults and children. In some series as many as 50% of children have a history of tuberculosis, whereas only 8% to 12% of adults have such a history. Hydrocephalus is a serious and potentially devastating complication that may develop in as many as 40% of children and be associated with a variety of focal neurologic signs including hemiparesis and blindness. Predisposing conditions in adults include alcohol abuse, intravenous drug abuse, immunosuppression due to steroid and other immunosuppressive treatments, HIV, and an assortment of underlying chronic illnesses. The differential diagnosis is quite wide and includes bacterial, viral, and fungal infections of the CNS as well as malignancies and noninfectious conditions such as CNS lupus (Table 12-6). Diagnosis Routine laboratory tests are not particularly helpful. The sedimentation rate varies considerably in series of patients with proven tuberculous meningitis, ranging from normal to as high as 90 mm/hr. Similarly, a syndrome of inappropriate antidiuretic hormone, manifested by hyponatremia and hypochloremia, has been observed in some cases, but this is by no means diagnostic and is affected by a variety of other symptoms such as vomiting and anorexia, which may accompany the disease. The chest radiograph is likewise nondiagnostic; however, 25% to 50% of adults may show evidence consistent with current or remote tuberculous infection. In children where tuberculous meningitis quite commonly follows on the heels of primary infection, chest radiographic evidence of tuberculosis has been observed in 50% to 80% of cases. Miliary disease was fairly commonly associated with tuberculous meningitis in the preantibiotic era. However, it is relatively rare at present. Skin testing is notoriously unreliable, ranging from 40% to 65% positive skin test results in adult series. Skin testing is more helpful among children, with positive results in the neighborhood of 85% to 90%.112 Perhaps the single most useful diagnostic procedure is examination of the CSF. Here the classic findings of elevated protein level, depressed glucose level, and elevated white blood cell count with a lymphocytic predominance, together with a chronic history, that is, weeks of illness as opposed to days as in acute bacterial meningitis, is strongly suggestive of a tuberculous or fungal etiology. Median CSF protein levels are generally 100 to 400 mg/dL with occasional levels as high as 1 to 2 g/dL, although levels that high usually suggest CSF block. The median white blood cell count generally runs between 100 to 200 WBC/mm3. The CSF glucose is less than
Table 12-6 Common Noninfectious Causes of Aseptic Meningitis Syndrome Drugs Nonsteroidal anti-inflammatory agents (ibuprofen, naproxen, tolmetin, diclofenac) Antibiotics (trimethoprim/sulfamethoxizole, trimethoprim, cephalosporins, penicillin, amoxicillin, isoniazid, ciprofloxacin, metronidazole) Intravenous immunoglobulin Muromonab-CD3 (OKT3) Azathioprine Carbamazepime Ranitidine Famotidine Indinivir Sulfasalazine Vaccines Measles, mumps, rubella (MMR), alone or in combination (pertussis—acute encephalopathy, not aseptic meningitis) Other conditions Intrathecal injections Neurosurgery-related procedures CNS tumors and cysts Carcinomatous meningitis Lymphoma Leukemia Systemic lupus erythematosus Sjögren’s syndrome Behçet’s disease Vogt-Koyanagi-Harada syndrome Mollaret’s meningitis* Sarcoidosis Vasculitis Kikuchi’s disease Relapsing polychondritis and aseptic meningitis Still’s disease Hypophysitis Uveomeningoencephalitis Steroid-responsive meningitis *Has been associated with herpes simplex virus.
45 mg/dL in 70% to 80% of patients. The glucose level tends to become progressively lower and the protein progressively higher as the duration of illness continues without treatment; when treatment is successful, the glucose level tends to return toward normal. Any one of these tests in spinal fluid may be completely normal, but it is extremely unusual for all three parameters to be completely normal in a patient with true tuberculous meningitis. If a patient has a completely normal spinal fluid and M. tuberculous is reported by the laboratory, the report should be considered suspect until proven otherwise. The spinal fluid acid-fast smear is positive in less than 25% of patients who ultimately have cultureproven tuberculous meningitis. It is also important to recognize that the culture itself is significantly less than 100% sensitive; in most series, the culture positivity ranges from 40% to 70%.108,113–117
W9418-12.qxd 11/3/03 2:58 PM Page 361
Chapter 12
Figure 12-6. Axial, low convexity, T1-weighted, post– gadolinium-enhanced, MR image demonstrates abnormal contrast enhancement involving the superficial ventral surface of the mesencephalon and also along the surface of the optic chiasm. Less pronounced changes are seen along the mesial surface of the left uncus. The enhancement appears thick as well. In addition, there is evidence of significant ventriculomegaly indicating concurrent external hydrocephalus. These findings are consistent with a more chronic type of granulomatous meningitis, most often seen in CNS tuberculosis or fungal infections of the meninges. (Courtesy of Ronald Quisling, MD, University of Florida.)
Newer diagnostic methods include measurements of mycobacterial antigens, tuberculostearic acid, and antibody to mycobacteria in spinal fluid. Although initial reports with all of these show up to 100% sensitivity and specificity, it is extremely unlikely that these spectacular results will withstand the test of time. PCR has likewise been applied to the diagnosis of tuberculosis meningitis, but it is probably even less sensitive than culture. Modern radiographic techniques such as CT, MRI with gadolinium enhancement, and MR angiography are extremely sensitive in delineating CNS involvement, readily demonstrating meningeal inflammation and entrapment of cranial nerves in the basilar tuberculous exudate (Fig. 12-6). In addition, MR angiography can detect characteristic vascular narrowing that accompanies tuberculous meningitis but is less common in other entities. Treatment Treatment of tuberculous meningitis consists of at least three, and usually four, drugs until the susceptibilities are
Central Nervous System Infections
361
established: isoniazid (INH), rifampin, ethambutol, and pyrazinamide are the standard, together with pyridoxine at a dose of 25 to 50 mg daily to prevent depletion by INH; many authors recommend dexamethasone for the first month to improve outcome. Treatment must be continued at least 1 year. INH is generally used at a dose of 300 mg/day in adults and 10 mg/kg/day in children. The most serious side effect is hepatitis, which ranges from asymptomatic enzyme elevations to fulminant hepatitic necrosis. This complication historically was observed in 0.5% to 2.0% of patients receiving INH, but recent data from over 11,000 persons suggest the true risk is approximately 0.1% among those receiving INH alone for prophylaxis and approximately 1% for those receiving INH as part of a treatment regimen for tuberculosis.118 The incidence of hepatotoxicity is higher in persons older than 35 years of age, as well as those with other conditions affecting liver function such as alcoholism and viral hepatitis. The dose of rifampin is 600 mg/day and is only infrequently associated with side effects such as a flu-like syndrome and a hypersensitivity reaction with renal, hepatic, and hematologic toxicity. Pyrazinamide is given in at a dosage of 25 mg/kg/day and has a relatively low incidence of side effects; there is little added toxicity when combined with INH and rifampin. The CSF penetration of pyrazinamide is excellent. Ethambutol is generally administered at a dose of 25 mg/kg for the first 1 to 2 months of treatment, with a reduction in dose to approximately 15 mg/kg/day because of the risk of optic neuritis that is seen in approximately 25% of patients. The first clue to the development of this complication is loss of red-green vision or diminished visual acuity; ophthalmologic consultation is suggested in these situations. Streptomycin was one of the first drugs found to be active against tuberculosis, and is commonly administered in a dose of 20 to 40 mg/kg/day for children and 1 g/day for adults. Unfortunately, the irreversible ototoxicity is so frequent that it is not advisable to use this agent unless absolutely necessary. Second-line antituberculous drugs such as para-aminosalicylic, cycloserine, ethionamide, kanamycin, and amikacin should only be used based on treatment failure with primary agents and antibiotic susceptibility studies. Of these agents, ethionamide and cycloserine penetrate well into the CNS. Because one cannot rule out tuberculosis based on all of the immediately available diagnostic modalities, empiric treatment may often have to be given while awaiting the results of cultures. Cryptococcal meningitis can be readily ruled out by a negative cryptococcal antigen, negative India ink, and no growth within the 10 to 14 days of culture. However, other fungal meningitides cannot be totally ruled out, and some patients may have to be placed empirically on treatment with both antituberculous medications and amphotericin B. Occasionally, patients treated in this fashion turn out to have noninfectious causes for their CNS symptomatology and careful radiographic studies or invasive diagnostic studies may be needed. In patients who show no
W9418-12.qxd 11/3/03 2:58 PM Page 362
362
Section I
Introduction
signs of improvement within the first week or so and particularly in the setting where the patient is known to have a malignancy, the spinal fluid should be reexamined for the possibility of meningeal carcinomatosis. Despite voluminous literature over the past 40 years, the place of steroids in the treatment of tuberculous meningitis remains unclear. Case reports and many numerous series demonstrate the often dramatic and immediate effects of steroids in terms of defervescence and clearing of the sensorium, even after a few doses. While there seems to be general agreement that survival from tuberculous meningitis is improved with the use of steroids, many of the survivors do so with severe sequelae. Therefore, most authorities currently recommend using steroids only for patients in certain specific situations such as extreme neurologic compromise, elevated ICP, impending herniation, or spinal block. The dose of prednisone is 60 mg/day or 1 mg/kg/day; dexamethasone may also be used at a dose of 8 to 16 mg/day in divided doses. Steroids are given for 3 to 6 weeks and then tapered over a 2- to 4-week period.112 Prognosis and Sequelae Before the availability of antibiotic treatment, survival from tuberculous meningitis was exceedingly rare. Survival rates are 70% to 80% in most recent series. Probably the most significant prognostic factor for survival is how advanced the disease is at the time of presentation. Other factors correlating with poor response to treatment are extremes of age and co-existent miliary disease. The earliest sign of response to therapy in most cases is reduction in peak daily temperatures within the first 1 to 2 weeks and subjective improvement in fatigue and malaise over the same time. However, early studies pointed out that it was not uncommon for some markers of disease, such as the presence of bacilli in a smear of the CSF or an increase in CSF protein, to occur shortly after the initiation of treatment. In general the glucose level in the CSF rises with successful treatment while the protein returns to normal more slowly, a process that may take as long as 6 months. In some series, up to 50% of survivors have a variety of neurologic deficits. As with survival itself, the more seriously ill the patient is on presentation, the more likely complications or sequelae are to occur. Among children, the most common of these complications are seizure disorders, ataxia, incoordination, persistent cranial nerve abnormalities, and spastic hemiparesis. Adults are most frequently left with chronic organic brain syndrome, with or without cranial nerve palsies, paraplegia, and hemiparesis. Optic atrophy can lead to varying degrees of visual impairment or blindness in both children and adults. Eighth cranial nerve abnormalities frequently reported in early series are probably the result of the use of streptomycin, which has largely been replaced.
Tuberculoma Tuberculomas are space-occupying mass lesions within the brain parenchyma ranging in size from less than 1 to more than 10 cm. Pathogenically, they arise by a similar mechanism to that of tuberculous meningitis in that a tubercle seeded during the time of primary infection breaks down but because of its location within the brain parenchyma, produces a mass lesion rather than meningitis. Clinically these lesions present with fever, headache, seizures, and other neurologic signs and symptoms that are related to their anatomic location. Generally, patients have a single lesion on presentation; however, autopsy series and sophisticated radiologic studies have shown that in up to 70% of patients, multiple lesions are present. The duration of symptoms prior to presentation is somewhat longer than in tuberculous meningitis, averaging weeks to months with occasional patients having symptoms for years prior to diagnosis. CT scanning and MRI have a virtually 100% diagnostic sensitivity; however, tissue confirmation is essential to rule out malignancies or other space-occupying lesions. Approximately 60% of the specimens from tuberculomas stain positive for acid fast on smear and approximately the same number ultimately grow in culture. Caseating granulomas are almost invariably seen histologically. Treatment is essentially the same as that for tuberculous meningitis. Spinal Tuberculosis Tuberculous spinal meningitis may accompany tuberculous meningitis or may occur as an isolated entity. The pathogenesis and pathologic findings consist of characteristic exudate surrounding many parts of the spinal cord with symptoms due to compression and vasculitic changes, as would be expected. Symptoms of transverse myelitis and spinal block, as well as nerve root pain, paresthesia, and motor weakness may be seen. The onset can be sudden or present as a slow ascending paralysis over several months to years. Fever and systemic symptoms occur in less than 50% of patients, but a sensory level was demonstrable in twothirds of patients with thoracic involvement in one series.119 Less commonly, intramedullary tumors of the spinal cord can present as a Brown-Sequard lesion. MRI is generally required for accurate diagnosis of spinal tuberculosis. However, biopsy is required to establish etiology. Treatment for these lesions may require surgical intervention to relieve compressive symptoms. Antibiotic therapy is based on that for tuberculous meningitis.
Fungal Meningitis The most common fungal pathogens of the central nervous system include the yeast Cryptococcus and the dimorphic fungi Histoplasma, Coccidioides, and Blastomyces. In
W9418-12.qxd 11/3/03 2:58 PM Page 363
Chapter 12
immunocompromised patients, Aspergillosis, Candida, and the Mucorales may be important pathogens as well. Cryptococcal Meningitis Cryptococcus neoformans is perhaps the most common fungal pathogen of the CNS. Cryptococcus has a worldwide distribution and is particularly associated with soil that has become contaminated with bird droppings. There are two varieties of C. neoformans: C. neoformans var neoformans and C. neoformans var gatti, which have a somewhat different distribution in nature. C. neoformans var neoformans is found worldwide and produces most of the infections in patients in the United States, while C. neoformans var gatti is a more common pathogen in Southeast Asia, Africa, Australia, and parts of Southern California. The most important determinant of CNS infection by Cryptococcus is the immune status of the host. This can be most easily demonstrated by the marked increase in the number of cases associated with HIV infection and the fact that as the CD4+ count decreases the incidence of cryptococcal infection increases markedly, particularly at CD4+ counts less than 200 cells/mm3. Although a variety of virulence factors have been described in Cryptococcus, such as the production of the pigment melanin, it is probably the production of the thick polysaccharide capsule that protects the fungus from phagocytosis by the host which is the most important parasiteassociated virulence factor. The initial infection with Cryptococcus is due to inhalation and the production of a pneumonitis, which is generally asymptomatic even in immunosuppressed patients. Symptomatic patients usually present with fever and cough, and a variety of chest radiographic findings, such as nodular, pleural-based lesions and lobar infiltrates. The initial pneumonia generally clears without treatment, even in immunosuppressed patients. Dissemination from the pneumonitis seeds many organs in the body particularly the central nervous system. The organism may remain latent in the lung and at other sites indefinitely. Thus, most patients presenting with cryptococcal meningitis have no evidence of a concurrent pulmonary disease. Clinically, the presentation of cryptococcal meningitis is indistinguishable from that of any other chronic meningitis due to pathogens such as Coccidioidomycosis, Histoplasmosis, or from tuberculous meningitis. Patients generally describe a history of 2 to 3 weeks of headache, fever, and stiff neck together with a variety of nonspecific symptoms, such as lethargy, confusion, nausea, vomiting, and relatively rarely, focal neurologic deficits. As the disease advances in untreated patients, evidence of increased ICP develops, cranial nerve palsies and seizures may be observed, and the disease progresses to obtundation and death. Papilledema may be seen in up to one third of cases and cranial nerve palsies may develop in approximately 20%. Occasionally, total visual loss develops secondary to fungal
Central Nervous System Infections
363
involvement of the optic tracts, as well as from adhesive arachnoiditis, chorioretinitis, or elevated ICP. Hydrocephalus frequently develops even in successfully treated patients. The diagnosis of cryptococcal meningitis is usually not difficult. The classic and time-honored method for diagnosis is demonstration of the yeast in spinal fluid by India ink stain. Microscopically, the India ink particles serve to outline the very large clear polysaccharide capsule surrounding the yeast. The test is positive in over 90% of patients with HIV, but in only 50% of patients with normal immunity. In the past 20 years, several latex agglutination tests have been developed that detect the excess polysaccharide capsule produced in the spinal fluid of patients with cryptococcal meningitis. This test result is positive in more than 95% of patients with the disorder. While the test is highly reliable, it is important to recognize that there can be both false-negative and false-positive reactions. For example, patients with HIV who have an overwhelming cryptococcal meningitis may have so much polysaccharide capsule in their CSF that a “prozone” effect occurs, resulting in the finding that undiluted CSF or CSF tested at a 1 : 2 dilution may appear negative. Generally, on dilution of the spinal fluid to 1 : 10 or greater, a positive reaction will be observed. Laboratories must be aware of this prozone effect and physicians who suspect cryptococcal meningitis in patients with patients should alert the laboratory to ensure that this possibility is not overlooked. In addition, cross-reactions with Trichosporon beigelii or Capnocytophaga may occasionally produce false-positive latex agglutination test results. The latex agglutination should be confirmed by growth of cryptococci in culture. If a positive cryptococcal antigen titer result is found in the spinal fluid from a patient whose CSF does not grow cryptococcus, the test cannot be assumed to be a false-positive result. A large volume of spinal fluid, 10 to 20 mL, should be obtained and recultured for cryptococci as well as used for repeating the cryptococcal antigen test. The higher the cryptococcal antigen titer, the less likely it is to be a false-positive result. Spinal fluid changes in cryptococcal meningitis generally parallel those of chronic tuberculous meningitis and other chronic meningitides (see Table 12-2). The CSF glucose level is generally below 40 mg/dL and the CSF protein elevated in the large majority of patients. CSF white blood cell count is also elevated with a lymphocytic predominance. Treatment of cryptococcal meningitis depends in part on the host susceptibility factors. In patients with no underlying chronic illness or HIV infection, amphotericin B at a dosage of 0.5 to 0.8 mg/kg/day IV together with flucytosine 37.5 mg/kg orally every 8 hours should be administered until the patient has become afebrile and culture negative; this takes approximately 6 weeks. This should probably be followed with treatment of fluconazole at a dose of 400 mg/day for an additional 2 to 3 months. Flucytosine may cause
W9418-12.qxd 11/3/03 2:58 PM Page 364
364
Section I
Introduction
severe leukopenia and thrombocytopenia, particularly in patients with impaired renal function that in turn may develop as a result of amphotericin therapy. Therefore, patients need to be watched extremely carefully for these particular toxic developments. Some authorities recommend measuring flucytosine levels and adjusting the dose to give a peak of 70 to 80 mg/L and a trough of 30 to 40 mg/L. However, these levels may not be readily available at many hospitals. Alternative treatments include the substitution of various lipid soluble amphotericin B preparations, which may be used for patients in whom nephrotoxicity develops. Additionally, if the patient is intolerant to flucytosine, treatment with amphotericin at a higher dose of 0.7 to 1 mg/kg/day for 6 to 8 weeks may be substituted. Fluconazole at a dose of 400 mg/day orally for 8 to 10 weeks may be curative, particularly in nonimmunocompromised patients who are less ill. However, it should be recognized that fluconazole alone is less effective than the combination of amphotericin B plus flucytosine, in that the duration of positive culture results of spinal fluid is longer and a higher rate of treatment failures have been associated with fluconazole regimens. Histoplasmosis Although found worldwide, histoplasmosis has a very distinct geographic distribution in the United States. Most cases occur in the Ohio and Mississippi River valleys in the Central Midwestern part of the United States, with extension as far east as Maryland, Delaware, and some parts of Georgia and Florida. The disease is rarely seen west of Texas, Oklahoma, and Kansas. Histoplasmosis belongs to a group of fungi known as the dimorphic fungi because they change their characteristic morphology depending on temperature. Along with Coccidioides immitis and Blastomyces dermatitidis, H. capsulatum exists as a mold at room temperature (approximately 23°C) and converts to a yeast form at a body temperature of 37°C. The organism is widely disseminated in nature in the soil and may reach high levels in areas where birds roost and in caves inhabited by large numbers of bats. In vitro, the mold phase is characterized by both macroconidia and microconidia. The microconidia are small, smooth oval bodies ranging in diameter from 2 to 5 microns and are believed to be the infective phase because their small size is such that they are readily carried down to the terminal bronchioles and alveoli. Once inhaled, the microconidia are ingested by alveolar macrophages and rapidly undergo conversion to the yeast phase. From the initial pulmonary foci, the yeast rapidly migrates to hilar lymph nodes from which they can disseminate to multiple foci in the body. As is the case with Cryptococcus and tuberculosis, an asymptomatic pulmonary infection frequently develops approximately 2 to 3 weeks after exposure. The development of the cellular immune response limits the spread of the organism and generally
clears the initial, early, pulmonary focus, leaving minimal to no calcifications in hilar lymph nodes and lung tissue. The disseminated lesions are most commonly manifest by widespread calcific lesions in the spleen and liver after they heal. Although the vast majority of primary infections in immunologically normal hosts resolve spontaneously, leaving the patient with a positive histoplasma skin test result, there are a number of unfavorable outcomes. Pulmonary disease may go on to produce chronic cavitary histoplasmosis that is radiographically identical to pulmonary tuberculosis. The initial pulmonary infection may result in an acute progressive disseminated infection that characteristically presents with fever, chills, weight loss, hepatosplenomegaly, and pancytopenia from bone marrow involvement. This form of disseminated histoplasmosis most often affects those who are highly immunosuppressed due to AIDS, lymphoma, or iatrogenic therapy. CNS involvement in this syndrome includes encephalitis, acute meningitis, and encephalopathy. Occasionally, histoplasmomas or mass lesions in the CNS are observed.120 A syndrome of disseminated histoplasmosis may also be observed in patients with normal immunity, which presents as a much lower grade chronic illness, rather than the acute presentation of disseminated histoplasmosis in immunosuppressed patients. CNS involvement manifests itself, therefore, as symptomatic intracranial mass lesions; isolated chronic meningitis, with or without other manifestations of disseminated histoplasmosis; and meningitis occurring in the presence of disseminated infection. Some patients may have CNS disease secondary to emboli from histoplasma endocarditis. In what is likely the most comprehensive series in the literature, Wheat and colleagues found that approximately 40% of patients with histoplasma meningitis had chronic meningitis as part of their disseminated disease.120 Approximately 25% to 30% had isolated chronic meningitis and the remainder presented with various forms of mass lesions and encephalitis. In general, the duration of symptoms of meningitis before diagnosis tends to be somewhat longer than for cryptococcal or tuberculous meningitis. In Wheat and coworkers’ series, approximately 30% of patients had duration of symptoms of less than a month, 44% a duration between 2 and 6 months, and another 27% had symptoms lasting for more than 6 months before the diagnosis was made. The diagnosis of histoplasma infection in the central nervous system is, to some extent, dependent on the presentation of the patient. Those who present with localizing CNS signs and symptoms should undergo CT and MRI to rule out mass lesions. Those with systemic manifestations may have the diagnosis made by culture or biopsy of an enlarged lymph node, liver, or bone marrow. CSF findings are typical for chronic fungal and tuberculous meningitis, with 90% of patients having abnormal leukocyte count in the CSF with lymphocytic predominance (see Table 12-2). At least 80% of patients will have an elevated protein level and a glucose level of less than 40 mg/dL. Serologic testing results for antibod-
W9418-12.qxd 11/3/03 2:58 PM Page 365
Chapter 12
ies to H. capsulatum in blood are generally positive in 60% to 90% of patients with CNS histoplasmosis. However, in an endemic area, serology is difficult to interpret because patients may have antibodies from prior exposure rather active infection. Likewise, skin testing in an endemic area is probably of no diagnostic value. In Wheat and associates’ series,120 CSF cultures were only positive in 26.7% of cases. A recently described histoplasma antigen test that detects the histoplasma polysaccharide capsular antigen is reported to have positive results in the urine in 90% of patients with disseminated histoplasmosis,120 but, positive results in the spinal fluid of only 40% of patients with CNS histoplasmosis.121 Treatment Amphotericin B at a dose of 0.7 to 1.0 mg/kg should be used to treat patients with histoplasma meningitis.120 After the first seven to ten days, this dose can be lowered to 0.8 mg/kg every other day; the lipid amphotericin formulation may be substituted if renal toxicity occurs. Therapy can be assessed with serial weekly or biweekly lumbar punctures, which will show a rise in the CSF glucose and fall in the CSF cell count and protein with successful treatment. The published data suggest that although approximately 80% of patients will respond to amphotericin, at least half of those initial responders will relapse and approximately 20% will die from the disease. Itraconazole and fluconazole are not useful in the treatment of CNS histoplasmosis, as they do not cross the blood-brain barrier particularly well. In non-CNS disease, treatment with amphotericin B should be continued for at least 8 to 12 weeks, to a total dose of approximately 30 to 35 mg/kg, in view of poor outcome among patients receiving less than this total dose.120 In patients with HIV, suppressive treatment with itraconazole, 200 mg orally every day, should begin after the initial course of amphotericin. Coccidioides immitis Infections with C. immitis have been recognized for longer than 100 years and were originally believed to be a nearly uniformly fatal infection. However, a widely publicized laboratory acquired infection in a young medical student in 1929 resulted only in pulmonary disease that spontaneously healed. This led to the recognition that this agent was capable of producing both fatal disseminated infection as well as the self-limited, well-recognized “Valley Fever” that was frequently observed in residents of the San Joaquin Valley in California. Like the other dimorphic fungi, H. capsulatum and B. dermatitidis, the natural habitat of C. immitis is soil. But unlike the other fungi, C. immitis is not distributed worldwide but limited to the lower Sonoran life zone, found primarily in the desert southwest of the United States, Mexico, and parts of South and Central America. The characteristics of this particular zone are an arid climate with a yearly rainfall of 5 to 20 inches, hot summers, warm winters,
Central Nervous System Infections
365
and an alkaline soil. In the mold phase of the fungus, which is the form found in soil and other environmental sources, the hyphae fragment into specific structures known as arthrospores, which are highly infectious and readily aerosolized when dust is produced. Thus, on occasion, dust storms have blown infectious spores of C. immitis as far north as Sacramento and beyond to produce outbreaks well outside the endemic area of the San Joaquin Valley. Transmission of the organism on fomites as far as the East Coast has produced occasional cases as well. The narrow environmental requirements of this fungus account for its endemic distribution within the United States. Epidemiologic studies of people who migrate into the Central Valley of California suggest that the annual risk of infection is approximately 15%. For example, Smith and co-workers showed that between 25% and 50% of military personnel stationed in the San Joaquin Valley had conversion of their skin test results within the first year of moving to that region.122 Pathogenesis As is the case with the other dimorphic fungi, the initial route of infection is inhalation with an early focus of infection in lung tissue. The arthrospores convert, through unknown mechanisms, into a structure called the spherule, which enlarges to approximately to 50 to 70 microns in diameter and subdivides into an extremely large number of locally invasive and infectious spores that are released upon spherule rupture. The outcome of this infection is highly variable. Approximately one-half to two-thirds of patients who are demonstrated to be infected with this agent show absolutely no, or very mild, initial pulmonary infection. The majority of patients who become symptomatic develop a mild selflimited respiratory infection manifested by fever, cough, malaise, arthralgias, weight loss, and in some patients, a striking clinical syndrome of erythema nodosum and erythema multiforme. The vast majority of these symptomatic cases resolve within 2 to 4 weeks, occasionally taking up to several months, without treatment. Complications occur in no more than about 10% of all patients with clinically symptomatic primary infections. Occasionally, patients will present with fulminant pneumonia and shock-like syndrome, possibly resulting from a particularly high inoculum or perhaps as a result of fungemia and miliary dissemination of the disease. Such a presentation is not uncommon among patients with HIV and severe depression of the CD4+ count. Other manifestations include pulmonary nodules, cavities, and chronic lung disease indistinguishable from chronic tuberculosis. The most common site of disseminated lesions is the skin where maculopapular lesions may progress to keratotic and verrucous ulcers with subcutaneous fluctuant abscesses. The most serious form of disseminated coccidioidal infection is coccidioidal meningitis. Without treatment it is nearly uniformly fatal within 2 years of diagnosis.123 Obser-
W9418-12.qxd 11/3/03 2:58 PM Page 366
366
Section I
Introduction
vational studies suggest about 80% of patients in whom meningitis develops become symptomatic within 6 months of the initial infection.124,125 The signs and symptoms of coccidioidal meningitis are very similar to those of other chronic fungal and tuberculous meningitides. Patients have headache, may or may not have fever, varying degrees of nuchal rigidity, nausea, vomiting, and altered mental status.123,124 A number of factors including pregnancy, immunosuppression due to HIV infection, steroids or other immunosuppressive drugs, and nonwhite race predispose to dissemination of C. immitis and the development of meningitis. Diagnosis of coccidioidal meningitis is based on analysis of the CSF, which shows typical findings of elevated white count, predominantly lymphocytic in nature, elevated protein, and depressed glucose (see Table 12-2). Occasionally, patients with this entity are found to have a striking CSF eosinophilia, which in one series, was seen in 70% of cases.126 Thus, in a patient who presents with a chronic meningitis together with typical CSF findings, a history of travel to or having lived in an endemic area must be carefully sought. Complications such as meningitis may occur as late as 2 years after the exposure, and the exposure may be exceedingly brief, as minimal as a drive through the California’s Central Valley.127 Although C. immitis grows well on typical fungal media, culture is often negative in spinal fluid. The most reliable method of diagnosis is the detection of complement-fixing antibodies in the spinal fluid. Although these results may be negative in a few patients early in the disease, after several months they are virtually 100% positive. In patients with pulmonary disease or extrapulmonary manifestations, biopsy with culture and histopathologic examination of involved tissue may give positive results. The demonstration of a spherule in tissue or a positive culture is diagnostically definitive. Treatment There is considerable disagreement among experts in regard to treatment of uncomplicated pulmonary disease in otherwise healthy patients, with some experts recommending amphotericin B for all patients while others reserve it for patients with risk factors such as HIV infection, immunosuppression, or non-white race. Fluconazole at a dose of 400 to 800 mg/day is currently recommended as the treatment of choice for coccidioidal meningitis because the response rate of approximately 70% is very close to that achieved with intrathecal amphotericin B, which was used previously. In patients who do not respond to the 400 mg/day dose of fluconazole, higher doses may be used. If there is no response to azole therapy, amphotericin B at a dose of 0.1 to 0.3 mg/day may be given intrathecally, optimally through an Omaya reservoir. With the demonstration of increasing CSF glucose, and decreasing cell count and protein, this schedule of amphotericin B administration may be decreased to three
times a week after 2 to 3 weeks of daily treatment, and maintenance may be achieved with once- or twice-weekly injections. Treatment with either oral fluconazole or intrathecal amphotericin B must be prolonged for at least 2 years after the spinal fluid becomes completely normal. In patients with HIV, treatment is lifelong. Patients who have disseminated extra-pulmonary disease in addition to meningitis should also receive systemic amphotericin at a dose of 0.6 to 1.0 mg/kg/day for 7 days followed by 0.8 mg/kg dose every other day, to a total dose of 2.5 to 3 g. The amphotericin should then be followed by a dosage of oral fluconazole at 400 mg/day for up to a year after the course of amphotericin. It should be noted that the response rate is by no means 100% in either meningitis or disseminated disease, and that relapses are not uncommon, even in patients who respond initially. As is the case with other fungi and tuberculosis, patients have presented occasionally with mass lesions in brain parenchyma due to C. immitis, which may require surgical drainage or excision. In addition, hydrocephalus is relatively common in patients with C. immitis meningitis, particularly in children, and must be managed with ventricular shunting. It should be pointed out that, on occasion, C. immitis may actually grow in the shunt and cause obstruction. In patients who have a ventricular shunt in place, intrathecal amphotericin B cannot be given into an Omaya reservoir, but rather must be administered intrathecally via intracisternal puncture or lateral neck injection under radiographic guidance. Although intrathecal therapy can be given via the lumbar route this inevitably leads to varying degrees of potentially severe and debilitating arachnoiditis after several weeks in most patients. Blastomyces dermatitidis B. dermatitidis is a dimorphic fungus that grows as a yeast form at 37°C and in a hyphal form at room temperature. It is believed that B. dermatitidis exists in nature in warm moist soils of wooded areas rich in organic debris, such as decaying vegetation. However, the reports of isolation of the organism in nature have been relatively few and somewhat inconsistent. Epidemiologically, the distribution appears to follow the distribution of the Mississippi River, and is most commonly reported in Louisiana, western Alabama, central Arkansas, Missouri, Kentucky, western Tennessee, and as far north as Minneapolis, Minnesota. However, the distribution also includes outbreaks along the St. Lawrence River in Canada and in many parts of North and South Carolina. Pathogenesis Pulmonary infection occurs by the inhalation of conidia, which convert to the yeast phase in the lung. From there, dissemination to skin and other organs may occur. Pulmonary manifestations vary from asymptomatic to multilobar over-
W9418-12.qxd 11/3/03 2:58 PM Page 367
Chapter 12
whelming infection. Chronic pulmonary infection resembling tuberculosis or other chronic fungal disease may develop. Dissemination to the skin produces a variety of manifestations such as chronic verrucous, giant keratoacanthoma, and even lesions mimicking pyoderma gangrenosa or squamous cell carcinoma. Pathologically, the histologic response is that of neutrophils and noncaseating granulomas with epithelioid and giant cells. Because of a vigorous proliferative response, the pseudoepitheliomatous changes seen on skin and mucosal surfaces in response to B. dermatitidis may resemble squamous cell carcinoma. Clinical Manifestations The vast majority of patients present with symptoms of pneumonia after an incubation period of 30 to 45 days. Symptoms are nonspecific, and include fever, cough, myalgias, arthralgias, and pleuritic chest pain. Chest examination may reveal lobar or segmental consolidation, including mass lesions mimicking carcinoma with hilar adenopathy. Suppurative granulomatous lesions of the skin and bone may develop and become chronic. CNS involvement occurs in between 6% and 33% of disseminated cases.128,129 Although CNS blastomycosis may develop in patients without evidence of disseminated disease, most cases do have evidence of systemic infection at the time of CNS involvement. In addition to meningitis, patients with blastomycosis may present more frequently with blastomycomas or mass lesions in the CNS than are seen with other fungal CNS infections. While steroids and immunosuppressive treatment do not predispose to meningitis per se, they may predispose to dissemination of pulmonary blastomycosis and, hence, increase the probability neurologic involvement. Diagnosis Because there is no highly specific serologic or diagnostic skin test material available, the diagnosis rests on culture of the organism from pulmonary secretions, skin biopsy material, or demonstration of the organism in biopsy tissue. Probably the best and the most widely available serologic diagnostic test is immunodiffusion, which has a sensitivity of 80%.130 Treatment Before the availability of amphotericin B and itraconazole, the disease was progressive and had a mortality rate greater than 60%. In view of this historical perspective, and the recognition that some patients with pulmonary disease have a self-limited infection, it is currently unclear whether the initial primary pulmonary infection should be treated. However, because of our inherent limited ability to predict which cases will disseminate, it certainly seems reasonable to treat most cases of pulmonary infection due to Blastomyces. In mild cases, the treatment of choice is itraconazole at a dose of 200 to 400 mg/day orally. In more serious cases, a
Central Nervous System Infections
367
dose of amphotericin B at 0.7 to 1.0 mg/kg/day should be used. Treatment should continue for 6 months with itraconazale, and up to a total dose of 1.5 to 2 grams with amphotericin B. If disease persists, itraconazole may be given as a follow-up to amphotericin B therapy. If the patient is immunosuppressed or has HIV infection, this follow-up treatment is certainly indicated. In those patients with lifethreatening disease, the total dose of amphotericin B must be increased. CNS disease should be treated with high dose amphotericin B and followed with longer term itraconazole suppression in patients with HIV. Candida Candida species are a part of the normal human mucosal flora and only rarely produce CNS disease. In general, CNS disease due to Candida species occurs as part of disseminated infection in hospitalized patients who have predisposing factors such prematurity, intravenous hyperalimentation, indwelling catheters, treatment with corticosteroids, neutropenia, diabetes, and/or broad spectrum antibiotic use. Typically, CNS microabscesses result from disseminated infections in these settings. Meningitis is rare except as an operative complication or complication of ventricular shunts. Neonates and premature infants seem to be at higher risk than are adults. Although C. albicans accounts for a majority of the infections, non-albicans species such as tropicalis, pseudotropicalis, guillermondii, krusei, and lusitaniae have been reported. Zygomycosis The genuses Rhizopus, Mucor, and Rhizomucor may invade the CNS by direct extension or from hematogenous spread. In patients with diabetes, this disease typically presents as rhinocerebral mucor (Figs. 12-7A and B). Patients are generally predisposed to this complication if their blood glucose has remained uncontrolled and they are acidotic for several weeks. Initial symptoms usually include sinus pain, headache, fever, and nasal stuffiness and discharge, which quickly progress to facial cellulitis, swelling, proptosis, cavernous sinus thrombosis and, if not treated, death. In addition to diabetics, patients who remain neutropenic for longer than 2 to 3 weeks are also subject to this particular infection. Disseminated disease occurring via the hematogenous route may also produce CNS mucormycosis resulting in infarction and abscess formation. Treatment requires amphotericin B in a dose of at least 1 mg/kg/day, together with vigorous surgical debridement of all involved tissue. Aspergillus Species Aspergillus fumigatus and Aspergillus flavus may also produce CNS disease, which is very similar to that described for
W9418-12.qxd 11/3/03 2:58 PM Page 368
368
Section I
Introduction
A
B
Figure 12-7. A, Coronal, T1-weighted post–gadolinium-enhanced MR image through the frontal temporal region demonstrates abnormal contrast enhancement affecting not only the meningeal surface above the fovea ethmoidalis and tuberculum sella but also affecting the pial surface of brain and portions of the brain parenchyma. This has an aggressive appearance. The parenchymal enhancement follows a distribution along the Virchow-Robin spaces representing either cerebritis or ischemic effects of perivascular infection. This appearance of multicompartment disease is consistent with invasive aspergillus or mucormycosis derived from associated frontal and/or ethmoid sinus disease. B, Oblique projection, right carotid arteriogram demonstrates fusiform aneurysmal dilatation of the mid portion of the first portion (A1 segment) of the right anterior cerebral artery. The parent artery both proximal and distal to the aneurysm is abnormal and displays inflammatory changes (regional vasculitis). This combination of fusiform aneurysm and regional vasculitis confirms the diagnosis of either an inflammatory or, as in this instance, an infectious (mycotic) aneurysm. Both aspergillosis and mucormycosis are known to cause such aneurysms. (Courtesy of Ronald Quisling, MD, University of Florida.)
mucormycosis. The majority of these cases occur in immunosuppressed transplant patients in the hospital who develop systemic disease, or who develop pulmonary disease and systemic spread to the brain via hematogenous dissemination. Direct extension from maxillary and ethmoid sinuses does occur with the production of cavernous sinus thrombosis, as seen with mucormycosis. However, Aspergillus sinus infection and extension to cavernous sinus thrombosis is generally seen in patients with prolonged neutropenia rather than in diabetes where mucor is the more common cause. Treatment requires surgical debridement of all involved tissue, plus voriconazole, caspofungin, or highdose amphotericin.
Central Nervous System Infections in Human Immunodeficiency Virus Infection Despite the marked improvement in and outlook for patients with HIV due to the introduction of highly active antiretroviral therapy (HAART) in 1995, there are approxi-
mately 40,000 new cases of HIV infection per year in the United States. Neurologic involvement in patients with HIV ranges from the traditional “big three” of Cryptococcus, toxoplasmosis, and primary B-cell lymphoma, to viral infections such as cytomegalovirus or progressive multifocal leukoencephalopathy (PML), as well as a variety of neuropathologic manifestations of HIV itself. Toxoplasma gondii T. gondii has a worldwide distribution, and infects numerous wild and domestic animals. Human infection generally occurs through the ingestion of raw or undercooked meat that contains cysts or through the ingestion of food or water contaminated by the oocysts shed in the stool of infected animals. In the United States, the major reservoir for this infection is the domestic cat. Infection in the cat results in the formation of oocysts in the intestine for approximately 3 weeks after initial infection; during this time, as many as 10 million oocysts may be shed daily. Oocysts require 1 to 5 days to become infectious after being shed by the cat, a process that depends on temperature and availability of
W9418-12.qxd 11/3/03 2:58 PM Page 369
Chapter 12
Central Nervous System Infections
369
lethargy, obtundation, and focal neurologic signs developing over 1 to 3 weeks. In some patients, the presentation can be abrupt, with seizures or cerebral hemorrhage. Hemiparesis and abnormalities of speech are quite common. Less commonly, brainstem involvement may result in cranial nerve deficits, while dyskinesias such as parkinsonism, dystonia, tremor, and hemiballismus may also accompany the presenting syndromes. Patients occasionally present with endocrine abnormalities due to involvement of the pituitary axis as well as psychiatric manifestations, such as psychoses and dementia.
Figure 12-8. Cyst of Toxoplasma gondii in brain tissue, original magnification approximately x1200, hematoxylin and eosin. (Courtesy of Anthony Yachnis, MD, University of Florida.)
oxygen.131 During intestinal infection, the tachyzoite form of the organism, a 2- to 4-micron wide by 4- to 8-micron long crescent-like structure, is produced and disseminates to many different areas in the host. In humans, as in many domestic and wild animals, this dissemination is asymptomatic and results in the formation of cysts in brain parenchyma as well as muscle and numerous other organs (Fig. 12-8). Transmission to humans is thus a result of exposure to excreta from cats, as well as ingested meat, particularly pork, that has not been fully cooked. In humans, primary infection with toxoplasma may, on occasion, produce a mononucleosis-like illness characterized by lymphadenopathy, fever, malaise, liver function abnormalities, and occasionally, myocarditis. However, in the large majority of cases, primary infection is asymptomatic and only becomes recognized under conditions of extreme immunosuppression, such as occurs in patients with HIV with CD4+ counts less than 100 cells/mm3 when the cysts break down and initiate symptomatic infection. Serologic surveys in the United States show that approximately 10% of the general population has been infected with toxoplasmosis at some time, with figures ranging from 3% to 35%, depending on the study.131,132 The prevalence is higher in less developed countries and, in certain parts of western Europe such as France, may range as high as 70% to 80% of the population. Although reactivation of toxoplasmosis in the brain generally leads to local replication with the production of single or multiple abscesses, hematogenous dissemination and infection in the lung and in other parts of the body has been documented.120 Clinical Manifestations Toxoplasma encephalitis generally presents with a syndrome of fever; headache; and varying degrees of confusion,
Diagnosis In a patient with HIV, the finding of single or multiple ringenhancing lesions on CT or MRI strongly suggests Toxoplasma encephalitis (Fig. 12-9). If the patient is also known to be, or is found to be, serologically positive for antibodies to toxoplasma, the combination is virtually diagnostic. The sensitivity of the CT scan in patients with HIV and multiple ring-enhancing lesions is estimated at 70% to 80%.131,132 The sensitivity is probably significantly higher with the use of MRI with gadolinium enhancement. In patients with a single
Figure 12-9. Axial, mid-convexity, T1-weighted post–gadoliniumenhanced MR image demonstrates multicentric, enhancing nodules with marginal enhancement and central necrosis. In addition to the nodules, there is further enhancement of the ependymal surface of the trigone of the right lateral indicating active ventriculitis. Cavitary multicentric lesions distributed mainly in the territories of end-arteries are indicative of a hematogenously disseminated process, in this case, toxoplasmosis. (Courtesy of Ronald Quisling, MD, University of Florida.)
W9418-12.qxd 11/3/03 2:58 PM Page 370
370
Section I
Introduction
lesion, the major differential diagnostic possibility is lymphoma, and a brain biopsy may be required to make a definitive diagnosis. A variety of newer imaging techniques, such as positron emission tomography and thallium 201 scanning,133–135 have been developed and appear promising in their ability to improve specificity of radiographic diagnosis of toxoplasmosis encephalitis. Examination of the spinal fluid shows nonspecific abnormalities such as increased white blood cell count and protein. Attempts to improve the sensitivity by the detection of toxoplasma oligoclonal antibody bands have been reported, but do not add any more diagnostic value than a positive serum antibody test. PCR for toxoplasmosis in CSF has been reported to have a sensitivity of 81% in untreated patients.136 Although a wide variety of antibody tests are available, including the original Sabin Feldman Dye Test, current commercial enzyme-linked immunosorbent assay, IgG and IgM tests, along with the indirect fluorescent antibody test, should be readily available and highly specific. Because the disease in patients with HIV is almost always due to reactivation of an old focus, the IgG test result should be positive and the IgM test result should be negative in these individuals. Treatment Treatment should be instituted empirically based upon the clinical setting of a patient with HIV and compatible clinical and radiographic findings. The standard drug regimens include pyrimethamine, 200 mg as a loading dose, followed by 75 to 100 mg/daily, together with a dose of either sulfadiazine at 1 to 1.5 g every 6 hours or clindamycin at 600 to 1200 mg IV every 6 hours. A dose of folinic acid (leucovorin) at 10 to 20 mg/day is given to reduce bone marrow toxicity. Alternative regimens include trimethoprimsulfamethoxazole, given either orally or IV, at a dose of 3 to 5 mg of the trimethoprim component per kg every 6 hours. In addition to pyrimethamine and folinic acid, if the patient is intolerant of sulfadiazine or clindamycin, doses of clarithromycin at 1 g orally every 12 hours, atovaquone at 700 mg orally every 6 hours, azithromycin at 1200 to 1500 mg/day orally, or dapsone at 100 mg/day may be substituted. Signs of improvement are generally seen in the level of consciousness and in a decrease of fever within 5 to 7 days, with over 90% generally responding by day 14.137 If there is no clinical or radiographic response within 10 days, alternative diagnoses and brain biopsy must be considered. Patients should be treated with the initial regimen for 3 to 6 weeks, depending upon the degree and rapidity of improvement, and then receive lifelong suppression with sulfadiazine at a dosage of 500 to 1000 mg orally four times daily, and with pyrimethamine at 25 to 75 mg and folinic acid at 10 to 20 mg, both by mouth daily. Relapse rates with pyrimethamine alone range from 10% to 28% at 50 mg/day, but decrease to 5% at 100 mg/day.138,139
Lymphoma As discussed previously, the major differential diagnostic consideration in toxoplasma encephalitis is CNS lymphoma. It has been estimated that systemic lymphoma will develop in as many of 5% of patients with HIV, and approximately one-third will present with neurologic disease.140 The isolated CNS lymphoma is almost always due to the Epstein-Barr virus–associated B cell type. Patients generally present with confusion, lethargy, memory loss, hemipareisis, speech and language disorders, seizures, and cranial nerve palsies.141 Involvement of the meninges is common and may be seen both in primary CNS lymphoma as well as spread to the meninges from systemic lymphoma. To date, there is nothing in the presentation or in the radiographic appearance that can distinguish CNS lymphoma from cerebral toxoplasmosis with complete certainty. Thus, patients who do not respond to empiric therapy for Toxoplasma encephalitis will need to undergo brain biopsy, in many cases, to establish a definite diagnosis. In patients with known systemic lymphoma and/or in patients who are known to be seronegative for toxoplasma, biopsy of the CNS lesion should be undertaken without waiting for an empiric response to treatment for Toxoplasma encephalitis. Primary CNS lymphoma does respond to therapy with whole brain irradiation and chemotherapy, but survival is generally less than 4 months.142,143
Cytomegalovirus In adults with HIV, serologic evidence of previous infection by cytomegalovirus (CMV) is present in over 90%. Serologic surveys in the general adult population show a range of CMV infection between 50% and 80%, depending on the socioeconomic status and the particular group studied. The seroprevalence of CMV increases with age. Approximately 10% of all infants born in the United States either have CMV infection at birth or acquire it within the neonatal period. The vast majority of these infections are asymptomatic and result from exposure to reactivated virus in seropositive mothers. Subsequently, both children and adults may become infected through exposure to infected urine, often from infants in day care; the virus may also be acquired from sexual contact, generally during the late teens and twenties. In nonimmunocompromised adults, primary CMV infection is clinically identical to that of infectious mononucleosis and runs a self-limited course of 2 to 3 weeks characterized by fever, fatigue, malaise, adenopathy, sore throat, and increased liver enzyme levels, together with atypical lymphocytes in the blood smear. In individuals with HIV, as long as cell-mediated immunity remains relatively intact, symptomatic reactivation of cytomegalovirus is not generally seen. However, as the disease progresses and the CD4+ count decreases to less than
W9418-12.qxd 11/3/03 2:58 PM Page 371
Chapter 12
50 cells/mm3, reactivation of CMV becomes particularly frequent. CMV retinitis is the most common form recognized clinically and in the pre-HAART era, up to 30% of patients wtih HIV showed evidence of CMV retinitis. The diagnosis of CMV retinitis is a based on the puffy white retinal infiltrates seen together with retinal hemorrhage. It may or may not be associated with systemic or other manifestations of cytomegalovirus. Either one or both eyes may be involved at a given time. CMV may also produce ulcerations at any level of the gastrointestinal tract, from the esophagus to the colon, and is probably the most frequent cause of gastrointestinal hemorrhage in patients with HIV. The diagnosis has to be made by colonoscopy or endoscopy with biopsy and demonstration of typical cytopathology, or culture of the virus. There are several forms of CNS involvement with CMV. The pathogenesis of CNS involvement with CMV appears to follow two different routes, first, via the ependymal cells in the ventricles, and second via the blood through capillary endothelial cells. Infection via the ependymal cells and CSF is manifest pathologically as necrotizing encephalitis limited to the periventricular areas, with numerous cytomegalic inclusions in and around the lesions. In contrast, infection acquired via hematogenous dissemination results in microglial nodular encephalitis, which is manifest pathologically as glial nodules formed by rod cells, few lymphocytes and macrophages. Very little tissue damage is associated with these lesions.144 The microglial nodular disease may involve multiple parts of the brain including the periventricular areas. As with CMV retinitis, these CNS complications of CMV occur very late in HIV infection, almost always in patients with CD4+ counts less than 50 cells/mm3. In the series reported by Grassi and colleagues, the mean CD4+ count for patients with microglial nodular encephalitis was 21 cells/mm3 and the average CD4+ count of patients with ventriculoencephalitis was 11 cells/mm3. Although there is considerable overlap of symptomatology between these two pathologic conditions, microglial nodular encephalitis is characterized by the onset of acute confusion associated with delirium and psychomotor agitation, whereas the onset of ventriculoencephalitis is insidious and generally characterized by cognitive disturbances, memory deficits, and mental sluggishness. Patients with either type of encephalopathy may complain of headaches and have seizures. CSF examination in these conditions reveals a slightly elevated protein, which tends to be lower in microglial nodular encephalitis, averaging approximately 60 mg/dL as compared with 171.5 mg/dL in ventriculoencephalitis. Glucose levels are normal and cells are generally, although not always, absent. Although MRI is more sensitive than CT in the diagnosis of these conditions, the MRI may be normal or show nonspecific changes.145 In patients with ventriculoencephalitis, typical MRI findings include a hyperintense ventricular rim on T2-weighted imaging or enhancements with gadolinium.
Central Nervous System Infections
371
Almost all cases show nonspecific cerebral atrophy. Systemic involvement with cytomegalovirus is more frequently associated with microglial nodular encephalitis, which may be documented with CMV antigenemia, PCR, or viral culture. Patients with ventriculoencephalitis are more likely to have the clinical syndrome of CMV radiculopathy, which is described next. CMV infection in patients with HIV may also involve the spinal cord. A syndrome characterized by ascending weakness of the lower extremities associated with loss of deep tendon reflexes progressing to loss of bowel and bladder control has been described.146 The syndrome may begin with low back pain with radiation down to the legs or into the groin or anal area, followed over 1 to 3 weeks by the development of progressive weakness. Pathologically, the CMV radiculopathy is characterized by mononuclear infiltration of the cauda equina and lumbar sacral nerve roots together with CMV inclusions seen in the Schwann’s cells and epithelial cells, leading to axonal destruction. Untreated, this condition generally progresses to irreversible paralysis. Analysis of spinal fluid in patients with this syndrome characteristically shows elevation of neutrophils in the spinal fluid, sometimes as high as 3000 to 5000 cells/mm3. Although the protein is only mildly elevated and the glucose generally normal, some patients may have marked hypoglycorrhachia with glucose levels as low as 5 to 10 mg/dL. Treatment Ganciclovir at a dose of 5 mg/kg IV every 12 hours should be given for 14 to 21 days, followed by maintenance or lower dose IV ganciclovir. Alternatively, a dose of foscarnet at 90 mg/kg, adjusted for renal function, can be given IV twice daily for 2 to 3 weeks. Following induction with either ganciclovir or foscarnet, HAART therapy can then be given in the hope of sustained improvement in CD4+ count and maintenance of a response to treatment. In the pre-HAART era, CMV encephalitis responded relatively poorly to ganciclovir, with general survival rates of only 3 to 4 months.
Human Immunodeficiency Virus Encephalopathy HIV encephalopathy has been observed in up to 7.3% of the 144,184 persons reported with AIDS to the CDC between 1987 and 1991. Only 2% to 4% of adult patients will manifest signs of HIV encephalopathy at diagnosis. However, this number increases as the disease progresses, so it is estimated that between 7% and 14% of patients per year will manifest signs of HIV encephalopathy. It has been estimated that approximately one-third of patients with AIDS will exhibit dementia at the time of death.147 This figure is well corroborated in a large autopsy series of 450 cases spanning the years 1984 to 1999, in which pathologic evidence of HIV
W9418-12.qxd 11/3/03 2:58 PM Page 372
372
Section I
Introduction
A
B
Figure 12-10. A, Axial, mid-convexity, T1-weighted, nonenhanced MR image demonstrating both ventriculomegaly and sulcal dilatation (indicating cortical brain atrophy). There is no transependymal fluid to suggest elevated CSF pressure. These findings are indicative of both central and cortical brain atrophy which, in this case, is out of proportion to chronological age. Generalized atrophy is one of the features of primary HIV involvement of brain. B, Axial T2-weighted high convexity MR image through the centrum semiovale demonstrates evidence of increased T2 signal. These changes are relatively symmetric and are consistent with nonfocal white matter, which is one of the main imaging features of primary brain HIV. (Courtesy of Ronald Quisling, MD, University of Florida.)
encephalopathy was found in 28.7% to 38%, depending on the year of diagnosis. Interestingly, no change in the incidence of HIV encephalopathy was seen between 1996 and 1999, during the era of HAART treatment.148 Pathologically, the most frequent findings are brain atrophy characterized by sulcal widening and ventricular dilatation together with varying degrees of meningeal fibrosis. The most distinctive histologic feature of this condition is white matter pallor, chiefly seen in a periventricular distribution together with microglial nodules, diffuse astrocytosis, and perivascular mononuclear inflammation. HIV can be readily demonstrated in these nodules by immunohistochemical techniques. Clinically, HIV encephalopathy presents with altered mental status characterized by mental slowing frequently accompanied by evidence of subcortical dementia such as bradykinesia, postural instability, slow and clumsy gait, and altered muscle tone. Radiographically, the most common features are generalized cerebral atrophy together with widespread relatively symmetric hyperintense white matter abnormalities seen on T2 imaging that generally have a periventricular distribution. Examples are shown in Figures
12-10A and B. CSF studies are nondiagnostic. A mononuclear pleocytosis is seen in approximately 20% of patients with cell counts generally less than 50 cells/mm3.149 Approximately two-thirds of patients will have an elevated protein, generally to levels less than 200 mg/dL. HIV may be detected by a variety of techniques in the spinal fluid, but is not diagnostic for HIV encephalopathy. HIV encephalopathy is treated by standard HAART therapy. Patients with are also at increased risk of CNS infection due to a variety of other microorganisms such as M. tuberculosis, atypical mycobacteria, syphilis, and Listeria, and are at higher risk for infection with common communityacquired agents such S. pneumoniae.
Parasitic Infections Malaria Despite its eradication from North America and Europe, it is estimated that malaria infects over 2.5 billion people worldwide and causes between one and three million deaths each year. The disease is caused by one of four different
W9418-12.qxd 11/3/03 2:58 PM Page 373
Chapter 12
species of Plasmodium: P. vivax, P. ovale, P. malariae, and P. falciparum. Essentially all deaths and all serious infections, especially those with CNS involvement, are caused by P. falciparum. Malaria sporozoites are transmitted from the female Anopheles mosquito to the patient at the time the mosquito bites. Sporozoites are carried rapidly to the liver where they multiply in approximately 1 week to become tissue schizonts or the dormant hypnozoites. Infected liver cells then burst, releasing thousands of merozoites, each of which in turn infects red blood cells in the bloodstream. Continued asexual replication in the bloodstream through repeated cycles of maturation and rupture of red blood cells with release of merozoites eventually results in symptomatic infection. During this process, some of the parasites develop into sexual forms called gametocytes that produce no symptoms themselves but that may circulate for a prolonged period. It is the ingestion of these gametocytes that leads to the sexual reproduction cycle in the Anopheles mosquito, resulting in the motile sporozoites that invade the mosquito salivary glands and can be transmitted back to humans at the time of the next feeding. Malaria is widely distributed in developing countries, particularly Central America, South America, Africa, the Middle East, the Far East, and Indonesia. The reader is strongly urged to access the CDC’s website for the most up to date availability on the distribution and drug resistance of malaria, especially P. falciparum, on a country-by-country basis. Pathogenesis Essentially, all serious infections and CNS involvement with malaria, as mentioned previously, are attributable to P. falciparum. As the P. falciparum trophozoites mature in the red blood cells, they induce changes in the surface of the cell and in the flexibility of the cell membrane so that red blood cells tend to stick to capillary endothelium, and because of their increased rigidity, tend to become trapped in capillaries. As a result, in severe cases of malaria in which the parasitemia level is high, capillary vascular obstruction ensues, preventing delivery of oxygen and nutrients to the tissues. In the CNS, this process results in impaired consciousness and tendency toward seizures. Systemic manifestations of severe falciparum malaria include anemia, lactic acidosis, hypoglycemia, pulmonary edema, and even disseminated intravascular coagulation. In a series of 158 cases of falciparum malaria, 83% had fever and chills, altered sensorium was seen in 48%, jaundice in 27%, anemia in 75%, cerebral malaria in 45%, thrombocytopenia in 40.5%, and renal failure in 25%. Overall mortality was 20%, most of which resulted from a delay in diagnosis and treatment. Adult respiratory distress syndrome was also a significant complication.150 The diagnosis of malaria is made from examination of the blood smear, provided that the diagnosis is considered as a differential diagnostic possibility in a patient with altered
Central Nervous System Infections
373
consciousness and fever. Examination of the blood smear is highly accurate. Because the level of parasitemia in falciparum malaria tends to be high, generally greater than 5%, it is not hard to find the ring forms characteristic of the disease. A travel history is critically important to elicit, as is the history of whether the patient took prophylaxis for malaria. The location of the travel is particularly critical because P. falciparum malaria is typically resistant to chloroquine, and resistance to trimethoprim-sulfamethoxazole, mefloquine, and other agents have been observed in many parts of the world, particularly Southeast Asia. Because there is no latent form of P. falciparum in the liver, as there is for P. vivax and P. ovale, all cases of P. falciparum malaria should become clinically evident within a month after leaving an endemic area. However, hypnozoites of P. vivax and P. ovale may survive in the liver for 6 to 11 months, particularly if the patient did not take a full course of prophylaxis on returning from overseas. It should be kept in mind that cases of malaria have been transmitted by transfusion and IV drug abuse. Finally, one must never to rule out the possibility of endemic transmission within the United States despite “eradication” of the disease. Treatment In cases of severe malaria with CNS involvement, the patient must be treated as though he or she has falciparum malaria, regardless of the preliminary interpretation of the blood smear. Standard treatment for adults is a dose of quinine sulfate at 650 mg orally every 8 hours for 3 to 7 days plus doxycycline 100 mg twice a day orally for 7 days. If the patient is intolerant of doxycycline, the quinine may be followed with pyrimethamine sulfadiazine, three tablets on the last day of quinine treatment. Alternatives include quinine followed by clindamycin at a dose of 900 mg three times daily for 5 days or followed by mefloquine 1250 mg as a single dose; quinine followed by halofantrine at 500 mg every 6 hours for three doses, repeated 1 week later, or followed by atovaquone at a dose of 1000 mg daily for 3 days plus proguanil 400 mg daily for 3 days after quinine treatment. If parenteral treatment is required, quinidine gluconate at a 10 mg salt/kg loading dose up to a maximum of 600 mg in normal saline is infused slowly over 1 to 2 hours, followed by a continuous infusion of 0.02 mg/kg/min until the patient is able to begin oral treatment.151 For P. vivax and P. malariae, a dose of 1000 mg chloroquine phosphate orally (600 mg base), followed by an additional 500 mg (300 mg base) dose 6 hours later and again on days 2 and 3 should result in a prompt therapeutic response. In the case of P. vivax and P. ovale, the treatment with chloroquine must be followed by primaquine phosphate 15.3 mg base (26.5 mg) phosphate salt (per day orally for 14 days) to clear the hepatic forms of the organism. With treatment, almost all patients with CNS malaria recover completely if they survive the acute episode. However, approximately 12% may have lasting neurologic complications.
W9418-12.qxd 11/3/03 2:58 PM Page 374
374
Section I
Introduction
Cysticercosis Neurocysticercosis in adults results from infection of the brain with the larval cysts of the cestode, T. solium, the pork tapeworm. Human infection occurs by two different mechanisms. Ingestion of the eggs leads to embryonization of the eggs and penetration of the intestinal wall with hematogenous transport of cysticerci to many different tissues, primarily muscle and brain, where they encyst and remain potentially infectious for long periods. Alternatively, ingestion of undercooked pork can result in ingestion of the cysticerci by humans. In this instance the cysticerci mature into the typical tapeworm, which attaches to the intestine and may grow to lengths of up to 3 m and live for as long as 25 years. During this time it produces egg-filled segments, called proglottids, which are excreted in the feces. It is ingestion of the eggs from these proglottids that leads to neurocysticercosis, occasionally even by autoinfection from the patient’s own intestinal tapeworm. Cysticercosis is widespread in areas such as Mexico, Central America, South America, Africa, Southeast Asia, India, the Philippines, and Southern Europe. Symptomatic neurocysticercosis results from the enlargement of the cysticercal cysts in the brain parenchyma over months to years. Patients in the United States may present with seizures due to neurocysticercosis after having left endemic areas for up to 30 years. Symptoms of disease depend on the location and size of cysts and most frequently include seizures and headaches. If the cysts block the flow of cerebrospinal fluid, signs of increased ICP, such as headache, nausea, vomiting, changes in vision, dizziness, ataxia, and confusion may result. When the cysts are located in the meninges, chronic meningitis may be seen. An unusual form called “racemose” cysticercosis due to the proliferation of cysts at the base of the brain can result in severe disease including mental deterioration and death. Intraspinal cysts are often found as well and may produce symptoms of spinal cord compression depending on their location and size. The disease is readily diagnosed with either CT or MRI, in which multiple cysts of varying sizes and stages are demonstrated. Serology is available with sensitivity as high as 94% in patients with multiple cysts, but significantly lower in those with single cysts or old calcified lesions. Serologic studies are generally available only through the CDC or at some national commercial laboratories. Treatment Based on MRI, CNS lesions can be classified into active and inactive neurocysticercosis (Fig. 12-11). Patients with inactive parenchymal neurocysticercosis have no evidence of viable or degenerating parasites and, thus, antiparasitic drugs have no useful role. However, these patients are at high risk of seizures, and standard anticonvulsive therapy with phenytoin, phenobarbital, or carbamazepine is indicated.
Figure 12-11. Axial, contrast-enhanced, mid-convexity CT scan demonstrates multicentric areas of focal brain abnormalities. Some of the lesions have lucent centers where an intracystic hyperdense object is evident. These areas represent neurocysticercosis in its early phase. The visualized objects are the scolices and indicate that organism is still viable. In the left anterior basal ganglia there is a calcified lesion which represents the chronic, nonviable state of neurocysticercosis. Other lesions of similar age are seen in other portions of brain. There is no hydrocephalus. The combination of both new and old lesions is typical of neurocysticercosis. (Courtesy of Ronald Quisling, MD, University of Florida.)
Patients in whom hydrocephalus develops need to be treated with ventricular peritoneal shunting. Virtually all patients with active neurocysticercosis have seizures that must be treated with anticonvulsants. Probably the majority of these patients can be treated symptomatically and observed with MRI because the cysticerci typically undergo complete degeneration over a 1- to 2-year period. This process results in either calcified inactive cysticerci, which will continue to require anticonvulsants or, in a majority, a normal MRI in which case anticonvulsants may be tapered as long as the patients remain seizure free. Antiparasitic treatment can be given with praziquantel at doses of 50 to 100 mg/kg/day for 15 to 30 days or albendazole at 10 to 15 mg/kg/day for 8 days. Although these agents apparently kill the cysticerci, their controlled trials have not shown any clinical benefit over symptomatic treatment alone.152–154 The main adverse side effect of praziquantel is worsening neurologic function, for example, headaches, dizziness, seizures, and increased ICP, probably as a result of an increase in the host inflammatory response to the dying parasites. Standard treatment of ventricular neurocysticer-
W9418-12.qxd 11/3/03 2:58 PM Page 375
Chapter 12
cosis has been the surgical removal of cysts, which block the flow of CSF. Recent studies have found fewer shunt failures when such patients are treated with antiparasitic drugs.155,156 Cysticercosis involving the basilar cisterns is associated with a prominent inflammatory arachnoiditis and can be complicated by both vasculitis, resulting in lacunar infarctions, and invasion of the cysticerci into larger vessels, resulting in strokes. Thus, some authors have recommended the addition of corticosteroids in the treatment of patients with cisternal cysticercosis.157 As with parenchymal disease, there is no definite clinical evidence that the addition of antiparasitic drugs improves the outcome compared with treatment of symptoms alone.
Central Nervous System Infections
375
develop as the parasite invades tissue. These structures also contain numerous protoscoleces. Because of the risk of spilling the contents in the cysts, which will result in potentially fatal local proliferation of the disease and possible dissemination, it is critical to remove the echinococcal cyst intact. Many surgeons use hypertonic saline or formalin injection to kill the protoscoleces prior to removal of the cyst. Although this technique is definitely safer for the patient, it literally destroys the morphology of the scoleces and renders histologic diagnosis impossible. Drug treatment with albendazole has been reported to show some activity both in animals and in humans.161,162 Strongyloidiasis
Echinococcus Echinococcal disease is caused by tapeworms, which commonly infect dogs, cats, wolves, and other carnivores. It is found worldwide but is particularly common in countries surrounding the Mediterranean, parts of East Africa, Russia, and South America. There is very little echinococcal disease in the United States, but it is important to recognize that bears, foxes, and wolves in Canada and Alaska are commonly infected with this parasite. Disease is produced when an egg from an infected animal is ingested and the oncosphere within it is activated, penetrates the gut wall, and travels via veins or lymphatics to various tissues in the body where cysts form. The most commonly involved areas are the liver in 50% to 70% of cases and the lungs in 20% to 30% of cases. The CNS is only involved in approximately 2% of cases.158 Four species infect humans; however, Echinococcus granulosus and E. multilocularis account for the vast majority of cases. The disease presents most frequently as solitary cysts of the liver and, in the case of CNS involvement, as a solitary cyst that slowly enlarges. Depending on the size and location of the cyst, the patient may be asymptomatic. However, as the CNS lesion enlarges, symptoms arise from the local effects of the lesion itself or secondary to increased ICP. Patients thus present with headache, papilledema, nausea and vomiting, seizures, hemiparesis, dysarthria, and cranial nerve palsies. The diagnosis may be suspected from the radiographic appearance of the cyst itself seen on CT or MRI. Classic radiographic features include a sharp, spherical border lacking a rim of enhancement or surrounding edema. Serologic testing is available from the CDC. Sensitivity of these tests varies from 60% to 90% so a negative test result does not absolutely rule out the disease.159,160 History of travel to or living in an endemic area, especially with exposure to sheep, also increases the likelihood of the diagnosis of echinococcus. In the case of E. granulosus, protoscoleces develop slowly on the inside of the capsule wall by budding while E. multilocularis forms aggregates of small grapelike cysts that
Strongyloides stercoralis is a small nematode with free-living forms found in soil, while parasitic forms, for example, the adult female, live within intestinal crypts in the duodenum and the jejunum. The eggs released from these organisms normally mature in the soil to produce more larvae that can directly penetrate the skins of humans and other mammals. Burrowing through the skin, the larvae enter lymphatics and, ultimately, the venous system where they are carried to the pulmonary capillaries. Here, they migrate out of the blood vessels into alveoli, up the airways and then down through the esophagus to reach the small bowel. Normally, the adult worms bore into the mucosa and produce eggs that pass out with the feces. For reasons not fully understood, in some patients the eggs hatch before being excreted and the larvae burrow through the intestinal wall and perianal skin to reinfect the patient. The phenomenon has been observed in World War II veterans who were in POW camps on the pacific front up to 30 years after their return to the United States. However, in immunosuppressed patients or patients on high-dose corticosteroids or with HIV, a cycle of autoinfection can reach such proportions that life-threatening pulmonary disease can develop, which is characterized by pulmonary infiltrates and adult respiratory distress syndrome. As part of this hyperinfection syndrome in immunocompromised patients, CNS involvement may be manifested by headache, altered mentation, meningismus, focal or generalized seizures, and motor weakness. The unique aspect of this process is that meningitis due to E. coli and other gram-negative enteric organisms may be observed. It is believed that enteric organisms are carried either on the larvae or within the gut of the larvae as they migrate through the tissues and thus causing meningitis when the CNS is invaded. Despite involvement by large numbers of migrating parasitic organisms, eosinophilia is almost never seen because of the setting of immunosuppression, (usually corticosteroids). The diagnosis can be readily made by a parasite examination of the stool. Treatment with thiabendazole at a dose of 25 mg/kg twice daily for 10 days has been effective for the treatment of this condition.
W9418-12.qxd 11/3/03 2:58 PM Page 376
376
Section I
Introduction
Toxocariasis Toxocara canis and T. cati are nematodes that infect the intestines of dogs and cats respectively. As a result of this infection in domestic animals, the eggs of these organisms are distributed widely in the soil to which humans may be exposed. Human infection occurs when eggs are ingested and hatch in the small intestine. Larvae then migrate through the intestinal wall and into various tissues into the body and most often manifest as visceral larvae migrans (VLM). Symptoms include abdominal pain, hepatomegaly, anorexia, nausea, vomiting, lethargy, behavioral changes, pneumonia, cough, wheezing, lymphadenopathy, and even fever. The hallmark of the disease is striking eosinophilia. When eosinophilia is seen in small children between the ages of 2 and 4 years, it readily suggests this diagnosis. Serologic surveys in the United States in the early 1970s showed that between 4.6% and 7.3% of children in varying parts of the United States have been infected with T. canis.163 A similar survey in New York in the 1980s showed a prevalence of 5.1% of female and 5.7% of male school children seropositive for T. canis.164 More recent serologic data from Ohio showed a seroprevalence of T. canis antibodies in 12% of children aged 4 to 10 years. No adverse outcome, as measured by intelligence testing, were noted once baseline differences in intelligence were taken into account.165 CNS involvement may occur in patients with VLM most commonly presenting as encephalopathy with seizures.166 Other manifestations include meningoencephalitis, transverse myelitis, and psychiatric disturbances. Treatment is diethylcarbamazine 2 mg/kg orally three times daily for 10 days or albendazole 400 mg orally twice a day for 5 days. Steroids are indicated for ocular disease and may be necessary for severe lung, heart, or CNS involvement. Differential diagnosis includes other migratory eosinophilia associated diseases such as Ancylostoma, Gnathostoma, or Spirometra.
Syphilis Syphilis is caused by the spirochete Treponema pallidum, belonging to the family Spirochaetaceae. The organisms are thin, tightly coiled bacteria that exhibit a characteristic undulating movement under direct dark-field observation. The clinical manifestations of the different stages of disease caused by T. pallidum have been recognized since the 1500s and controversy persists to this day as to whether it was transported de novo to Europe from contact with the indigenous population of South and Central America or whether it originated from closely related species of Treponemes found in the Near East. When serologic surveys first became available at the end of the 19th and the beginning of the 20th centuries between 8% and 14% of adults in major cities such as Paris, Berlin, and New York had positive test results.167 Before the availability of penicillin, cases of the primary and
secondary syphilis reached a high of approximately 100,000 per year in the 1940s, and fell rapidly with the introduction of antibiotics after World War II. Perhaps coincident with the introduction of oral contraceptives and the “sexual revolution,” the incidence of primary and secondary syphilis increased from the range of approximately 20,000 cases per year with a gradual rise to as high as 40,000 cases a year in the late 1980s during the early stages of the HIV epidemic. Since 1990, new cases of primary and secondary syphilis have dramatically declined to 6000 to 7000 cases per year between 1999 and 2000.168 Although syphilis can be transmitted from any skin or mucous membrane lesion containing infectious spirochetes, the vast majority of cases are transmitted through genital sexual contact. The initial manifestation is the chancre, which begins at the site of inoculation as a painless papule and rapidly ulcerates into a relatively painless ulcer that is dark-field positive. Regional lymphadenopathy frequently accompanies primary chancre, and the chancre generally heals spontaneously in 3 to 6 weeks. For a period of two to eight weeks following the chancre, a systemic illness known as secondary syphilis may develop in as many as 25% of untreated patients. The classic manifestations are macular papular, papular, or pustular skin lesions, which are distributed over the entire body and include the palms and soles, characteristic findings seen in very few other conditions. Genital ulcerations develop in between 20% to 35% of those with clinically evident secondary syphilis. Approximately two-thirds of patients will develop systemic manifestations including fever, malaise, pharyngitis, anorexia, weight loss, arthralgias. Some evidence of CNS infection develops in 8% to 40% of patients, including meningitis, headache, decreased vision, tinnitus, vertigo, and even cranial nerve involvement. Direct dark-field examination of the CSF in these patients rarely reveals spirochetes. Using rabbit inoculation, Lukehart found that 12 of 40 patients (30%) with primary and secondary syphilis had viable treponemes in CSF.169 Like the primary chancre, secondary syphilis resolves spontaneously without antibiotic therapy. After secondary syphilis, the patient is once again is asymptomatic until manifestations of tertiary syphilis appear 5 years to longer than 40 years after the primary infection. Without treatment, at least 25% of infected patients will develop some manifestations of tertiary syphilis. Clinically, tertiary syphilis is divided into three general types: neurosyphilis, cardiovascular, and gummatous syphilis. The classical manifestation of cardiovascular syphilis is an aneurysm of the ascending aorta, which may dissect down into the aortic valve ring with distortion of the valve cusps and resultant aortic insufficiency. Pathogenically, this is due to syphilitic inflammatory involvement of the vasa vasorum which leads to destruction of the elastic tissue and saccular dilatation of the aortic route. Cardiovascular syphilis will develop in approximately 10% of untreated patients.
W9418-12.qxd 11/3/03 2:58 PM Page 377
Chapter 12
Syphilitic gummas are progressive granulomatous tumorlike lesions primarily involving skin, mucous membranes, and bone but which can develop in any organ in the body including the brain. Localized findings range from small superficial nodules to large radiating lesions. Symptoms depend on the location of the lesions. Neurosyphilis is generally divided into four syndromes. Syphilitic meningitis most often occurs within the first two years after infection and results from small vessel arteritis in the meninges, which leads to typical symptoms of headache, nausea, and vomiting seen in approximately 90% of patients. In addition, up to 45% of patients with syphilitic meningitis may have cranial nerve palsies. Seizures have been reported in 17% and fever occurs in less than 50%.170 This condition tends to occur as part of secondary syphilis and, therefore, may resolve even without treatment, with the exception of cranial nerve abnormalities that may not fully recover, particularly eighth cranial nerve lesions. The CSF white blood cell count is almost invariably abnormal, but there is only a mild decrease in CSF glucose. Although there is some overlap in terms of symptoms, meningovascular syphilis presents with findings of meningitis together with focal neurologic findings due to syphilitic arteritis. The peak incidence of this condition is approximately seven years after acquisition of syphilis and accounts for approximately 10% to 12% of patients with CNS involvement.171–173 Patients with meningovascular syphilis generally present with several weeks to months of prodromal signs and symptoms such as headache, vertigo, personality changes, behavioral changes, insomnia or seizures, and stroke-like neurologic deficits most frequently involving the distribution of vessels in the territory of the middle cerebral artery followed by that of the basilar artery. Thus, while the distribution of strokes in such patients may be similar to that of the patient with atherosclerotic disease, the occlusive symptoms develop gradually over a period of time in meningeal vascular syphilis as opposed to suddenly in patients with atherosclerotic strokes.131,170,174 The majority of neurologic manifestations of tertiary syphilis are known as parenchymatous neurosyphilis and include the two classical syndromes, tabes dorsalis and general paresis. In contrast to the pathogenesis of meningovascular syphilis or syphilitic meningitis, these syndromes result from progressive neuronal destruction rather than ischemic damage from vasculitis. General paresis, also known as general paralysis of the insane, is a chronic progressive meningoencephalitis with a peak incidence 10 to 20 years after acquisition of syphilis. It generally presents with a gradual deterioration of mental functioning characterized by difficulties in concentration, irritability, and deficits of higher cognitive function. As the condition progresses, these manifestations become more obvious and symptoms may mimic psychiatric disease. Difficulties with motor control then develop with a loss of facial muscle and extremity tone, loss of fine motor control, and the development of tremors
Central Nervous System Infections
377
and dysarthria. Subsequently, patients may have seizures, loss of bowel and bladder control, and paralysis. Untreated, the disease follows a progressive or subacute course over 3 to 4 years. Pathologically there is diffuse cortical atrophy, dilatation of the ventricles, and neuronal dropout with accompanying glioses. Spirochetes can be demonstrated in 25% to 40% of patients with silver stain. The diagnosis is established with a combination of the clinical presentation, positive serologic test results, and elevated CSF white count and protein. Tabes dorsalis, like general paresis, results from progressive neuronal degeneration. It has a peak incidence that is slightly later than that of paresis, approximately 15 to 20 years after infection, and the progression of this condition is somewhat slower than that of paresis. The classical early symptom is “lightning pains” in the distribution of nerve roots.170 These pains are described as lancinating, lasting for minutes to hours, and most often involve the lower extremities. In 10% to 20% of cases, patients with tabes may also present with episodic attacks of abdominal pain. In addition to pain, some patients experience parethesias, which also occur episodically. Ultimately, patients experience progressive loss of sensation and proprioception, particularly in the lower extremities. As a result, the patients exhibit a characteristic broad-based shuffling gait and Charcot joints may develop. Muscular atrophy develops in approximately 20% of patients. The Argyll Robertson pupil is one in which one or both pupils constrict with accommodation, but do not react directly to light; this is a characteristic feature of tabes dorsalis. Pathologically there is atrophy of the posterior columns of the spinal cord with inflammatory infiltrates and loss of neurons. In contrast to paresis, it is unusual to be able to stain the spirochete in nerve tissue. The diagnosis is readily made from the characteristic neurologic findings together with positive serologic test results. However, CSF leukocytosis is observed in only 50% of patients, and protein elevation is seen in approximately 53% of patients, as well.170 Finally, gummas may arise in almost any part of the CNS, most often associated with the pia mater and consisting of rubbery masses varying in size from several millimeters to centimeters.175 Serologic Testing Laboratory diagnosis of syphilis depends on the stage of the disease and clinical manifestations. Patients with lesions on moist skin or mucous membranes during either primary or secondary syphilis can usually be diagnosed by the demonstration of treponemes on dark-field microscopy. While treponemes can be demonstrated in dry lesions or lymph nodes by biopsy or saline aspiration, the yield is considerably lower. Serologic tests for syphilis are generally divided into two different types: treponemal tests and nontreponemal tests. The nontreponemal tests are IgG or IgM antibodies that are
W9418-12.qxd 11/3/03 2:58 PM Page 378
378
Section I
Introduction
directed at lipid antigens, which were originally extracts of cardiolipin from beef heart or liver. Currently, a standardized mixture of cardiolipin, cholesterol, and lecithin, which has fewer false-positive test reactions, is used and forms the basis of today’s standardized tests. The most common of these are the classic VDRL, in which agglutination of the cardiolipin, cholesterol, and lecithin antigen is done on a slide using heated serum, the rapid plasma reagin (RPR) card test, the automated reagin test, or the toluidine red unheated test. Because of the stringency of the technical requirements for the test, the VDRL is generally performed only on CSF and serum screening tests are generally done with the RPR and its variants. Specific treponemal tests for syphilis include the fluorescent treponemal antibody absorption (FTA-ABS), Treponema pallidum immobilization, microhemagglutination Treponema pallidum, hemagglutination treponemal test for syphilis (HATTS), as well as newer tests such as the Serodia Treponema pallidum particle agglutination (TP-PA) and the Captia Syphilis-G-enzyme immunoassay. These tests use treponemal antigens as the target antigen and use either indirect immunofluorescence, ELISA, or hemagglutination formats. It is important to understand that the serologic response to syphilis increases gradually over the course of the primary infection, so that when the chancre is first observed no more than 10% to 20% of patients may be seropositive by any method. However, this will increase with the duration of the chancre during primary infection to approximately 70% for both the treponemal and nontreponemal tests by the time the chancre heals. During secondary syphilis serologic test results, whether treponemal or nontreponemal, are positive in almost 100% of patients. Early treatment of the primary infection should render the patient seronegative within a year. Treatment of syphilis during the secondary and latent stages will generally result in a significant decrease in the titer of the nontreponemal tests. These test results should be negative within 1 year in a patient treated for primary syphilis or within 2 years for a patient treated for secondary syphilis. Patients who remain seropositive by nontreponemal tests after treatment probably have either persistent infection or the so-called biologic “false-positive” result sometimes seen with HIV infection.176,177 In general, once the treponemal test results become positive they remain so for life, even if the patient has been successfully treated. Serologic tests can be used to diagnose neurosyphilis during the latent and late, latent stages by testing the spinal fluid for VDRL antibodies. With the exception of rare false-positive results, possibly resulting from blood contamination, a positive CSF VDRL is diagnostic of neurosyphilis. Unfortunately, the CSF VDRL has been reported to be negative in up 30% of patients with neurosyphilis, thus a negative test result does not exclude the diagnosis. Unfortunately, one cannot use the more sensitive treponemal antibody tests with CSF to diagnosis neurosyphilis because these antibodies cross the blood-brain barrier and are present essentially in all patients who are seropositive for
treponemal tests for syphilis. Thus, they do not provide any more information than testing serum alone. For example, in one study the CSF-FTA-ABS test result was positive in 48 patients, 15 of whom had clinical neurosyphilis.178 While one can resort to sophisticated methods of CSF analysis, such as levels of CSF treponemal antibody compared with serum levels adjusted for changes in the blood-brain barrier by using serum/CSF albumin ratios, and demonstrating a significantly higher than expected CSF level of specific treponemal antibody, it is probably safer to treat patients who have a positive serum test result if they have any clinical signs of neurosyphilis. If the patient has no clinical manifestations of neurosyphilis and preventive treatment for latent neurosyphilis is being considered, treatment should be based on the presence of an abnormal number of white blood cells in the spinal fluid rather than trying to fine tune the serologic diagnosis. Recent studies have been published using PCR showing that DNA from T. pallidum can be detected in CSF. However, it is unclear whether a positive test result means that the patient has to be treated for latent neurosyphilis or that a negative test result excludes the diagnosis.169,179–183 Treatment The preferred treatment for all manifestations of neurosyphilis is intravenous aqueous crystalline penicillin G 12 to 24 million units per day given in six divided doses for 10 to 14 days. Alternatively, 2.4 million units of procaine penicillin G can be given intramuscularly (IM) together with 500 mg per day of probenecid four times a day for 10 to 14 days. In penicillin-allergic patients, a dose of doxycycline 200 mg orally each day for 21 days, or ceftriaxone 1 g IM or IV for 14 days has been recommended. However, treatment failures have been documented with ceftriaxone, especially in patients with HIV. Treatment of syphilitic meningitis or meningovascular syphilis is generally very good with the exception of focal cranial nerve abnormalities sometimes associated with syphilitic meningitis and larger ischemic defects associated with meningovascular syphilis. For patients with tabes dorsalis or paresis, improvement approaching cure is relatively uncommon and, in fact, a majority of patients actually continue to progress despite “adequate” penicillin treatment.184 Patients with asymptomatic neurosyphilis, that is, positive syphilis serologic test results together with CSF abnormalities, appear to respond very well to treatment. In one study, 89% of 454 patients who initially had a minimum of 10 leukocytes/mm3 of CSF had normalized their cell counts at a one year follow-up, as had 69% of those with abnormal protein before treatment.185 Because patients with primary and secondary syphilis are curable with standard treatment of benzathine penicillin G 2.4 million units IM weekly for three weekly doses, the question of proper treatment always arises when an asymptomatic patient is found to have a positive VDRL result, RPR
W9418-12.qxd 11/3/03 2:58 PM Page 379
Chapter 12
test result, or other screening test for syphilis. The problem is that this regimen does not reliably provide CSF levels in excess of 0.018 mg/mL of CSF in all patients, which is believed to be necessary to kill spirochetes within the CNS.186–188 To rule out neurosyphilis in an asymptomatic patient, examination of the spinal fluid is required and if no abnormalities of protein or cells are found, weekly treatment can be safely administered. It is also recommended that all patients found to have serologic evidence of syphilis, assuming that false-positive test results such as those due to pregnancy and other intercurrent illness can be ruled out, should have serologic testing for HIV because the ability to eradicate syphilis is considerably lower in patients with HIV and retreatment may be necessary.
Lyme Disease Lyme disease was first recognized in the United States in the early 1970s when Dr. Allen Steere at Yale University investigated an outbreak of juvenile rheumatoid arthritis in the small towns of Lyme, Old Lyme, and East Haddum, Connecticut. In the initial report, they found 39 children and 12 adults who had a classic, characteristic, remitting, relapsing oligoarticular arthritis generally with onset in the summer or early fall. All of these patients lived in rural areas and half of the patients lived on two adjacent country roads. In addition, 13 of these patients had noted an unusual skin lesion an average of 4 months before the onset of the arthritis.189 Subsequent prospective studies then defined neurologic abnormalities such as Bell’s palsy, sensory radiculoneuritis, lymphocytic meningitis, and cardiac conduction abnormalities also associated with Lyme disease. These studies showed that at least a quarter of the patients remembered a tick bite at the site of the initial skin lesion, and based on examination of the actual tick from one of these patients, the vector was identified as a tick from the Ixodes family.190–192 The agent of Lyme disease was finally isolated by Dr. Willie Burgdorfer, from the Rocky Mountain Laboratory in Hamilton, Montana, when he was searching for evidence of Rocky Mountain spotted fever in ticks isolated from New York State. No rickettsia were found; however, spirochetes were seen in stains of the insect’s digestive tract.193 The earliest manifestations of Lyme disease occur at the site of the tick bite, beginning as a red macule or papule that then expands to an area as large as 10 to 15 cm with red outer borders and partial central clearing. The lesion develops as early as 3 days and as late as 30 days following the bite and generally lasts 3 to 4 weeks. It is most commonly located on the thigh or groin and develops in approximately 80% of patients.194–198 Dissemination of the spirochete occurs during the development of this initial lesion, and many patients will have multiple secondary annular lesions that are similar to the primary site. Systemic symptoms of fatigue, lethargy, and malaise along with generalized lymphadenopathy, meningis-
Central Nervous System Infections
379
mus, encephalopathy, migratory musculoskeletal pain, splenomegaly, sore throat, and cough may develop in varying degrees during this early disseminated phase.195 For the most part, the systemic manifestations as well the erythema chronicum migrans lesions themselves usually resolve without treatment in 3 to 4 weeks. Untreated, however, the Lyme disease spirochete becomes sequestered and persists in certain tissues such as the nervous system, joints, heart, and even the skin. In some patients with meningitis, particularly in Europe where the disease is caused by a different species of spirochete, significant neurologic abnormalities develop including cranial neuritis that can present as a isolated facial palsy (motor and sensory), radicular mononeuritis multiplex, or myelitis alone or in varying combinations. Although these patients may have some neck stiffness on extreme flexion, typical Kernig’s and Brudzinski’s signs are not present. Patients may complain of excruciating headache as well as severe musculoskeletal pain. Early on, examination of the spinal fluid may be normal but patients generally develop a lymphocytic pleocytosis of approximately 100 cells/mm3 with a normal glucose level. It is during this early dissemination stage that cardiac manifestations are generally seen, usually consisting of atrial-ventricular block with varying degrees of other forms occasionally noted including complete heart block, which rarely persists for more than a week and generally does not require the insertion of a pacemaker.199–201 Occasional patients have been described with osteomyelitis, myositis, panniculitis, eosinophilic fasciitis, conjunctivitis, or even deeper involvement of the orbital structures including panophthalmitis and choroid retinitis with exudative retinal detachment or interstitial keratitis. The third stage of Lyme disease is characterized by arthritis, which develops in approximately 60% of untreated patients. Symptoms include intermittent attacks of pain particularly involving the large joints, such as the knee, in an asymmetric pattern. Attacks generally last weeks to months followed by periods of remission. Joint fluid counts range from 500 to 100,000 cells/mm3 with a high percentage of polymorphonuclear leukocytes. Even untreated, this condition resolves gradually over a period of years. The late manifestations of CNS involvement of Lyme disease generally develop a year or more after the onset of illness and generally do not improve spontaneously. In both North American and European forms of this disease, persistence of the Borrelia burgdorferi spirochete has been demonstrated in CSF and in brain parenchyma up to 9 years after the onset of illness.202–204 Symptoms of late progressive Lyme encephalomyelitis may develop either acutely or gradually, and once started they worsen progressively over months to years. Progression may be gradual or stepwise characterized by sudden deterioration and only partial improvement between episodes. The most common neurologic symptoms are speech abnormalities, limb weakness,
W9418-12.qxd 11/3/03 2:58 PM Page 380
380
Section I
Introduction
gait difficulties, ataxia, bladder dysfunction, visual changes, hearing loss, and poor memory and concentration.203,205–216 Headaches, nausea, vomiting, and neck stiffness are reported but occur less often, while mental deficits such as behavioral changes and poor memory and concentration are common. More severe changes including confusion, disorientation, dementia, delirium, and somnolence can occur. Symptoms such as apraxia, monoclonus, hemiparesthesia, and visual field abnormalities have been reported. Spinal cord involvement is common, and myelitis may present as progressive paraparesis or quadriparesis that can be quite severe. Approximately 45% of patients have cranial nerve palsies, most frequently unilateral facial nerve involvement and bilateral hearing loss. Optic nerve involvement is also common and abnormalities of the ocular motor nerves as well as cranial nerves and have all been reported.193,217–219 Peripheral radiculoneuritis occurs in less than 10% of patients. Cerebrospinal fluid is abnormal in almost all cases. Generally, there is CSF pleocytosis, predominantly monocytic in the range of 100 to 200 cells/mm3, although levels as high as 2300 cells/mm3 have been reported. Protein concentrations are usually in the range of 100 to 200 mg/dL, but concentrations levels as high as 1800 mg/dL have been reported. The glucose concentration is generally normal to low. Oligoclonal bands specific for B. burgdorferi have been reported. Both EEG and CT abnormalities including infarcts in the internal capsule, thalamus, and lentiform nucleus, and hydrocephalus and cerebral atrophy have been reported.205,207,214–216,220,221 MRI shows additional lesions such as multifocal white matter abnormalities, infarcts, periventricular and subinsular cavities, and atrophy of the pons and medulla.207,214,216,220 MRI in patients with myelitis have shown diffuse or focal signal abnormalities in relevant parts of the spinal cord.207,220 In contrast to the dramatic and objectively documented neurologic abnormalities and syndromes seen in a small number of patients with Lyme encephalopathy, a certain number of North American patients have reported the development of a less dramatic, but nonetheless disabling CNS symptom complex. These patients complain of overwhelming fatigue, accompanied by loss of memory and concentration, and almost always without physical neurologic abnormalities. Psychological testing shows abnormalities of immediate and delayed memory, ability to learn information, attention span, concentration, problem solving, perceptual motor performance, and verbal fluency. Depression and irritability are frequently reported, but in general fatigue is the most overriding complaint. Many of these patients fit the definition of the chronic fatigue syndrome as defined by the CDC. Objective laboratory and radiographic signs of infection are generally absent. CSF pleocytosis is present in less than 5% of the cases, and CSF protein is elevated in only a minority of patients. Oligoclonal bands for Borrelia are
absent and patients with these complaints may or may not have antibodies to Lyme disease in the serum. Some of these patients have been reported to have MRI abnormalities such as focal areas of increased signal in deep cerebral white matter.215,222,223 In general, symptoms in these patients do not improve spontaneously and a variable number apparently do respond to antibiotic treatment, although it may require 6 months or more.215,224 In addition, a number of North American patients have developed a mild multifocal polyneuropathy distinct from the meningopolyneuritis of early disseminated Lyme disease as a manifestation of late Lyme disease. Intermittent tingling and paresthesias of the extremities are the most common symptoms, occurring in approximately 50% of patients with this form of late Lyme disease. The onset is generally eight months to several years after the initial infection. The symptoms are usually distal, may be symmetric or asymmetric, and can involve both arms and legs. About one-quarter of patients present with carpal tunnel syndrome or develop it at some point. Radicular pain occurs in 25% to 50% of those with this syndrome and is intermittent, asymmetric, and multifocal, typically radiating from the spine into the limbs or trunk.73,215,224–226 Sensory changes such as mild stocking and glove distal sensory loss, as well as distal asymmetrical or truncal sensory loss also occur. 73,215,224–226 Objective evidence of organic disease is much more common in these patients than in those reporting symptoms of chronic fatigue, with up to 83% of patients having electromyographic abnormalities demonstrable particularly among those with distal paresthesias. In addition, CSF abnormalities, mostly in the form of increased CSF protein concentration and intrathecal antibody synthesis specific for B. burgdorferi, are found in up to 70% of patients with these symptoms.215,225 Treatment with antibiotics does improve the paresthesias and electrophysiologic conduction abnormalities, but may require 3 to 7 months.75,215,224–226 Improvement among patients with radicular pain is less frequent and is only seen in about 50%.227 The diagnosis of Lyme disease can be made with reasonable assurance by observation of typical skin lesions of erythema chronicum migrans together with a well-documented history of a tick bite. However, during this early stage of the disease, only 30% to 40% of patients will have a positive serologic test result for Lyme disease in an acute serum specimen, and only 60% to 70% of these patients will have positive results in the convalescent sera 2 to 4 weeks later. In general, serologic testing is done using an ELISA to screen for the presence of antibody, together with confirmation by Western blot assay. Both IgG and IgM antibodies are formed. However, persistence of the IgM antibody alone in the absence of an IgG response after the first month of illness may represent a false-positive reaction. After the first 1 to 2 months of infection, over 90% of patients will have a specific IgG antibody response to the spirochete. It has been
W9418-12.qxd 11/3/03 2:58 PM Page 381
Chapter 12
noted that patients treated effectively early in the course of erythema chronicum migrans may never develop a humoral immune response, although cellular immunity may be demonstrated and may persist for years. Treatment of early dissemination and localized erythema chronicum migrans consists of doxycycline at a dose of 100 mg twice a day for 20 to 30 days or amoxicillin at 500 mg three times daily for 20 to 30 days, with some experts recommending the addition of probenecid at 500 mg three times daily to the amoxicillin regimen. Cefuroxime axetil at a dose of 500 mg orally three times daily for 2 to 3 weeks has also been recommended, as has erythromycin at 400 kg/mg/day in four divided doses for 2 to 4 weeks in children. For patients with arthritis, treatment with doxycycline or amoxicillin is extended to 1 to 2 months, and IV ceftriaxone at 2 g/day for 14 to 30 days is also recommended. For patients with early or late neurologic abnormalities, ceftriaxone at 2 g/day IV for 2 to 4 weeks is generally recommended. Alternatives include doses of penicillin G at 20 million units IV in four divided doses daily for up to 30 days, as well as doxycycline at 100 mg orally three times per day for 14 to 30 days. Treatment failures have been reported for all of these regimens and treatment may need to be repeated. Cardiac abnormalities are treated as for early infection in those patients with first-degree atrioventricular (AV) block; intravenous ceftriaxone or penicillin is used for higher degrees of AV block. In patients with neurologic manifestations of early disseminated Lyme disease, intravenous treatment with penicillin or ceftriaxone can lead to a mild Herxheimer-like reaction with worsening of pain and fever during the first 18 to 20 hours199,220,228 However, in general, meningismus, radicular pain, and systemic symptoms improve within days although residual fatigue, arthralgias, and muscular skeletal pain can persist for some time thereafter. Motor deficits improve more slowly, over 2 to 3 months, and sometimes never fully recover. Central nervous system abnormalities usually stop progressing and begin to improve slowly, but residual deficits may remain.220,228–232 CSF cell counts respond over the course of treatment but may not return to normal for several months, and the protein concentration falls even more slowly, remaining elevated for as long as a year in some patients. If the patient does not respond by the end of the second week, treatment should be extended for at least another 2 weeks. The severe abnormalities of late Lyme disease generally respond well to high dose penicillin, doxycycline or ceftriaxone.205,207,216,220,221,229,233 Altogether 80% to 90% of patients improve with IV cephalosporins, but recovery is slow and often incomplete, with little change occurring during the treatment itself and only developing over the subsequent weeks after treatment has stopped. At this point, use of steroids is not recommended because a controlled trial showed that patients responded as well to antibiotic treat-
Central Nervous System Infections
381
ment alone as to the combination of antibiotic with steroids.234
Miscellaneous Infectious Agents Rickettsia and Ehrlichia A wide variety of infectious agents invade the CNS and produce symptoms there as part of their systemic infection. For example, symptoms such as severe headache, photophobia, stiff neck, confusion, lethargy, restlessness, insomnia, and vertigo are relatively common, nonspecific symptoms seen in Rocky Mountain spotted fever and in patients with human monocytic or granulocytic ehrlichiosis. Rickettsia and Ehrlichia are small gram-negative, obligate intracellular organisms transmitted by tick bites. Despite the name, most infections with Rocky Mountain spotted fever occur in the inland parts of North Carolina, Virginia, and other southeastern States. As the name implies, the disease is also seen in the Rocky Mountain states such as Colorado and Wyoming. Infections generally follow the distribution of the main vectors Dermacentor andersoni in the west and Dermacentor variabilis in the east. Infections are limited to warmer months. The primary pathogenesis of rickettsial infection, particularly that of Rocky Mountain spotted fever, is a vasculitis characterized by endothelial swelling, necrosis, and mononuclear cell infiltrate. This is in contrast to the infiltrate of polymorphonuclear leukocytes seen in typical immunocomplex vasculitis. As a result of the variability of the vessels involved in the central nervous system, a wide range of neurologic findings are occasionally seen in Rocky Mountain spotted fever such as seizures, deafness, facial diplegia, gaze palsies, nystagmus, ataxia, dysphasia, transverse myelitis, neurogenic bladder, hemiplegia, and paraplegia, or quadriplegia. Approximately 25% of cases have significant alteration of consciousness and thus present as encephalitis. The diagnosis of Rocky Mountain spotted fever is fundamentally a clinical diagnosis, and depends on the history of a tick bite together with a compatible systemic illness with or without a rash. Although serologic diagnosis is highly accurate, it is not readily available in most institutions and therefore treatment must be started empirically. In adults and children older than 8 years of age, recommended treatment is a dose of doxycycline at 100 mg twice daily. For children younger than 8 years of age, chloramphenicol at 50 mg/kg/day in four divided doses is recommended because of the tooth discoloration associated with tetracyclines. Most patients show improvement within 2 days, but may require up 7 to 10 days in severe cases. Despite the use of modern antibiotic treatment the fatality rate is still in the range of 2% to 6%.235,236 Even after recovery, CNS abnormalities may persist in a significant number of patients with Rocky Mountain spotted fever, including intellectual defects, impaired fine motor skills, aphasia, and EEG changes.237
W9418-12.qxd 11/3/03 2:58 PM Page 382
382
Section I
Introduction
Human granulocytic ehrlichiosis and human monocytic ehrlichiosis are caused by two recently described rickettsia of the genus Ehrlichia: E. chaffeensis, which causes human monocytic ehrlichiosis (HME); and an unnamed Ehrlichia species associated with infection of granulocytes (HGE). Both diseases are spread by tick bites and HME is relatively common in the South and Southeast, while most cases of human HGE so far have been recorded from the upper Midwest. DNA studies of the agent of HGE show that it is mostly closely related to E. canis, but probably represents a different subspecies. Symptoms of ehrlichiosis include high fever, chills, severe headache, and myalgias, and may include confusion, disorientation, obtundation, and ataxia in some patients.238,239 CSF and laboratory studies are nondiagnostic, and as with Rocky Mountain spotted fever, treatment with doxycycline or chloramphenicol must be given empirically. Serologic confirmation of illness is highly accurate, but available only from state and other reference laboratories. Cat Scratch Disease The agent of cat scratch disease, Bartonella henselae, may on occasion involve the CNS. While the typical case develops impressive fluctuant localized adenopathy 1 to 7 weeks after inoculation of the infectious agent, systemic symptoms, such as fever and malaise, will develop in approximately one third of patients; 10% may actually have a variety of CNS symptoms including encephalopathy, retinitis, or Parinaud’s ocular glandular syndrome.240,241 Although most cases of cat scratch disease in the normal host are self-limited and resolve over 3 to 8 weeks, treatment is recommended and azithromycin is probably the treatment of choice since macrolides as well tetracyclines and rifampin appear to significantly reduce the levels of infectious organisms in tissue.242 Whipple’s Disease Whipple’s disease is a slowly progressive infection, which primarily involves the gastrointestinal tract of middle-aged men and presents with arthralgias, abdominal pain, fever, diarrhea, malabsorption, and weight loss. The disease is caused by Tropheryma whippleii. CNS involvement is one of the more frequent complications of this disease and can, on occasion, occur in the absence of recognized gastrointestinal or systemic involvement. The disease is most often diagnosed through biopsy of intestine or a mesenteric lymph node revealing the presence of PAS-positive material, with occasional bacilliform organisms being stainable. Pathologically, involvement of the brain is most often manifested by chalky, yellowish white, 1- to 2-mm nodules distributed diffusely throughout the cortical and subcortical gray matter of both the cerebrum and the cerebellum. The most frequently involved sites are the temporal, periventricular, and periaqueductal grey matter as well as the hippocampus, hypo-
thalamus, and basal ganglia. Histologically, the nodules are made up of microglia that stain strongly positive with PAS. The most frequent neurologic manifestations of Whipple’s disease are dementia, ophthalmoplegia, myoclonus, and hypothalamic dysfunction. The dementia progresses slowly and is characterized by memory impairment, confusion, personality change, paranoia, emotional instability, and depression.243–249 An unusual syndrome of synchronized eye movements and contractions of the jaw, known as oculomasticatory myorrhythmia, is sometimes seen and may be unique to Whipple’s disease.250,251 Although PCR diagnosis of Whipple’s disease is possible, it is available only on a research basis and, because of the wide range in the differential diagnosis of degenerative CNS diseases, tissue biopsy is generally required for diagnosis. CNS radiographic studies consistent with Whipple’s disease occurring in the setting of biopsy proven systemic Whipple’s disease are also sufficiently diagnostic. The initial treatment of CNS Whipple’s disease is ceftriaxone at a dose of 2 g IV twice daily, with streptomycin at 1 gram IM daily for 14 days, followed by trimethoprim/ sulfamethoxazole at 160 mg to 800 mg orally two to three times daily for at least 1 year.252,253 Intravenous trimethoprim/sulfamethoxazole at a dose of 960 mg twice daily for 2 weeks can be used in place of ceftriaxone.197 Two weeks of daily IM procaine penicillin G, one to two million units, together with streptomycin at 1 g IM daily has also been recommended,254 but the superior CNS penetration of ceftriaxone and trimethoprim/sulfamethoxazole would seem critical. Some clinicians continue treatment for 2 years to life. Unfortunately, the overall prognosis of CNS Whipple’s disease is not good. With antibiotic therapy progression can usually be stopped, but significant clinical improvement is limited 255 and relapses are frequent. Amebic Encephalitis A dramatic and almost uniformly fatal primary meningoencephalitis can be seen with infection caused by the free-living ameba Naegleria fowleri. These free-living amebae are widespread in nature, particularly in the upper surface layers of lakes in warm climates. In animal models it can be shown that the amebae are capable of invading through the nasal cavity along the blood vessels associated with the olfactory nerves and reach the frontal lobes and the surrounding meninges where they rapidly produce a highly necrotizing destructive encephalitis. Clinically, patients present with sudden onset of high fever, photophobia, headaches, with progression to obtundation relatively quickly. There is classically a history of swimming in warm, freshwater lakes. Because of the olfactory involvement, there may be alterations of smell or taste. Otherwise nonspecific symptoms such as confusion, irritability, restlessness, and seizures with rapid progression to delirium, stupor, and coma unfortunately are the rule. Examination of the spinal fluid shows leukocytosis with neutrophil predominance, low glucose
W9418-12.qxd 11/3/03 2:58 PM Page 383
Chapter 12
levels, and elevated protein. However, Gram stain and culture results are negative, but if the diagnosis is suspected one can examine the CSF unstained with a slide warmer to look for the typical, mobile, ameboid motion of the trophozoites. Radiographically the frontal lobe involvement is readily seen on MRI but the diagnosis is unfortunately often delayed because of the rarity of the condition and paucity of immediately diagnostic signs. Only a small number of patients have been reported to have survived and all received amphotericin B to which the Naegleria are susceptible in vitro.256–258 The optimal treatment regimen is not known for this condition and some authors recommend maximal systemic doses of amphotericin B with intracisternal amphotericin B and concomitant rifampin or tetracycline.255 Progressive Multifocal Leukoencephalopathy Progressive multifocal leukoencephalopathy (PML) is a progressive, severe, usually fatal demyelinating disease that is primarily seen in highly immunosuppressed patients. Before the HIV epidemic, the reported incidence was 1.5 cases per 10 million people. By 1987, with the HIV epidemic, the incidence had increased to 6.1 cases per 10 million persons259 and the risk of PML in HIV infected persons was estimated to be 80- to 100-fold higher than that of the general population.260 PML is caused by JC virus, a small nonenveloped DNA virus, closely related to two other polyomavirus, BK virus and the simian virus SV40. Both BK and JC were isolated in the early 1970s, JC from a patient with PML and BK from the urine of renal transplant patients.261,262 These viruses share approximately 75% DNA homology and 70% homology with SV40.263,264 Based on serologic studies, it appears that approximately 60% to 80% of adults in the United States have been infected with one or both and that infection with BK virus occurs in early childhood at approximately age 3 to 4 years while infection with JC virus occurs in the age range of 10 to 14 years.53,265,266 It is believed that both viruses remain clinically latent in renal tissue and other tissues in normal persons following primary infection. Approximately 30% to 50% of normal persons have detectable BK or JC virus sequences demonstrable by PCR and other methods in renal tissue obtained at surgery or autopsy.267–269 Asymptomatic viruria with BK virus has been detected in up to 3% of pregnant woman and increases significantly in immunosuppressed patient populations. JC virus has been detected in the blood of up to 22% of immunosuppressed patients in the absence of PML.270,271 Even normal brain tissue from immunologically normal patients with no evidence of PML has been found to contain JC virus DNA by sensitive PCR methods.272–275 Thus, it is believed that under conditions of immunosuppression, replication of JC virus in the CNS increases and depending on factors not understood at this time, may ultimately result in clinical disease.
Central Nervous System Infections
383
PML characteristically presents with progressive focal neurologic defects, primarily hemiparesis, visual field defects, and cognitive deterioration.276–288 As the disease progresses, aphasia, ataxia, and cranial nerve defects may occur, ultimately resulting in cortical blindness, quadriparesis, profound dementia, and coma. The average duration of survival is approximately 4 months, although a subset of 5% to 10% of patients, even with HIV, lives for a year or more. The diagnosis is made from a combination of the clinical picture with characteristic findings on CT or MRI (Figs. 12-12A and B) Definitive diagnosis requires brain biopsy and histologic examination, which should demonstrate characteristic cytopathic changes in the oligodendrocytes. These cells typically contain a homogeneous basophilic nuclear inclusion and virus can be demonstrated by in situ hybridization. Involvement of the oligodendrocytes ultimately leads to widespread, but patchy, multifocal demyelination which correlates with the patients clinical status. Virus particles can be demonstrated in the inclusions by electronmicroscopy.289,290 Not surprisingly, many studies have attempted to use PCR diagnostically to detect JC virus DNA by less invasive means, using CSF rather than brain biopsy. JC virus can be detected in the CSF of most patients with PML, whether immunosuppressed or not.271,291–293 Caution must be observed in interpreting these results. For example, McGuire and colleagues291 found that CSF of 24 of 26 patients with HIV-1 and PML had JC virus sequences present, but so did ten of 114 patients with HIV-1 without evidence of neurologic disease. In general, patients with PML are highly likely to have JC virus DNA in their CSF as compared with normal or immunosuppressed patients without PML.294–296 However, one recent study found JC virus DNA in the CSF of 12 of 12 patients with HIV-1 and no evidence of PML, but zero of 11 patients with multiple sclerosis (MS), if the test was made sufficiently sensitive.297 Despite the sensitivity of PCR testing for JC virus, occasional cases of PML have been observed in which no JC virus can be found even in brain tissue at autopsy.297 It is possible these cases may be due to the closely related polyoma virus, SV40.298 A relatively new agent for the treatment of CMV infection, cidofovir, does have activity against polyoma viruses in vitro and in animal models, and may be of benefit in the treatment of PML.299 There is no proven effective therapy for PML at this time. Initial reports of success with cytosine arabinoside (ara-C) were not supported by a recent clinical trial.300 In fact, a recent nonblinded, multicenter trial showed that one year survival with PML was 61% in patients with HIV-1 who received HAART with cidofovir, compared with 29% in those who received HAART without cidofovir.301 Thus although blinded studies are needed to confirm this observation, cidofovir should be strongly considered in the empiric treatment of PML, keeping in mind its significant renal toxicity. Other agents such as camptothecin and analogs, which inhibit topoisomerase I, are currently under consideration.
W9418-12.qxd 11/3/03 2:58 PM Page 384
384
Section I
Introduction
A B Figure 12-12. A, Axial, high convexity, T2-weighted FLAIR image demonstrates focally increased T2signal within the deep white matter bilaterally. The effects are confined predominantly to white matter with minimal cortical involvement and, as seen in (B), little or no enhancement. There is little or no mass effect. B, Axial, high-convexity, gadolinium-enhanced MR image demonstrates low-density changes (cystic leukomalacia) within the white matter of the frontal regions bilaterally. There is no significant enhancement of the lesion or the margins surrounding the lesion. These two findings are indicative of a white matter destructive process, which in this case, is related to progressive multifocal leukoencephalopathy associated with a JC (papova) virus encephalitis in an immunosuppressed host. (Courtesy of Ronald Quisling, MD, University of Florida.)
Spinal Epidural Abscess Epidemiology Spinal epidural abscess (SEA) is a rare clinical entity with a prevalence of between 0.2 and 2.0 per 10,000 hospital admissions.302–305 SEA is more common in the elderly, with a peak incidence during the 6th and 7th decades.302,303 The disease is rare among children and typically affects patients whose comorbid conditions predispose them to immunocompromised states.302–304,306 A recent increase in the incidence of SEA has been noted and is thought to be secondary to the aging population, increased numbers of IV drug abusers, the prevalence of AIDS, and the increased number of spinal surgical procedures.303,304 Risk Factors The majority of patients who develop an SEA have a recognizable risk factor.305,307 In one large review of the literature, encompassing 254 cases, the following frequencies of comorbid conditions were observed: osteomyelitis/diskitis 18%;
diabetes 16%; degenerative joint disease of the spine 11%; IV drug abuse 7%; alcoholism 4%, and cancer 4%.305 The spectrum of risk factors is fairly consistent between reports of other large case series.302–304,307 Certain comorbid states such as diabetes, chronic renal failure, and alcoholism result in an immunodeficient state that predisposes a patient to the development of a spinal abscess.304,308 Other risk factors have a more direct role in the development of SEAs. Diskitis and the bacteremia associated with IV drug abuse directly seed the epidural space with pathogens and result in the epidural infection.302 Pathophysiology The formation of SEAs can be spontaneous or secondary to direct inoculation of pathogen. The most common cause of the spontaneous variety is hematogenous spread from infections of the skin, respiratory tract, urinary tract, or oral cavity.302 Other causes of spontaneous abscess formation include extension from pre-existing diskitis/vertebral osteomyelitis or extension of a paraspinal abscess.302 Secondary causes include postoperative infections (16% of all
W9418-12.qxd 11/3/03 2:58 PM Page 385
Chapter 12
SEAs) and infections associated with epidural anesthesia catheters.302 Some series suggest that blunt spinal trauma precedes the development of SEA formation in 15% to 35% of cases.302,309 The authors propose that the trauma results in an epidural hematoma that then becomes secondarily infected.302 A relationship exists between the underlying cause of the epidural abscess and its location within the spinal canal.302–304 The majority of spontaneous epidural abscesses are the result of hematogenously spread bacteria and the resultant SEA is usually located posteriorly within the spinal canal.302–304 Conversely, epidural abscesses secondary to preexisting vertebral osteomyelitis are confined to the anterior spinal canal.302–304 The segregation and isolation of abscesses to the anterior or posterior spinal canal is thought to be secondary to septations within the epidural fat.303 These septations not only divide the epidural space into an anterior and posterior compartment but also divide the space longitudinally.303 The longitudinal septations usually limit the extent of epidural abscess formation to 3 to 4 vertebral levels.303 Postsurgical SEAs often involve multiple compartments, extending several levels and circumferentially around the spinal cord, secondary to disruption of the epidural septations.304 The neurologic compromise caused by epidural abscess formation can be slowly progressive or dramatically acute in nature with complete paralysis in a matter of hours.303,304,308,309 The underlying cause of neurologic injury is thought to be secondary to both compression of the neural elements and vascular thrombosis.303,304,308–310 Animal studies suggest that early neurologic dysfunction is the result of neural compression, with irreversible vascular compromise occurring later.310 Clinical Features and Diagnostic Considerations The classic clinical presentation of an SEA is back pain and fever associated with nerve root symptoms followed by limb weakness.302 In reality, the presentation is highly variable and most patients are initially misdiagnosed.305 The most common symptom associated with an SEA is back pain, present in 90% of patients.302,304,305,307,308 Other common findings include fever (61%), paresis (53%), bowel or bladder dysfunction (36%), sepsis (17%), radiculopathy (12%) and plegia (14%).307 Point tenderness over the involved vertebral levels is present in about one quarter of patients and is associated with underlying bony involvement.305 The most common location of SEA formation is in the lumbar region, but thoracic and cervical involvement is not uncommon.302 Time between symptom onset and presentation is highly variable and does not correlate well with intra-operative findings.311 Neurologic deficits are present in the majority of patients at the time of presentation.302,305 Neurologic decline can occur chronically over months or precipitously over a few hours.302,304,305
Central Nervous System Infections
385
The most consistent laboratory abnormality is an elevated erythrocyte sedimentation rate, which is present virtually 100% of the time.304,305,311 Leukocytosis is found in roughly 75% of patients.305 Results of cerebrospinal fluid analysis are variable, ranging from normal to frank pus.304 Blood cultures have been reported to be positive in up to 60% of cases.303 When the diagnosis of SEA is suspected, an imaging study of the spine should be performed.302 The imaging study of choice is a contrast enhanced MRI scan, which gives superior soft tissue resolution when compared to CT myelography and is noninvasive.302,309 After radiologic diagnosis, consultation with a neurosurgeon should be undertaken.304 Microbiology The spectrum of microbial pathogens capable of causing SEA formation is wide.302 However, Staph. aureus is responsible for the overwhelming majority of cases, with a reported incidence of 60% to 70% of culture positive abscesses.302,304,305,307,309 In one large review of pyogenic infections, Staph. aureus accounted for 62%, other gram-positive cocci 10%, gram-negative species 18%, and anaerobes 2% of culture positive spinal abscesses.302 Mycobacterium has been reported as the cause in up to 25% of infections in some series.302 Other less common causes include Brucella, Actinomyces, and fungal etiologies such as Cryptococcus and Aspergillus.302 Treatment Most authorities consider the treatment of SEAs to be surgical evacuation followed by prolonged parenteral antibiotics.302–304,307,308 SEA carries with it a high mortality rate and a significant long-term neurologic morbidity rate.305,307 The mortality rate is estimated to be 14%, and 35% to 40% of patients will have residual neurologic deficits.302,303,305 Prognosis depends on the clinical and neurologic condition of the patient at the time of presentation.302 Patients presenting with sepsis or plegia have higher mortality and long-term morbidity rates.302,305 Some authors have reported success treating SEAs with medical therapy alone.309 The majority of these patients fell into one of three categories: (1) panspinal infections not amenable to drainage, (2) poor operative candidates secondary to poor health, or (3) complete paralysis for greater than 24 hours.308 In order to treat patients nonoperatively, the clinician must obtain the organism by another means such as blood culture or needle aspirate, be willing to perform serial neurologic exams, or monitor the response to therapy with serial MRI scans.308 Caution is warranted in treating spinal epidural abscesses nonoperatively.309 Culture specific antibiotic therapy failed to protect 9 of 39 patients in one series from developing an acute irreversible neurologic deficit.304
W9418-12.qxd 11/3/03 2:58 PM Page 386
386
Section I
Introduction
Vertebral Osteomyelitis Epidemiology Vertebral osteomyelitis (VO) is another relatively uncommon clinical entity with an estimated population incidence of 1/250,000.312 Pyogenic spinal infections account for approximately 2% to 4% of all cases of osteomyelitis, ranking third behind infections of the femur and tibia in adults.313,314 VO is the most common hematogenously acquired osteomyelitis.313,314 The incidence of VO appears to have increased in proportion to the increase in the number of immunocompromised patients over the past decade.313,315,316 VO is characteristically a disease of older men, occurring with a two-fold higher incidence in men, with 50% of cases occurring in patients older than fifty years.314,315,317 VO occurring in conjunction with IV drug abuse accounts for the majority of cases in younger patients.314,315 There are a number of known risk factors for the development of spontaneous VO, which include diabetes mellitus, systemic steroid use, history of a genitourinary infection or procedure, bacteremia, protein calorie malnutrition, IV drug abuse, and malignancy.313,314,317–319 Advanced age may be an independent risk factor for the development of the disorder.320 The common denominator in each of these predisposing conditions appears to be a deficiency in some aspect of cellular or humoral immunity.319 Postsurgical VO accounts for approximately 2.5% of all spinal osteomyelitis and has been shown to have an increased incidence in malnourished patients, diabetics, and patients on steroid therapy.317,318,321 The overall incidence of diabetes in patients with VO is between 18% and 25%.313 As the prevalence of AIDS has increased, there has been resurgence in the incidence of spinal tuberculosis and the emergence of other fungal causes of VO.319 IV drug abusers have accounted for over half the cases of pyogenic spinal osteomyelitis in some large series of patients.315 Prolonged steroid use is a risk factor not only for the more common bacterial VO, but is also a risk factor for atypical mycobacterium causes.319,322 In summary, patients at risk for VO are typically of advanced age, diabetic (18% to 25%), have other immunodeficient states, have a history of a recent bacteremic state (50%), or urinary tract infection or procedure (30%).313,314 An exception to the typical presentation is the young male IV drug abuser.315 Pathogenesis Spontaneous VO results from hematogenous spread of organisms through the segmental spinal arteries to the subchondral plate region of the vertebral body adjacent to the disk space.318,320 In adults, the nidus of infection begins in the vertebral bodies at the level of the end arteriolar arcades and, after endplate destruction, spreads secondarily
into the avascular disk space.313,320 In children, the disk space contains vascular channels that allow primary seeding of the intervertebral disk.313,320 Segmental spinal arteries typically bifurcate to supply adjacent vertebral segments; this bifurcation is thought to account for the fact that VO typically involves two adjacent vertebrae and the intervening disk space.320 Postoperative VO results from direct inoculation at the time of surgery.313,318 Surgical teams and the patient’s skin flora are the principal sources of wound contamination.318 Unlike spontaneous VO, the nidus of infection in postoperative patients is often the disk space.313,318 Other forms of direct inoculation can result in VO and include decubitus ulcers and trauma.317 Clinical Presentation and Diagnostic Considerations The most common presentation of VO is pain, occurring in more than 90% of patients.314,319 The pain is localized, continuous, and classically unrelated to movement or position.314 Nearly all spinal infections are associated with tenderness to palpation over the involved level.314,319 The pain is most commonly over the lumbar spine due to the fact that nearly 50% of VO cases are localized to the lumbar vertebrae.313,314 There is an increased incidence of cervical VO in IV drug abusers and thoracic disease in tuberculous osteomyelitis.313 The presence of radiculopathy, positive straight leg raise, and neurologic deficit (4% to 16%), are less reliable and often indicate the presence of epidural involvement.314,319 Fever is found in approximately half of patients with pyogenic spinal osteomyelitis.318 Constitutional symptoms, including malaise, night sweats, and anorexia have been reported.314 VO has an insidious onset and has proven to be a diagnostic challenge.320 In several large patient series, the delay between the onset of symptoms and eventual diagnosis has ranged from 3 weeks to 3 months.314,315,320,323 These delays in diagnosis have resulted in significant neurologic morbidity.320 When VO is suspected, a diagnostic work-up should proceed in a logical manner. Basic laboratory work-up should include a complete cell count, ESR, and blood cultures. A leukocytosis is found in less than half of patients with VO.314 The most common laboratory abnormality is an elevated ESR, with 90% of cases being between 20 and 100 mm/hr.314 Blood cultures are positive in 25% of cases and when positive may mitigate the need for an invasive diagnostic biopsy.314,317 The principal diagnostic modality in VO is spinal imaging. Plain radiologic findings include disk space narrowing and vertebral endplate changes that usually become apparent 2 to 4 weeks after symptom onset in approximately 80% of patients.313,314,318 Technetium99m bone scanning combined with gallium scanning has a 90% sensitivity and specificity for VO.324 MRI imaging is the most sensitive and
W9418-12.qxd 11/3/03 2:58 PM Page 387
Chapter 12
specific imaging modality for VO and has the added benefit of providing detail regarding the presence of an epidural abscess, paraspinal abscess, or other sources of neural compromise.313,325 Characteristic MRI changes include decreased T1 signal and increased T2 signal of the involved vertebral endplates and disk space.325 Once a radiographic diagnosis of VO has been made, optimal pharmacologic treatment requires that the infecting microbe be identified and antibiotic susceptibilities be determined.314 A positive blood culture is present 25% of the time, correlates strongly with the responsible spinal pathogen, and can direct antibiotic therapy without the need for invasive procedures.314,315 When blood cultures are negative, biopsy of the infected vertebra is mandatory.313,314,318 A CT-guided needle biopsy can be used to make a definitive microbiologic diagnosis in most cases and this technique is successful 60% to 90% of the time.314 An open biopsy should be considered when both blood cultures and CT aspirates are negative.318 A skin test for tuberculosis should be undertaken in all patients and CT aspirates or biopsy material should be sent for routine fungal stains and cultures.318,323
Microbiology There exists in the literature an enormous array of bacterial and nonbacterial pathogens that have been reported to cause VO.316,317,319,323,326,327 The most common pathogen is Staph. aureus, found in up to 60% of cases.313,317 In one large review of the literature, the etiology of culture positive pyogenic VO was found to be due to gram-positive aerobic cocci 68% of the time, gram-negative aerobic bacilli in 29% of the patients, and the remaining 3% were due to anaerobic bacteria.317 Tuberculous osteomyelitis is often indistinguishable on a radiologic basis from bacterial infection and must be in the differential of spinalosteomyelitis.316 Coccidioides immitis, Blastomyces dermatiditis, Cryptococcus neoformans, Aspergillus species, and other less common fungi have all been associated with VO.317 Stratification of patients by known risk factors has revealed some interesting correlations between predisposing conditions and the microbiology of VO.319 Pseudomonas and Staph. aureus are the most prevalent pathogens in IV drug abusers and occur at roughly equal rates.317 In elderly males with urinary tract infections or following invasive urological procedures, the most common pathogens are E. coli and Proteus species317; postoperative spinal infections are usually caused by Staph. aureus.317 There is a high prevalence of mycobacterial infections in countries where tuberculosis is endemic and in the growing AIDS population.317,319 Patients on long term steroid treatment are susceptible to infections caused by atypical mycobacterium and Aspergillus.319,322,326,327 VO is overwhelmingly caused by a single organism; however, decubitus ulceration leading to direct spread of infection is often polymicrobial in nature.317,319
Central Nervous System Infections
387
Treatment Focused antibiotic therapy, spinal immobilization, and when necessary surgical interventions are the mainstays of treatment for VO.313,314,317 Seventy-five percent of patients with spinal osteomyelitis can be managed without surgical intervention.313 Prolonged courses of parental antibiotics directed by specific culture results and antibiotic susceptibilities are the rule.317 One large review of the literature has shown that four weeks of high dose IV antibiotics is sufficient to treat pyogenic spinal osteomyelitis as long as the following criteria are met: (1) there are no undrained abscesses; (2) the patient is clinically stable; and (3) the ESR has decreased to one half its original value.317 Tuberculous VO requires an average of 12 months treatment with a combination of isoniazid, rifampin, ethambutol, and pyrazinamide depending on regional susceptibility.313 Treatment of other less common bacterial and fungal causes of VO should be tailored to the individual pathogen and to regional susceptibilities. The indications for surgical intervention in spinal osteomyelitis are the presence of a neurologic deficit, spinal instability, unresponsiveness to medical therapy, or a nondiagnostic CT-guided biopsy.315 In addition, surgery is recommended for the drainage of epidural or paraspinal abscesses that often accompany VO.317 The goals of surgery are decompression of the neural elements, correction of spinal deformity, debridement of necrotic tissue, and the promotion of long term stability.313,315,320,323 A variety of surgical approaches have been described in the literature, each with its own advantages and disadvantages.313,315,323 Most recent surgical reviews recommend early instrumentation and fusion at the time of the initial operation to facilitate ambulation and avoid the complications of prolonged bedrest.313,315,323
External Ventricular Drainage Infectious Considerations External ventricular drainage (ventriculostomy) devices are an integral aspect of the intensive care management of neurosurgical patients. Common indications for their use include management of hydrocephalus, elevated intracranial pressure, intracranial hemorrhage, and the administration of intrathecal medications.328–330 External ventricular drains (EVDs) provide diagnostic information as well as providing therapeutic cerebrospinal fluid drainage.328–330 The most common complication involving the use of EVDs is infection. The reported infection rate is 0% to 27%.328,329 Risk factors for the development of an EVD– associated infection include hemorrhage, neurosurgical operation, and irrigation or manipulation of the drainage system.328,330 Conflicting data exist regarding the association between the duration of ventricular drainage and the rate of
W9418-12.qxd 11/3/03 2:58 PM Page 388
388
Section I
Introduction
EVD–associated infections.328,329,331 Some clinical series have shown a linear relationship between infection and duration of ventricular drainage.332,333 Several of these authors have advocated the routine changing of ventricular catheters after day five in order to lower the risk of infection.333 Other studies have suggested that the duration of monitoring is not a risk factor associated with infection. In contrast, these later studies have shown a constant daily rate of infection that actually decreases after day 10 or 11 of drainage.334 The role of perioperative and prophylactic antibiotics in the prevention of EVD–associated infections has been extensively debated. Alleyne and colleagues reviewed over 300 patients who received either daily prophylactic antibiotics or perioperative antibiotics alone and found no significant difference in the infection rate between the two groups.328
Low infection rates associated with prolonged EVD placement have been reported.328,329,331 Factors thought to be important in preventing infection include perioperative antibiotic administration prior to skin incision, tunneling of the ventricular catheter, surgical skin prep, closed ventricular drainage system, and meticulous sterile nursing care.328–331,335 Perioperative antibiotics should be chosen for their efficacy against skin flora, keeping in mind that Staph. epidermidis is the most frequent cause of EVD–associated infections.330,335 Recently, antibiotic impregnated catheters have been manufactured for a variety of applications.336 Preliminary use of ventriculostomy catheters has been promising but further study is needed to validate the efficacy of this emerging technology.
P earls 1. S. pneumoniae is the most common cause of bacterial meningitis in the United States today in all age groups except infants in the immediate neonatal period. 2. Pneumococcal meningitis is the most common form of recurrent meningitis in patients who have CSF leaks. 3. Following the introduction of the H. influenzae type b vaccine there has been a profound reduction in the number of invasive infections due to H. influenzae in the United States. For example, Murphy and colleagues5 found a reduction of 85% to 92% in the incidence of invasive H. influenzae type b disease between 1983 to 1984 and 1991 after widespread use of the vaccine. 4. . . . . gram-negative meningitis is highly significant in hospital acquired cases of meningitis. The vast majority of these cases are seen following neurocranial surgery, spinal surgery, and in patients who have suffered head trauma. 5. On entering into the subarachnoid space, bacterial replication proceeds virtually unchecked by host defense mechanisms. By virtue of the blood-brain barrier, both immunoglobulin and complement levels are far lower in CSF than in serum and interstitial fluid. In addition, leukocyte proteases derived from an initial influx of leukocytes have actually been shown to degrade complement components in CSF from patients with meningitis. 6. In a review of 493 episodes, Durand and associates37 found that 95% of the patients with bacterial meningitis had fever greater than 37.7°C on admission; neck stiffness was present in 88% of patients. Only 22% were alert, while 51% were confused or lethargic and 22% were responsive only to pain. Within the
7.
8.
9.
10.
11.
first 24 to 48 hours of onset, 29% had focal seizures and/or focal neurologic findings. Glucose enters the cerebrospinal fluid by transport through the choroid plexus and endothelium of the capillaries in the subarachnoid space. CSF levels of glucose are thus a function of active transport of glucose and the rate of glucose consumption within the CNS. The level of CSF glucose under normal conditions in normal subjects is 60% to 70% of the blood glucose level. However, a study by Skipper and Davis44 showed this CSF/serum ratio was only accurate when the serum glucose was between 89 and 115 mg/dL. St. Louis encephalitis is found throughout the Midwest and South, as far north as New York and Michigan with cases even reported on the West Coast. Eastern equine encephalitis, however, is essentially confined to the Southeast, and California encephalitis and La Crosse encephalitis tend to be seen primarily in the northern Midwest. At least 12 persons were hospitalized with confirmed West Nile virus infection in the summer of 2000, most of them in New York City. The clinical presentation of viral meningitis includes fever, stiff neck, photophobia, and varying degrees of nonspecific symptoms such as malaise, myalgias, nausea, vomiting, abdominal pain, or diarrhea. The presence of impairment of consciousness such as obtundation, disorientation, seizures or localized neurologic signs or symptoms should suggest brain parenchymal involvement and a diagnosis of encephalitis or meningoencephalitis. Because of the rarity of complications from acyclovir it is difficult to argue with presumptive treatment and brain biopsy certainly is not justified to prove the presence of herpes encephalitis before treatment.
W9418-12.qxd 11/3/03 2:58 PM Page 389
Chapter 12
12. Despite its chronicity and high level of bacteremia, endocarditis accounts for only 1% to 5% of cases of brain abscess.96 A significant number of brain abscesses are associated with penetrating trauma such as gunshot wounds, depressed skull fractures with retained bone fragments, cranial penetration from objects such as pencils, animal bites, or even as a complication of cervical traction associated with pin site infection. In approximately 25% of cases no underlying etiology can be found. 13. The most common fungal pathogens of the central nervous system include the yeast Cryptococcus and the dimorphic fungi Histoplasma, Coccidioides, and Blastomyces. In immunocompromised patients, Aspergillosis, Candida, and the Mucorales may be important pathogens as well. 14. The diagnosis of cryptococcal meningitis is usually not difficult. The classic and time-honored method for diagnosis is demonstration of the yeast in spinal fluid by India ink stain. Microscopically, the India ink particles serve to outline the very large clear polysaccharide capsule surrounding the yeast. The test result is positive in over 90% of patients with HIV, but in only 50% of patients with normal immunity. 15. The most serious form of disseminated coccidioidal infection is coccidioidal meningitis. Without treatment it is nearly uniformly fatal within 2 years of diagnosis.123 Observational studies suggest about 80% of patients in whom meningitis develops become symptomatic within 6 months of the initial infection. 16. The genuses Rhizopus, Mucor, and Rhizomucor may invade the CNS by direct extension or from hematogenous spread. In patients with diabetes, this disease typically presents as rhinocerebral mucor.
References 1. Ryan MW, Antonelli PJ: Pneumococcal antibiotic resistance and rates of meningitis in children. Laryngoscope 2000;110(6):961–964. 2. Schuchat A, Robinson K, Wenger JD, et al: Bacterial meningitis in the United States in 1995. Active surveillance team. N Engl J Med 1997;337(14):970–976. 3. Sherry B, Emanuel I, Kronmal RA, et al: Interannual variation of the incidence of haemophilus influenzae type b meningitis. JAMA 1989;261(13):1924–1929. 4. Fothergill LD, Wright J. Influenzal meningitis: Relation of age incidence to bacterial power of blood against causal organism. J Immunol 1933;24:273–284. 5. Murphy TV, White KE, Pastor P, et al: Declining incidence of haemophilus influenzae type b disease since introduction of vaccination [see comments]. JAMA 1993;269(2):246–248. 6. Unhanand M, Mustafa MM, McCracken GH Jr, Nelson JD: Gramnegative enteric bacillary meningitis: A twenty-one-year experience. J Pediatr 1993;122(1):15–21. 7. Sarff LD, McCracken GH, Schiffer MS, et al: Epidemiology of escherichia coli K1 in healthy and diseased newborns. Lancet 1975;1(7916):1099–1104.
17.
18.
19.
20.
Central Nervous System Infections
389
Patients are generally predisposed to this complication if their serum glucose has remained uncontrolled, and they are acidotic for several weeks. Initial symptoms usually include sinus pain, headache, fever, and nasal stuffiness and discharge, which quickly progresses to facial cellulitis, swelling, proptosis, cavernous sinus thrombosis and, if not treated, death. Interestingly, no change in the incidence of HIV encephalopathy was seen between 1996 and 1999, during the era of HAART treatment. The diagnosis of malaria is made from examination of the blood smear, provided that the diagnosis is considered as a differential diagnostic possibility in a patient with altered consciousness and fever. Examination of the blood smear is highly accurate. Because the level of parasitemia in falciparum malaria tends to be high, generally greater than 5%, it is not hard to find the ring forms characteristic of the disease. Serologic tests for syphilis are generally divided into two different types: treponemal tests and nontreponemal tests. The nontreponemal tests are IgG or IgM antibodies that are directed at lipid antigens, which were originally extracts of cardiolipin from beef heart or liver. If left untreated, the Lyme disease spirochete becomes sequestered and persists in certain tissues such as the nervous system, joints, heart, and even the skin. In some patients with meningitis, particularly in Europe where the disease is caused by a different species of spirochete, significant neurologic abnormalities develop including cranial neuritis, which can present as a isolated facial palsy (motor and sensory), radicular mononeuritis multiplex, or myelitis alone or in varying combinations.
8. Sarff LD, Platt LH, McCracken GH Jr: Cerebrospinal fluid evaluation in neonates: Comparison of high-risk infants with and without meningitis. J Pediatr 1976;88(3):473–477. 9. Smith AL: Neonatal bacterial meningitis. In Scheld WM, Whitley RJ, Durack DT (eds): Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, pp 313–334. 10. Dawson KG, Emerson JC, Burns JL: Fifteen years of experience with bacterial meningitis. Pediatr Infect Dis J 1999;18(9):816–822. 11. Musser JM, Mattingly SJ, Quentin R, Goudeau A, Selander RK: Identification of a high-virulence clone of type iii streptococcus agalactiae (group b streptococcus) causing invasive neonatal disease. Proc Natl Acad Sci USA 1989;86(12):4731–4735. 12. Kallstrom H, Liszewski MK, Atkinson JP, Jonsson AB: Membrane cofactor protein (MCP or CD46) is a cellular pilus receptor for pathogenic neisseria. Mol Microbiol 1997;25(4):639–647. 13. Virji M, Alexandrescu C, Ferguson DJ, Saunders JR, Moxon ER: Variations in the expression of pili: The effect on adherence of neisseria meningitidis to human epithelial and endothelial cells. Mol Microbiol 1992;6(10):1271–1279. 14. Virji M, Watt SM, Barker S, Makepeace K, Doyonnas R: The Ndomain of the human CD66a adhesion molecule is a target for opa proteins of neisseria meningitidis and neisseria gonorrhoeae. Mol Microbiol 1996;22:929–939.
W9418-12.qxd 11/3/03 2:58 PM Page 390
390
Section I
Introduction
15. Mulks MH, Plaut AG: IgA protease production as a characteristic distinguishing pathogenic from harmless neisseriaceae. N Engl J Med 1978;299(18):973–976. 16. Vitovski S, Read RC, Sayers JR: Invasive isolates of neisseria meningitidis possess enhanced immunoglobulin a1 protease activity compared to colonizing strains. FASEB J 1999;13(2):331–337. 17. Klein NJ, Ison CA, Peakman M, et al: The influence of capsulation and lipooligosaccharide structure on neutrophil adhesion molecule expression and endothelial injury by neisseria meningitidis. J Infect Dis 1996;173(1):172–179. 18. Densen P: Interaction of complement with neisseria meningitidis and neisseria gonorrhoeae. Clin Microbiol Rev 1989;2[Suppl]:S11–S17. 19. Fijen CA, Kuijper EJ, Hannema AJ, Sjoholm AG, van Putten JP: Complement deficiencies in patients over ten years old with meningococcal disease due to uncommon serogroups [see comments]. Lancet 1989;2(8663):585–588. 20. Fijen CA, Kuijper EJ, Tjia HG, Daha MR, Dankert J: Complement deficiency predisposes for meningitis due to nongroupable meningococci and neisseria-related bacteria. Clin Infect Dis 1994;18(5):780– 784. 21. Derkx HH, Kuijper EJ, Fijen CA, Jak M, Dankert J, van Deventer SJ: Inherited complement deficiency in children surviving fulminant meningococcal septic shock [see comments]. Eur J Pediatr 1995; 154(9):735–738. 22. Djupesland G, Gedde-Dahl TW: Sequelae of meningococcal disease. NIPH Ann 1983;6(1):85–90. 23. Pfister HW, Fontana A, Tauber MG, Tomasz A, Scheld WM: Mechanisms of brain injury in bacterial meningitis: Workshop summary. Clin Infect Dis 1994;19(3):463–479. 24. Mason EO Jr, Kaplan SL, Wiedermann BL, Norrod EP, Stenback WA: Frequency and properties of naturally occurring adherent piliated strains of haemophilus influenzae type b. Infect Immun 1985;49(1): 98–103. 25. Pichichero ME, Loeb M, Anderson, Smith DH: Do pili play a role in pathogenicity of haemophilus influenzae type B? Lancet 1982; 2(8305):960–962. 26. Austrian R: The pneumococcus at the millennium: Not down, not out. J Infect Dis 1999;179(Suppl 2):S338–S341. 27. Finland M: Excursions into epidemiology: Selected studies during the past four decades at Boston city hospital. J Infect Dis 1973;128(1):76– 124. 28. Kim KS: E. coli invasion of brain microvascular endothelial cells as a pathogenetic basis of meningitis. In Oelschlaeger, Hacker (eds): Subcellular Biochemistry, vol. 33: Bacterial Invasion into Euykaryotic Cells. New York, Plenum, 2000, pp 47–59. 29. Prasadarao NV, Wass CA, Weiser JN, Stins MF, Huang SH, Kim KS: Outer membrane protein a of escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect Immun 1996;64(1): 146–153. 30. Whittle HC, Greenwood BM: Cerebrospinal fluid immunoglobulins and complement in meningococcal meningitis. J Clin Pathol 1977;30(8):720–722. 31. Tuomanen E, Tomasz A, Hengstler B, Zak O: The relative role of bacterial cell wall and capsule in the induction of inflammation in pneumococcal meningitis. J Infect Dis 1985;151(3):535–540. 32. Rietschel ET, Wollenweber HW, Russa R, Brade H, Zahringer U: Concepts of the chemical structure of lipid A. Rev Infect Dis 1984; 6(4):432–438. 33. Morrison DC, Ryan JL: Bacterial endotoxins and host immune responses. Adv Immunol 1979;28:293–450. 34. Urbaschek R, Urbaschek B: Tumor necrosis factor and interleukin 1 as mediators of endotoxin-induced beneficial effects. Rev Infect Dis 1987;9[Suppl 5]:S607–S615. 35. Tracey KJ, Lowry SF, Cerami A: Cachectin: a hormone that triggers acute shock and chronic cachexia. J Infect Dis 1988;157(3): 413–420.
36. Carpenter RR, Petersdorf RG: The clinical spectrum of bacterial meningitis. Am J Med 1962;33:262–275. 37. Durand ML, Calderwood SB, Weber DJ, et al: Acute bacterial meningitis in adults. A review of 493 episodes [see comments]. N Engl J Med 1993;328(1):21–28. 38. Gorse GJ, Thrupp LD, Nudleman KL, Wyle FA, Hawkins B, Cesario TC: Bacterial meningitis in the elderly. Arch Intern Med 1984; 144(8):1603–1607. 39. Verghese A, Gallemore G: Kernig’s and Brudzinski’s signs revisited. Rev Infect Dis 1987;9(6):1187–1192. 40. DeMyer W: Examination of the patient who has a disorder of consciousness. In DeMyer W (ed): Technique of the Neurologic Examination, 3rd ed. New York, McGraw-Hill, 1980, pp 359–404. 41. Shetty AK, Desselle BC, Craver RD, Steele RW: Fatal cerebral herniation after lumbar puncture in a patient with a normal computed tomography scan. Pediatrics 1999;103(6 Pt 1):1284–1287. 42. Herndon RM, Brumback RA: The Cerebrospinal Spinal Fluid. Boston, Kluwer, 1989. 43. Fishman RA: Cerebrospinal Fluid in Diseases of the Central Nervous System. Philadelphia, WB Saunders, 1980. 44. Skipper BJ, Davis LE: Ascertaining hypoglycorrhachia in an acute patient. Am J Emerg Med 1997;15(4):378–380. 45. Negrini B, Kelleher KJ, Wald ER: Cerebrospinal fluid findings in aseptic versus bacterial meningitis. Pediatrics 2000;105(2):316–319. 46. Feigin RD, Shackelford PG: Value of repeat lumbar puncture in the differential diagnosis of meningitis. N Engl J Med 1973;289(11):571– 574. 47. Gray LD, Fedorko DP: Laboratory diagnosis of bacterial meningitis. Clin Microbiol Rev 1992;5(2):130–145. 48. Roos KL, Tunkel AR, Scheld M: Acute bacterial meningitis in children and adults. In Scheld WM, Whitley RJ, Durack DT (eds). Infections of the Central Nervous System, 2nd ed. Philadelphia, LippincottRaven, 1997, pp 335–401. 49. Hayden RT, Frenkel LD: More laboratory testing: Greater cost but not necessarily better. Pediatr Infect Dis J 2000;19(4):290– 292. 50. Mein J, Lum G: CSF bacterial antigen detection tests offer no advantage over gram’s stain in the diagnosis of bacterial meningitis. Pathology 1999;31(1):67–69. 51. Gray BM, Simmons DR, Mason H, Barnum S, Volanakis JE: Quantitative levels of C-reactive protein in cerebrospinal fluid in patients with bacterial meningitis and other conditions. J Pediatr 1986;108(5 Pt 1):665–670. 52. Committee on Infectious Diseases; American Academy of Pediatrics; Pickering LK, Peter G, Baker CJ, et al: 2000 Red Book: Report of the Committee on Infectious Diseases, 25th ed. Elk Grove Village, IL, American Academy of Pediatrics, 2000. 53. Rziha HJ, Belohradsky BH, Schneider U, Schwenk HU, Bornkamm GW, zur Hausen H: BK virus: II. Serologic studies in children with congenital disease and patients with malignant tumors and immunodeficiencies. Med Microbiol Immunol (Berl) 1978;165(2): 83–92. 54. Padgett BL, Walker DL: Prevalence of antibodies in human sera against jc virus, an isolate from a case of progressive multifocal leukoencephalopathy. J Infect Dis 1973;127(4):467–470. 55. Ohga S, Okada K, Ueda K, et al: Cerebrospinal fluid cytokine levels and dexamethasone therapy in bacterial meningitis. J Infect 1999; 39(1):55–60. 56. Ahmed A, Jafri H, Lutsar I, et al: Pharmacodynamics of vancomycin for the treatment of experimental penicillin. Antimicrob Agents Chemother 1999;43(4):876–881. 57. McIntyre PB, Berkey CS, King SM, et al: Dexamethasone as adjunctive therapy in bacterial meningitis: A meta-analysis of randomized clinical trials since 1988. JAMA 1997;278(11):925–931. 58. de Gans J, van de Beek D: Dexamethasone in adults with bacterial meningitis. N Engl J Med 2002;347:1549–1556.
W9418-12.qxd 11/3/03 2:58 PM Page 391
Chapter 12 59. Gilbert DN, Moellering RC, Sande M: The Sandford Guide to Antimicrobial Therapy, 33rd ed. Hyde Park, VT, Antimicrobial Therapy, 2003, p 4. 60. Dacey RG: Monitoring and treating increased intracranial pressure. Pediatr Infect Dis J 1987;6:1161–1163. 61. Roos KL, Scheld WM: The management of fulminant meningitis in the intensive care unit. Crit Care Clin 1988;4(2):375–392. 62. Tunkel AR, Scheld WM: Acute meningitis. In Mandel GA, Bennett JE, Dolan R (eds): Principles and Practice of Infectious Diseases, 4th ed. New York, Livingston, 1995, pp 831–865. 63. Meldrum BS, Vigouroux RA, Brierley JB: Systemic factors and epileptic brain damage: Prolonged seizures in paralyzed, artificially ventilated baboons. Arch Neurol 1973;29(2):82–87. 64. Prevention and control of meningococcal disease in college students, no. RR-7. MMWR 2000;49. 65. Immunization of college students against meningococcal disease. Med Lett Drugs Ther 2000;42:69–70. 66. CDC: Arboviral Infections of the Central Nervous System—United States, 1996–1997. Atlanta, CDC, 2000, report no. 47(25). 67. Asnis DS, Conetta R, Teixeira AA, Waldman G, Sampson BA: The West Nile virus outbreak of 1999 in New York: The Flushing Hospital experience [published erratum appears in Clin Infect Dis 2000;30(5):841]. Clin Infect Dis 2000;30(3):413–418. 68. Rappole JH, Derrickson SR, Hubalek Z: Migratory birds and spread of West Nile virus in the western hemisphere. Emerg Infect Dis 2000;6(4):319–328. 69. Update: West Nile activity—northeastern United States 2000, no. 36. MMWR 2000;49. 70. Weekly update: West Nile virus activity—United States, October 31–November 6, 2001, report no. 50. MMWR 2001;44:983. 71. Human rabies prevention: United States 1999: Recommendations of the advisory committee on immunization practices, No. RR-1. MMWR 1999;48. 72. Sharpe AH, Fields BN: Pathogenesis of viral infections. Basic concepts derived from the reovirus model. N Engl J Med 1985;312(8):486–497. 73. Lentz TL: The recognition event between virus and host cell receptor: A target for antiviral agents. J Gen Virol 1990;71(Pt 4):751–766. 74. Hogle JM, Chow M, Filman DJ: Three-dimensional structure of poliovirus at 2.9 A resolution. Science 1985;229(4720):1358–1365. 75. Griffin DE: Encephalitis, myelitis and neuritis. In Mandell, Bennett, Dolan (eds): Principles and Practice of Infectious Diseases, 5th ed. Philadelphia, Churchill Livingstone, 2000, pp 1009–1019. 76. Cassady K, Kevin A, Whitley RJ: Pathogenesis and pathophysiology of viral infections of the central nervous system. In Scheld WM, Whitley RJ, Durack DT (eds): Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, pp 7–22. 77. Esiri MM: Herpes simplex encephalitis: An immunohistological study of the distribution of viral antigen within the brain. J Neurol Sci 1982;54(2):209–226. 78. Noah DL, Drenzek CL, Smith CS, et al: Epidemiology of rabies in the United States, 1980 to 1986. Ann Intern Med 1998;128:922–930. 79. Jolles S, Sewell WA, Leighton C: Drug-induced aseptic meningitis: Diagnosis and management. Drug Saf 2000;22(3):215–226. 80. Whitley RJ, Soong SJ, Linneman C Jr, Liu C, Pazin G, Alford CA: Herpes simplex encephalitis. Clinical assessment. JAMA 1982; 247(3):317–320. 81. Mitchell PS, Espy MJ, Smith TF, et al: Laboratory diagnosis of central nervous system infections with herpes simplex virus by PCR performed with cerebrospinal fluid specimens. J Clin Microbiol 1997;35(11):2873–2877. 82. Revello MG, Baldanti F, Sarasini A, Zella D, Zavattoni M, Gerna G: Quantitation of herpes simplex virus DNA in cerebrospinal fluid of patients with herpes simplex encephalitis by the polymerase chain reaction. Clin Diagn Virol 1997;7(3):183–191. 83. Lakeman FD, Whitley RJ: Diagnosis of herpes simplex encephalitis: Application of polymerase chain reaction to cerebrospinal fluid from
84.
85.
86. 87. 88.
89. 90. 91. 92.
93. 94.
95. 96.
97.
98. 99. 100. 101.
102.
103.
104.
105.
106.
107.
Central Nervous System Infections
391
brain-biopsied patients and correlation with disease. National Institute of Allergy and Infectious Diseases Collaborative Antiviral Study Group [see comments]. J Infect Dis 1995;171(4):857–863. Whitley RJ, Alford CA, Hirsch MS, et al: Vidarabine versus acyclovir therapy in herpes simplex encephalitis. N Engl J Med 1986; 314(3):144–149. Bryceson AD, Greenwood BM, Warrell DA, et al: Demonstration during life of rabies antigen in humans. J Infect Dis 1975;131(1):71– 74. Canale DJ: William Macewen and the treatment of brain abscesses: Revisited after one hundred years. J Neurosurg 1996;84(1):133–142. Osma U, Cureoglu S, Hosoglu S: The complications of chronic otitis media: Report of 93 cases. J Laryngol Otol 2000;114(2):97–100. Nicolosi A, Hauser WA, Musicco M, Kurland LT: Incidence and prognosis of brain abscess in a defined population: Olmsted County, Minnesota, 1935-1981. Neuroepidemiology 1991;10(3):122–131. Morgan H, Wood MW, Murphey F: Experience with 88 consecutive cases of brain abscess. J Neurosurg 1973;38(6):698–704. Samson DS, Clark K: A current review of brain abscess. Am J Med 1973;54(2):201–210. O’Donoghue MA, Green HT, Shaw MD: Cerebral abscess on Merseyside 1980-1988. J Infect 1992;25(2):163–172. Nielsen H, Gyldensted C, Harmsen A: Cerebral abscess. aetiology and pathogenesis, symptoms, diagnosis and treatment. A review of 200 cases from 1935-1976. Acta Neurol Scand 1982;65(6):609– 622. Small M, Dale BA: Intracranial suppuration 1968-1982—a 15 year review. Clin Otolaryngol 1984;9(6):315–321. Garvey G: Current concepts of bacterial infections of the central nervous system. Bacterial meningitis and bacterial brain abscess. J Neurosurg 1983;59(5):735–744. Cohen DJ: Lung abscess—Back for an encore? Postgrad Med 1982;72: 215–218. Wispelwey B, Dacey RG, Scheld M: Brain abscess. In Scheld WM, Whitley RJ, Durack DT: Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, p 463. Mamelak AN, Obana WG, Flaherty JF, Rosenblum ML: Nocardial brain abscess: Treatment strategies and factors influencing outcome. Neurosurgery 1994;35(4):622–631. Mathisen SM: Brain abscess and cerebritis. Rev Infect Dis 1984;6(Suppl 1):S101–S106. Chun CH, Johnson JD, Hofstetter M, Raff MJ: Brain abscess. A study of 45 consecutive cases. Medicine (Baltimore) 1986;65(6):415–431. Arseni C, Ciurea AV: Cerebellar abscesses. A report on 119 cases. Zentralbl Neurochir 1982;43(4):359–370. Moltas S, Tornquist C, Cronquist S: Diagnostic difficulties in computed tomography of brain abscesses. Comput Assist Tomogr 1982;6:683–688. Miller ES, Dias PS, Uttley D: CT Scanning in the management of intracranial abscess: A review of 100 cases. Br J Neurosurg 1988;2(4): 439–446. Desprechins B, Stadnik T, Koerts G, Shabana W, Breucq C, Osteaux M: Use of diffusion-weighted MR imaging in differential diagnosis between intracerebral necrotic tumors and cerebral abscesses [see comments]. AJNR 1999;20(7):1252–1257. Nerlich AG, Haas CJ, Zink A, Szeimies U, Hagedorn HG: Molecular evidence for tuberculosis in an ancient Egyptian mummy [letter; comment]. Lancet 1997;350(9088):1404. Salo WL, Aufderheide AC, Buikstra J, Holcomb TA: Identification of mycobacterium tuberculosis DNA in a pre-Columbian Peruvian mummy. Proc Natl Acad Sci USA 1994;91(6):2091–2094. Centers for Disease Control: Expanded Tuberculosis Surveillance and Tuberculosis Mortality—United States, 1993. Atlanta, CDC, 1994, report no. 43. Hernandez-Pando R, Jeyanathan G, Mengistu G, Aguilar D, Orozco H, Harboe M: Persistence of DNA from mycobacterium tuberculosis
W9418-12.qxd 11/3/03 2:58 PM Page 392
392
108.
109. 110.
111.
112.
113.
114. 115. 116.
117. 118.
119.
120.
121. 122.
123.
124.
125.
126.
127.
128.
129.
130.
Section I
Introduction
in superficially normal lung tissue during latent infection. Lancet 2000;356:2133–2138. Ogawa SK, Smith MA, Brennessel DJ, Lowy FD: Tuberculous meningitis in an urban medical center. Medicine (Baltimore) 1987;66(4): 317–326. Kennedy DH, Fallon RJ: Tuberculous meningitis. JAMA 1979; 241(3):264–268. Klein NC, Damsker B, Hirschman SZ: Mycobacterial meningitis. Retrospective analysis from 1970 to 1983. Am J Med 1985;79(1):29– 34. Kent SJ, Crowe SM, Yung A, Lucas CR, Mijch AM: Tuberculous meningitis: A 30-year review [see comments]. Clin Infect Dis 1993; 17(6):987–994. Zuger A, Lowy FD: Tuberculosis. In Scheld WM, Whitley RJ, Durack DT (eds). Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, pp 417–443. Alvarez S, McCabe WR: Extrapulmonary tuberculosis revisited: A review of experience at boston city and other hospitals. Medicine (Baltimore) 1984;63(1):25–55. Barrett-Connor E: Tuberculous meningitis in adults. South.Med J 1967;60(10):1061–1067. Lincoln EM, Sewell EM: Tuberculosis in Children. New York, McGraw Hill, 1963. Sumaya CV, Simek M, Smith MH, Seidemann MF, Ferriss GS, Rubin W: Tuberculous meningitis in children during the isoniazid era. J Pediatr 1975;87(1):43–49. Traub M, Colchester AC, Kingsley DP, Swash M: Tuberculosis of the central nervous system. Q J Med 1984;53(209):81–100. Nolan CM, Goldberg SV, Buskin SE: Hepatotoxicity associated with isoniazid preventive therapy: A 7-year survey from a public health tuberculosis clinic [see comments]. JAMA 1999;281(11):1014– 1018. Dastur DK, Wadia NH: Spinal meningitides with radiculomyelopathy: part 2. Pathology and pathogenesis. J Neurol Sci 1969;8:261– 293. Wheat LJ, Batteiger BE, Sathapatayavongs B: Histoplasma capsulatum infections of the central nervous system. A clinical review. Medicine (Baltimore) 1990;69(4):244–260. Wheat LJ, French ML, Wass JL: Sarcoidlike manifestations of histoplasmosis. Arch Intern Med 1989;149(11):2421–2426. Smith CE, Beard RR, Whiting EG, et al: Varieties of coccidiodial infection in relation to the epidemiology and control of the disease. Am J Public Health 1946;36:1394–1402. Bouza E, Dreyer JS, Hewitt WL, Meyer RD: Coccidioidal meningitis. An analysis of thirty-one cases and review of the literature. Medicine (Baltimore) 1981;60(3):139–172. Vincent T, Galgiani JN, Huppert M, Salkin D: The natural history of coccidioidal meningitis: VA-armed forces cooperative studies, 1955– 1958. Clin Infect Dis 1993;16(2):247–254. Einstein HE, Holeman CW Jr, Sandidge LL, et al: Coccidioidal meningitis. The use of amphotericin B in treatment. Calif Med 1961;94:339– 343. Ragland AS, Arsura E, Ismail Y, Johnson R: Eosinophilic pleocytosis in coccidioidal meningitis: Frequency and significance. Am J Med 1993;95(3):254–257. Calgioni J: Coccidiodes immitis. In Mandell GL, Bennett JE, Dolan R (eds). Principles and Practices of Infectious Diseases, 5th ed. Philadelphia, Churchill Livingstone, 2000, pp 1394–1402. Gonyea EF: The spectrum of primary blastomycotic meningitis: a review of central nervous system blastomycosis. Ann Neurol 1978; 3(1):26–39. Buechner HA, Clawson CM: Blastomycosis of the central nervous system. II. A Report of nine cases from the Veterans Administration cooperative study. Am Rev Respir Dis 1967;95(5):820–826. Turner S, Kaufman L: Immunodiagnosis of blastomycosis. Semin Respir Infect 1986;1(1):22–28.
131. Montoya JG, Remington JS: Toxoplasma gondii. In Mandell GL, Bennett JE, Dolan R (eds). Principles and Practices of Infectious Diseases, 5th ed. Philadelphia, Churchill-Livingstone, 2000, pp 2858– 2892. 132. Smith KL, Wilson M, Hightower AW, et al: Prevalence of toxoplasma gondii antibodies in US military recruits in 1989: Comparison with data published in 1965. Clin Infect Dis 1996;23(5):1182–1183. 133. Lorberboym M, Wallach F, Estok L, et al: Thallium-201 retention in focal intracranial lesions for differential diagnosis of primary lymphoma and nonmalignant lesions in AIDS patients. J Nucl Med 1998;39(8):1366–1369. 134. Ramsey RG, Gean AD: Neuroimaging of AIDS. I. Central nervous system toxoplasmosis. Neuroimaging Clin North Am 1997;7(2):171– 186. 135. O’Doherty MJ, Barrington SF, Campbell M, Lowe J, Bradbeer CS: PET scanning and the human immunodeficiency virus-positive patient. J Nucl Med 1997;38(10):1575–1583. 136. Rodriguez JC, Martinez MM, Martinez AR, Royo G: Evaluation of different techniques in the diagnosis of toxoplasma encephalitis. J Med Microbiol 1997;46(7):597–601. 137. Luft, Heffner: Toxoplasmic encephalitis in patients with the acquired immunodeficiency syndrome. N Engl J Med 1993;329:995–1000. 138. Maslow C, Matheron S, Saimtog: Cerebral Toxoplasmosis: Assessment of Maintenance Therapy VIII. International Conference on AIDS. Amsterdam, 1992, B123. 139. de Gans J, Portegies P, Reiss P, Troost D, van Gool T, Lange JM: Pyrimethamine alone as maintenance therapy for central nervous system toxoplasmosis in 38 patients with AIDS. J Acquir Immune Defic Syndr 1992;5(2):137–142. 140. Levine AM: Epidemiology clinical characteristics and management of AIDS related lymphoma. Hematol Oncol Clin North Am 1991;5:331– 342. 141. Rosenblum ML, Levy RM, Bredesen DE, So YT, Wara W, Ziegler JL: Primary central nervous system lymphomas in patients with AIDS. Ann Neurol 1988;23[Suppl]:S13–S16. 142. Baumgartner JE, Rachlin JR, Beckstead JH, et al: Primary Central Nervous system lymphomas: Natural history and response to radiation therapy in 55 patients with acquired immunodeficiency syndrome. J Neurosurg 1990;73(2):206–211. 143. Remick SC, Diamond C, Migliozzi JA, et al: Primary central nervous system lymphoma in patients with and without the acquired immune deficiency syndrome. A retrospective analysis and review of the literature. Medicine (Baltimore) 1990;69(6):345–360. 144. Grassi MP, Clerici F, Perin C, et al: Microglial nodular encephalitis and ventriculoencephalitis due to cytomegalovirus infection in patients with AIDS: Two distinct clinical patterns. Clin Infect Dis 1998; 27(3):504–508. 145. Clifford DB, Arribas JR, Storch GA, Tourtellote W, Wippold FJ: Magnetic resonance brain imaging lacks sensitivity for AIDS associated cytomegalovirus encephalitis. J Neurovirol 1996;2(6):397–403. 146. Eidelberg D, Sotrel A, Vogel H, Walker P, Kleefield J, Crumpacker CS III: Progressive polyradiculopathy in acquired immune deficiency syndrome. Neurology 1986;36(7):912–916. 147. McCarther JC HIV Associated CNS syndromes. In Bellman A (ed). Director Course 140: AIDS in the Central Nervous Systems. New York, American Academy of Neurology, 1993, p 5. 148. Jellinger KA, Setinek U, Drlicek M, Bohm G, Steurer A, Lintner F: Neuropathology and general autopsy findings in AIDS during the last 15 years [in process citation]. Acta Neuropathol (Berl) 2000;100(2): 213–220. 149. Navia BA, Jordan BD, Price RW: The AIDS dementia complex: I. clinical features. Ann Neurol 1986;19(6):517–524. 150. Murthy GL, Sahay RK, Srinivasan VR, Upadhaya AC, Shantaram V, Gayatri K: Clinical profile of falciparum malaria in a tertiary care hospital [in process citation]. J Indian Med Assoc 2000;98(4):160–162, 169.
W9418-12.qxd 11/3/03 2:58 PM Page 393
Chapter 12 151. Krogstad DJ: Plasmodium species (malaria). In Mandell GL, Bennett JE, Dolan R (eds). Principles and Practices of Infectious Diseases, 5th ed. Philadelphia, Churchill Livingstone, 2000, pp 2817–2831. 152. Carpio A, Santillan F, Leon P, Flores C, Hauser WA: Is the course of neurocysticercosis modified by treatment with antihelminthic agents? [see comments]. Arch Intern Med 1995;155(18):1982–1988. 153. Padma MV, Behari M, Misra NK, Ahuja GK: Albendazole in single CT ring lesions in epilepsy. Neurology 1994;44(7):1344–1346. 154. Padma MV, Behari M, Misra NK, Ahuja GK: Albendazole in neurocysticercosis. Natl Med J India 1995;8(6):255–258. 155. White AC Jr: Neurocysticercosis: a major cause of neurological disease worldwide. Clin Infect Dis 1997;24(2):101–113. 156. Shandera WX, White AC Jr, Chen JC, Diaz P, Armstrong R: Neurocysticercosis in Houston, Texas. A report of 112 cases. Medicine (Baltimore) 1994;73(1):37–52. 157. Del Bruto OH, Soltelo J, Roman GC: Neurocysticercosis. In Sweats and Zylinger (eds): A Clinical Handbook, 5th ed. Philadelphia, Churchill Livingstone, 1998. 158. Arana-Iniguez R, Lopez-Fernandez JR: Parasitosis of the nervous system, with special reference to echinococcosis. Clin Neurosurg 1966;14:123–144. 159. Diaz JF, Verastegui M, Gilman RH, et al: Immunodiagnosis of human cysticercosis (taenia solium): A field comparison of an antibodyenzyme-linked immunosorbent assay (ELISA), an antigen-ELISA, and an enzyme-linked immunoelectrotransfer blot (EITB) assay in Peru. The Cysticercosis Working Group in Peru (CWG). Am J Trop Med Hyg 1992;46(5):610–615. 160. Zarzosa MP, Orduna Domingo A, Gutierrez P, et al: Evaluation of six serological tests in diagnosis and postoperative control of pulmonary hydatid disease patients. Diagn Microbiol Infect Dis 1999;35(4):255– 262. 161. Okelogba: Hydatid disease research and control in turkana III albendazole in the treatment of inoperable hydatid disease in Kenya: A report on 12 cases. Trans R Soc Tropl Med Hyg 1986;80:193– 195. 162. Davis A, Dixon H, Pawlowski ZS: Multicentre clinical trials of benzimidazole-carbamates in human cystic echinococcosis (phase 2). Bull World Health Organ 1989;67(5):503–508. 163. Herrmann N, Glickman LT, Schantz PM, Weston MG, Domanski LM: Seroprevalence of zoonotic toxocariasis in the United States: 1971– 1973. Am J Epidemiol 1985;122(5):890–896. 164. Marmor M, Glickman L, Shofer F, et al: Toxocara canis infection of children: Epidemiologic and neuropsychologic findings. Am J Public Health 1987;77(5):554–559. 165. Nelson S, Greene T, Ernhart CB: Toxocara canis infection in preschool age children: Risk factors and the cognitive development of preschool children. Neurotoxicol Teratol 1996;18(2):167–174. 166. Mikhael NZ, Montpetit VJ, Orizaga M, Rowsell HC, Richard MT: Toxocara canis infestation with encephalitis. Can J Neurol Sci 1974; 1(2):114–120. 167. Tramont EC: Treponema pallidum. In Mandell, Bennett, Dolan (eds). Principles and Practice of Infectious Disease, 5 ed. Philadelphia, Churchill-Livingstone, 1998, pp 2474–2490. 168. Notfiable diseases/deaths in selected cities information. MMWR 2001;49:51. 169. Lukehart SA, Hook EW III, Baker-Zander SA, Collier AC, Critchlow CW, Handsfield HH: Invasion of the central nervous system by treponema pallidum: Implications for diagnosis and treatment. Ann Intern Med 1988;109(11):855–862. 170. Merritt HH, Adams RD, Solomon HC: Neurosyphilis, 5th ed. New York, Oxford University Press, 1946. 171. Simon RP: Neurosyphilis. Arch Neurol 1985;42(6):606–613. 172. Hooshmand H, Escobar MR, Kopf SW: Neurosyphilis. A study of 241 Patients. JAMA 1972;219(6):726–729. 173. Hotson JR: Modern neurosyphilis: A partially treated chronic meningitis. West J Med 1981;135(3):191–200.
Central Nervous System Infections
393
174. Holmes MD, Brant-Zawadzki MM, Simon RP: Clinical features of meningovascular syphilis. Neurology 1984;34(4):553–556. 175. Horowitz HW, Valsamis MP, Wicher V, et al: Brief report: cerebral syphilitic gumma confirmed by the polymerase chain reaction in a man with human immunodeficiency virus infection [see comments]. N Engl J Med 1994;331(22):1488–1491. 176. Brown ST, Zaidi A, Larsen SA, Reynolds GH: Serological response to syphilis treatment. A new analysis of old data. JAMA 1985;253(9): 1296–1299. 177. Fiumara NJ: Treatment of primary and secondary syphilis. Serological response. JAMA 1980;243(24):2500–2502. 178. Davis LE, Schmitt JW: Clinical significance of cerebrospinal fluid tests for neurosyphilis [see comments]. Ann Neurol 1989;25(1):50–55. 179. Hook EW: Syphilis. In Scheld WM, Whitley RJ, Durack DT (eds): Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997. 180. Merritt HH: The early clinical and laboratory manifestations of syphilis of the central nervous system. N Engl J Med 1940;(223):446– 450. 181. Hahn RD, Clark EG: Asymptomatic neurosyphilis: A review of the literature. Am J Syph Gonorrhea Vener Dis 1946;(30):305–316. 182. Moore JE, Hopkins HH: Asymptomatic neurosyphilis. VI. The prognosis of early and late asymptomatic neurosyphilis. JAMA 1937; (95):1637–1641. 183. Moore JE, Faupel M: A comparison of early and late asymptomatic neurosyphilis. Arch Dermatol Syph 1928. 184. Wilner E, Brody JA: Prognosis of general paresis after treatment. Lancet 1968;2(7583):1370–1371. 185. Hahn RD, Cutler JC, Curtis AC, Gammon G, Heyman A, Johnwick E: Penicillin treatment of asymptomatic central nervous system syphilis. I: Probability of progression to symptomatic neurosyphilis. Arch Dermatol 1956;(74):367–377. 186. Dunlop EM, al Egaily SS, Houang ET: Penicillin levels in blood and CSF achieved by treatment of syphilis. JAMA 1979;241(23):2538– 2540. 187. Mohr JA, Griffiths W, Jackson R, Saadah H, Bird P, Riddle J: Neurosyphilis and penicillin levels in cerebrospinal fluid. JAMA 1976;236(19):2208–2209. 188. Schoth PE, Wolters EC: Penicillin concentrations in serum and CSF during high-dose intravenous treatment for neurosyphilis. Neurology 1987;37(7):1214–1216. 189. Steere AC, Malawista SE, Snydman DR, et al: Lyme arthritis: An epidemic of oligoarticular arthritis in children and adults in three connecticut communities. Arthritis Rheum 1977;20(1):7–17. 190. Steere AC, Malawista SE, Hardin JA, Ruddy S, Askenase W, Andiman W: A. Erythema chronicum migrans and lyme arthritis. The enlarging clinical spectrum. Ann Intern Med 1977;86(6):685–698. 191. Steere AC, Broderick TF, Malawista SE: Erythema chronicum migrans and lyme arthritis: Epidemiologic evidence for a tick vector. Am J Epidemiol 1978;108(4):312–321. 192. Wallis RC, Brown SE, Kloter KO, Main AJ Jr: Erythema chronicum migrans and lyme arthritis: Field study of ticks. Am J Epidemiol 1978;108(4):322–327. 193. Burgdorfer W: Discovery of the lyme disease spirochete and its relation to tick vectors. Yale J Biol Med 1984;57(4):515–520. 194. Asbrink E, Hovmark A: Early and late cutaneous manifestations in ixodes-borne borreliosis (erythema migrans borreliosis, lyme borreliosis). Ann NY Acad Sci 1988;539:4–15. 195. Steere AC, Bartenhagen NH, Craft JE, et al: The early clinical manifestations of Lyme disease. Ann Intern Med 1983;99(1):76–82. 196. Steere AC: [Lyme disease—a tick-transmitted spirochete infection]. Ter Arkh 1987;59(4):32–34. 197. Berger BW: Erythema chronicum migrans of lyme disease. Arch Dermatol 1984;120(8):1017–1021. 198. Berger BW: Dermatologic manifestations of Lyme disease. Rev Infect Dis 1989;11(Suppl 6):S1475–S1481.
W9418-12.qxd 11/3/03 2:58 PM Page 394
394
Section I
Introduction
199. Steere AC: Lyme disease [see comments]. N Engl J Med 1989; 321(9):586–596. 200. Vlay SC: Cardiac manifestations. In Coyle PK (ed). Lyme Disease. St. Louis, Mosby-Year Book, 1993, pp 68–92. 201. van der Linde MR, Ballmer PE: Lyme carditis. In Weber K, Burgdorfer W (eds). Aspects of Lyme Borreliosis. Berlin, SpringerVerlag, 1993, pp 131–151. 202. Preac-Mursic V, Weber K, Pfister HW, et al: Survival of borrelia burgdorferi in antibiotically treated patients with lyme borreliosis. Infection 1989;17(6):355–359. 203. Keller TL, Halperin JJ, Whitman M: PCR detection of borrelia burgdorferi dna in cerebrospinal fluid of lyme neuroborreliosis patients [see comments]. Neurology 1992;42(1):32–42. 204. Shadick NA, Phillips CB, Logigian EL, et al: The long-term clinical outcomes of Lyme disease. A population-based retrospective cohort study [see comments]. Ann Intern Med 1994;121(8):560–567. 205. Ackermann R, Rehse-Kupper B, Gollmer E, Schmidt R: Chronic neurologic manifestations of erythema migrans borreliosis. Ann NY Acad Sci 1988;539:16–23. 206. Kuntzer T, Bogousslavsky J, Miklossy J, Steck AJ, Janzer R, Regli F: Borrelia rhombencephalomyelopathy. Arch Neurol 1991;48(8):832– 836. 207. Kohler J, Kern U, Kasper J, Rhese-Kupper B, Thoden U: Chronic central nervous system involvement in lyme borreliosis. Neurology 1988;38(6):863–867. 208. Pachner AR: borrelia burgdorferi in the nervous system: the new “great imitator”. Ann NY Acad Sci 1988;539:56–64. 209. Hanny PE, Hauselmann HJ: [Lyme disease from the neurologist’s viewpoint]. Schweiz Med Wochenschr 1987;117(24):901–915. 210. Meurers B, Kohlhepp W, Gold R, Rohrbach E, Mertens HG: Histopathological findings in the central and peripheral nervous systems in neuroborreliosis. A report of three cases. J Neurol 1990;237(2):113–116. 211. Kohlhepp W, Mertens HG, Oschmann P, Rohrbach E: [Acute and chronic diseases in transmitted borreliosis by tick bite]. Nervenarzt 1987;58(9):557–563. 212. Rafto SE, Milton WJ, Galetta SL, Grossman RI: Biopsy-confirmed CNS lyme disease: MR appearance at 1.5 T. AJNR 1990;11(3):482–484. 213. Mokry M, Flaschka G, Kleinert G, Kleinert R, Fazekas F, Kopp W: Chronic lyme disease with an expansive granulomatous lesion in the cerebellopontine angle. Neurosurgery 1990;27(3):446–451. 214. Martin R, Kohlhepp W, Mertens HG: Chronic central nervous system involvement. In Weber K, Burgdorfer W (eds). Aspects of Lyme Borreliosis. Berlin, Springer-Verlag, 1993, pp 205–218. 215. Logigian EL, Kaplan RF, Steere AC: Chronic neurologic manifestations of lyme disease [see comments]. N Engl J Med 1990;323(21):1438– 1444. 216. Reik L Jr: Stroke due to lyme disease. Neurology 1993;43(12):2705– 2707. 217. Afzelius A: Erythema chronicum migrans. Acta Derm Venereol (Stockh) 1921;2:120–125. 218. Lennhoff C: Spirochaetes in aetiologically obscure diseases. Acta Derm Venereol (Stockh) 1948;28:295–324. 219. Hollstrom E: Successful treatment of erythema chronicum migrans afzelius. Acta Derm Venereol (Stockh) 1951;31:235–243. 220. Reik L: Lyme Disease and the Nervous System. New York, Thieme Medical Publishers, 1991. 221. Weder B, Wiedersheim P, Matter L, Steck A, Otto F: Chronic progressive neurological involvement in borrelia burgdorferi infection. J Neurol 1987;234(1):40–43. 222. Halperin JJ, Luft BJ, Anand AK, et al: Lyme neuroborreliosis: central nervous system manifestations [see comments]. Neurology 1989;39(6):753–759. 223. Halperin JJ, Krupp LB, Golightly MG, Volkman DJ: Lyme borreliosisassociated encephalopathy [see comments]. Neurology 1990; 40(9):1340–1343.
224. Halperin JJ, Pass HL, Anand AK, Luft BJ, Volkman DJ, Dattwyler RJ: Nervous system abnormalities in lyme disease. Ann NY Acad Sci 1988;539:24–34. 225. Logigian EL, Steere AC: Clinical and electrophysiologic findings in chronic neuropathy of lyme disease. Neurology 1992;42(2):303– 311. 226. Halperin JJ, Luft BJ, Volkman DJ: Lyme neuroborreliosis: peripheral nervous system manifestations. Brain 1990;(113):1207–1221. 227. Halperin JJ, Volkman DJ, Luft BJ, Dattwyler RJ: Carpal tunnel syndrome in lyme borreliosis. Muscle Nerve 1989;12(5):397–400. 228. Steere AC, Pachner AR, Malawista SE: Neurologic abnormalities of Lyme disease: Successful treatment with high-dose intravenous penicillin. Ann Intern Med 1983;99(6):767–772. 229. Pfister H-W, Kristoferitsch W, Skoldenberg B: Therapy of Lyme Neuroborreliosis. Aspects of Lyme borreliosis. Berlin, Springer-Verlag, 1993, pp 328–339. 230. Pfister HW, Preac-Mursic V, Wilske B, Einhaupl KM: Cefotaxime vs penicillin g for acute neurologic manifestations in lyme borreliosis. A prospective randomized study. Arch Neurol 1989;46(11):1190–1194. 231. Pfister HW, Preac-Mursic V, Wilske B, Schielke E, Sorgel F, Einhaupl KM: Randomized comparison of ceftriaxone and cefotaxime in lyme neuroborreliosis. J Infect Dis 1991;163(2):311–318. 232. Karlsson M, Hammers-Berggren S, Lindquist L, Stiernstedt G, Svenungsson B: Comparison of intravenous penicillin G and oral doxycycline for treatment of lyme neuroborreliosis. Neurology 1994;44(7):1203–1207. 233. Kornblatt AN, Steere AC, Brownstein DG: Infection in rabbits with the lyme disease spirochete. Yale J Biol Med 1984;57(4):613–616. 234. Pfister H-W, Einhaupl KM, Franz P, Garner C: Corticosteroids for radicular pain in bannwarth’s syndrome: a double-blind, randomized, placebo-controlled trial. Ann NY Acad Sci 1988;(539):485–487. 235. Woodward TE: Rocky Mountain spotted fever: epidemiological and early clinical signs are keys to treatment and reduced mortality. J Infect Dis 1984;150(4):465–468. 236. Helmick CG, Bernard KW, D’Angelo LJ: Rocky Mountain spotted fever: Clinical, laboratory, and epidemiological features of 262 cases. J Infect Dis 1984;150(4):480–488. 237. Marrie TJ, Raoult D: Rickettsial infections of the central nervous system. Semin Neurol 1992;12(3):213–224. 238. Bakken JS, Dumler JS, Chen SM, Eckman MR, Van Etta LL, Walker D: H. Human granulocytic ehrlichiosis in the upper midwest United States. A new species emerging? [see comments]. JAMA 1994;272(3): 212–218. 239. Everett ED, Evans KA, Henry RB, McDonald G: Human ehrlichiosis in adults after tick exposure. Diagnosis using polymerase chain reaction. Ann Intern Med 1994;120(9):730–735. 240. Carithers HA, Margileth AM: Cat-scratch disease. Acute encephalopathy and other neurologic manifestations. Am J Dis Child 1991; 145(1):98–101. 241. Margileth AM, Wear DJ, English CK: Systemic cat scratch disease: Report of 23 patients with prolonged or recurrent severe bacterial infection. J Infect Dis 1987;155(3):390–402. 242. Koehler JE, Sanchez MA, Garrido CS, et al: Molecular epidemiology of bartonella infections in patients with bacillary angiomatosispeliosis [see comments]. N Engl J Med 1997;337(26):1876–1883. 243. Pollock S, Lewis PD, Kendall B: Whipple’s disease confined to the nervous system. J Neurol Neurosurg Psychiatry 1981;44(12):1104– 1109. 244. Kitamura T: Brain involvement in Whipple’s disease: A case report. Acta Neuropathol (Berl) 1975;33(3):275–278. 245. Romanul FC, Radvany J, Rosales RK: Whipple’s disease confined to the brain: A case studied clinically and pathologically. J Neurol Neurosurg Psychiatry 1977;40(9):901–909. 246. Wroe SJ, Pires M, Harding B, Youl BD, Shorvon S: Whipple’s disease confined to the CNS presenting with multiple intracerebral mass lesions. J Neurol Neurosurg Psychiatry 1991;54(11):989–992.
W9418-12.qxd 11/3/03 2:58 PM Page 395
Chapter 12 247. Gartner J: Whipple’s disease of the central nervous system, associated with ophthalmoplegia externa and severe asteroid hyalitis. A clinicopathologic study. Doc Ophthalmol 1980;49(1):155–187. 248. Halperin JJ, Landis DM, Kleinman GM: Whipple disease of the nervous system. Neurology 1982;32(6):612–617. 249. Stoupel N, Monseu G, Pardoe A, Heimann R, Martin JJ: Encephalitis with myoclonus in whipple’s disease. J Neurol Neurosurg Psychiatry 1969;32(4):338–343. 250. Schwartz MA, Selhorst JB, Ochs AL, et al: Oculomasticatory myorhythmia: A unique movement disorder occurring in Whipple’s disease. Ann Neurol 1986;20(6):677–683. 251. Wietholter H, Dichgans J: Diagnosis of cerebral Whipple’s disease by cerebrospinal fluid cytology. Arch Psychiatry Nervenkr 1982; 231(3):283–287. 252. Ratnaike RN: Whipple’s disease. Postgrad Med J 2000;76(902): 760–766. 253. Keinath RD, Merrell DE, Vlietstra R, Dobbins WO III: Antibiotic treatment and relapse in Whipple’s disease. Long-term follow-up of 88 patients. Gastroenterology 1985;88(6):1867–1873. 254. Singer R: Diagnosis and treatment of Whipple’s disease. Drugs 1998;55(5):699–704. 255. Durack DT: Amebic infections. In Scheld WM, Whitley RJ, Durack DT (eds). Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, pp 831–844. 256. Seidel JS, Harmatz P, Visvesvara GS, Cohen A, Edwards J, Turner J: Successful treatment of primary amebic meningoencephalitis. N Engl J Med 1982;306(6):346–348. 257. Brown RL: Successful treatment of primary amebic meningoencephalitis [see comments]. Arch Intern Med 1991;151(6):1201– 1202. 258. Poungvarin N, Jariya P: The fifth nonlethal case of primary amoebic meningoencephalitis. J Med Assoc Thailand 1991;74(2):112–115. 259. Holman RC, Janssen RS, Buehler JW, Zelasky MT, Hooper WC: Epidemiology of progressive multifocal leukoencephalopathy in the United States: Analysis of national mortality and AIDS surveillance data. Neurology 1991;41(11):1733–1736. 260. Berger JR, Simpson DM: Neurologic complications of AIDS. In Scheld WM, Whitley RJ, Durack DT (eds). Infections of the Central Nervous System, 2nd ed. Philadelphia, Lippincott-Raven, 1997, pp 255–271. 261. Gardner SD, Field AM, Coleman DV, Hulme B: New human papovavirus (B.K.) isolated from urine after renal transplantation. Lancet 1971;1(7712):1253–1257. 262. Padgett BL, Walker DL, ZuRhein GM, Eckroade RJ, Dessel BH: Cultivation of papova-like virus from human brain with progressive multifocal leucoencephalopathy. Lancet 1971;1(7712):1257–1260. 263. Yang RC, Wu R: BK virus DNA: Complete nucleotide sequence of a human tumor virus. Science 1979;206(4417):456–462. 264. Frisque RJ, Bream GL, Cannella MT: Human polyomavirus JC virus genome. J Virol 1984;51(2):458–469. 265. Flaegstad T, Ronne K, Filipe AR, Traavik T: Prevalence of anti BK virus antibody in Portugal and Norway. Scand J Infect Dis 1989;21(2):145– 147. 266. Dei R, Marmo F, Corte D, Sampietro MG, Franceschini E, Urbano P: Age-related changes in the prevalence of precipitating antibodies to BK virus in infants and children. J Med Microbiol 1982;15(3):285– 291. 267. Chesters PM, Heritage J, McCance DJ: Persistence of DNA sequences of BK virus and JC virus in normal human tissues and in diseased tissues. J Infect Dis 1983;147(4):676–684. 268. Tominaga T, Yogo Y, Kitamura T, Aso Y: Persistence of archetypal JC virus DNA in normal renal tissue derived from tumor-bearing patients. Virology 1992;186(2):736–741. 269. Heritage J, Chesters PM, McCance DJ: The persistence of papovavirus BK DNA sequences in normal human renal tissue. J Med Virol 1981;8(2):143–150.
Central Nervous System Infections
395
270. Dubois V, Dutronc H, Lafon ME, et al: Latency and reactivation of JC virus in peripheral blood of human immunodeficiency virus type 1infected patients. J Clin Microbiol 1997;35(9):2288–2292. 271. Koralnik IJ, Boden D, Mai VX, Lord CI, Letvin NL: JC virus DNA load in patients with and without progressive multifocal leukoencephalopathy. Neurology 1999;52(2):253–260. 272. Elsner C, Dorries K: Evidence of human polyomavirus BK and JC infection in normal brain tissue. Virology 1992;191(1):72–80. 273. Quinlivan EB, Norris M, Bouldin TW, et al: Subclinical central nervous system infection with JC virus in patients with AIDS. J Infect Dis 1992;166(1):80–85. 274. Ferrante P, Caldarelli-Stefano R, Omodeo-Zorini E, Vago L, Boldorini R, Costanzi G: PCR detection of JC virus DNA in brain tissue from patients with and without progressive multifocal leukoencephalopathy. J Med Virol 1995;47(3):219–225. 275. White FA III, Ishaq M, Stoner GL, Frisque RJ: JC virus DNA is present in many human brain samples from patients without progressive multifocal leukoencephalopathy. J Virol 1992;66(10):5726–5734. 276. Astrom K-E, Mancall EL, Richardson EP Jr: Progressive multifocal leukoencephalopathy: A hitherto unrecognized complication of chronic lymphatic leukemia and Hodgkin’s disease. Brain 1958;(81): 93–110. 277. Richardson EP Jr: Progressive multifocal leukoencephalopathy. N Engl J Med 1961;(265):815–823. 278. Newton P, Aldridge RD, Lessells AM, Best PV: Progressive multifocal leukoencephalopathy complicating systemic lupus erythematosus. Arthritis Rheum 1986;29(3):337–343. 279. Rockwell D, Ruben FL, Winkelstein A, Mendelow H: Absence of immune deficiencies in a case of progressive multifocal leukoencephalopathy. Am J Med 1976;61(3):433–436. 280. Bolton CF, Rozdilsky B: Primary progressive multifocal leukoencephalopathy. A case report. Neurology 1971;21(1):72–77. 281. Fermaglich J, Hardman JM, Earle KM: Spontaneous progressive multifocal leukoencephalopathy. Neurology 1970;20(5):479–484. 282. Holman RC, Janssen RS, Buehler JW, Zelasky MT, Hooper WC: Epidemiology of progressive multifocal leukoencephalopathy in the United States: Analysis of national mortality and AIDS surveillance data [see comments]. Neurology 1991;41(11):1733–1736. 283. Berger JR, Kaszovitz B, Post MJ, Dickinson G: Progressive multifocal leukoencephalopathy associated with human immunodeficiency virus infection. A review of the literature with a report of sixteen cases. Ann Intern Med 1987;107(1):78–87. 284. Brooks BR, Walker DL: Progressive multifocal leukoencephalopathy. Neurol Clin 1984;2(2):299–313. 285. Gillespie SM, Chang Y, Lemp G, et al: Progressive multifocal leukoencephalopathy in persons infected with human immunodeficiency virus, San Francisco, 1981–1989. Ann Neurol 1991;30(4):597–604. 286. von Einsiedel RW, Fife TD, Aksamit AJ, et al: Progressive multifocal leukoencephalopathy in AIDS: A clinicopathologic study and review of the literature. J Neurol 1993;240(7):391–406. 287. Hair LS, Nuovo G, Powers JM, Sisti MB, Britton CB, Miller JR: Progressive multifocal leukoencephalopathy in patients with human immunodeficiency virus. Hum Pathol 1992;23(6):663–667. 288. Krupp LB, Lipton RB, Swerdlow ML, Leeds NE, Llena J: Progressive multifocal leukoencephalopathy: Clinical and radiographic features. Ann Neurol 1985;17(4):344–349. 289. Silverman L, Rubinstein LJ: Electron microscopic observations on a case of progressive multifocal leukoencephalopathy. Acta Neuropathol (Berl) 1965;5(2):215–224. 290. ZuRhein GM, Chou S-M: Particles resembling papova viruses in human cerebral demyelinating disease. Science 1965;(148):1477– 1479. 291. McGuire D, Barhite S, Hollander H, Miles M: JC virus DNA in cerebrospinal fluid of human immunodeficiency virus-infected patients: predictive value for progressive multifocal leukoencephalopathy. Ann Neurol 1995;37(3):395–399.
W9418-12.qxd 11/3/03 2:58 PM Page 396
396
Section I
Introduction
292. Vago L, Cinque P, Sala E, et al: JCV-DNA and BKV-DNA in the CNS tissue and CSF of AIDS patients and normal subjects. Study of 41 cases and review of the literature. J Acquir Immune Defic Syndr Hum Retrovirol 1996;12(2):139–146. 293. Hammarin AL, Bogdanovic G, Svedhem V, Pirskanen R, Morfeldt L, Grandien M: Analysis of PCR as a tool for detection of JC virus DNA in cerebrospinal fluid for diagnosis of progressive multifocal leukoencephalopathy. J Clin Microbiol 1996;34(12):2929–2932. 294. Cinque P, Vago L, Dahl H, et al: Polymerase chain reaction on cerebrospinal fluid for diagnosis of virus-associated opportunistic diseases of the central nervous system in HIV-infected patients. AIDS 1996;10(9):951–958. 295. De Luca A, Cingolani A, Linzalone A, et al: Improved detection of JC virus DNA in cerebrospinal fluid for diagnosis of AIDS-related progressive multifocal leukoencephalopathy. J Clin Microbiol 1996;34(5):1343–1346. 296. Fong IW, Britton CB, Luinstra KE, Toma E, Mahony JB: Diagnostic value of detecting JC virus DNA in cerebrospinal fluid of patients with progressive multifocal leukoencephalopathy. J Clin Microbiol 1995;33(2):484–486. 297. Dorries K, Arendt G, Eggers C, Roggendorf W, Dorries R: Nucleic acid detection as a diagnostic tool in polyomavirus JC induced progressive multifocal leukoencephalopathy. J Med Virol 1998;54(3):196– 203. 298. Tognon M, Martini F, Iaccheri L, Cultrera R, Contini C: Investigation of the simian polyomavirus SV40 as a potential causative agent of human neurological disorders in AIDS patients. J Med Microbiol 2001;50(2):165–172. 299. Dodge P: A case study: The use cidofovir for management of progressive multifocal leukoencephalopathy. J Assoc Nurses AIDS Care 1999;10(4):70–74. 300. Hall CD, Dafni U, Simpson D, et al: Failure of cytarabine in progressive multifocal leukoencephalopathy associated with human immunodeficiency virus infection. AIDS clinical trials group 243 team [see comments]. N Engl J Med 1998;338(19):1345–1351. 301. De Luca A, Giancola ML, Ammassari A, et al: Potent anti-retroviral therapy with or without cidofovir for AIDS-associated progressive multifocal leukoencephalopathy: Extended follow-up of an observational study. J Neurovirol 2001;7(4):364–368. 302. Mackenzie AR, Laing RBS, Smith CC, Kaar GF, Smith FW: Spinal epidural abscess: The importance of early diagnosis and treatment. J Neurol Neurosurg Psychiatry 1998;65:209–212. 303. Martin RJ, Yuan HA: Neurosurgical care of spinal epidural, subdural, and intramedullary abscesses and arachnoiditis. Spinal Infections 1996;27:25–136. 304. Hlavin ML, Kaminski HJ, Ross JS, Ganz E: Spinal epidural abscess: A ten year perspective. Neurosurgery 1990;27:177–184. 305. Maslen DR, Jones SR, Crislip MA, Bracis R, Dworkin RJ, Flemming JE: Spinal epidural abscess: Optimizing patient care. Arch Intern Med 1993;153:1713–1721. 306. Auletta JJ, Chandy JC: Spinal epidural abscesses in children: A 15-year experience and review of the literature. Clin Infect Dis 2001;32:9–16. 307. Khanna RK, Ghaus MM, Rock JP, Rosenblum ML: Spinal epidural abscess: Evaluation of 5 factors influencing outcome. Neurosurgery 1996;39:958–964. 308. Obrador GT, Levenson DJ: Spinal epidural abscess in hemodialysis patients: Report of three cases and review of the literature. Am J Kidney Dis 1996;27:75–83. 309. Wheeler D, Keiser P, Rigamonti D, Keay S: Medical management of spinal epidural abscesses: Case report and review. Clin Infect Dis 1992;15:22–27. 310. Feldenzer JA, McKeever PE, Schaberg DR, Campbell JA, Hoff JT: The pathogenesis of spinal epidural abscess: Microangiopathic studies in an experimental model. J Neurosurg 1998;69:110–114. 311. Curling OD, Gower DJ, McWhorter JM: Changing concepts in spinal epidural abscess: A report of 29 cases. Neurosurgery 1990;27:185–192.
312. Digby JM, Kersley JB: Pyogenic non-tuberculous spinal infection: an analysis of thirty cases. J Bone Joint Surg [Br] 1979;61:47–55. 313. Khan IA, Vaccaro AR, Zlotolow DA: Management of vertebral diskitis and osteomyelitis. Orthopedics 1999;22:758–765. 314. Strausbaugh LJ: Vertebral osteomyelitis: How to differentiate it from other causes of back and neck pain. Postgrad Med J 1995;97:147– 154. 315. Rezai AR, Woo HH, Errico TJ, Cooper PR: Contemporary management of spinal osteomyelitis. Neurosurgery 1999;44:1018–1025. 316. Nussbaum ES, Rockswold GL, Bergman TA, Erickson DL, Seljeskog EL: Spinal tuberculosis: A diagnostic and management challenge. J Neurosurg 1995:83:243–247. 317. Sapico FL: Microbiology and antimicrobial therapy of spinal infections. Orthopedic Clinics of North America 1996;27:9–13. 318. Ozuna RM, Delamarter RB: Pyogenic vertebral osteomyelitis and postsurgical disc space infections. Orthopedic Clinics of North America 1996;27:87–94. 319. Broner FA, Garland DE, Zigler JE: Spinal infections in the immunocompromised host. Orthopedic Clinics of North America 1996;27:37–46. 320. Cahill DW, Love LC, Rechtine GR: Pyogenic osteomyelitis of the spine in the elderly. J Neurosurg 1991;74:878–886. 321. Klein JD, Garfin SR: Nutritional status in the patient with spinal infection. Orthopedic Clinics of North America 1996;27:33–36. 322. Sarria JC, Chutkan NB, Figueroa JE, Hull A: Atypical mycobacterial vertebral osteomyelitis: Case report and review. Clin Infect Dis 1998:26:503–505. 323. Rath SA, Neff U, Schneider O, Richter H: Neurosurgical management of thoracic and lumbar vertebral osteomyelitis and discitis in adults: A review of 43 consecutive surgically treated patients. Neurosurgery 1996;38:926–933. 324. Turpin S, Lambert R: Role of scintigraphy in musculoskeletal and spinal infections. Radiologic Clinics of North America 2001;39:169–189. 325. Vaccaro AR, Shah SH, Schweiter ME, Rosenfeld JF, Cotler JM: MRI description of vertebral osteomyelitis, neoplasm, and compression fracture. Orthopedics 1999;22:67–73. 326. Martinez M, Lee AS, Hellinger WC, Kaplan J: Vertebral Aspergillus osteomyelitis and acute diskitis in patients with chronic obstructive pulmonary disease. Mayo Clin Proc 1999;74:579–583. 327. Vinas FC, King PK, Diaz FG: Spinal Aspergillus osteomyelitis. Clin Infect Dis 1999;28:1223–1229. 328. Alleyne CH, Hassan M, Zambramski JM: The efficacy and cost of prophylactic and periprocedural antibiotics in patients with external ventricular drains. Neurosurgery 2000;47:1124–1127. 329. Khanna RK, Rosenblum ML, Rock JP, Malik GM: Prolonged external ventricular drainage with percutaneous long-tunnel ventriculostomies. J Neurosurg 1995;83:791–794. 330. Cummings R: Understanding external ventricular drainage. J Neurosci Nurs 1992;24:84–87. 331. Chan K, Mann KS: Prolonged therapeutic external ventricular drainage: A prospective study. Neurosurgery 1988;23:436–438. 332. Narayan RK, Kishore PRS, Becker DP, et al: Intracranial pressure: To monitor or not to monitor? A review of our experience with severe head injury. J Neurosurg 1982;56:650–659. 333. Mayhill CG, Archer NH, Lamb VA, et al: Ventriculostomy-related infections: A prospective epidemiologic study. N Engl J Med 1984;310:553–559. 334. Winfield JA, Rosenthal P, Kanter RK, Casella G: Duration of intracranial pressure monitoring does not predict daily risk of infectious complications. Neurosurgery 1993;33:424–431. 335. Rodvold KA: Therapeutic considerations for infections caused by Staphylococcus Epidermidis. Pharmacotherapy 1988(Suppl 8):14S–18S. 336. Darouiche RO, Raad II, Heard SO, et al: A comparison of two antimicrobial-impregnated central venous catheters. Catheter Study Group. N Engl J Med 1999;340:1–8.
W9418-13.qxd 11/3/03 2:51 PM Page 397
Chapter 13 Acute Stroke and Other Neurologic Emergencies David M. Greer, MD, MA
Intensive Care Management of Cerebrovascular Disorders Introduction The focus of this section will be cerebrovascular disorders necessitating intensive care unit (ICU) level of care, including a discussion of ischemic stroke pathophysiology and subtypes, relative outcomes, and potential treatment modalities. Consideration will be given to current therapies in ischemic stroke, including intravenous and intra-arterial thrombolysis, blood pressure management, and anticoagulation. Next will be a review of the clinical features, diagnosis, and treatment of primary intracerebral hemorrhage, including deep, lobar, and primary intraventricular hemorrhage. Finally, the current diagnosis and treatment options for cerebral venous thrombosis will be discussed. Hemorrhagic cerebrovascular disease is reviewed elsewhere (see Chapter 6). Ischemic Stroke Pathophysiology Ischemic stroke can be separated into three broad categories (Fig. 13-1; Table 13-1): large vessel atherothrombosis, small vessel or lacunar infarction, and cerebral embolism, up to two thirds of which may have an unknown clinical source (cryptogenic embolus) (Fig. 13-2). The determination of the etiology in the acute setting is imperative, because progression of stroke and clinical worsening can occur within minutes to hours of presentation, depending on the cause. For example, a patient presenting with subtle speech changes
and right hand weakness due to a small left middle cerebral artery (MCA) infarct may experience a subsequent severe aphasia and dense right hemiparesis if the underlying lesion is a critical stenosis of the left internal carotid artery (ICA) with recurrent embolism or progressive hemispheric hypoperfusion. On the other hand, if the infarct is secondary to a small cardiogenic embolus to the hand motor cortex region, the symptoms will likely be maximal at the onset, with a relatively low probability for progression. Large vessel atherothrombosis encompasses approximately 15% of all strokes; 9% are of extracranial ICA origin, and 6% are due to intracranial atheromatous disease.1 The typical clinical presentation is a sudden onset of focal neurologic deficits, with potential stepwise progression of symptoms referable to the same arterial distribution. Events in the anterior circulation present with cortical surface symptoms (aphasia, apraxia, neglect) and/or deep white matter or basal ganglia symptoms (weakness, sensory changes, movement disorders). Posterior circulation events present with brainstem (cranial nerve, motor), cerebellar (ataxia), thalamic (sensory, aphasia), or occipital symptoms (vision changes, personality changes, memory disturbance). In stroke secondary to large vessel disease, the pathophysiology is secondary to atherosclerosis in a major extracranial or intracranial artery, with subsequent ulcer formation and thrombosis. This leads to recurrent thromboemboli, as well as arterial stenosis or occlusion.2 Stroke occurs secondarily either by artery-to-artery embolism or by hypoperfusion causing a low flow state, in which there may be blood pressure–sensitive fluctuations in the clinical examination, causing progressing or regressing stereotyped symp397
W9418-13.qxd 11/3/03 2:51 PM Page 398
398
Section I
Introduction
Figure 13-1. Classification of stroke by mechanism, with frequency estimates of the abnormalities. Note that about 30% of stroke is cryptogenic. (From Albers GW, et al: Chest 2001;119:300S, with permission.)
toms. The most common site of extracranial stenosis is the origin of the ICA, followed by the origin of the great vessels, with the vertebral artery origin more common than the common carotid origin.3,4 In the intracranial circulation, the most common sites for atherosclerotic lesions are the distal vertebral artery, the proximal to midbasilar artery, the siphon portion of the ICA, and the MCA stem. Small vessel or lacunar infarction encompasses approximately 25% of all ischemic strokes. Patients present with focal neurologic deficits in the territory of a single penetrating artery arising from the distal vertebral artery, basilar artery, MCA stem, or the arteries of the circle of Willis. The mechanism of vascular compromise in this subtype is thought to be progressive lipohyalinosis, in which the midportion of the penetrating artery is concentrically occluded.5 Hypertension is a leading factor in the development of lacunae, with possible contributions from diabetes mellitus, dyslipidemia, and smoking. The infarcts are by definition
less than 1.5 cm in horizontal diameter. Classic lacunar syndromes include pure motor hemiparesis (caused by infarction in the internal capsule or basis pontis), pure hemisensory symptoms (caused by infarction in the ventroposterolateral [VPL] thalamic nucleus), dysarthria-clumsy hand syndrome (with pontine or internal capsule infarcts), and ataxia-hemiparesis (pontine infarct).
Table 13-1 Broad Categorization of Stroke Types* Type of Stroke Large-vessel atherothrombotic Due to internal carotid artery stenosis Small-vessel (lacunar) Embolic Due to atrial fibrillation Other (Due to dissection or other causes)
Proportion of Strokes (%) 15 9 25 60 15 3
*The data are from the Stroke Data Bank of the National Institute of Neurological and Communicative Disorders and Stroke and the Framingham Study. The percentages do not total 100 because of a modification of the categories of stroke used. From Kistler JP, Furie KL: N Engl J Med 2000;342:1743, with permission.
Figure 13-2. Common sites of arterial and cardiac lesions causing ischemic stroke. (From Albers GW, et al: Chest 2001;119:300S, with permission.)
W9418-13.qxd 11/3/03 2:51 PM Page 399
Chapter 13
An important consideration in the stroke patient presenting with a lacunar syndrome is the possibility of atherothrombotic disease in the parent vessel, causing occlusion of the penetrator vessel by thrombotic occlusion at the origin of the penetrator artery. These patients merit early evaluation of their intracranial circulation to exonerate the parent vessel in question. In addition, lacunar infarction may be the initial presentation of an embolic event, and a search for an embolic source should be undertaken. In a study by Ay and associates,6 patients with acute stroke presenting with a classic lacunar syndrome were studied with diffusionweighted magnetic resonance imaging (MRI), and 16% (10 of 62) were found to have evidence of multiple emboli in addition to the suspected syndrome-causing lacune, or “index lesion.” As a general rule, patients with small vessel infarcts have a relatively better clinical outcome than those with strokes secondary to large vessel atherothrombosis. However, in some cases, depending on the location of the stroke, patients may be left with pronounced motor and/or cranial nerve deficits. The mainstay of prevention is the control of risk factors, such as hypertension and smoking, as well as use of antiplatelet agents. Studies evaluating the use of oral anticoagulant therapy in lacunar infarction failed to show any benefit.7 Approximately 60% of all ischemic strokes are caused by cerebral embolism, only one third of which have known clinical sources.8 Cerebral emboli present with sudden, focal neurologic deficits in the territory of an extracerebral artery. They may present with cortical and/or subcortical symptoms, as well as cerebellar hemispheric or diencephalic symptoms. By definition, there is a lack of intrinsic pathology in the parent artery or arteries, allowing a direct conduit from a more proximal source, such as the heart or aortic arch. The clinical outcome depends on the size and location of the infarct, and cerebral emboli are thought to have a higher rate of hemorrhagic conversion in comparison to ischemic stroke caused by other subtypes.9 Cerebral emboli may originate from multiple sources,10 including atrial fibrillation, intrinsic or mechanical valvular disease, intracardiac thrombus (atrial or ventricular), atrial myxoma, dilated cardiomyopathy, patent foramen ovale and/or atrial septal aneurysm, aortic arch athromatous disease, and marantic or bacterial endocarditis. It is important to seek out the definitive source, because it may have a profound impact on the management of secondary stroke prevention. For example, most sources of cardiogenic emboli are treated with oral anticoagulation; however, this may be contraindicated in a number of conditions such as with bacterial endocarditis,11 atrial myxoma,12 and potentially aortic arch atheromatous disease, because these conditions may have a higher rate of hemorrhagic conversion. Despite the number of potential mechanisms by which a cerebral embolus can arise, approximately two thirds of embolic infarcts remain cryptogenic. The optimal treatment of cryptogenic emboli has been a controversial subject for
Acute Stroke and Other Neurologic Emergencies
399
years. Although the recent Warfarin vs. Aspirin in Recurrent Stroke Study (WARSS)7 showed no significant benefit for warfarin over aspirin for the primary endpoint, it provided some suggestion of potential benefit in favor of oral anticoagulant therapy over aspirin in the selected subpopulation of patients with cryptogenic emboli and a cortical element to their infarct. However, due to the relatively small number of patients with this stroke subtype, as well as the design of this study, this finding did not reach statistical significance in the WARSS study, and further studies of this subgroup are needed. Several other etiologies of ischemic stroke bear mentioning. Arterial dissection is an important consideration, especially in young patients or following a traumatic head or neck injury.13 Dissections typically arise at the petrous portion of the ICA, or at the C1 to C2 level in the vertebral artery. Dissections in the petrous carotid artery can give rise to pain involving the ipsilateral eye and posterior pharynx, as well as a Horner’s syndrome. Vertebral artery dissections cause posterior cervical and auricular pain, and may present with a stroke in the distribution of the posterior inferior cerebellar artery, also at times causing a Horner’s syndrome. Thrombus may form at the site of an intimal tear, extending into the media, with subsequent artery-to-artery embolism. Low-flow or “watershed” infarction can also occur with severe vessel narrowing or occlusion. Dissection may also lead to subarachnoid hemorrhage when a pseudoaneurysm forms after the artery passes intradurally.14 Cerebral angiitis is another rare cause of stroke, and may cause lesions that are ischemic, hemorrhagic, or both.15 The etiology may be as part of a systemic angiitis (e.g., with SLE), as a drug-induced phenomenon (e.g., with cocaine or marijuana), or as a manifestation of primary cerebral angiitis. Consideration should also be given for a hypercoagulable state, especially in a young patient, or one with a known or occult carcinoma. Multiple prothrombotic conditions may currently be detected, including protein C and S deficiency, antithrombin III deficiency, activated protein C resistance, antiphospholipid antibodies, prothrombin G20210A gene mutation, and factor V Leiden mutation.16 Evaluation The most crucial step in the evaluation of a patient presenting with an acute neurological change suggestive of ischemic stroke is to exclude an intracerebral hemorrhage. This is usually most easily done with computed tomography (CT), which is the safest, most convenient and readily available modality.17 MRI is also useful to evaluate for hemorrhage, but is not as readily available or convenient, and is more cumbersome for monitoring patients with acute stroke, many of whom may be unstable from a respiratory or hemodynamic standpoint.18–20 Based on the findings on a noncontrast-enhanced CT scan, consideration may be given for possible thrombolytic therapies. Exclusion criteria for thrombolysis based on
W9418-13.qxd 11/3/03 2:51 PM Page 400
400
Section I
Introduction
the CT scan include the presence of hemorrhage, a wellestablished acute infarct, a brain tumor (other than small meningiomas), a cerebral abscess, or a vascular malformation. Further CT evaluation in the acute stroke setting may include CT angiography (CTA; Fig. 13-3), involving an intravenous bolus of iodinated contrast, which may give additional information regarding the extracranial and intracranial circulation.21 This may be especially important for patients in whom intra-arterial thrombolysis is considered. This modality is currently available mostly in specialized care centers, and must be used with caution in patients with renal insufficiency (due to nephrotoxicity) or with known hypersensitivity to contrast material. The initial evaluation of a patient with acute stroke must allow for the exclusion of other conditions that can mimic a stroke presentation. These include seizure (with subsequent Todd’s paresis or other postictal syndromes), cerebral neoplasm, encephalitis, complex migraine, and hypoglycemia. Clinicians must also be vigilant for stroke as a presentation of bacterial endocarditis. Exclusion of these stroke mimics may save patients from unnecessary and potentially harmful therapies. MRI may be useful in the evaluation of the patient with acute stroke, especially when trying to determine if a patient has the potential to benefit from intra-arterial thrombolysis or hypertensive therapy. Diffusion-weighted imaging gives useful information regarding tissue that is reversibly or irreversibly ischemic.22 Perfusion-weighted imaging, using a timed bolus of intravenous gadolinium contrast material, gives further insight to tissue destined for
infarction, as well as tissue at risk for ischemic injury, the so-called ischemic penumbra.23 Furthermore, MRA can provide useful anatomic information regarding the extraand intracranial vasculature, and the sensitivity of this modality may be enhanced by the use of gadolinium contrast material.24 Other noninvasive measures may be used in the acute and subacute settings to evaluate stroke patients. Duplex carotid ultrasound provides information regarding the status of the extracranial circulation, including the carotid system as well as the extracranial vertebral arteries. Carotid duplex can evaluate the degree of stenosis, the presence of turbulent flow, and plaque morphology. Furthermore, transcranial Doppler (TCD) can be used to evaluate the intracranial circulation, including the vessels of the Circle of Willis as well as the distal vertebral and basilar system. TCD can be used continuously in the acute stroke setting to evaluate the efficacy of thrombolysis.25 The benefit of conventional cerebral angiography has diminished with the introduction of less invasive diagnostic modalities.26 However, it is still used in some circumstances, including during the performance of intra-arterial thrombolysis, in defining a questionable area of carotid stenosis, in diagnosing cerebral angiitis, and for exclusion of mycotic aneurysms in the setting of bacterial endocarditis. It is still considered the “gold standard” for diagnosing arterial dissection, but advances in MRI and CT angiography have diminished its necessity. In particular, MRI of the head and neck vessels using fat-saturated axial T1 images has emerged as a sensitive noninvasive means of evaluating for dissection.14
Figure 13-3. Contrast-enhanced CT and MR scans of the head showing the effects of a right internal carotid occlusion.
W9418-13.qxd 11/3/03 2:51 PM Page 401
Chapter 13
Other modalities may be used in the subacute evaluation of stroke patients. Cardiac monitoring (24- to 48-hour Holter monitoring) can detect paroxysmal atrial or ventricular dysrhythmias. Transthoracic echocardiography helps to delineate cardiac valvular abnormalities, wall dyskinesis, atrial or ventricular aneurysms, and patent foramen ovale (PFO) or atrial septal defect with the use of agitated saline contrast. Trans-esophageal echocardiography provides greater detail to these abnormalities, and is better suited to look for intracardiac thrombi and aortic atheromatous disease. However, it is a more invasive and expensive modality, and at present is only used in the minority of cases. Management Patients with acute stroke should be evaluated initially as potentially medically unstable. Therefore, the first priority should be to assess cardiovascular and respiratory function. An electrocardiograph (ECG) should be obtained shortly after admission to the emergency department, as stroke can commonly accompany an acute myocardial infarction.27 In addition, a secure airway should be established for patients with a depressed level of consciousness, as seen in patients with intracerebral hemorrhage or major strokes involving the vertebrobasilar system. An extensive and rapid laboratory evaluation should be undertaken, including a complete blood cell count with platelet count, basic metabolic and hepatic panel, glucose, prothrombin time (PT) and partial thromboplastin time (PTT), fibrinogen, erythrocyte sedimentation rate, urine human chorionic gonadotropin concentration in women of childbearing age, type and cross match, and in specific settings, a toxicology screen and hypercoagulability panel. Furthermore, prompt grading of the stroke severity is vital to choosing appropriate therapy, and standard stroke scales, such as the National Institutes of Health Stroke Scale (NIHSS) (Table 13-2)28 or the Scandinavian Stroke Scale (SSS) (Table 13-3),29 can help to exclude a patient from a potentially harmful therapy on the basis of the stroke being too small or too severe. Equally crucial for therapy is the proper determination of the exact time of onset of the stroke. The patient must be witnessed to have had an abrupt change in neurologic status by a reliable observer; otherwise the time of onset, by default, must be the last time the patient was seen at his or her baseline level of neurologic function. At this point, only the United States and Canada have regulatory approval for recombinant tissue plasminogen activator (rt-PA) use in stroke. Tissue-PA is endogenously synthesized and secreted by endothelial cells in physiologic concentrations. It specifically targets fibrin clots, with minimal consumption of circulating coagulation factors. It has a serum half-life of four to six minutes and a rapid onset of action.30,31 Free rt-PA is rapidly inactivated by plasminogen activator inhibitor type-1 (PAI-1) in circulation, but fibrin-bound rt-PA is only slowly inhibited by PAI-1. Given the mechanism of rt-PA thrombolysis, it is not surprising that platelet-rich clots can be resistant to rt-PA therapy.32–34
Acute Stroke and Other Neurologic Emergencies
401
To date, four large trials of intravenous rt-PA have been performed: the NINDS (National Institute of Neurologic Disorders and Stroke) recombinant tPA study,35–37 the European Cooperative Acute Stroke Study (ECASS)-I,38 ECASS-II,39 and the ATLANTIS (Alteplase Thrombolysis for Acute Noninterventional Therapy in Ischemic Stroke) rt-PA (Alteplase) Acute Stroke Trial (parts A and B).40,41 The NINDS rt-PA Acute Stroke Study was a randomized, doubleblind, placebo-controlled study of rt-PA given within 3 hours of clearly defined symptom onset. The study enrolled 624 patients and the NIHSS, modified Rankin score (mRS), Barthel Index (BI) and Glasgow Outcome Scale (GOS) were measured at 24 hours and 3 months. The results of this study were positive in favor of rt-PA. The global odds ratio (OR) for favorable outcome in the rt-PA–treated group was 1.7 (95% CI, 1.2 to 2.6) (Table 13-4). Patients treated with rt-PA were at least 30% more likely to have minimal or no disability at 3 months compared to patients treated with placebo (Fig. 13-4). The rt-PA treatment group had an 11% to 13% absolute increase in the number of patients with good outcomes, defined as a mRS score of 0 or 1. A similar degree of benefit was seen for all stroke subtypes. The two groups had no statistical difference in mortality rates at 3 months (17% in the rt-PA–treated group vs. 21% in the placebotreated group, P = .30), but symptomatic hemorrhage occurred more often in the rt-PA–treated group (6.4% vs. 0.6% in the placebo-treated group, P < .001). The benefit seen by the rt-PA–treated group existed regardless of patient age, stroke location, or stroke subtype. Severity of stroke and brain edema or mass effect on the pretreatment CT scan were associated with an increased risk of intracerebral hemorrhage in the rt-PA–treated group, although both groups still were more likely to have favorable outcomes if treated with rt-PA (although the difference was not statistically significant for the group with brain edema or mass effect). The benefits of rt-PA were durable over 1 year, with an odds ratio for favorable outcome of 1.7 (95% CI, 1.2 to 2.3), and the rt-PA–treated patients were at least 30% more likely to have minimal or no disability than the placebo-treated patients.42 The results of the other major trials of rt-PA in acute ischemic stroke (ECASS-I, ECASS-II, and ATLANTIS) all showed a similar benefit for the rt-PA treated group when treated within three hours of symptom onset. However, this benefit was not sustained in patients treated within three to six hours (ECASS-I and II), or 3 to 5 hours (ATLANTIS). At present, intravenous rt-PA cannot be recommended beyond the three-hour time window, although some suggest it may be useful in a carefully selected patient population within 3 to 6 hours.43 Additionally, in patients beyond the 3-hour time window, consideration may be given for intra-arterial thrombolysis in certain instances (see following section on Intra-arterial Thrombolysis). The NINDS study helped to establish strict inclusion criteria for the administration of thrombolytic therapy in acute stroke patients: (1) symptoms consistent with acute brain
W9418-13.qxd 11/3/03 2:51 PM Page 402
Table 13-2 NIH Stroke Scale: “Quick and Easy” Version Category
Description
1a. Level of consciousness (LOC) (Alert, drowsy, etc)
Alert Drowsy Stuporous Coma Answers both correctly Answers 1 correctly Incorrect Obeys both correctly Obeys 1 correctly Incorrect Normal Partial gaze palsy Forced deviation No visual loss Partial hemianopsia Complete hemianopsia Bilateral hemianopsia Normal Minor Partial Complete No drift Drift Can’t resist gravity No effort against gravity No movement Amputation, joint fusion (explain) No drift Drift Can’t resist gravity No effort against gravity No movement Amputation, joint fusion (explain) No drift Drift Can’t resist gravity No effort against gravity No movement Amputation, joint fusion (explain) No drift Drift Can’t resist gravity No effort against gravity No movement Amputation, joint fusion (explain) Absent Present in 1 limb Present in 2 limbs Normal Partial loss Severe loss No aphasia Mild to moderate aphasia Severe aphasia Mute Normal articulation Mild to moderate dysarthria Near to indecipherable or worse Intubated or other physical barrier No neglect Partial neglect Complete neglect
1b. LOC questions (Month, age) 1c. LOC commands (Open, close eyes, make fist, let go) 2. Best gaze (Eyes open—patient fallows examiner’s finger or face) 3. Visual (Introduce visual stimulus/threat to patient’s visual field quadrants) 4. Facial palsy (Show teeth, raise eyebrows, and squeeze eye lids) 5a. Motor arm–left (Elevate extremity to 90° and score drift/movement)
5b. Motor arm–right (Elevate extremity to 30° and score drift/movement)
6a. Motor leg–right (Elevate extremity to 30° and score drift/movement)
6b. Motor leg–right (Elevate extremity to 30° and score drift/movement)
7. Limb ataxia (Finger-nose, heel to shin) 8. Sensory (pin prick to face, arm, trunk, and leg—compare side to side) 9. Best language (Names items, describe a picture, and read sentences) 10. Dysarthria (Evaluate speech clarity by patient repeating listed words) 11. Extinction and inattention (Use information from prior testing to identify neglect or double simultaneous stimuli testing)
From NNDS rt-PA Stroke Study Group: N Engl J Med 1995;333:1581, with permission.
Score 0 1 2 3 0 1 2 0 1 2 0 1 2 0 1 2 3 0 1 2 3 0 1 2 3 4 9 0 1 2 3 4 9 0 1 2 3 4 9 0 1 2 3 4 9 0 1 2 0 1 2 0 1 2 3 0 1 2 9 0 1 2
Baseline Date/Time
Date/Time
W9418-13.qxd 11/3/03 2:51 PM Page 403
Chapter 13
Table 13-3 Scandinavian Stroke Scale Individual Administering Scale 1. Consciousness Fully conscious Somnolent, can be awakened to full consciousnes Reacts to verbal command but is not fully conscious Stupor (reacts to pain only) Coma 2. Orientation Correct for time, place, and person Two of these (time, place, person) One of these Completely disoriented 3. Speech No aphasia Impairment of comprehension or expression disability More than yes/no or less 4. Eye movement No gaze palsy Gaze palsy present Forced lateral gaze 5. Facial palsy None/dubious/slight Present 6. Gait Walks at least 5 m without aids Walks with aids Walks with help of another person Sits without support Bedridden/wheelchair 7. Arm, motor power (assessed only on affected side) Raises arm with normal strength Raises arm with reduced strength Raises arm with flexion in elbow Can move but not against gravity Paralysis 8. Hand, motor power (assessed only on affected side) Normal strength Reduced strength in full range Some movement, fingertips do not reach palm Paralysis 9. Leg, motor power (assessed only on one side) Normal strength Raises straight leg against resistance with reduced strength Raises leg with flexion of knee against gravity Can move but not against gravity Paralysis 10. Foot paresis None Present
Score 6 4 2 0 0 6 4 2 0 10 6 0 4 2 0 2 0 12 9 6 3 0 6 5 4 2 0 6 4 2 0 6 5 4 2 0 2 0
From Scandinavian Stroke Study Group: Stroke 1985;16:885, with permission.
infarction, with a clearly defined onset of less than 3 hours before rt-PA will be given (if the onset was not witnessed, the ictus is measured from the time the patient was last seen to be at baseline); (2) a significant neurologic deficit expected to result in long-term disability; and (3) a noncontrast-enhanced CT with no evidence of hemorrhage or well-established infarction.
Acute Stroke and Other Neurologic Emergencies
403
Absolute exclusion criteria have been established as well. These include (1) mild or rapidly improving deficits; (2) hemorrhage on CT, well-established acute infarct on CT, or any other CT diagnosis that contraindicates treatment, including abscess or tumor, excluding small meningiomas; (3) a known CNS vascular malformation or tumor; and (4) bacterial endocarditis. The rationale for excluding patients with improving symptoms is to not give potentially harmful treatments to patients with postictal presentations or with arterial occlusions secondary to clot that is already dissolving on its own. In addition, there is a long list of relative contraindications to thrombolytic therapy, including • significant trauma within 3 months, including cardiopulmonary resuscitation with chest compressions within 10 days • ischemic stroke within 3 months • history of intracranial hemorrhage, or symptoms suggestive of subarachnoid hemorrhage • major surgery within 14 days • minor surgery within 10 days, including liver and kidney biopsy, thoracentesis, and lumbar puncture • arterial puncture at a noncompressible site within 14 days • pregnancy, and up to 10 days postpartum • gastrointestinal, urologic, or respiratory hemorrhage within 21 days • known bleeding diathesis, including renal or hepatic insufficiency • peritoneal dialysis or hemodialysis • PTT greater than 40 seconds • INR > 1.7 • platelet count less than 100,000/mm3 • seizure at onset of stroke (this relative contraindication is intended to prevent treatment of patients with a deficit due to postictal Todd’s paralysis or with seizure due to some other CNS lesion that precludes thrombolytic therapy—if rapid diagnosis of vascular occlusion can be made, treatment may be given) • glucose concentration less than 50 or greater than 400 mg/dL (this relative contraindication is intended to prevent treatment of patients with focal deficits due to hypoglycemia or hyperglycemia; if the deficit persists after correction of the serum glucose, or if rapid diagnosis of vascular occlusion can be made, treatment may be given) • systolic blood pressure greater than 180 mm Hg or diastolic blood pressure greater than 110 mm Hg, despite basic measures to lower it acutely. • Consideration should be given to the increased risk of hemorrhage in patients with severe deficits (NIHSS >20), older than 75 years, or early edema with mass effect on CT. The dose of rt-PA is 0.9 mg/kg, with a maximum dose of 90 mg. Ten percent is given as a bolus over 1 minute. This is
W9418-13.qxd 11/3/03 2:51 PM Page 404
404
Section I
Introduction
Table 13-4 Final Multivariate Model for Predicting Benefit from rt-PA*
From NINDS rt-PA Stroke Trial: Stroke 1997;28:2119, with permission.
followed by a continuous infusion of the remaining 90% over 60 minutes. Following treatment, patients should be monitored in an ICU for at least 24 hours. Vital signs should be checked every 15 minutes for the first 2 hours, then every 30 minutes for 6 hours, then every hour for 16 hours. Blood pressure should be strictly controlled for 24 hours, keeping the systolic blood pressure less than 180 mm Hg and the diastolic blood pressure less than 110 mm Hg. Labetolol is recommended for control of hypertension; 10 mg should be given intravenously over 1 to 2 minutes, and then the dose repeated or doubled every 5 to 15 minutes, up to a total of 150 mg. If the blood pressure remains refractory despite these measures, consideration can be given to a continuous infusion of sodium nitroprusside. With refractory hyperten-
sion, the risk of hemorrhage increases, and withholding therapy might be in the patient’s best interest. Neurologic evaluations should be performed every hour. Oxygenation should be checked by continuous pulse oxymetry. An oxygen cannula or mask should be used to keep oxygen saturation greater than 95%. Acetaminophen, 650 mg every 4 hours orally or rectally, should be given for any temperature greater than 37.4°C, and a cooling blanket used for temperatures greater than 38.9°C. Antiplatelet or anticoagulant therapies should be avoided for the first 24 hours. No Foley catheter, nasogastric tube, arterial catheter, or central venous catheter should be placed for 24 hours unless absolutely necessary, due to hemorrhage risk. Emergent head CT should be performed for any neurologic worsening.
Figure 13-4. Graph showing an odds ratio estimate model for favorable outcome at 3 months in rtPA–treated patients as compared to placebo-treated patients, from time of stroke onset to time of treatment. An OR greater than 1 (solid line) suggests there are greater odds that a rt-PA–treated patient will have a favorable outcome, compared to patients receiving placebo, at 3 months. With 95% confidence intervals (dotted lines). (From Marler, et al: Neurology 2000;55:1649, with permission.)
W9418-13.qxd 11/3/03 2:51 PM Page 405
Chapter 13
If an intracerebral hemorrhage develops following thrombolysis, several steps must be taken emergently. Neurosurgery should be contacted for possible hematoma evacuation. Blood should be sent immediately for complete blood cell count, PT, PTT, platelets, fibrinogen, and D-dimer (this should be repeated every 2 hours until bleeding is controlled). Two units of fresh frozen plasma should be given every 6 hours for 24 hours after the thrombolytic agent was given. Cryoprecipitate (20 U) should be given; if the fibrinogen level is less than 200 mg/dL at 1 hour, repeat the cryoprecipitate dose. Four units of platelets should be given. Protamine sulfate (1 mg per 100 U of heparin given in the past 3 hours) should be given; a test dose of 10 mg slow IV push over 10 minutes should be given while observing for anaphylaxis, and then the remaining dose by slow IV push, up to a maximum dose of 50 mg. Institute frequent neurologic checks, as well as management of increased intracranial pressure, as needed. Aminocaproic acid (Amicar) can be given as a last resort, in a dose of 5 g in 250 cc normal saline IV over 1 hour. Intra-Arterial Thrombolysis The use of thrombolytic interventions outside of the 3-hour time window is controversial. Trials that have extended the therapeutic window beyond 3 hours for intravenous therapy have failed to show convincing benefit, including the ECASS I and II studies and the ATLANTIS trial. However, there have been several attempts to prove the benefit of catheterdirected therapy via an intra-arterial approach for focal clot lysis. Theoretically, there are several potential benefits to treatment of stroke via an intra-arterial approach. First, with angiographic confirmation of vessel occlusion, patients can be protected from receiving an unnecessary and potentially harmful therapy. Second, high concentrations of thrombolytic agents can be given directly at the site of thrombosis, thus minimizing systemic concentrations. Third, the response to lysis can be monitored by direct visualization, and thus discontinued if quickly successful. Fourth, mechanical disruption of the clot (e.g., via balloon angioplasty) may aid in clot disruption and thrombolysis. To date, there have been two randomized trials using recombinant pro-urokinase (r-pro-UK) in catheter-directed clot lysis of the middle cerebral artery stem. PROACT I (Prolyse in Acute Cerebral Thromboembolism) compared rpro-UK versus placebo.44 Patients displaying TIMI grade 0 or 1 occlusion of the M1 or an M2 branch of the MCA were randomized 2 : 1 to receive either r-pro-UK 6 mg or placebo over 120 minutes into the proximal end of the thrombus. At 24 hours, the researchers assessed recanalization efficacy and symptomatic intracerebral hemorrhage. In total, 40 patients were treated, 26 of whom received r-pro-UK. The median time to treatment from symptom onset was 5.5 hours. Recanalization occurred at a significantly higher rate in the r-pro-UK-treated group (P = .0085). However, intracerebral hemorrhage occurred in 15.4% of the r-pro-UK–treated
Acute Stroke and Other Neurologic Emergencies
405
patients vs. 7.1% of the placebo-treated patients, although this did not reach statistical significance. In patients who also received adjuvant higher-dose heparin, the recanalization rate was 81.8%, vs. 40% in the low-dose heparin group (the dose was lowered due to concerns raised during the study by the safety committee). The mortality rate appeared to be lower in the r-pro-UK–treated group, but this result was not statistically significant. Based on the suggestion of possible benefit from the PROACT I study, a second study was designed to address the efficacy of intra-arterial thrombolysis, PROACT II.45 In PROACT II, 180 patients were randomized 2 : 1 to receive either treatment with 9 mg r-pro-UK over 2 hours plus the PROACT I lower dose of heparin (2000 IU bolus followed by a 500 IU/hour continuous infusion for 4 hours) vs. heparin alone. Again, the inclusion criteria mandated new neurologic signs attributable to a MCA stem occlusion within 6 hours (i.e., allowing initiation of treatment by 6 hours). Mechanical clot disruption (e.g., with balloon angioplasty) was not permitted in this study. The primary outcome was the modified Rankin Scale (mRS) of 2 or less at 90 days. Secondary outcomes included recanalization rates (TIMI 2 and 3), symptomatic ICH, and overall mortality. Forty percent of the r-pro-UK–treated patients and 25% of the heparin-only patients had a mRS of 2 or less, translating into an absolute benefit of 15%, relative benefit of 58%, and a number needed to treat (NNT) of 7 (P = .04) (Fig. 13-5). The recanalization rate was 66% in the r-pro-UK group vs. 18% in the control group (P < .001), and the TIMI 3 recanalization rate in the r-pro-UK group was 19% vs. 2% in the control group (P < .003). Overall mortality was 25% in the r-pro-UK group and 27% in the control group (P = NS). Early symptomatic ICH occurred only in patients with NIHSS scores greater than 11 within 24 hours; these
Figure 13-5. Modified Rankin Scale Score in patients treated with either recombinant pro-urokinase or as controls. There was a significant outcome difference between treated and control subjects. (From Furlan, et al: JAMA 1999;282:2003, with permission.)
W9418-13.qxd 11/3/03 2:51 PM Page 406
406
Section I
Introduction
A B Figure 13-6. A, Pretreatment angiogram showing absence of middle cerebral artery (MCA) flow. B, Post-treatment angiogram, showing significant improvement in flow.
occurred in 10.2% of the r-pro-UK patients vs. 2% of the control patients. Based on the currently available clinical trials, the U.S. Food and Drug Administration (FDA) has not approved intra-arterial thrombolysis as a therapeutic modality for acute stroke, except in experimental conditions. Given the increased risk of intracerebral hemorrhage, informed consent must always be obtained prior to intra-arterial therapy. The increased hemorrhage rate is at least partially explained by the severity of the strokes admitted to the trials. The average baseline NIHSS in PROACT II was 17 vs. 11 in ECASS II and ATLANTIS and 14 in NINDS. Furthermore, in the PROACT II study, patients were given a fixed dose of 4.5 mg r-pro-UK per hour in 2 sequential hours, regardless of whether recanalization was seen between doses. Additionally, no mechanical disruption of clot was allowed, perhaps hindering the abilities of the interventional neuroradiologist to achieve clot lysis, further predisposing the patient to either hemorrhage or poor outcome. The protocol for patient management prior to performing intra-arterial thrombolysis is similar to that for IV thrombolysis, with notable exceptions. The inclusion criteria for IA thrombolysis mandate that there be a proximally occlusive clot involving the internal carotid artery, the middle cerebral artery stem, or the basilar artery. If there is
clearly going to be a delay prior to initiation of therapy, many clinicians advocate the interim use of intravenous heparin, although this remains quite controversial. Hypertension is permitted to preserve penumbral tissue perfused via collateral circulation; at our institution, we allow for a blood pressure up 220/110 mm Hg. If angiographic recanalization (Fig. 13-6) is achieved (TIMI 2 or 3 flow), the patient’s blood pressure is then lowered, because this will minimize the risk of hemorrhagic conversion of the infarct. The treatment of patients following IA thrombolysis is also controversial. We adopted a general practice of discontinuing all anticoagulation in patients in whom there is complete recanalization of the vessel, with no evidence of intimal injury or disruption. Frequently, however, complete recanalization is not achieved, and guidelines for the use of heparin, intravenous glycoprotein IIb/IIIa inhibitors, or oral antiplatelet agents in patients with only partial recanalization have not been established. If a stent is placed during the procedure, antiplatelet agents are typically used to prevent thrombosis. A third project, PROACT III, has been planned, with an attempt to implement magnetic resonance imaging to aid with proper patient selection.46 Unfortunately, r-pro-UK was removed from the market in 1999 due to concerns with its preparation, and as a result rt-PA is the thrombolytic agent
W9418-13.qxd 11/3/03 2:51 PM Page 407
Chapter 13
Acute Stroke and Other Neurologic Emergencies
407
Table 13-5 Time to Treatment of Various Stroke Studies
From Ernst, et al: Stroke 2000;31:2552, with permission.
currently under investigation. In our experience at the Massachusetts General Hospital, rt-PA has been less satisfactory than r-pro-UK had been in terms of recanalization results as well as clinical outcomes. One possible explanation is that rt-PA may actually be prothrombotic at the site of intra-arterial infusion, due to platelet aggregation. With this in mind, we evaluated the efficacy of adding a glycoprotein IIb/IIIa inhibitor (eptifibatide), 90 mcg/kg bolus followed by 0.5 to 2.0 mcg/kg/min infusion, in the background of applying rt-PA intra-arterially.47 In a series of 37 patients, 13 of whom were treated with intra-arterial (IA) rt-PA and 24 of whom were treated with combination therapy, we found a trend toward better revascularization (TIMI 2 or 3 flow) in the combination therapy group (58% vs. 31%). However, this trend disappeared when balloon angioplasty was performed in cases resistant to thrombolytics alone. There was one symptomatic hemorrhage in the combination therapy group, and two in the rt-PA-alone group. Thirty-eight percent of patients in the combination therapy group had a good outcome (mRS 0-2) vs. 27% in the rt-PA-alone group. This study is promising, but prospective data on larger numbers of patients will be necessary to convincingly advocate this approach. Eckert and colleagues recently reported three cases of acute basilar artery occlusion treated successfully by a similar combination of IA rt-PA and a glycoprotein IIb/IIIa receptor inhibitor.48 The Emergency Management of Stroke Bridging Trial49 was a multicenter trial designed to evaluate the safety and potential efficacy of combined IV and IA rt-PA when initiated within 3 hours of stroke onset. Thirty-five patients were randomized to receive either placebo or an adjusted IV rtPA dose (0.6 mg/kg, up to a maximum of 60 mg; 15% given as a bolus, followed by the remainder over 30 minutes). All patients subsequently underwent angiography, where, if an appropriate clot was seen (correlating to the patient’s symptoms), 2 mg of rt-PA was injected directly into the clot, followed by continuous direct infusion of IA rt-PA at 10 mg/hour for up to 2 hours. This study had remarkably quick treatment times: the mean time to IV treatment from stroke onset was 2 hours and 30 minutes (±32 minutes), and to IA treatment was 4 hours and 10 minutes. In total, there were nine patients who received IV + IA therapy, with a baseline NIHSS of 17.2. Of these, six patients (66%) achieved a modified Rankin Score of 0 to 2 at 3 months. Of the six patients treated with only IA rt-PA (with an average NIHSS of 11.6),
5 (83%) achieved an mRS of 0 to 2 at 3 months. Although the numbers of patients in this study were too small to reach statistical significance, the authors postulated that the surprisingly high number of favorable outcomes could be at least partially attributable to the rapid time to treatment. Based on the safety and the positive trend in the Emergency Management of Stroke Bridging Trial (EMS) study, Ernst and co-workers50 studied an additional 20 patients with combination therapy, using a similar protocol. The median NIHSS for the patients in this study was higher than in EMS, at 21 (range, 11 to 31). Again, a distinguishing feature of this study was the rapidity of treatment (Table 13-5): IV therapy was started at a median of 2 hours and 2 minutes, and IA therapy was initiated at a median of 3 hours and 30 minutes. Of the 20 patients, 13 (65%) recovered to a mRS of 0 to 2 at 2 months. Of the remainder, five patients had a mRS of 4 or 5, a symptomatic intracerebral hemorrhage (thought to be secondary to labile blood pressure postprocedure) developed in one patient who eventually died, and another patient died of complications from the stroke (Fig. 13-7). Although the data from these studies are encouraging, improved efficacy of combined IV-IA therapy over IV
Figure 13-7. Percentage of patients from various stroke studies recovering to a modified Rankin Scale Score of 0 to 2 (favorable outcome with slight or limited disability). Controls are from the PROACT II study treated only with intravenous heparin. (From Ernst, et al: Stroke 2000;31:2552, with permission.)
W9418-13.qxd 11/3/03 2:51 PM Page 408
408
Section I
Introduction
therapy alone will need to be proven in a larger randomized prospective study. Additional Techniques and Future Directions Newer agents for chemical and mechanical treatment of acute stroke have recently been implemented in small studies. Alteplase, which is the wild-type recombinant t-PA, consists of a fibronectin finger-like domain, which binds to fibrin. It also contains the kringle-1 domain with receptor binding, the kringle-2 domain with low affinity fibrin binding, and a protease domain with specific binding to plasminogen. In addition, it has an epidermal growth factor domain, which is thought to be responsible for hepatic binding, and thus accelerated hepatic clearance, with a halflife of 3 to 5 minutes.51 Retavase is a more fibrin-specific molecule, in which the finger domain, epidermal growth factor domain, and the kringle-1 region of the wild-type t-PA have been deleted. The retention of the kringle-2 domain allows the molecule to keep its fibrin specificity. It is produced in Escherichia coli, and thus does not contain carbohydrate side chains. Its halflife is 18 minutes. Pro-urokinase (Prolyse) is a glycosylated 411 amino acid single-chain pro-enzyme precursor of urokinase. It is derived from murine hybridoma cells. Single chain prourokinase is activated by fibrin-bound plasmin at the thrombus surface to form two-chain urokinase. The half-life of pro-urokinase is approximately 20 minutes. In comparison, 5 mg of Alteplase is equivalent to one unit of Retavase and 9 mg of pro-urokinase. Despite the success seen in the PROACT II study and other trials of IA thrombolysis using chemical methods, a large proportion of occlusive cerebrovascular disease is refractory to these measures, and 25% to 40% of attempts using chemical methods alone are unsuccessful. In refractory cases, the angiographer must be prepared to consider additional techniques to achieve clot lysis. Mechanical disruption of the clot may aid in clot lysis by means of increasing surface area on which chemical thrombolytic agents can act, or by fragmenting the clot itself.52,53 Certain risks accompany these techniques, including distal embolization of clot, vessel rupture or dissection, or failure of disruption. The approach is limited by the ability of the angiographer to navigate toward the target vessel. Perhaps the most commonly used tool for mechanical disruption of clots is angioplasty. The angioplasty balloons currently in use are either polytetrafluoroethylene- or silicone-based, and are intended for atheromatous plaque or vasospasm, respectively. Stent placement has traditionally been reserved for the internal carotid artery, but more recently the distal circulation has become experimentally amenable to stenting with the advent of super-flexible balloon-mounted stents. Newer techniques are rapidly being introduced into the angiography suite in an attempt to safely mechanically
disrupt clots, but to date all of these techniques remain unproven and experimental. With laser-induced microcavitation, a low energy laser is used as a transducer, creating an ultra short-lived microbubble that collapses and induces a series of waves that act to gently agitate the clot surface. The goal is to create clot disruption, but with such low energy that surrounding tissues are not harmed. Endovascular photoacoustic recanalization (EPAR) creates a microcavitation effect by using a neodymium yttrium-argon-garnet laser with similarly low energy levels. In vivo studies with this technique have demonstrated a primary fragment peak at the 3-mm particle size, which is considerably smaller than the diameter of capillaries. Furthermore, ultrasonic cavitation of clots has been investigated, and this technique similarly creates a cavitation phenomenon through the use of ultrasonic energy. Ultrasonic energy may be generated using either a piezoelectric driver (Ekos) or with acoustic horn technology (ACS/Guidant). All of these techniques are limited by the ability of the fiberoptic assembly to navigate to the target vessel safely. Other techniques have aimed at mechanical extraction or physical removal of the offending clot from the vessel, without creating distal emboli. These include a “flowering” type catheter with distal enclosures designed to openly approach the clot face, following which an extraction compartment is intended to surround and entrap the clot (Interventional Innovations); a “corkscrew” device that screws into a clot, and then may be removed into a protective guiding catheter (MIS); and a “snare” that has been employed much like a lasso, encircling and withdrawing a clot into a protective guiding catheter (Wikholm). Additional techniques include the Possis Angiojet, which uses a high-pressure microstream to create a distal Venturi suction, so that the approached clot face is gently agitated and the fragments are sucked into the catheter (Fig. 13-8). Future acute stroke therapies that may move from the laboratory to the emergency room include hyperoxygenation, aimed at increasing the oxygen carrying capacity in the circulation to penumbral tissue; retroperfusion, in which the venous system is used to temporarily provide oxygenated blood into an ischemic capillary bed; hypothermia, which is also being tested in other forms of brain injury, including traumatic brain injury, subarachnoid hemorrhage, and cardiac arrest; neuroprotective agents; and growth factors aimed at angiogenesis and neurogenesis. Presently, all of these techniques remain experimental. Therapies in Patients not Eligible for Thrombolysis For patients who are not eligible for IV or IA thrombolysis, there is considerable debate regarding the use of heparin and antiplatelet agents (Table 13-6). Patients are considered at risk for neurological worsening during the acute stroke period, either from extension of thrombosis or from recurrent embolism. Large artery atherosclerotic lesions may be at the highest risk of worsening in the acute period, with risk
W9418-13.qxd 11/3/03 2:51 PM Page 409
Chapter 13
Acute Stroke and Other Neurologic Emergencies
409
Figure 13-8. The AngioJet catheter. (From Chow, et al: Stroke 2000;31:1420, with permission.)
of deterioration approaching 30% during the initial hospitalization.54 Neurologic worsening may occur by several mechanisms, including progressive hypoperfusion secondary to inadequate collateral circulation, progression of an existing thrombus, or recurrent thromboembolism. With this in mind, anticoagulation with heparin is often undertaken, but with inadequate supportive data.55 The data to
support the use of heparin in atrial fibrillation are even less convincing.56,57 The International Stroke Trial (IST) evaluated the possibility of benefit of subcutaneous heparin in two different doses (5000 U or 12,500 U twice daily), with or without aspirin, versus aspirin alone.58 This trial included a placebo arm as well. The results of this study suggested that the
Table 13-6 Agents Used in Patients for Whom IV or IA Thrombolysis Is Inappropriate
From Coull et al: Neurology 2002;59:13, with permission.
W9418-13.qxd 11/3/03 2:51 PM Page 410
410
Section I
Introduction
higher dose of heparin was accompanied by excessive bleeding risk, but that the lower dose may have been effective in preventing recurrent stroke or PE. The Trial of ORG 10172 in Acute Stroke Treatment (TOAST) compared a lowmolecular-weight heparin with placebo in acute stroke, and failed to demonstrate a statistical difference in the primary endpoint of favorable outcome at 3 months.59 A subgroup analysis, however, suggested a benefit in favor of the danaparoid group for the subset of patients with stroke caused by larger artery atherosclerosis. A second study, evaluating dalteparin in acute ischemic stroke patients with atrial fibrillation, failed to show a benefit in favor of heparin over aspirin (160 mg/day).60 Currently in progress, the Rapid Anticoagulation Preventing Ischemic Damage (RAPID) study is a randomized, multicenter trial comparing the safety and efficacy of IV unfractionated heparin with aspirin. They are enrolling patients with acute, nonlacunar ischemic strokes within 12 hours of symptom onset. Aspirin therapy in acute stroke has been evaluated in two large randomized trials. In the International Stroke Trial,58 the patients who were randomized to aspirin had significantly fewer recurrent ischemic strokes vs. placebo (2.8% vs. 3.9%), had no increase in hemorrhagic strokes (0.9% vs. 0.8%), and had a nonsignificant trend toward a reduction in death or dependence at 6 months (61.2% vs. 63.5%). The Chinese Acute Stroke Trial (CAST)61 randomized 21,106 patients to receive either aspirin (160 mg/day) or placebo within 48 hours of symptom onset. Significant reductions were found in the aspirin-treated group for both recurrent ischemic strokes (1.6% vs. 2.1%) and for early mortality (3.3% vs. 3.9%). There was a nonsignificant trend toward
decreased death or dependence in the aspirin-treated group as well (30.5% vs. 31.6%). Management of Specific Stroke Emergencies The neurointensivist must be aware of at least two additional ischemic stroke emergencies beyond thrombolytic therapies: large cerebellar infarcts and large cerebral hemispheric infarcts. Larger cerebellar hemispheric infarctions present with lethargy, confusion, decreased spontaneous activity, and profound ipsilateral ataxia. The cerebellopontine angle can become compromised, leading to reduced ipsilateral corneal reflex and ipsilateral seventh cranial nerve palsy. As a general rule, infarct volumes of greater than one third of a cerebellar hemisphere carry a more ominous prognosis, and more readily progress to an edematous state, causing mass effect and fourth ventricular compression, leading to hydrocephalus (Figs. 13-9, 13-10).62 The lower brainstem tegmentum can be directly compressed by the expanding mass, and the cerebellar tonsils may be forced into the foramen magnum. Upward herniation can also occur, with distortion of the midbrain and cerebral aqueduct, as well as buckling of the quadrigeminal plate.63 Given the insensitivity of CT for detecting early infarction in the posterior fossa, we recommend using MRI with diffusion-weighted imaging to outline the volume of injury. If the volume is greater than one third of a cerebellar hemisphere, early surgical decompression should be strongly considered. Medical measures that may be used before surgery include hyperosmotic therapy with mannitol, glycerol, or hypertonic saline; however, these patients can acutely decompensate due to rapidly progressive edema causing
Figure 13-9. Noncontrast CT of brain showing right cerebellar hemispheric infarction with edema and compression of the fourth ventricle.
W9418-13.qxd 11/3/03 2:51 PM Page 411
Chapter 13
Acute Stroke and Other Neurologic Emergencies
411
Figure 13-10. MR and CT images of the brain showing multiple bilateral cerebellar and occipital lobe infarctions, with incipient hydrocephalus due to fourth ventricular compression.
acute hydrocephalus, and thus these medical measures are considered only temporizing prior to definitive surgical decompression. Postoperatively, patients may have continued edema and fourth ventricular compromise, and many advocate the use of a ventriculostomy in the management, even in patients in whom there is not yet evidence of hydrocephalus. In addition to frequent serial clinical examinations, serial CT scans may be checked every one to two days in the early postoperative period to evaluate the degree of edema, fourth ventricular compression, and hydrocephalus. The patient should be observed in the ICU until it is clear that the edema is abating, and that close neurologic monitoring is no longer necessary. The data to support hemicraniectomy in large hemispheric ischemic lesions are less robust. To date, there have been no prospective, controlled trials of hemicraniectomy in large hemispheric ischemic stroke patients. The idea behind hemicraniectomy is to limit secondary ischemic injury that can occur due to either increased ICP or secondary infarction due to compression of the anterior cerebral or posterior cerebral arteries as they are impinged upon by the expanding MCA lesion. As a general rule, younger patients are at the greatest risk of secondary injury given the relatively preserved baseline brain volume, in contrast to older patients in whom generalized cerebral atrophy decreases their risk of herniation from a mass lesion. Hemicraniectomy prior to secondary injury and/or herniation can be a life-saving procedure, but selecting the correct patient population remains problematic. The largest series in the evaluation of early hemicraniectomy in large hemispheric ischemic strokes was performed
by Schwab and colleagues.64 Sixty-three patients prospectively underwent hemicraniectomy with acute complete middle cerebral artery infarction. Seventy-three percent survived, and none of the survivors had residual complete hemiplegia or were permanently wheelchair bound. In patients with dominant hemispheric strokes (n = 11), their degree of aphasia was graded as only “mild to moderate.” The authors postulated that one of the reasons for the exceptionally good outcomes in their study might have been that the patients were operated on early, before the onset of secondary injury. Common pitfalls with this approach include (1) the inability to recognize which patients will fail medical therapy alone and (2) the creation of an inadequate hemicraniectomy window to allow sufficient room for brain expansion and relief from edema. Studies are ongoing using bilateral ICP monitors to predict which patients will necessitate hemicraniectomy, using standardized surgical technique to ensure adequate windows for expansion. Intracerebral Hemorrhage Approximately 10% of all primary cerebrovascular events are spontaneous intracerebral hemorrhages (ICH). They can impact neurologic functioning based on dissection of brain tissue, the development of a mass lesion, the formation of edema, or by causing hydrocephalus. An intracerebral hemorrhage can occur in the setting of trauma (see Chapter 8), illicit drug use (e.g., marijuana- or cocaine-induced vasculitis, heroin-associated embolus/bacterial endocarditis), or over-the-counter medications (e.g., phenylpropanolamine),65 excessive alcohol consumption,66,67 due to an underlying vascular abnormality (e.g., arteriovenous mal-
W9418-13.qxd 11/3/03 2:51 PM Page 412
412
Section I
Introduction
Table 13-7 Risk Factors for Intracerebral Hemorrhage
From Monforte, et al: Stroke 1990;21:1529, with permission.
formation, cerebral aneurysm), secondary to an underlying brain tumor (primary or secondary), or with a bleeding diathesis (Table 13-7). In addition, primary intracerebral hemorrhages can occur in the absence of pre-existing intracranial pathology. The location of the hemorrhage (Fig. 13-11) and the age of the patient can give clues regarding the etiology of the ICH. Spontaneous intracerebral hemorrhages that occur in the basal ganglia (Fig. 13-12), thalamus, cerebellum, and pons are typically attributed to hypertension, and are less commonly seen in the elderly population. A theoretical explanation for why hypertensive hemorrhages occur in these locations is that they arise in the distribution of the lenticulostriate and paramedian vessels, which have thinner walls in comparison to cortical vessels, and are exposed to higher intravascular pressures due to their proximity to main vascular trunks.68,69
Figure 13-11. Location of lobar hemorrhage on initial CT scan of the head. (From Flemming, et al: J Neurol Neurosurg Psych 1999;66:600, with permission.)
In contrast, lobar intracerebral hemorrhages occur more commonly in the elderly population (Fig. 13-13), and are often associated with cerebral amyloid angiopathy. They occur in patients without documented hypertension, and recur in 10% of cases.70,71 Diagnosis is confirmed only with biopsy or postmortem examination, but probable diagnosis can be made in an elderly patient in whom there are multiple recurrent lobar hemorrhages without an underlying cause.72 MRI with susceptibility-weighted imaging may be helpful in looking for evidence of previous, perhaps silent, lobar microhemorrhages.73 More rarely, hemorrhage can occur directly into the ventricular system, sparing the brain parenchyma. Primary intraventricular hemorrhage (PIVH) accounts for approximately 3% of all intracerebral hemorrhages (Fig. 13-14). The acute form of this disease can have a very high mortality rate, especially if signs of brainstem dysfunction are present on the initial examination. There are a variety of purported causes, the most common of which are likely hypertensivetype hemorrhages arising from tissues adjacent to the ventricular system (caudate nucleus, thalamus, cerebellar vermis). MRI can help to distinguish in these patients whether an intraparenchymal component is present. Additional causes include arteriovenous malformations (AVMs), especially in younger, nonhypertensive patients; systemic bleeding disorders; intracranial aneurysms; brain tumors; and venous thrombosis. PIVH is historically associated with a high mortality rate, and ventriculostomy is often of little benefit unless performed very early after onset to relieve the sudden, massive increase in ICP.74 If patients survive the initial insult, control of hypertension is an additional early necessary step to minimize ongoing bleeding. More recent studies have suggested that prognosis for survival may be improving.75,76 Recent studies have suggested that the use of intraventricular urokinase and hematoma drainage may be helpful, but this needs to be studied in a larger, more controlled fashion.77
W9418-13.qxd 11/3/03 2:51 PM Page 413
Chapter 13
Figure 13-12. CT scan of the head showing hemorrhage of the left basal ganglia.
Acute Stroke and Other Neurologic Emergencies
413
Figure 13-14. MR image of a primary intraventricular hemorrhage in an infant.
Outcomes in intracerebral hemorrhage are dependent on a number of factors, including the location and size of the hemorrhage (Fig. 13-15),78 the age of the patient, the Glasgow Coma Score (GCS)79 and the etiology of the hemorrhage. Furthermore, the presence of intraventricular blood greatly influences outcome,80 as does the volume of blood within the ventricular system.81 Hydrocephalus can also portend a poor outcome in ICH patients82 (Table 13-8), and the placement of an extraventricular drain alone has had little impact on outcome.83 Several small studies have evaluated the use of intraventricular hemorrhage with thrombolytic medications, with promising results,84 but this has yet
to be studied in randomized, controlled fashion. Other factors associated with a poor outcome include hyperglycemia,85 anemia or hypoxia,86 midline shift,85 active bleeding,87 and marked hypertension.85 Patients with lobar hemorrhage may have an additional set of factors that portend deterioration, including a decreased level of consciousness on presentation and the presence of shift on CT (Table 13-9).88 Interestingly, some propose that the poor outcomes associated with ICH may be a “self-fulfilling prophesy,” in that patients may have care withdrawn when the outcome is presumed to be poor, and thus long-term clinical data about the
Figure 13-13. Graph showing age-specific incidence rates of intracerebral hemorrhage by location. (From Broderick, et al: Stroke 1993;24:49, with permission.)
Figure 13-15. Volume of hemorrhage on initial CT scan of the head. The y-axis is number of patients. (From Flemming, et al: J Neurol Neurosurg Psych 1999;66:600, with permission.)
W9418-13.qxd 11/3/03 2:51 PM Page 414
414
Section I
Introduction
Table 13-8 Impact of Hydrocephalus on Hospital Outcome in ICH
From Diringer, et al: Stroke 1998;29:1352, with permission.
true course of the illness and possible recovery may be lacking.89 The initial steps in the treatment of patients with an intracerebral hemorrhage must always include a rapid assessment of the level of consciousness, the security of the airway, the blood pressure, and the detection of any cardiac dysrhythmias. A quick assessment of the neurologic examination should be performed. If intubation is necessary, it should be performed carefully, with mindful attention to
Table 13-9 CT Characteristics
From Diringer, et al: Stroke 1998;29:1352, with permission.
provide maximal pre-oxygenation, and to avoid administering drugs that might cause reflex arrhythmias or major blood pressure fluctuations, including atropine, thiopental, midazolam, propofol, or succinylcholine. Once the patient is stabilized, a noncontrast-enhanced head CT should be performed urgently to verify brain hemorrhage. CT may also be helpful in detecting aneurysms, AVMs, underlying tumors or abscesses, especially when contrast is administered to suspected abnormalities. Becker and colleagues found that contrast extravasation into the hematoma correlated with poor outcome in ICH patients, and that the risk of contrast material extravasation increased with extreme hypertension, depressed level of consciousness, and larger hemorrhages.89 This study provides more evidence for aggressive blood pressure control in the acute treatment of ICH to limit ongoing bleeding and edema formation. MRI is also sensitive in the evaluation of hemorrhage, but is more cumbersome to perform and is potentially risky for an acutely ill patient, given the limitations of monitoring ability in the MRI suite. MRI may be especially useful in dating the time course of the hemorrhage, in detecting areas of prior hemorrhage (with the use of gradient-echo imaging),90 and in diagnosing cavernous malformations. We recommend a follow up MRI at approximately three months after the hemorrhage if the cause remains unclear. This should include the use of intravenous gadolinium to evaluate for an underlying mass lesion, MRA to evaluate for an underlying vascular malformation or aneurysm, and gradient-echo sequencing to evaluate for evidence of hemorrhage, if not done previously. Angiography should be considered in a young patient with an intracerebral hemorrhage with no history of hypertension, as this may reveal an underlying AVM or aneurysm not readily appreciated on CT. Halpin and associates91 performed angiography on 38 patients with suspicious CT findings, including the presence of subarachnoid or intraventricular hemorrhage, abnormal intracranial calcifications, prominent vascular structures, or an abnormal site
W9418-13.qxd 11/3/03 2:51 PM Page 415
Chapter 13
of hemorrhage. Eighty-four percent of the cases were positive for an underlying abnormality, either an AVM or aneurysm. The timing of angiography should take into account the patient’s clinical state and the urgency of surgery, if surgery is anticipated.92 The patient’s coagulation status, in particular the platelet count, PT, and PTT, should be checked immediately, and corrected if abnormal. If the patient recently received heparin, 1 mg protamine sulfate per 100 units of heparin should be administered with caution, given that protamine can cause hypotension. Patients taking coumadin who have an elevated INR on admission should be reversed with vitamin K (10 mg subcutaneously for three doses), as well as fresh-frozen plasma to control the PT to less than 14.5 during the time it takes for vitamin K to take effect. The mainstays of medical treatment of acute intracerebral hemorrhage are correction of any coagulopathy and avoidance of severe hypertension, although no studies of medical (or surgical) therapy have shown definitive benefit.92 Most data suggest that the bulk of hematoma expansion occurs within the first several hours of onset, and that this is the crucial time in which to control any elevations of blood pressure that may be present (Table 13-10).93,94 Beyond the first 4 to 6 hours, lowering the blood pressure is of questionable benefit. Careful consideration should also be given to the patient’s baseline blood pressure, possible raised intracranial pressure, the presumed etiology of the hemorrhage, and any vascular stenoses that may be present, so as not to incur additional injury by hypoperfusion. In patients with large intracerebral hemorrhages, ICP may be such an issue that overly aggressive lowering of blood pressure theoretically may cause a decrease in cerebral perfusion pressure (CPP) below an acceptable level. Qureshi and colleagues found a possible correlation between rapid blood pressure decline within the first 24 hours of ICH patients and mortality, suggesting that the rate of blood pressure control may influence secondary injury by possible hypoperfusion due to inadequate cerebral perfusion pressure.95 Several studies have evaluated PET and CT perfusion to document an ischemic penumbra around a hematoma, but these have been unconvincing to date. Qureshi and coworkers used a dog model of ICH to determine whether blood pressure reduction caused regional hypoperfusion or significant ICP reductions, and found no adverse effect.96 Subsequently, Powers and colleagues performed positron emission tomography studies in humans with spontaneous ICH at an average of 15 hours from symptom onset, before and after pharmacologic lowering of blood pressure.97 They found no significant impairment in cerebral autoregulation in the peri-hematoma region with blood pressure lowering. The major limitation of this study was that the hematoma sizes were small, and may not have had as significant an influence as a larger mass lesion.98 Furthermore, Rosand and colleagues, using dynamic single-section CT perfusion (CTP) imaging found a rim of decreased cerebral blood flow
Acute Stroke and Other Neurologic Emergencies
415
Table 13-10 Blood Pressure Management in ICH
Modified from Broderick, et al: Stroke 1999;30:905, with permission.
in the immediate perihematoma region, but it was unclear whether this area was critically ischemic.99 The guidelines for management of intracerebral hemorrhage set forth by Broderick and associates recommend that “blood pressure levels be maintained below a mean arterial pressure of 130 mm Hg in persons with a history of hypertension,”and that “in patients with elevated ICP who have an ICP monitor, cerebral perfusion pressure (= MAP - ICP) should be kept >70 mm Hg.”92 The first-line intravenous agents to be used in control of hypertension should be the alpha- and beta-blocker labetalol or the beta-blocker esmolol, given the theoretical risk of increased ICP with the use of nitroprusside or nitroglycerine, which are potent vasodilators. This potential risk has not been proven in a clinical study. In addition to the initial cerebral insult caused by the hemorrhage, secondary injury can occur by a number of means, including seizures, hydrocephalus, and edema, all of which can lead to a further increase in ICP. Prophylactic anti-
W9418-13.qxd 11/3/03 2:51 PM Page 416
416
Section I
Introduction
convulsant medications may be considered for patients with lobar or superficial subcortical bleeding, especially if they have had a seizure. Phenytoin remains the preferred agent, given that it may be administered intravenously and its relative lack of impact on level of consciousness. We recommend a loading dose of 15 to 18 mg/kg. The duration of therapy with anticonvulsants is unclear, but if the patient remains seizure-free, and has never had a seizure, antiepileptic medications are typically withdrawn after 4 to 6 weeks of therapy. The role for prophylactic anticonvulsant therapy in patients with a deeper hemorrhage, with or without an intraventricular component, is more controversial, and not generally recommended. Another potentially catastrophic secondary complication of ICH, especially in the right hemisphere or insula region, is the higher propensity for causing abnormal cardiac electrical activity, and “cerebrogenic sudden death.”100 These patients should be monitored closely in the intensive care unit during their first several days after hemorrhage. ICP monitors should be considered for patients with larger hematomas with suspected increases in ICP, or in patients with a deteriorating neurological examination, especially with a declining level of consciousness. An intraventricular drain should be placed in patients in whom hydrocephalus is present, or who are considered to be at high risk for developing hydrocephalus. ICP monitoring was studied in a systematic fashion by Fernandes and associates101 and was thought to play a significant role in predicting delayed deterioration, death, and Glasgow Outcome Scale (GOS) at discharge. The study went on to suggest that ICP monitoring may be a useful guide to determining which patients would eventually require surgical evacuation of their hematomas. The principles for control of ICP are discussed in detail in Chapter 25, but the initial mainstays of treatment include hyperventilation in the acutely decompensating patient (as a temporizing measure), followed by hyperosmotic therapy (using mannitol, hypertonic saline, or glycerol), and perhaps barbiturate coma, although this has never proven definitively to be of benefit in this patient population. In a study by Poungvarin and co-workers102 there was no benefit to ICH patients who received steroids vs. placebo in terms of outcome, and there was a higher rate of infectious complications in the steroid-treated group. We do not recommend the use of corticosteroids in the setting of ICH, other than for the treatment of unrelated disorders as needed. Further medical measures that are commonly implemented in the care of ICH patients include head of bed elevation, cautious sedation to avoid agitation with associated elevations in ICP, avoidance of hyponatremia and hyperthermia, and maintenance of a euvolemic or slightly hypovolemic state. Patients with ICH and immobility are at high risk for developing deep venous thromboses (DVT) and subsequent pulmonary embolism. In a small study by Boeer and colleagues,103 22 ICH patients were treated with subcutaneous heparin (5000 U three times a day), starting day 2 after
Table 13-11 Recommendations for Surgical Treatment of ICH
From Broderick, et al: Stroke 1999;30:905, with permission.
ICH. This significantly lowered the incidence of PE compared with initiation of heparin at a later time (day 4 or day 10), with no increase in the number of patients with increased hematoma size. To date, no larger randomized study has been performed, but these results suggest that early anticoagulation may be safe and effective in this patient population. Patients with ICH may require neurosurgical consultation (Table 13-11). This is especially true in the case of cerebellar hemorrhages, in which mortality is high due to a rapidly expanding mass lesion with edema in the smaller posterior fossa. Patients with a cerebellar hemorrhage greater than 3 cm in diameter should be considered for emergency decompression, especially if there are signs of brainstem compression, hydrocephalus, or neurologic deterioration.93 There is a relatively low morbidity associated with the procedure. The data to support surgical evacuation of hemorrhages in the anterior fossa are less convincing. Patients are typically treated medically with anterior fossa hemorrhages, unless they are showing clear signs of clinical deterioration despite maximal medical therapy. In addition, there may be significant postoperative morbidity associated with hematoma evacuation in the dominant hemisphere, and most clinicians are hesitant to push for surgery in these instances. Furthermore, patients with lobar hemorrhages secondary to amyloid
W9418-13.qxd 11/3/03 2:51 PM Page 417
Chapter 13
angiopathy have exceptionally friable cortical blood vessels, and control of bleeding around the hematoma can be quite difficult. In general, patients with a GCS of 4 or less have a uniformly poor outcome (either death or severely disabled) regardless of whether or not surgery is performed, and thus these patients should be treated medically. The Surgical Trial in Intracerebral Hemorrhage (STICH) is underway to assess whether early surgical evacuation of ICH is superior to conservative medical management in improving patient outcome. The STICH study pilot trial104 was a single-center, randomized trial comparing standard craniotomy performed within 12 hours of symptoms onset vs. best medical therapy. Although the results did not reach statistical significance, the mortality in the surgical group was 17.6% vs. 23.5% in the medically treated group. The study was criticized for having unequal numbers of lobar hemorrhage in each group, as well as the small numbers of patients. The current STICH study is a multicenter randomized trial, with the goal of enrolling 1000 patients.105 Other surgical approaches have been studied as well. Auer and associates105 randomized 100 patients to either best medical therapy or burr hole with endoscopic drainage. In this procedure, the hematoma was continuously irrigated with artificial CSF at a pressure of 10 to 15 mm Hg, followed by suction removal of fluid and clot at regular intervals. Continued bleeding within the hematoma cavity was laser coagulated under direct visual control. The 6-month mortality in the surgical group was 42%, compared with 70% in the medical group (P < .01). In this study, the patients with larger hematoma volumes had a significantly lowered mortality rate, but without an improvement in quality of life. The benefits seen were primarily in patients with smaller hemorrhages (which were lobar) and in the younger patients in the study. Other methods that have been employed to aid in hematoma evacuation include an Archimedes screw inside a cannula,106 an ultrasonic aspirator,107 a modified nucleotome,107,108 a double track aspiration,109 and direct instillation of chemical thrombolytics into a partially evacuated hematoma.108 At this time, there have been no surgical techniques that have been studied adequately enough to provide definitive recommendations.92 Cerebral Venous Thrombosis Cerebral venous thrombosis (CVT) is a disease state in which there is thrombotic occlusion of one or several cerebral veins. This leads to a process of venous backflow or congestion, with associated increase in intracranial pressure and occasionally venous infarction, which commonly has a hemorrhagic component. The most common conditions that lead to CVT are disorders of hypercoagulability (Table 13-12). Numerous studies have suggested that the Factor V Leiden and the Prothrombin 20210A gene mutations are
Acute Stroke and Other Neurologic Emergencies
417
Table 13-12 Causal Factors in the Pathogenesis of Cerebral Venous Thrombosis in Adults Prothrombotic states Pregnancy; puerperium Hereditary coagulopathies Protein S deficiency Antithrombin III deficiency Factor II (prothrombin) gene mutations (20210 G Æ A) Factor V gene mutations (factor V Leiden) Von Willebrand’s disease 5, 10 methylene tetrahydrofolate reductase mutation (677C Æ T) Homocystinuria Familial thrombophilia of unknown nature Coagulopathies secondary to blood dyscrasia Thrombocythaemia Primary polycythaemia Paroxysmal nocturnal haemoglobinuria Iron deficiency anaemia Sickle cell disease Disseminated intravascular coagulation After bone marrow transplantation Coagulopathies secondary to systemic disease Behçet’s disease Carcinoma (breast, prostate) Lymphoma Systemic lupus erythematosus Nephrotic syndrome Vasculitis Ulcerative colitis, Crohn’s disease Antiphospholipid antibodies Coagulopathies caused by drugs Oral contraceptives (3rd generation >2nd) Corticosteroids Dihydroergotamine Androgens “Ecstasy” (3,4-methylenedioxymethamphetamine) Coagulopathies secondary to local infection or infiltration Otitis Sinusitis Dental abscess Tonsillitis Obstruction by tumour Coagulopathies secondary to general infections or infiltration Uveomeningitis Sarcoidosis Chronic meningitis Subdural empyema Carcinomatous meningitis Dural puncture Epidural anaesthesia Metrizamide myelography Diagnostic tap Trauma Unknown (20%) From Van Gijn, et al: J Royal Soc Med 2000;93:230, with permission.
W9418-13.qxd 11/3/03 2:51 PM Page 418
418
Section I
Introduction
associated with CVT.110-113 Other combinations of hypercoagulable states have been implicated: peripartum mothers with protein S deficiency,114 Behçet’s disease while pregnant,115 and oral contraceptive use with the presence of factor V Leiden mutation.116 Other than hypercoagulable states, several other conditions can lead to CVT, including infection (e.g., adjacent bacterial meningitis), infiltration of the venous system by a neoplastic process, direct trauma to the venous system, and dehydration. The etiology remains idiopathic in 20% of patients. The clinical presentation of patients with CVT includes progressive generalized headache, vision changes, and occasionally seizures, lethargy, and focal deficits. Patients frequently are found to have papilledema, and in a young, obese person it may be difficult to distinguish between CVT and benign intracranial hypertension, or “pseudotumor cerebri.” Patients with CVT may have sixth cranial nerve palsy or other cranial nerve palsies, secondary to increased ICP. Venous infarctions develop in the areas adjacent to the venous thrombosis (e.g., parasagittally with superior sagittal sinus thrombosis), and the focal deficits are consistent with the location of the infarction. They are frequently hemorrhagic, and commonly lead to seizures. Seizures or focal deficits may be the presenting feature in 10% to 15% of patients.117 Prognosis depends on the extent of venous thrombosis, the amount of parenchymal damage, and the response to therapy. Additionally, Stolz and colleagues found that intracranial venous hemodynamics, as measured by transcranial duplex sonography, may aid in prognostication.118 The mortality rate is from 5% to 30% in different studies.119-121 The main causes of death include herniation due to increased ICP with a hemorrhagic mass lesion, status epilepticus, and medical complications.117) Features associated with a poor outcome in CVT include rapid onset of symptoms, early low GCS, focal neurologic signs, seizures, and a concomitant infection.122
Neuroimaging used in the diagnosis of CVT includes CT and MRI. With CT, there are two signs that are thought to be pathognomonic for the disease. The first is the “empty delta sign” (Fig. 13-16), with a naturally enhancing dura surrounding a nonenhancing thrombus in the sagittal sinus.123 The second sign is the “cord” or “dense triangle” sign (Fig. 13-17), which signifies an acute thrombus within a vein or sinus, respectively.124 With the use of helical scanning and high-dose intravenous contrast material, CT venography can greatly add to the ability to detect thrombus in the venous system.125 MRI with MR venography may also be used in the diagnosis of CVT, but may be limited in the detection in the acute phase.126 MRI, in comparison to CT, may be more sensitive for detecting early parenchymal changes (including microhemorrhages) that may accompany CVT, especially with the use of susceptibility-weighted127,128 and diffusionweighted imaging.129 The appropriate therapy for CVT remains controversial. Numerous retrospective studies have evaluated anticoagulation therapy130,131 and, although suggestive of a benefit in favor of anticoagulation, were for the most part inconclusive. One small prospective randomized trial was published in 1991.130 This trial was stopped early after 20 patients were enrolled to either intravenous heparin therapy or placebo, as the heparin-treated group appeared to have a significantly lower morbidity and mortality rate (zero deaths versus three in the placebo group), but this trial drew abundant criticism132 due to its small numbers, unclear clinical criteria, and late entry into the trial (over 1 month for the heparin-treated group). A larger prospective study of anticoagulation was performed by De Bruijn and colleagues.133 In all, 59 patients were randomized to either low-molecular-weight heparin (approximately 180 anti-factor-Xa U/kg per 24 hours, divided into twice daily dosing) or placebo. After three weeks, the heparin-treated group received oral coumadin
Figure 13-16. CT scan of head showing empty delta sign in cerebral venous thrombosis.
W9418-13.qxd 11/3/03 2:51 PM Page 419
Chapter 13
Acute Stroke and Other Neurologic Emergencies
419
Figure 13-17. CT venography showing the dense triangle sign that signifies acute cerebral venous thrombosis.
anticoagulation, and the placebo group received nothing. The outcome measures suggested a benefit in favor of the heparin-treated group, but this did not reach statistical significance. In spite of inconclusive studies thus far, heparin anticoagulation is generally regarded as safe and effective therapy of patients with CVT, even with associated intracerebral hemorrhage.134–136 Intravenous thrombolytic therapy has been attempted, but has been unconvincing as a successful therapy thus far.137,138 Only small numbers of patients have been evaluated, with great concern regarding expansion of hematoma volumes.139 Catheter-directed local thrombolysis may be used in selected cases. Many patients are maintained on heparin, and thus one indication for proceeding to the more invasive procedure could be progression of disease, either clinically or radiographically, despite adequate anticoagulation. Other considerations include the clinical state (coma or obtundation) and the degree of clot burden. The data to support the use of local thrombolytic agents are sparse140,141 and studies are limited to small numbers of patients. Additionally, mechanical disruption of clot, as with intra-arterial thrombolysis, may aid in recanalization efforts, and two case reports have illustrated the efficacy of a rheolytic catheter device.142–144 At present, the use of catheter-directed local thrombolysis is limited to specialized care centers and is still considered experimental.
Intensive Care Management of Neuromuscular Weakness Introduction This section first discusses Guillain-Barré syndrome (GBS) and its various subtypes, reviewing etiology, presentation, and management implications, especially from an intensive care standpoint. Second, it presents the modern approaches to and management of myasthenia gravis (MG). Third, guidelines to the approach for the patient who develops neuromuscular weakness in the intensive care unit are provided. Finally, there is a brief discussion of neuroleptic malignant syndrome. Guillain-Barré Syndrome Presentation and Diagnosis GBS encompasses a collection of related illnesses that manifest as a self-limited, acute-to-subacute onset of an inflammatory demyelinating polyneuropathy. New diagnostic and classification criteria were set forth in 2001, delineating specific subsets as acute inflammatory demyelinating polyneuropathy (AIDP) or motor-sensory GBS, pure motor GBS, the Miller-Fisher variant (MFS), bulbar variant, and primary axonal GBS.145 The heterogeneity of presentations, as well as the numerous specific autoantibodies that have been con-
W9418-13.qxd 11/3/03 2:51 PM Page 420
420
Section I
Introduction
nected with specific syndromes, led to the subclassification system. The incidence of GBS is approximately 1 to 2 per 100,000, with speculation that the rising incidence may simply be a reflection of improvements in diagnosis. It is frequently precipitated by an illness, including Campylobacter jejuni,146 cytomegalovirus (CMV),147 herpes simplex virus, and upper respiratory infections and immunizations (Table 13-13).148 However, in up to one third of patients no precipitating cause is found. The underlying pathophysiology is thought to be complement activation triggering myelin destruction in the peripheral nervous system. An infecting organism may induce a cellular and humoral immune response, leading to cross-reactivity with ganglioside surface components of peripheral nerves with similar epitopes (molecular mimicry).149 The complement cascade is triggered by the binding of antibodies to Schwann cells, leading to vesicular myelin degeneration. The axon may also be involved in up to 15% of cases, most typically secondary to Campylobacter infection. The clinical presentation (Table 13-14) commonly consists of the subacute onset of migratory weakness, sensory dysesthesias, and hyporeflexia. A nonspecific but frequently accompanying symptom is back pain.150 The necessary criteria for the clinical diagnosis of GBS were set forth by Van der Meché and colleagues:145 (1) subacutely developing flaccid paralysis; (2) weakness starts on both sides of the body, with a strong tendency to do so symmetrically; (3) deep tendon reflexes decrease and usually disappear altoTable 13-13 Events That May Precede Onset of GBS Infection Campylobacter jejuni Mycoplasma pneumoniae Cytomegalovirus Epstein-Barr virus Human immunodeficiency virus Lyme disease Neoplasia Hodgkin’s disease Other lymphoma Vaccination Rabies vaccine Flu vaccine Tetanus vaccine Drug Zimelidine Penicillamine Streptokinase Captopril Danazol Heroin GBS, Guillain-Barré syndrome. From Fulgham, Wijdicks, et al: Crit Care Clin 1997;13:1, with permission.
Table 13-14 Clinical Presentation of GBS Motor power Upper limb weakness Lower limb weakness Asymmetry in the minority Reflexes Complete areflexia Partial areflexia/hyporeflexia Cranial nerves Ophthalmoplegia, complete or partial Ptosis Facial weakness Bulbar dysfunction Decreased or absent gag reflex Dysarthria Palatal weakness Tongue weakness Dysphagia Weak cough Sensory impairment Light touch Pin prick Vibratory sensation Proprioception GBS, Guillain-Barré syndrome. From Fulgham, Wijdicks, et al: Crit Care Clin 1997;13:1, with permission.
gether transiently; and (4) other causes for a rapidly developing flaccid paralysis are ruled out based on the clinical history and additional tests, as needed. Clinical variants include MFS, in which the cardinal features are ataxia, ophthalmoplegia, and hyporeflexia without appendicular weakness.151 Of significant note, GBS can present with rapidly progressive symptoms, to the point at which patients can appear brain dead.152 In its most severe form, patients can appear comatose, with completely flaccid extremities and absent brainstem reflexes. GBS should be considered in patients who appear comatose or even brain dead, but in whom there is no immediately apparent cause for their clinical state. The differential diagnosis (Table 13-15) for the patient presenting with an acute-to-subacute flaccid paralysis should include: myasthenia gravis and other myasthenia-like syndromes (e.g., Lambert-Eaton syndrome), polymyositis, botulism, tick paralysis, organophosphate poisoning, snake or spider bites, porphyria, glue sniffing, and poisoning by ingestion of puffer fish or fruit of the buckthorn shrub in places where these species are endemic. The evaluation of GBS primarily comprises the clinical examination with supportive ancillary tests, including CSF evaluation and electromyography/nerve conduction studies (EMG/NCS). Neuroimaging is of little use, as the MRI is typically normal. There have been reports of pronounced gadolinium enhancement of the spinal nerve roots and cauda equina153,154 as well as multifocal white matter lesions
W9418-13.qxd 11/3/03 2:51 PM Page 421
Chapter 13
Table 13-15 Differential Diagnosis in GBS Brainstem (pontine) infarction Acute myelopathy/skull base lesions Polymyositis Defects at the neuromuscular junction (myasthenia gravis, Lambert-Eaton myasthenic syndrome, black widow spider venom) Tick paralysis Acute porphyria Critical illness polyneuropathy Polyradiculoneuropathy (Lyme disease, Epstein-Barr virus, hepatitis) GBS, Guillain-Barré syndrome. From Fulgham, Wijdicks, et al: Crit Care Clin 1997;13:1, with permission.
in the periventricular area and brainstem in chronic inflammatory demyelinating inflammatory polyneuropathy,155 but these are of uncertain significance. Typical CSF findings include an elevated protein early in the course of the illness with normal cell counts (albumino-cytologic dissociation), but the protein may be normal in the first week of the illness.156 In clinical variants of GBS, the CSF protein tends to be normal. If a marked pleocytosis (>20 cells) is present, an evaluation for human immunodeficiency virus157 and Lyme disease158 should be undertaken. EMG/NCS tend to be the most useful tests in the evaluation of GBS. Characteristic findings include motor nerve conduction block, prolonged distal conduction, and slowing of nerve conduction. An important early finding is prolongation, dispersion, or absence of F waves, suggesting root demyelination.159 In more severely affected patients, early extensive fibrillations, multifocal conduction block, or inexcitable motor responses suggest a more protracted course.160 Phrenic nerve conduction studies can help to establish the diagnosis, but are of little predictive value for the need for mechanical ventilation.161 Grand’Maison eloquently outlined the usefulness, as well as the pitfalls, of EMG/NCS testing in the intensive care unit.162 Antiganglioside antibody testing has become more frequently employed in the past several years. Antibodies that can be investigated in GBS include GQ1b Ab (associated with MFS),163,164 GM1 Ab, GM2 Ab, GD1a (associated with axonal forms)165 and antibodies to C. jejuni and CMV. The prognosis for patients with GBS is generally good, with 65% achieving a complete or almost complete cure at the end of 1 year, to the point at which they can perform manual work. Of the remainder, up to 8% may die in the acute phase from medical complications, including cardiac arrhythmias and pulmonary emboli.166 Furthermore, Lawn and Wijdicks167 retrospectively analyzed 320 patients with GBS, 14 of whom died. The patients who died were older (mean age 75 years in the fatal group, 55 in the nonfatal group, P = .006), and were more likely to have underlying pulmonary disease (six, or 43% in the fatal group; ten, or 10% in the nonfatal group, P = .004). This study may have
Acute Stroke and Other Neurologic Emergencies
421
been subject to a referral bias, being performed at a tertiary care center where more severely affected and elderly patients were treated. Clearly, intensive physical therapy is essential in the recovery process, and guidelines have been set forth regarding specific physical therapy methods and concerns.168 Management The initial steps in the management should include the assessment of respiratory function, including the patients’ ability to protect their own airway from secretions; management of volume status; and detection of autonomic instability. Indeed, for as much attention as has been paid to the specific therapies for the underlying illness, it is clear that troubleshooting the medical complications is perhaps equally important as a focus of care, if not more so.169 Indications for intubation (Table 13-16) should include vital capacity less than 15 mL/kg and maximum negative inspiratory pressure of less than or equal to 25 mm Hg.170 Patients who do not meet these criteria should undergo incentive spirometry every hour to prevent atelectasis. Other clinical markers that may be used include restlessness, tachycardia, tachypnea, staccato speech, use of accessory muscles, paradoxical breathing, and sweating. Patients who are rapidly developing symptoms and signs of respiratory distress should be electively intubated, rather than waiting for the crisis of respiratory decompensation. Lawn and associates171 retrospectively analyzed 114 patients, measuring the clinical and electrophysiologic features in 60 patients with GBS who received mechanical ventilation, and 54 who did not. They found that the clinical features that were associated with a high likelihood of need for mechanical ventilation included rapid disease progression, bulbar dysfunction, bilateral facial weakness, and dysautonomia. Respiratory parameters included a vital capacity less than 10 mL/kg, maximal inspiratory pressure of less than 30 cm H2O, maximal expiratory pressure less than 40 cm H2O, or a reduction of more than Table 13-16 Clinical and Laboratory Criteria for Mechanical Ventilation in Patients with GBS Clinical
Restlessness, anxiety Tachycardia Tachypnea Staccato speech Inability to count to 20 on one breath Use of accessory muscles (heightened activity of sternocleidomastoid muscles) Parodoxic breathing
Laboratory
VC £15 mL/kg* Pi max £ -25 mm Hg Hypoxemia Respiratory acidosis on arterial blood gases (late)
*Or 50% decline from baseline. GBS, Guillain-Barré syndrome; VC, vital capacity. From Fulgham, Wijdicks, et al: Crit Care Clin 1997;13:1, with permission.
W9418-13.qxd 11/3/03 2:51 PM Page 422
422
Section I
Introduction
30% in any of those three parameters. The presence of a chest radiograph abnormality, upper limb paresis, age, and preexisting pulmonary disease all showed trends for intubation, but did not reach statistical significance. It is important to stress the precarious situation of the patient with bulbar weakness with GBS, with or without appendicular weakness. These patients are particularly susceptible to aspiration, and need close monitoring for the need for mechanical ventilation. Respiratory rate, adequacy of coughing, the ability to count to 20 on one breath, and other methods of measuring bulbar function should be employed.170 Patients are typically placed on an SIMV setting of six to ten breaths per minute, with tidal volumes of 10 to 12 mL/kg and 5 cm H2O CPAP to prevent atelectasis. Patients with adequate triggering may perform well on pressure support ventilation (PSV), and this is the more comfortable setting. Early tracheostomy can be considered in patients with an expected protracted course, such as those with a rapidly progressive quadriplegia; those with EMG evidence of widespread fibrillations, low amplitudes or absent responses; and those with no response after a course of IVIG or plasma exchange. Aggressive pulmonary toilet with frequent suctioning is paramount in these patients to prevent atelectasis and pneumonia. Diaphragmatic weakness can improve significantly earlier than weakness in the extremities, and thus respiratory parameters should be observed as the indication for extubation. Extubation should be delayed in patients with ongoing dysautonomia, as the stress of weaning can cause dramatic fluctuations in blood pressure, as well as cardiac arrhythmias. Weaning typically occurs 20 to 30 days after intubation, and tracheostomy should be delayed during this time to await the potential effectiveness of specific therapy. Weaning entails reducing the IMV and switching to PSV. When the vital capacity has reached 15 mL/kg, the patient can produce tidal volumes of 10 to 12 mL/kg, and the pressure support has been reduced to 5 cm H2O or a level at which imposed work of breathing is zeroed out, extubation can be attempted. It is important to institute rest periods during the weaning period, especially when nearing the time of extubation. Some clinicians prefer weaning with a T-piece and humidified oxygen, but this may place additional stress on the patient. Dysautonomia is one of the most significant causes of morbidity and mortality in patients with GBS. The autonomic nervous system can be involved in patients with varying degrees of weakness, but is most typically seen in patients with rapidly progressive symptoms/signs, bulbar weakness, and ophthalmoplegia, as well as those with severe motor weakness and respiratory failure.172,173 The cardinal manifestations of dysautonomia are blood pressure fluctuations and cardiac dysrhythmias. Hypotension in patients with GBS can be caused by sepsis, pulmonary embolus, venous pooling or severe electrolyte disturbances, but wide fluctuations of blood pressure over minutes is suggestive of dysautonomia. The presumed cause is impaired barorecep-
tor buffering.174 Hypotension may also follow the administration of vasoactive medications by the mechanism of denervation hypersensitivity175 with vagal stimulation during tracheal suctioning or ocular pressure. Most patients will show some spontaneous fluctuations of blood pressure that quickly self-correct, and these are best left untreated. Persistent hypotension is treated by placing the patient in the Trendelenburg position and administering a fluid bolus. Alpha-agonists, such as phenylephrine, can be used, but can cause hypertension by overcompensation. Hypertension can also be present, even without the use of vasopressor medications, because increased sympathetic outflow and activity are common in patients with GBS.176,177 Persistent hypertension may lead to congestive heart failure and/or cardiac ischemia in patients with underlying coronary artery disease or systolic dysfunction, and can be treated with a morphine bolus, or with careful administration of intravenous antihypertensive medications with a short half-life, such as sodium nitroprusside. Beta-blockers should be avoided, as their use has been linked with cardiac arrest in patients with dysautonomia and profound bradycardia. The short-acting (T1/2 ~9 minutes) beta-blocker esmolol has been used successfully in a pediatric patient with GBS.178 Careful attention to volume status may aid in treating hypotension, and a pulmonary artery catheter may be helpful in selected cases. Cardiac dysrhythmias are most often insignificant, primarily manifesting as sinus bradycardia or sinus tachycardia. Sinus bradycardia, sinus arrest, and atrioventricular block typically occur in the setting of intubation and with tracheal suctioning. A frequently reported cause of sudden death in patients with GBS who experience dysautonomia is complete heart block. This is emergently treated with a temporary pacemaker. Few patients ultimately require permanent pacing.179 Untreated vagal activity can progress to profound bradycardia and asystole.180,181 On the other hand, insufficient vagal tone can lead to tachyarrhythmias, including sinus tachycardia, atrial fibrillation, atrial flutter, and ventricular tachycardia.182,183 Identifying patients with GBS at risk of developing dysrhythmias is of paramount importance.184 Other consequences of the dysautonomic state include gastroparesis and bladder paralysis. All patients should be given stool softeners, and many will require bowel motility agents. Ileus develops in 3% to 5% of patients, and is treated with continuous suctioning, intravenous hydration, and placement of a flatus tube. If ileus persists beyond three days, assuming normal nutritional status at disease onset, parenteral nutrition should be considered. For patients with bladder paralysis, intermitted catheterization can be attempted, but most will require indwelling catheter placement. Patients should continue to take in enteral nutrition for as long as possible. For those with bulbar weakness and swallowing difficulties, a nasogastric tube should be placed, and enteral feeding begun. Attention must be paid to the devel-
W9418-13.qxd 11/3/03 2:51 PM Page 423
Chapter 13
opment of ileus, and parenteral nutrition, either peripheral or central, should be instituted as noted previously. General medical measures should also be instituted in the care of GBS, as with other prone patients. This should include DVT prophylaxis with subcutaneous heparin (5000 U twice daily) or low-molecular-weight heparin, as well as intermittent pneumatic compression devices. Patients should be turned frequently to prevent pressure sores. Proper positioning is important to prevent pressure palsies at peripheral nerves, including the ulnar nerve at the elbow, and the peroneal nerve at the fibular head. Psychological support is needed, with reassurance that many patients with GBS have a full recovery. Pain is a common component of the disease, but tricyclic antidepressant medications, with their anticholinergic side effects, must be used with great caution in this patient population prone to autonomic dysfunction. Other agents that may be used in pain control, especially that of a neuropathic quality, include antiepileptic medications, such as valproic acid and gabapentin. Nonsteroidal anti-inflammatory drugs are also useful, but many patients may require narcotic medications, such as morphine and its longer acting forms. Specific Therapy Given the underlying pathophysiology of an autoimmune reaction, directed treatment approaches in GBS have focused on immunomodulatory actions, including plasma exchange, intravenous immunoglobulin (IVIG), and steroids. Recently, a review was performed for all three modalities in the treatment of GBS, using the Cochrane Neuromuscular Trial Register.185–187 It concluded that plasma exchange is the “first and only treatment that has been proven to be superior to supportive treatment alone in Guillain-Barré syndrome.” The benefit was seen for mild, moderate, and severe forms of the disease. Furthermore, continuous flow plasma exchanges were thought to be superior to intermittent treatments, and albumin was felt to be safer as the replacement fluid than fresh-frozen plasma. Plasma exchange was felt to be most beneficial when started within the first seven days of presentation, but could still be beneficial in patients treated up to 30 days. Plasma exchange is considered the “gold standard” against which all other treatments are compared. Relative contraindications to treatment with plasma exchange include sepsis, myocardial infarction within 6 months, marked dysautonomia, and active bleeding. Side effects from the treatment may include vasovagal reactions, hypovolemia, anaphylaxis, hemolysis, air embolism, hematoma formation, hypocalcemia, thrombocytopenia, hypothermia, hypokalemia, and postpheresis infection. The Cochrane Review did not find sufficient evidence based on adequate clinical trials in the past to determine a benefit for IVIG over placebo (supportive therapy). They did find that IVIG had a similar ability to plasma exchange in speeding the recovery from GBS, but that IVIG following plasma exchange was not significantly better than plasma
Acute Stroke and Other Neurologic Emergencies
423
exchange alone. It is unclear from the trials thus far performed whether IVIG is helpful in milder forms of the disease, or when the symptoms have lasted more than two weeks. Two small studies have suggested a higher relapse rate with IVIG than plasma exchange188,189 but this has yet to be demonstrated in a larger trial. IVIG is typically less expensive than plasma exchange, and does not require the placement of a central venous catheter. The standard dose is 0.4 g/kg/day for 5 days. Side effects include aseptic meningitis, anaphylaxis, acute renal failure, and thromboembolic events (including ischemic stroke).190 Patients who relapse with either therapy will typically respond favorably to a second course with the same therapy.191 Of note, a randomized, multicenter study comparing IVIG, plasma exchange, and immune adsorption was attempted.192 In total, 67 patients were randomized to the three groups, and there were no statistical differences between the groups in either outcomes or adverse events. Unfortunately, the study was discontinued prematurely, secondary to a declining referral base in Europe in the late 1990s with the publication of favorable results with the use of IVIG, as small hospitals were subsequently treating their patients locally with the less invasive therapy. The Cochrane Review also evaluated six randomized or quasi-randomized clinical trials of corticosteroids in GBS, and found it to be of no benefit over controls. They concluded that it does not have a primary role in the treatment of GBS, but if the patient with GBS required corticosteroid treatment for some other reason it would probably not be harmful to their recovery. On the other hand, corticosteroids have been well established to be effective in the treatment of chronic inflammatory demyelinating polyneuropathy,193 but these patients rarely require an ICU level of care. Myasthenia Gravis Presentation and Diagnosis Myasthenia gravis (MG) is an autoimmune disease characterized by an autoantibody reaction at the antigen epitopes of the acetylcholine receptor, leading to destruction and simplification of the junctional fold, with subsequent widening of the synaptic cleft. The prevalence of the disease is 50 to 125 cases per million.194 It occurs at all ages, with a peak in women in the second and third decades, and a peak in men in the sixth and seventh decades. The most common reason for ICU admission is myasthenic crisis, in which respiratory decompensation is the most threatening complication. Triggers for a myasthenic crisis include recent upper respiratory infection, surgery, delivery, and use of an exacerbating medication. In most patients, symptoms become most severe within the first 3 years of onset. The natural history is of spontaneous, typically short remissions. There are some reports of spontaneous remissions lasting more than 10 years. Patients
W9418-13.qxd 11/3/03 2:51 PM Page 424
424
Section I
Introduction
typically present with weakness and easy fatigability of the skeletal muscles, primarily affecting the proximal musculature more than the distal. The weakness characteristically is exacerbated by repeated activity, and alleviated by rest. Ptosis and ophthalmoparesis occur commonly early in the course of the disease, and the weakness can remain localized to the extraocular and levator palpebrae muscles in 15% of patients.195 When the facial musculature is prominently involved, there is a tendency for other elements of bulbar weakness, including hypophonic or nasal speech, as well as difficulty with chewing and swallowing. Generalized weakness develops in approximately 85% of patients,195 often affecting the diaphragm. If the weakness affecting respiration is severe enough to require mechanical ventilation, the patient is in myasthenic crisis by definition. It is important to objectively quantify the degree of weakness to establish the clinical course, as well as the response to treatment. In particular, bulbar function must be assessed early in order to determine the necessity for elective intubation. The degree of ptosis is established by having the patient maintain an upward gaze for three minutes. Jaw opening is typically stronger than jaw closure, and forceful biting on a tongue depressor for 20 seconds can establish the degree of closure weakness. Nasopharyngeal weakness can be determined by listening to the voice for slurring or nasal tone. Dysphagia is assessed by having the patient attempt to take a few sips of water. Respiratory failure can present with staccato-type speech, restlessness, diaphoresis, tachycardia and tachypnea. The tachypnea typically allows the patient to have a normal pCO2 despite low tidal volumes. The degree of respiratory weakness can be measured at the bedside by having the patient count to 20 on one breath, or by performing sequential vital capacity measurements. Isolated respiratory failure has been reported as the presenting symptom in a myasthenic patient.196 In patients with a known diagnosis of myasthenia who present with a respiratory decompensation, it is important to evaluate for cholinergic crisis as the cause, which can occur when patients are overmedicating. Clues are excessive salivation, thick bronchial secretions, and diarrhea. The diagnosis of MG is typically made on clinical grounds. The edrophonium test is helpful in establishing the diagnosis, but should only be performed in a well-monitored setting, such as the emergency room or ICU. Markers for improvement can include improvement in upward gaze, ptosis, and dynamometry of a muscle or muscle group in the extremity. The dose is 1 mL of edrophonium in a 10 mg/mL solution. One-tenth of a milliliter (1 mg) is given as a test dose, waiting 30 seconds for excessive muscarinic effects. The remainder (9 mg) is given over 1 minute. Edrophonium has a rapid onset (30 seconds) and a short duration of action (2 to 20 minutes). The test result is considered positive if there is unequivocal improvement in an objectively tested weak
muscle.197 Atropine at a dose of 0.5 mg IV bolus should be given if abdominal cramps, bronchospasm, vomiting or bradycardia occurs. If the bradycardia persists and is accompanied by hypotension, an additional 1 mg dose of atropine should be given. Edrophonium can also be given when cholinergic crisis is considered, but in a smaller dose (1 mg). EMG/NCS are useful in establishing the diagnosis of MG as well. The patient should be without anticholinesterase medications for 12 hours before testing. Any significant improvement with plasmapheresis or IVIG is delayed by 2 to 3 days, and thus testing during this period is valid. Surface electrodes are used for repetitive stimulation at a rate of 2 to 5 Hz before and after maximal voluntary contraction of the tested muscle. An abnormal result is defined as a 15% or greater reduction of the compound muscle action potential (CMAP) amplitude between the first and fourth responses with supramaximal stimulation.198 Single-fiber EMG shows increased jitter and blocking. In more severe disease, needle EMG can show fibrillations, indicating functional denervation of the muscle fibers.199 Several autoantibodies have been shown to be useful in the diagnosis. Acetylcholine-receptor antibodies are positive in approximately 85% of all patients with MG, but in a lower proportion (approximately 50%) in patients with isolated ophthalmoparesis or bulbar weakness.200–202 Straitional antibodies are present in 80% of patients with thymoma. In seronegative patients, antibodies against thyroid or gastric parietal cells can support the diagnosis.203 All patients with MG should have a chest CT or MRI performed to detect the presence of a thymoma or enlargement of the thymus gland.204 If a thymoma is detected, it is an absolute indication for removal unless the patient is a poor operative candidate, or if the tumor is significantly involving the mediastinum. For tumors that cannot be completely resected at the time of surgery, focused radiation can be carried out postoperatively. Patients between the ages of puberty and 60 should have a surgical thymectomy,205 because removal of the thymus can induce remission in a significant proportion of patients.206 However, patients may be weaker following their thymectomy, either from the stress of surgery or from the loss of a suppressive effect from the thymoma,207 and may require further immunosuppressive therapy. The differential diagnosis for MG should include Lambert-Eaton myasthenic syndrome, which may be the initial presentation of an occult malignancy208,209; congenital myasthenic syndromes210; Graves disease; botulism; progressive external ophthalmoplegia211 and intracranial mass lesions.212 Lambert-Eaton syndrome is distinguished from MG on the basis of the clinical presentation, in which there are commonly seen autonomic and sensory symptoms, as well as with EMG/NCS, in which there is an improvement in response to repetitive stimulation.
W9418-13.qxd 11/3/03 2:51 PM Page 425
Chapter 13
Another highly important consideration in the differential diagnosis is drug-induced or exacerbated MG.213–215 The medications that can aggravate MG include certain antibiotics, cardiovascular agents, anti-epileptic medications, and psychotropic agents. The prognosis with MG has dramatically improved, with a mortality rate of essentially zero.206 Some patients are able to achieve a complete remission with thymectomy or induction with immunosuppressive agents, but the majority of patients must take immunosuppressive medications indefinitely. There are reports of spontaneous remissions lasting more than 10 years.216 Management The cause for ICU admission for patients with MG is usually myasthenic crisis with respiratory decompensation. This can be precipitated by a recent infection (especially an upper respiratory tract infection), new insulting medication, initial administration of a large dose of corticosteroids, recent tapering of steroids, hyperthyroidism or hypothyroidism, surgery, or postpartum state. Patients with early signs of respiratory distress should be admitted to the ICU for close observation, with consideration for early intubation before decompensation. Although many find it useful to perform serial bedside measures of pulmonary function, including negative inspiratory force and vital capacity, these measures have recurrently failed to be adequate predictors of the need for mechanical ventilation in these patients.217,218 Normally, when the vital capacity has fallen to less than 15 mL/kg, or is less than 25% of the predicted value, ventilatory failure is considered imminent.219 However, the performance on these respiratory tests may be hampered by the myasthenic patient’s inability to form a tight seal on the mouthpiece due to weakened facial musculature, giving misleading measurements. Further complicating factors for respiration in myasthenics include difficulty with controlling airway secretions due to bulbar weakness, a situation exacerbated at times by anticholinesterase medications that can increase secretions. Aggressive pulmonary toilet and frequent suctioning may aid in the prevention of aspiration and subsequent pneumonia. Hypoxemia precedes hypercarbia, and is the harbinger for the incipient need for mechanical ventilation. Once intubated, patients are most comfortable when on pressure support ventilation, provided that they have adequate triggering of the ventilator. Often, SIMV may be used at night to provide additional rest. The patient is continued on mechanical ventilation until it is clear that their respiratory mechanics are improving, they are medically stable, and they have completed several days of definitive therapy for their disease. Again, patients may often worsen shortly after the institution of corticosteroid therapy, and some suggest starting therapy at lower levels or on alternating dosing schedules.220
Acute Stroke and Other Neurologic Emergencies
425
Specific therapy of MG includes immunomodulation and enhancement of neuromuscular transmission with anticholinesterase agents. Plasma exchange is thought by many to be more effective than IVIG, but the treatment must be tailored to the individual patient.221 The typical course is removal of 2 to 3 L of plasma three times a week, with replacement with albumin to preserve oncotic pressure. This is repeated until improvement plateaus, which is usually after five to six exchanges. Improvement typically begins within the first 48 hours of the first exchange, and can last for weeks to months.222 IVIG has also been shown to be effective in the treatment of myasthenia, and the typical protocol is 0.4 g/kg/day for 5 days.223 There may be a role for IVIG following plasma exchange in the future as a means of prophylaxis against relapse, but no clinical trial has been conducted to date. Additional agents used in the treatment of MG include anticholinesterase medications and corticosteroids, but their use in the acute setting is a matter of continued debate. A recent retrospective study by Berrouschot and co-workers224 compared the use of pyridostigmine alone, pyridostigmine plus prednisolone, and plasma exchange, and suggested that none demonstrated a significant superiority in outcome. This study had numerous limitations, however, with variable treatments used, as well as an unusually high mortality rate of 17%. The concern raised by using anticholinesterase medications in the acute setting pertains to the increased secretions caused, which may lead to respiratory worsening with atelectasis and pneumonia. Previous reports suggested a benefit in favor of a continuous infusion of pyridostigmine in patients with myasthenic crisis.225 Corticosteroids are typically continued for 1 month, and there is general consensus that they aid in keeping patients in remission. Other modes of therapy are considered second line, and are reserved for patients who have poor tolerance or contraindications to corticosteroid therapy. These include azathioprine,226 cyclosporine,227 and immunoadsorption therapy. In known myasthenic patients with a new respiratory decompensation, it is important to distinguish whether they are having an exacerbation of their underlying illness (myasthenic crisis) or are overdosing on their anticholinesterase medication (cholinergic crisis). Patients with progressive weakness may incrementally increase their anticholinesterase medication in an effort to avoid a myasthenic crisis, but this may lead to progressive respiratory difficulties due to the increased secretions. Other signs that help to distinguish cholinergic crisis include miosis (patients with myasthenic crisis typically have mydriasis), muscle fasciculations, abdominal cramping, diarrhea, and excessive sweating and tearing. The edrophonium test may be helpful in distinguishing the two as well, as there may be worsening in the patient with cholinergic crisis, versus possible improvement in the patient with myasthenic crisis.220 Treatment consists
W9418-13.qxd 11/3/03 2:51 PM Page 426
426
Section I
Introduction
of withholding or weaning the anticholinesterase medication, being mindful of potential rapid worsening of weakness and respiratory distress.228 Thymectomy is recommended for most patients with MG, but perioperative management is complicated, as clinical worsening is common.207 Approximately 50% of patients show clinical improvement with thymectomy, with the prospect of achieving a complete remission.229 Strong indications for surgery include severe generalized MG, age less than 60 years, patients with a progressively decreasing response to medication, and patients with repeated episodes of myasthenic crisis.205 Transcervical or transsternal thymectomy may be performed. Transcervical thymectomy may be associated with damage to the phrenic nerve, parathyroid gland, or recurrent laryngeal nerve, as well as pneumothorax and hemothorax.230 Extended thymectomy via transsternal approach is recommended for patients with presumed thymoma, especially if considered invasive.231 Anticholinergic medications are discontinued on the morning of surgery. Atropine or glycopyrrolate is used to decrease secretions. In patients on corticosteroid maintenance, stress dose steroids are given (typically hydrocortisone 125 mg IV every 8 hours for 3 days postoperatively). Some also advocate the use of plasma exchange before surgery. D’Empaire and associates232 retrospectively evaluated 37 patients with myasthenia who underwent thymectomy. For the 11 patients who underwent prethymectomy plasma exchange, they found a significantly decreased time on mechanical ventilation and a shorter stay in the ICU, compared with the 26 patients who did not receive plasma exchange. Intensive Care Unit Weakness A common challenge in the surgical or medical ICU is determining the cause of weakness in critically ill patients. Although GBS and MG can occur in the ICU setting and should be considered as potential etiologies, other conditions are more common, including critical illness polyneuropathy (CIP), critical illness myopathy, and deconditioning syndrome. Other less common entities include acute rhabdomyolysis and central pontine myelinolysis. Often there are multiple causes at play, and it may be erroneous to try to label a patient with a particular syndrome when overlap is so common.233 Critical illness polyneuropathy typically appears in elderly, severely ill patients, often with sepsis syndrome.234,235 It is a self-limited process, and recovery is often quite good, if the underlying critical condition can be adequately treated. In some studies it has been found to develop in 70% to 80% of patients with severe sepsis and multiple organ failure.236,237 The exact cause is unknown, but some have postulated the role of systemic inflammatory factors involving the peripheral nerves.238 Additional risk factors that have been postulated include duration of mechanical ventilation,
hyperosmolality, parenteral nutrition, nondepolarizing neuromuscular blockers and neurologic failure,239 as well as the APACHE III score.240 Examination in CIP reveals that both motor and sensory systems are affected, with flaccid tetraparesis and muscle atrophy. Deep tendon reflexes are typically reduced or absent, although they may be preserved in up to one third of patients.241 EMG/NCS reveal a distal axonal sensorimotor polyneuropathy, with fibrillation and positive sharp waves in the proximal and distal muscles, and with relative sparing of the facial muscles.242 Spinal cord injuries must also be ruled out if the patient has undergone a significant traumatic injury. Nerve biopsy reveals predominantly axonal degeneration,243 as well as denervation atrophy of both proximal and distal muscles, chromatolysis of anterior horn cells, and loss of dorsal root ganglia in more severe cases.244 Suxamethonium should not be given to patients with CIP given their risk of developing hyperkalemic cardiac arrest.245 Outcome is guarded, and closely tied in with the underlying critical illness. The course of recovery is protracted, often with a less than favorable ultimate functional status.246,247 Patients with slowing of nerve conduction may have a particularly poor prognosis. Some argue that use of intermitted bolus pancuronium may be less prone to inciting the syndrome than a continuous infusion, but this has yet to be proven definitively.248 Critical illness myopathy is often difficult to distinguish from CIP, and there is commonly an overlap of the two. In fact, some have proposed the term critical illness polyneuromyopathy, or CIPNM.249 Some argue that myopathy is even more common than polyneuropathy in critical illness.250 A proposed mechanism for injury is the generalized catabolic state induced by the systemic inflammatory response syndrome, with a coincident catabolic state.251 Corticosteroids are thought to play a vital role in the development or exacerbation of the syndrome, as is the use of neuromuscular blocking agents.252,253 Pathologic changes include abnormal fiber size, atrophy, angulated fibers, internalized nuclei, rimmed vacuoles, fatty degeneration, fibrosis, and single fiber necrosis.253 All patients with neuromuscular weakness being evaluated for possible myopathy should undergo repetitive stimulation testing during EMG to rule out an underlying neuromuscular junction process (MG), which could be unmasked due to the use of neuromuscular blocking agents. Thick filament myopathy is a variant that is often seen in patients who have received corticosteroids for acute severe asthma or for organ transplantation, with or without concomitant neuromuscular blocking agents.234 Its appearance may be potentiated by disuse.254 Still another variant is necrotizing myopathy, which is distinguished by often highly elevated serum creatine kinase. No specific therapies for these different myopathies has been found to be helpful, other than minimizing offending agents, such as neuromuscular blockade, sepsis, and corticosteroids, as well as proper
W9418-13.qxd 11/3/03 2:51 PM Page 427
Chapter 13
supportive care of patients, including positioning to avoid damage due to pressure. There are ongoing arguments as to whether patients with suspected critical illness myopathy should undergo muscle biopsy, as it rarely impacts treatment. The case has been made that muscle biopsy should be reserved for patients in whom an inflammatory myopathy might be present, as this would clearly impact therapy.255 Some advocate the use of EMG/NCS early in the ICU course, to predict which patients are likely to develop CIP or CIM, so that potentially offending agents may be minimized or withheld.256 Deconditioning syndrome occurs in immobile, poorly nourished patients, who reach a highly catabolic state. It is exacerbated by prolonged, dense neuromuscular blockade. On examination, patients have diffuse weakness as well as atrophy. EMG/NCS are often normal. Muscle biopsy reveals predominantly type II fiber atrophy. If the underlying cause of immobility can be corrected, the clinical course is often favorable, with intensive physical therapy. Acute rhabdomyolysis occurs with traumatic crush injuries, drug overdose, toxin exposure, severe metabolic abnormalities, and infections.257 Patients have swollen, tender muscles, with localized or diffuse weakness. There is breakdown of skeletal muscle, with leakage of its intracellular contents, causing secondary organ damage. The creatinine phosphokinase concentration is elevated, and there may also be a leukocytosis. Urine myoglobin is present, and acute renal failure may ensue. Other metabolic abnormalities include hyperkalemia, hyperuricemia, hypocalcemia or hypercalcemia, hyperphosphatemia, lactic acidosis, thrombocytopenia, and disseminated intravascular coagulation (DIC). EMG/NCS show increased spontaneous activity with myopathic changes. Muscle biopsy reveals diffuse muscle fiber necrosis. Management consists of controlling the renal failure, which may often require temporary hemodialysis (either standard intermittent dialysis or continuous venovenous hemodialysis), as well as correcting metabolic abnormalities and DIC as possible. The clinical course can be favorable, provided that the associated injuries are not severe, and the underlying pathology can be corrected. Central pontine myelinolysis (CPM) can occur in the ICU setting, where there are often rapid alterations in serum electrolytes, particularly sodium. The syndrome can occur with alcoholism, dehydration, or serious systemic illness, often following rapid correction of hyponatremia.258,259 It has also been described in burn patients who terminally develop severe hypernatremia, having previously been normonatremic, suggesting that the sudden rise in sodium or osmolality is the precipitating mechanism.260 Patients with CPM typically have an impaired level of consciousness, ranging from confusion to coma.261 Paresis involves the upper extremities more than the lower extremities, and sixth cranial nerve palsies and rigidity are common. Other ocular abnormalities include miotic or mydriatic pupils, conjugate gaze palsies, and ocular bobbing.261 Larger lesions can cause
Acute Stroke and Other Neurologic Emergencies
427
a “locked-in syndrome,” with preserved consciousness but profound tetraparesis and loss of horizontal eye movements. EMG/NCS are normal, and there is no underlying muscle or peripheral nervous system pathology. Neuroimaging and pathology reveal extensive myelinolysis in the central pontine region. The reason why the pons is so dramatically affected in this disorder is unclear, but it has been postulated that the oligodendrocytes in the pons are located close to the highly vascularized gray matter, causing it to be particularly susceptible to damage from vasogenic edema and leakage of myelinotoxic substances from the vessel.262 Other areas of brain that can be affected include the midbrain, basal ganglia, white matter of the folia cerebelli, and in the deep layers of the cerebral cortex and adjacent white matter.263–265 Treatment is limited to supportive care. Steroids have been found to be of no benefit.261 There is some suggestion of benefit from use of thyrotropin-releasing hormone, but this has been limited to case reports primarily.266,267 The outcome is generally quite poor, with many patients left with little or no improvement.261,268 Neuroleptic Malignant Syndrome Neuroleptic malignant syndrome (NMS) is a rare neuromuscular disorder, originally described by Delay in 1960,269 that occurs in the setting of neuroleptic medication. It is more common in males than females, and in two prospective studies the incidence rates were 0.07% and 0.9%.270,271 The disease has quite serious potential consequences, with a mortality rate ranging from 20% to 30%.272,273 The cardinal features are severe muscle rigidity with increased heat production, leading to hyperthermia and profound autonomic instability, secondarily resulting in hypertension, tachycardia, and tachypnea. Alteration of consciousness is common, ranging from confusion to coma. The hyperthermia is likely related to increased heat production with the increased muscle rigidity, to the point at which physiologic mechanisms to reduce heat (sweating, panting) are overwhelmed.274 Hyperthermia results in a hypermetabolic state, with subsequent tachycardia and tachypnea. Tachypnea can also result from decreased chest wall movement with the muscle rigidity. The autonomic dysfunction is theoretically possible on the basis of either the heightened peripheral hyperadrenergic state, or due to central effects of neuro-leptic medication. However, given the rapid resolution of autonomic dysfunction with appropriate therapy with dantrolene to decrease muscle rigidity, the first hypothesis is more likely.275 Despite numerous reports of the condition, it is still considered a relatively rare complication of treatment with antipsychotic medications, and given the patient population in which the alteration in consciousness appears (psychotic or agitated patients with an already altered sensorium), the diagnosis is considered quite difficult to make.276 The criteria for diagnosing NMS were established by Pope and associates in 1986.277 The first requirement is hyperthermia, with
W9418-13.qxd 11/3/03 2:51 PM Page 428
428
Section I
Introduction
an oral temperature of at least 38°C without another known cause. Second, at least two of the following extrapyramidal signs must be present: lead pipe muscle rigidity, trismus, pronounced cogwheel rigidity, dysphagia, sialorrhea, choreiform movements, oculogyric crisis, dyskinetic movements, retrocollis, festinating gait, opisthotonus, and flexor/extensor posturing. Third, autonomic dysfunction must be manifested by two of the following: hypertension (defined as at least 20 mm Hg diastolic above baseline), tachycardia (defined as at least 30 beats above baseline), tachypnea (defined as at least 25 respirations per minute), prominent diaphoresis, and incontinence. Additional features that may support the diagnosis include alteration in consciousness, leukocytosis of greater than 15 ¥ 109/L, and creatine kinase greater than 1000 U/L. Essentially all neuroleptics can produce NMS, including atypical antipsychotics.278,279 Metoclopramide, an antiemetic medication, as well as tricyclic antidepressants have also caused NMS.280,281 Other agents can cause an NMS-like syndrome, including anticholinergics,282 selective serotonin reuptake inhibitors,283 and cocaine.284 The presumed pathologic mechanism underlying NMS is sudden and profound central dopamine blockade in the setting of receiving neuroleptic medications, particularly affecting the basal ganglia and hypothalamus.285,286 Some postulate that there may be an individual or genetic predisposition to developing the syndrome.287 Certain psychiatric states may also predispose toward the development of NMS, including catatonia288 and affective disorder.289 Other predisposing factors include the physical state of the patient, particularly the volume status, with dehydration playing a prominent role, as well as exhaustion and concurrent organic brain disease.290 Additionally, the potency of the neuroleptic used and its route have vital importance, with high potency neuroleptics given intramuscularly or intravenously carrying significant risk.290 Iron deficiency may also predispose the patient to developing NMS.291 Pathologically, the diagnosis can be suggested by exposing muscle tissue to caffeine or halothane in vitro, which has induced a hypercontractile state in patients with NMS, but only in a small case-control study.292 In another small series of three cases, pathologic features that were described included significant edema of muscle fibers, vacuolization in the sarcoplasm, and
contraction bands separating some myofibrils.293 There have been no large postmortem analyses, and thus the diagnosis remains clinical. A number of laboratory studies are helpful in the evaluation of a patient with suspected NMS, although none are specific for the disease. Creatine kinase is typically elevated above 1000 U/L. The white blood cell count is typically greater than 10,000 cells/mL. As with other patients with delirium, additional laboratory tests may be helpful, including urinalysis, arterial blood gas analysis, electrolytes, hepatic and renal function tests, thyroid function tests, and a cortisol level. Blood and urine cultures should be considered, as well as a lumbar puncture. Toxicology screens may be indicated for certain patients. Other tests to consider include EEG, CT, or MRI. The treatment of NMS first consists of stabilizing the patient from a cardiovascular and respiratory standpoint. All patients with a possible diagnosis of NMS should be admitted to an intensive care unit. Fluids should be administered to correct for dehydration, and all neuroleptic medications should be discontinued, as should lithium.294 Antipyretic medication should be administered, and a cooling blanket placed on the patient. Patients with severe tachypnea secondary to extreme muscle rigidity may require mechanical ventilation. Due to a decreased gag response, NMS patients are at high risk for aspiration pneumonia as well, which may also lead to the need for mechanical ventilation. Patients should be given subcutaneous heparin or low-molecularweight heparin to prevent deep venous thrombosis during immobility. Specific therapy for the syndrome typically follows conservative therapy, because patients typically improve within 2 to 14 days with removal of the offending medications.295 Bromocriptine and other dopamine agonists are reasonable options, at doses of 2.5 to 7.5 mg three times daily, up to a total of 45 mg/day. Dantrolene may be added, at 1 to 10 mg/kg intravenously, or divided oral doses of 50 to 600 mg/day, but with careful attention to possible hepatitis with this medication.296 Electroconvulsive therapy has also been advocated in the treatment of NMS, and this may have the added benefit of treating the underlying condition as well.297,298 Recently, clonidine has been suggested as acute treatment in the ICU.299
W9418-13.qxd 11/3/03 2:51 PM Page 429
Chapter 13
Acute Stroke and Other Neurologic Emergencies
429
P earls 1. Large vessel atherothrombosis encompasses approximately 15% of all strokes. Of these, 9% are of extracranial ICA origin, and 6% are due to intracranial atheromatous disease. 2. Small vessel or lacunar infarction encompasses approximately 25% of all ischemic strokes. 3. In a study by Ay and colleagues6 patients with acute stroke presenting with a classic lacunar syndrome were studied with diffusion-weighted MRI, and 16% (10 of 62) were found to have evidence of multiple emboli in addition to the suspected syndromecausing lacune, or “index lesion.” 4. Approximately 60% of all ischemic strokes are caused by cerebral embolism, only one third of which has known clinical sources. 5. Consideration should also be given for a hypercoagulable state, especially in a young patient or one with a known or occult carcinoma. 6. The initial evaluation of a patient with an acute stroke must also include the exclusion of other conditions that can mimic a stroke presentation. These include seizure (with subsequent Todd’s paresis or other postictal syndromes), cerebral neoplasm, encephalitis, complex migraine, and hypoglycemia. 7. Diffusion-weighted imaging gives useful information regarding tissue that is reversibly or irreversibly ischemic.22 Perfusion-weighted imaging, using a timed bolus of intravenous gadolinium contrast material, gives further insight to tissue destined for infarction, as well as tissue at risk for ischemic injury, the so-called ischemic penumbra. 8. An ECG should be performed shortly after admission to the emergency department, because stroke can commonly accompany an acute myocardial infarction. 9. The benefits of rt-PA were durable over one year, with an odds ratio for favorable outcome of 1.7 (95% CI, 1.2 to 2.3), and the rt-PA—treated patients were at least 30% more likely to have minimal or no disability than the placebo-treated patients. 10. With laser-induced microcavitation, a low energy laser is used as a transducer, creating an ultra short-lived microbubble that collapses and induces a series of waves that act to gently agitate the clot surface. 11. The Chinese Acute Stroke Trial (CAST)61 randomized 21,106 patients to receive either aspirin (160 mg/day) or placebo within 48 hours of symptom onset. Significant reductions were found in the aspirin-treated group for both recurrent ischemic strokes (1.6% vs. 2.1%) and for early mortality (3.3% vs. 3.9%). 12. The location of the hemorrhage and the age of the patient can give clues regarding the etiology of the ICH.
13. Primary intraventricular hemorrhage (PIVH) accounts for approximately 3% of all intracerebral hemorrhages. 14. Interestingly, some propose that the poor outcomes associated with ICH may be a “self-fulfilling prophesy,” in that patients may have care withdrawn when the outcome is presumed to be poor, and thus longterm clinical data about the true course of the illness and possible recovery may be lacking.89 15. Prophylactic anticonvulsant medications may be considered for patients with lobar or superficial subcortical bleeding, especially if they have had a seizure. Phenytoin remains the preferred agent. 16. Another potentially catastrophic secondary complication of ICH, especially in the right hemisphere or insula region, is the higher propensity for causing abnormal cardiac electrical activity, and “cerebrogenic sudden death.” 17. Patients with ICH and immobility are at high risk for developing deep venous thromboses (DVT) and subsequent pulmonary embolism. 18. In general patients with a GCS of 4 or less have a uniformly poor outcome (either death or severely disabled), regardless of whether or not surgery is performed, and thus these patients with ICH should be treated medically. 19. Features associated with a poor outcome in CVT include rapid onset of symptoms, early low GCS, focal neurologic signs, seizures, and a concomitant infection. 20. The incidence of GBS is approximately 1 to 2 per 100,000, with speculation that the rising incidence may simply be a reflection of improvements in diagnosis. 21. EMG/NCS tend to be the most useful tests in the evaluation of GBS. Characteristic findings include motor nerve conduction block, prolonged distal conduction, and slowing of nerve conduction. An important early finding is prolongation, dispersion or absence of F waves, suggesting root demyelination. 22. Hypotension in patients with GBS can be caused by sepsis, pulmonary embolus, venous pooling, or severe electrolyte disturbances, but wide fluctuations of blood pressure over minutes is suggestive of dysautonomia. The presumed cause is impaired baroreceptor buffering. 23. A frequently reported cause of sudden death in patients with GBS who experience dysautonomia is complete heart block. This is emergently treated with a temporary pacemaker. Few patients ultimately require permanent pacing.179 Untreated vagal activity can progress to profound bradycardia and asystole.
Continued
W9418-13.qxd 11/3/03 2:51 PM Page 430
430
Section I
Introduction
24. Studies concluded that plasma exchange is the “first and only treatment that has been proven to be superior to supportive treatment alone in Guillain-Barré syndrome.” 25. Myasthenia gravis (MG) is an autoimmune disease characterized by an autoantibody reaction at the antigen epitopes of the acetylcholine receptor, leading to destruction and simplification of the junctional fold, with subsequent widening of the synaptic cleft. 26. Generalized weakness develops in approximately 85% of patients with MG, often affecting the diaphragm. 27. The edrophonium test is helpful in establishing the diagnosis of MG, but should only be performed in a well-monitored setting, such as the emergency department or ICU.
References 1. Kistler JP, Furie KL: Carotid endarterectomy revisited. N Engl J Med 2000;342:1743–1745. 2. Sacco RL, Ellenberg JH, Mohr JP, et al: Infarcts of undetermined cause: The NINDS stroke data bank. Ann Neurol 1989;25:382–390. 3. Fisher CM: Occlusion of the carotid arteries: Further experiences. Arch Neurol Psychiatry 1954;72:187–204. 4. Fisher CM, Gore I, Okabe M, et al: Atherosclerosis of the carotid and vertebral arteries—extracranial and intracranial. J Neuropathol Exp Neurol 1965;24:455–476. 5. Fisher CM: The arterial lesions underlying lacunes. Acta Neuropathol 1969;12:1–15. 6. Ay H, Oliveira–Filho J, Buonanno FS, et al: Diffusion–weighted imaging identifies a subset of lacunar infarction associated with embolic source. Stroke 1999;30:2644–2650. 7. Mohr JP, Thompson JL, Lazar RM, et al: A comparison of warfarin and aspirin for the prevention of recurrent ischemic stroke. N Engl J Med 2001;345:1444–1451. 8. Bogousslavsky J, Van Melle G, Regli F: The Lausanne Stroke Registry: Analysis of 1000 consecutive patients with first stroke. Stroke 1988;19:1083–1092. 9. Beghi E, Bogliun G, Cavaletti G, et al: Hemorrhagic infarction: Risk factors, clinical and tomographic features, and outcome: A case-control study. Acta Neurol Scand 1989;80:226–231. 10. Mohr JP, Caplan LR, Melski JW, et al: The Harvard cooperative stroke registry: A prospective registry of cases hospitalized with stroke. Neurology 1978;28:754–762. 11. Pruitt AA, Rubin RH, Karchmer AW, et al: Neurologic complications of bacterial endocarditis. Medicine 1978;57:329–343. 12. Roeltgen DP, Weimer GR, Patterson LF: Delayed neurologic complications of left atrial myxoma. Neurology 1981;31:8–13. 13. Bogousslavsky J, Pierre P: Ischemic stroke in patients under age 45. Neurol Clin 1992;10:113–123. 14. Schievink WI: Spontaneous dissection of the carotid and vertebral arteries. N Engl J Med 2001;344:898–906. 15. Moore PM: Vasculitis of the central nervous system. Semin Neurol 1994;14:307–312. 16. Seligsohn U, Lubetzky A: Genetic susceptibility to venous thrombosis. N Engl J Med 2001;344:1222–1231. 17. Lansberg MG, Albers GW, Beaulieu C, et al: Comparison of diffusion–weighted MRI and CT in acute stroke. Neurology 2000;54: 1557–1561.
28. Lambert-Eaton syndrome is distinguished from MG on the basis of the clinical presentation, in which there are commonly seen autonomic and sensory symptoms, as well as with EMG/NCS, in which there is an improvement in response to repetitive stimulation. 29. Critical illness myopathy is often difficult to distinguish from CIP, and there is commonly an overlap of the two. In fact, some have proposed the term critical illness polyneuromyopathy, or CIPNM. 30. Deconditioning syndrome occurs in immobile, poorly nourished patients, who reach a highly catabolic state.
18. Hacke W, Warach S: Diffusion weighted MRI as an evolving standard of care in acute stroke. Neurology 2000;54:1548–1549. 19. Albers GW, Lansberg MG, Norbash AM, et al: Yield of diffusion weighted MRI for detection of potentially relevant findings in stroke patients. Neurology 2000;54:1562–1567. 20. Hasegawa Y, Fisher M, Latour LL, et al: MRI diffusion mapping of reversible and irreversible ischemic injury in focal brain ischemia. Neurology 1994;44:1484–1490. 21. Wildermuth S, Knauth M, Brandt T, et al: Role of CT angiography in patient selection for thrombolytic therapy in acute hemispheric stroke. Stroke 1998;29:935–938. 22. Schaefer PW: Applications of DWI in clinical neurology. J Neurol Sci 2001;186:S25–S35. 23. Calamante F, Gadian DG, Connelly A: Quantification of perfusion using bolus tracking magnetic resonance imaging in stroke: Assumptions, limitations and potential implications for clinical use. Stroke 2002;33:1146–1151. 24. Okumura A, Araki Y, Nishimura Y, et al: The clinical utility of contrast–enhanced 3D MR angiography for cerebrovascular disease. Neurol Res 2001;23:767–771. 25. Alexandrov AV, Demchuk AM, Wein TH, et al: Yield of transcranial doppler in acute cerebral ischemia. Stroke 1999;30:1604–1609. 26. Hankey GJ, Warlow CP, Sellar RJ: Cerebral angiographic risk in mild cerebrovascular disease. Stroke 1990;21:209–222. 27. Johannessen KA, Nordrehaug JE, von der Lippe G, et al: Risk factors for embolization in patients with left ventricular thrombi and acute myocardial infarction. Br Heart J 1988;60:104–110. 28. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group: Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333:1581–1587. 29. Scandinavian Stroke Study Group: Multi-center trial of hemodilution in ischemic stroke. Stroke 1985;16:885–890. 30. Rijkin DC, Hoylaerts M, Collen D: Fibrinolytic properties of one-chain and two-chain human extrinsic (tissue-type) plasminogen activator. J Biol Chem 1982;257:2920–2925. 31. Garabedian HD, Gold HK, Leinbach RC, et al: Comparative properties of two clinical preparations of recombinant human tissue-type plasminogen activator in patients with acute myocardial infarction. J Am Coll Cardiol 1987;9:599–607. 32. Wagner OF, de Vries C, Hohmann C, et al: Interaction between plasminogen activator inhibitor type 1 (PAI-1) bound to fibrin and either tissue-type plasminogen activator (t-PA) or urokinase-type plasminogen activator (u-PA). Binding of t-PA/PAI-1 complexes to fibrin
W9418-13.qxd 11/3/03 2:51 PM Page 431
Chapter 13
33.
34.
35.
36. 37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
mediated by both the finger and the kringle-2 domain of t-PA. J Clin Invest 1989;84:647–655. Potter van Loon BK, Rijken DC, Brommer EJ, et al: The amount of plasminogen, tissue-type plasminogen activator and plasminogen activator inhibitor type 1 in human thrombi and the relation to ex-vivo lysibility. Thromb Haemost 1992;67:101–105. Levy DE, Brott TG, Haley EC Jr, et al: Factors related to intracranial hematoma formation in patients receiving tissue-type plasminogen activator for acute ischemic stroke. Stroke 1994;25:291–297. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333:1581–1587. Generalized efficacy of t-PA for acute stroke: Subgroup analysis of the NINDS t-PA Stroke Study Group. Stroke 1997;28:2119–2125. The NINDS t-PA Stroke Study Group. Intracerebral hemorrhage after intravenous t-PA therapy for ischemic stroke. Stroke 1997;28: 2109–2118. Hacke W, Kaste M, Fieschi C, et al: Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke: The European Cooperative Stroke Study (ECASS). JAMA 1995;274:1017–1025. Hacke W, Kaste M, Fieschi C, et al: Randomized double–blind placebo controlled trial of thrombolytic therapy with intravenous alteplase in acute ischemic stroke (ECASS II): Second European–Australian Acute Stroke Study Investigators. Lancet 1998;352:1245–1251. Clark WM, Wissman S, Albers GW, et al: Recombinant tissue-type plasminogen activator (Alteplase) for ischemic stroke 3–5 hours after symptom onset. JAMA 1999;282:2019–2026. Clark WM, Albers GW, Madden KP, et al: The rt-PA (Alteplase) 0-to-6-hour acute stroke trial, part A (A0276g): Results of a double-blind, placebo-controlled multicenter study. Thrombolytic therapy in a acute stroke study investigators. Stroke 2000;31:811– 816. Kwiatowski TG, Libman RB, Frankel M, et al: Effects of tissue plasminogen activator for acute ischemic stroke at one year. N Engl J Med 1999;340:1781–1787. The European Stroke Initiative: Recommendations for stroke management. Cerebrovasc Dis 2000;10[Suppl 3]:1–34. Also, Ringleb PA, Schellinger PD, Schranz MD, et al: Thrombolytic therapy within 3 to 6 hours after onset of ischemic stroke: Useful or harmful? Stroke 2002;33:1437–1441. del Zoppo GJ, Higashida RT, Furlan AJ, et al: PROACT: A phase II randomized trial of recombinant pro-urokinase by direct arterial delivery in acute middle cerebral artery stroke. Stroke 1998;29:4–11. Furlan A, Higashida R, Wechler L, et al: Intra–arterial prourokinase for acute ischemic stroke. The PROACT II study: A randomized controlled trial. Prolyse in Acute Cerebral Thromboembolism. JAMA 1999;282:2003–2001. Schellinger PD, Feibach JB, Mohr A, et al: Thrombolytic therapy for ischemic stroke-A review. Part II—Intra-arterial thrombolysis, vertebrobasilar stroke, phase IV trials, and stroke imaging. Crit Care Med 2001;29:1819–1825. McDonald CT, O’Donnell J, Bemporad J, et al: The clinical utility of intravenous integrilin combined with tissue plasminogen activator in acute ischemic stroke: The MGH experience. Presented at the 27th Annual Stroke Conference, Ft. Lauderdale, FL, 2001. Eckert B, Koch C, Götz T, et al: Acute basilar artery occlusion treated with combined intravenous abciximab and intra-arterial tissue plasminogen activator. Stroke 2002;33:1424–1427. Lewandowski CA, Frankel M, Tomsick TA, et al: Combined intravenous and intra-arterial t-PA versus intra-arterial therapy of acute ischemic stroke: Emergency Management of Stroke (EMS) Bridging Trial. Stroke 1999;30:2598–2605. Ernst R, Pancioli A, Tomsick T, et al: Combined intravenous and intraarterial recombinant tissue plasminogen activator in acute ischemic stroke. Stroke 2000;31:2552–2557.
Acute Stroke and Other Neurologic Emergencies
431
51. Smalling RW: Molecular biology of plasminogen activators: What are the clinical implications of drug design? Am J Cardiol 1996;78:2–7. 52. Mitchel JF, Shwedick M, Alberghini TA, et al: Catheter–based local thrombolysis with urokinase: Comparative efficacy of intraluminal clot lysis with conventional urokinase infusion techniques in an in vivo porcine thrombus model. Cath Cardivasc Diag 1997;41:293–302. 53. Shangguan HQ, Gregory KW, Casperson LW, et al: Enhanced laser thrombolysis with photomechanical drug delivery: An in vitro study. Lasers Surg Med 1998;23:151–160. 54. Sacco RL, Foulkes MA, Mohr JP, et al: Determinants of early recurrence of cerebral infarction: The Stroke Data Bank. Stroke 1989 ;20:983–989. 55. Haley EC, Kassell NF, Torner JC: Failure of heparin to prevent progression in progressing ischemic infarction. Stroke 1988;19:10–14. 56. Sherman DG, Hart HG, Easton JD: The secondary prevention of stroke in patients with atrial fibrillation. Arch Neurol 1986;43:68–70. 57. Cerebral Embolism Study Group: Immediate anticoagulation of embolic stroke: Brain hemorrhage and management options. Stroke 1984;15:779–789. 58. International Stroke Trial Collaborative Group: The International Stroke Trial (IST): A randomized trial of aspirin, subcutaneous heparin, both, or neither among 19,435 patients with acute ischemic stroke. Lancet 1997;349:1569–1581. 59. Publications Committee for the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) Investigators: Low molecular weight heparinoid, ORG 10172 (danaparoid), and outcome after acute ischemic stroke: A randomized controlled trial. JAMA 1998;279:1265–1272. 60. Berge E, Abdelnoor M, Nakstad PH, et al: Low molecular-weight heparin versus aspirin in patients with acute ischemic stroke and atrial fibrillation: A double-blind randomised study (HAEST Study Group). Lancet 2000;355:1205–1210. 61. Chinese Acute Stroke Trail (CAST) Collaborative Group: CAST: A randomized placebo–controlled trial of early aspirin use in 20,000 patients with acute ischemic stroke. Lancet 1997;349: 1641–1649. 62. Greenberg J, Skubick D, Shenken H: Acute hydrocephalus in cerebellar infarct and hemorrhage. Neurology 1979;9:409. 63. Cuneo R, Caronna J, Pitts L, et al: Upward transtentorial herniation. Arch Neurol 1989;36:618. 64. Schwab S, Steiner T, Aschoff A, et al: Early hemicraniectomy in patients with complete middle cerebral artery infarction. Stroke 1998;29:1888–1893. 65. Kernan WN, Viscoli Catherine M, Brass LM: Phenylpropanolamine and the risk of hemorrhagic stroke. N Engl J Med 2000;343: 1826–1832. 66. Monforte R, Estruch R, Graus F, et al: High ethanol consumption as a risk factor for intracerebral hemorrhage in young and middle–aged people. Stroke 1990;21:1529–1532. 67. Sacco RL, Elkind M, Boden–Albala B, et al: The protective effect of moderate alcohol consumption on ischemic stroke. JAMA 1999;281: 53–60. 68. Castel JP, Kissel P: Spontaneous intracerebral and infratentorial hemorrhage. In Youmans JR (ed): Neurological Surgery. Philadelphia, WB Saunders, 1990, pp 1890–1917. 69. Caplan LR: Intracerebral hemorrhage revisited. Neurology 1988;38:624–627. 70. Vinters HV: Cerebral amyloid angiopathy: A critical review. Stroke 1987;18:311–324. 71. Kalyan-Raman UP, Kalyan Raman K: Cerebral amyloid angiopathy causing intracranial hemorrhage. Ann Neurol 1984;16:321–329. 72. Knudsen KA, Rosand J, Karluk D, Greenberg SM: Clinical diagnosis of cerebral amyloid angiopathy: Validation of the Boston criteria. Neurology 2001;56:537–539. 73. Tsushima Y, Tanizaki Y, Aoki J, et al: MR detection of microhemorrhages in neurologically healthy adults. Neuroradiology 2002;44: 31–36.
W9418-13.qxd 11/3/03 2:51 PM Page 432
432
Section I
Introduction
74. Caplan LR: Primary intraventricular hemorrhage. In Kase CS, Caplan LR (eds): Intracerebral Hemorrhage. London, ButterworthHeinemann, 1994, pp 383–401. 75. Marti-Fabregas J, Piles S, Guardia E, et al: Spontaneous primary intraventricular hemorrhage: Clinical data, etiology and outcome. J Neurol 1999;246:287–291. 76. Agelopoulos M, Gupta SR, Azar Kia B: Primary intraventricular hemorrhage in adults: Clinical features, risk factors, and outcome. Surg Neurol 1995;44:433–437. 77. Naff NJ, Carhuapoma JR, Williams MA, et al: Treatment of intraventricular hemorrhage with urokinase. Stroke. 2000;31:841–847. 78. Broderick JP, Brott TG, Duldner JE, et al: Volume of intracerebral hemorrhage. A powerful and easy-to-use predictor of 30-day mortality. Stroke 1993;24:987–993. 79. Tuhrim S, Dambrosia JM, Price TR, et al: Prediction of intracerebral hemorrhage survival. Ann Neurol 1988;24:258–263. 80. Tuhrim S, Horowitz DR, Sacher M, et al: Validation and comparison of models predicting survival following intracerebral hemorrhage. Crit Care Med 1995;23:950–954. 81. Tuhrim S, Horowitz DR, Sacher M, et al: Volume of ventricular blood is an important determinant in outcome in supratentorial intracerebral hemorrhage. Crit Care Med 1999;27:617–621. 82. Diringer MN, Edwards DF, Zazulia AR: Hydrocephalus: A previously unrecognized predictor of poor outcome from supratentorial intracerebral hemorrhage. Stroke 1998;29:1352–1357. 83. Adams RE, Diringer MN: Response to external ventricular drainage in spontaneous intracerebral hemorrhage with hydrocephalus. Neurology 1998;50:519–523. 84. Naff NJ, Carhuapoma JR, Williams MA, et al: Treatment of intraventricular hemorrhage with urokinase: Effects on 30-day survival. Stroke 2000;31:841–847. 85. Fogelholm R, Avikainen S, Murros K: Prognostic value and determinants of first-day mean arterial pressure in spontaneous supratentorial intracerebral hemorrhage. Stroke 1997;28:1396–1400. 86. Lee E-J, Hung Y-C: Marked anemic hypoxia deteriorates cerebral hemodynamics and brain metabolism during massive intracerebral hemorrhage. J Neurol Sci 2001;190:3–10. 87. Becker KJ, Baxter AB, Bybee HM, et al: Extravasation of radiographic contrast is an independent predictor of death in primary intracerebral hemorrhage. Stroke 1999;30:2025–2032. 88. Flemming KD, Wijdicks EFM, St. Louis EK, Li H: Predicting deterioration in patients with lobar hemorrhage. J Neurol Neurosurg Psychiatry 1999;66:600–605. 89. Becker KJ, Baxter AB, Cohen WA, et al: Withdrawal of support in intracerebral hemorrhage may lead to self-fulfilling prophesies. Neurology 2001;56:766–772. 90. Greenberg SM, Finklestein SP, Schaefer PW. Petechial hemorrhages accompanying lobar hemorrhage: Detection by gradient-echo MRI. Neurology. 1996;46:1751–1754. 91. Halpin SF, Britton JA, Byrne JV, et al: Prospective evaluation of cerebral angiography and computed tomography in cerebral hematoma. J Neurol Neurosurg Psychiatry 1994;57:1180–1186. 92. Broderick JP, Adams HP, Barsan W, et al: Guidelines for the management of spontaneous intracerebral hemorrhage: A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke 1999;30: 905–915. 93. Broderick JP, Brott T, Tomsick T, et al: Ultra-early evaluation of intracerebral hemorrhage. J Neurosurg 1990;72:195–199. 94. Kazui S, Naritomi H, Yamamoto H, et al: Enlargement of spontaneous intracerebral hemorrhage. Incidence and time course. Stroke 1996; 27:1783–1787. 95. Qureshi AI, Bliwise DL, Bliwise NG, et al: Rate of 24-hour blood pressure decline and mortality after spontaneous intracerebral hemorrhage. A retrospective analysis with a random effects regression model. Crit Care Med 1999;27:480–485.
96. Qureshi AI, Wilson DS, Hanley DF, et al: Pharmacologic reduction of mean arterial pressure does not adversely affect regional cerebral blood flow and intracranial pressure in experimental intracerebral hemorrhage. Crit Care Med 1999;27:965–971. 97. Powers WJ, Zazulia AR, Videen TO, et al: Autoregulation of cerebral blood flow surrounding acute (6 to 22 hours) intracerebral hemorrhage. Neurology 2001;57:18–24. 98. Morgenstern LB, Yonas H: Lowering blood pressure in acute intracerebral hemorrhage. Safe, but will it help? Neurology 2001;57:5–6. 99. Rosand, J, Eskey C, Chang Y, et al: Dynamic single-section CT demonstrates reduced cerebral blood flow in acute intracerebral hemorrhage. Cerebrovasc Dis 2002;14(3–4):214–220. 100. Cheung RTF, Hachinski V: The insula and cerebrogenic sudden death. Arch Neurol 2000;57:1685–1688. 101. Fernandes HM, Siddique S, Banister K: Continuous monitoring of ICP and CPP following ICH and its relationship to clinical, radiological and surgical parameters. Acta Neurochir Suppl 2000;76:463–466. 102. Poungvarin N, Bhoopat W, Viriyavejakul A, et al: Effects of dexamethasone in primary supratentorial intracerebral hemorrhage. N Engl J Med 1987;316:1229–1233. 103. Boeer A, Voth E, Henze T, et al: Early heparin therapy in patients with spontaneous intracerebral hemorrhage. J Neurol Neurosurg Psychiatry 1991;54:466–467. 104. Morgenstern LB, Frakowski RF, Shedden P, et al: Surgical treatment for intracerebral hemorrhage (STICH): A single-center, randomized clinical trial. Neurology 1998;51:1359–1363. 105. Auer L, Deinsberger W, Niederkorn K, et al: Endoscopic surgery versus best medical treatment for spontaneous intracerebral hematoma: A randomized study. J Neurosurg 1989;70:530–535. 106. Backlund EO, von Holst H: Controlled subtotal evacuation of intracerebral haematomas by stereotactic technique. Surg Neurol 1978;9: 99–101. 107. Nguyen JP, Decq P, Brugieres P, et al: A technique for stereotactic aspiration of deep intracerebral hematomas under computed tomographic control using a new device. Neurosurgery 1992;31: 330–335. 108. Kaufman H: Stereotactic aspiration with fibrinolytic and mechanical assistance. In Kaufman H, ed: Intracerebral Hematomas. New York: Raven Press, 1992, pp 181–185. 109. Niizuma H, Suzuki J: Stereotactic aspiration of putaminal hemorrhage using a double track aspiration technique. Neurosurgery 1988;22:432–436. 110. Martinelli I, Landi G, Merati G, et al: Factor V gene mutation is a risk factor for cerebral venous thrombosis. Thromb Haemost 1996;75: 393–394. 111. Zuber M, Toulon P, Marnet L, et al: Factor V leiden mutation in cerebral venous thrombosis. Stroke 1996;27:1721–1723. 112. Lùldemann P, Nabavi DG, Junker R, et al: Factor V leiden mutation is a risk factor for cerebral venous thrombosis—A case-control study of 55 patients. Stroke 1998;29:2507–2510. 113. Biousse V, Conard J, Brouzes C, et al: Frequency of the 20210 GÆA mutation in the 3¢–untranslated region of the prothrombin gene in 35 cases of cerebral venous thrombosis. Stroke 1998;29:1398–1400. 114. Galan HL, McDowell AB, Johnson PR, et al: Puerperal cerebral venous thrombosis associated with decreased free protein S–A case report. J Reprod Med 1995;40:859–856. 115. Wechsler B, Généreau T, Biousse V, et al: Pregnancy complicated by cerebral venous thrombosis in Behçet’s disease. Am J Obstet Gynecol 1995;173:1627–1629. 116. Debruijn SFTM, Stam J, Koopman MMW, et al: Cerebral venous thrombosis study. Case-control study of risk of cerebral sinus thrombosis in oral contraceptive users who are carriers of hereditary prothrombotic conditions. BMJ 1998;316:589–592. 117. Cantú C, Barinagarrementeria F: Cerebral venous thrombosis associated with pregnancy and puerperium. Review of 67 cases. Stroke 1993;24:1880.
W9418-13.qxd 11/3/03 2:51 PM Page 433
Chapter 13 118. Stolz E, Gerriets T, Bödecker RH: Intracranial venous hemodynamics is a factor related to a favorable outcome in cerebral venous thrombosis. Stroke 2002;33:1645–1650. 119. Ameri A, Bousser MG: Cerebral venous thrombosis. Neurol Clin 1992;10:87–111. 120. Daif A, Awada A, Al-Rajeh S, et al: Cerebral venous thrombosis in adults: A study of 40 cases from Saudi Arabia. Stroke 1995;26: 1193–1195. 121. Tsai FY, Wang AM, Matovich VB, et al: MR staging of acute dural sinus thrombosis: Correlation with venous pressure measurements and implications for treatment and prognosis. Am J Neuroradiol 1995; 16:1021–1029. 122. Smith AG, Cornblath WT, Deveikis JP: Local thrombolytic therapy in deep cerebral venous thrombosis. Neurology 1997;48:1613–1619. 123. Buonanno FS, Moody DM, Ball MR, et al: Computer cranial tomographic findings in cerebro sino-venous occlusion. J Computer Assisted Tomogr 1978;2:281–290. 124. Patronas NJ, Duda EE, Mirfakhraee M, et al: Superior sagittal sinus thrombosis diagnosed by computed tomography. Surg Neurol 1981;15:11–14. 125. Casey SO, Alberico RA, Patel M, et al: Cerebral CT venography. Radiology 1996;198:163–170. 126. Lafitte F, Boukobza M, Guichard JP, et al: MRI and MRA for diagnosis and follow up of cerebral venous thrombosis (CVT). Clin Radiol 1997;52:672–679. 127. Selim M, Fink J, Linfante I: Diagnosis of cerebral venous thrombosis with echo-planar T2*-weighted magnetic resonance imaging. Arch Neurol 2002;59:1021–1026. 128. Tsai FY, Wang AM, Matovich V, et al: MR staging of acute dural sinus thrombosis: Correlation with venous pressure measurements and implications for treatment and prognosis. Am J Neuroradiol 1995;16: 1021–1029. 129. Ducreux D, Oppenheim C, Vandamme X, et al: Diffusion-weighted imaging patterns of brain damage associated with cerebral venous thrombosis. Am J Neuroradiol 2001;22:261–268. 130. Einhäupl KM, Villringer A, Meister W, et al: Heparin treatment in sinus venous thrombosis. Lancet 1991;338:597–600. 131. Brucker AB, Vollert-Rogenhofer H, Wagner M, et al: Heparin treatment in acute cerebral sinus venous thrombosis: A retrospective clinical and MR analysis of 42 cases. Cerebrovasc Dis 1998;8: 25–30. 132. Enevoldson TP, Ross Russell RW: Heparin treatment in sinus venous thrombosis. Lancet 1991;338:1153–1154. Also, Stam J, Lensing AWA, Vermeulen M, et al: Heparin treatment for cerebral venous and sinus thrombosis. Lancet 1991;338:1154. 133. De Bruijn SFTM, Stam J: Randomized placebo–controlled trial of anticoagulant treatment with low-molecular-weight heparin for cerebral sinus thrombosis. Stroke 1999;30:484–488. 134. Bousser MG: Cerebral venous thrombosis: Nothing, heparin, or local thrombolysis? Stroke 1999;30:481–483. 135. Fink JN, McAuley DL: Safety of anticoagulation for cerebral venous thrombosis associated with intracerebral hematoma. Neurology 2001;57:1138–1139. 136. Buonanno FS: Cerebral sinovenous thrombosis. Curr Treat Opt Cardiovasc Med 2001;3:417–427. 137. Vines FS, Davis DO: Clinical–radiological correlation in cerebral venous occlusive disease. Radiology 1971;98:9–22. 138. Di Rocco C, Iannelli A, Leone G, et al: Heparin-urokinase treatment in aseptic dural sinus thrombosis. Arch Neurol 1981;38:431–435. 139. Gettelfinger DM, Kokmen E: Superior sagittal sinus thrombosis. Arch Neurol 1977;34:2–6. 140. Scott JA, Pascuzzi RM, Hall PV, et al: Treatment of dural sinus thrombosis with local urokinase infusion. J Neurosurg 1988;68:284–287. 141. Yamini B, Loch Macdonald R, Rosenblum J: Treatment of deep cerebral venous thrombosis by local infusion of tissue plasminogen activator. Surg Neurol 2001; 55:340–346.
Acute Stroke and Other Neurologic Emergencies
433
142. Opatowsky MJ, Morris PP, Regan JD, et al: Rapid thrombectomy of superior sagittal sinus and transverse sinus thrombosis with a rheolytic catheter device. Am J Neuroradiol 1999;20: 414–417. 143. Dowd CF, Malek AM, Phatouros CC, et al: Application of a rheolytic thrombectomy device in the treatment of dural sinus thrombosis: A new technique. Am J Neuroradiol 1999;20:568–570. 144. Chow K, Pierre Gobin Y, Saver J, et al: Endovascular treatment of dural sinus thrombosis with rheolytic thrombectomy and intra-arterial thrombolysis. Stroke 2002;31:1420–1425. 145. Van der Meché FGA, Van Doorn PA, Meulstee J, et al: Diagnostic and classification criteria for the Guillain-Barré syndrome. Eur Neurol 2001;45:133–139. 146. Hughes RAC, Rees JH: Clinical and epidemiological features of Guillain-Barré syndrome. J Infect Dis 1997;176:S92–S98. 147. Winer JB, Hughes RAC, Anderson MJ, et al: A prospective study of acute idiopathic neuropathy. II: Antecedent events. J Neurol Neurosurg Psychiatry. 1988;51:613–618. 148. Raschetti R, Maggini M, Popoli P, et al: Gangliosides and GuillainBarré syndrome. J Clin Epidemiol 1995;48:1399–1405. 149. Hughes RAC, Gray IA, Gregson NA, et al: Immune responses to myelin antigens in Guillain-Barré syndrome. J Neuroimmunol 1984;6:303–312. 150. Wilmshurst JM, Thomas NH, Robinson RO, et al: Lower limb and back pain in Guillain-Barré syndrome and associated contrast enhancement in MRI of the cauda equina. Acta Pediatr 2001;9: 691–694. 151. Fisher M: Syndrome of ophthalmoplegia, ataxia and areflexia. N Engl J Med 1956;255:57–65. 152. Vargas F, Hilbert G, Gruson D, et al: Fulminant Guillain-Barré syndrome mimicking cerebral death: Case report and review of the literature. Intens Care Med 2000;26:623–627. 153. Byun WM, Park WK, Park BH, et al: Guillain-Barré syndrome: MR imaging findings of the spine in eight patients. Radiology 1998;208: 137–141. 154. Garcia–Rivera CA, Zhou D, Allahyari P, et al: Miller Fisher syndrome: MRI findings. Neurology 2001;57:1755. 155. Hawke SH, Hallinan J, McLeod JG: Cranial magnetic resonance imaging in chronic demyelinating polyneuropathy. J Neurol Neurosurg Psychiatry 1990;53:794–796. 156. Hughes RAC: Guillain-Barré syndrome. London, Springer-Verlag, 1990. 157. Thornton CA, Latif AS, Emmanuel JC: Guillain-Barré syndrome associated with human immunodeficiency virus in Zimbabwe. Neurology 1991;41:812–815. 158. Shapiro EE: Guillain-Barré syndrome in a child with serologic evidence of Borrelia burgdorferi infection. Pediatr Infect Dis J 1998;17: 264–265. 159. Kimura J: Proximal versus distal slowing of motor nerve conduction velocity in the Guillain-Barré syndrome. Ann Neurol 1978;3: 344–350. 160. Ropper AH, Wijdicks EFM, Shahani BT: Electrodiagnostic abnormalities in 113 consecutive patients with Guillain-Barré syndrome. Arch Neurol 1990;47:881–887. 161. Zifko U, Chen R, Remtulla H, et al: Respiratory electrophysiological studies in Guillain-Barré syndrome. J Neurol Neurosurg Psychiatry 1996;60:191–194. 162. Grand’Maison F: Methods of testing neuromuscular transmission in the intensive care unit. Can J Neurol Sci 1998;25:S36–S39. 163. Chiba A, Kusunoki S, Obata H, et al: Serum anti-GQ1b IgG antibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barré syndrome: clinical and immunohistochemical studies. Neurology 1993;43:1911–1917. 164. Willison HJ, Veitch J, Paterson G, et al: Miller Fisher syndrome is associated with serum antibodies to GQ1b ganglioside. J Neurol Neurosurg Psychiatry 1993;56:204–206.
W9418-13.qxd 11/3/03 2:51 PM Page 434
434
Section I
Introduction
165. Ho TW, Willison HJ, Nachamkin I, et al: Anti-GD1a antibody is associated with axonal but not demyelinating forms of Guillain-Barré syndrome. Ann Neurol 1999;45:168–173. 166. Rees JH, Thompson RD, Smeeton NC, et al: Epidemiological study of Guillain-Barré syndrome in south east England [see comments]. J Neurol Neurosurg Psychiatry 1998;64:74–77. 167. Lawn ND, Wijdicks EFM: Fatal Guillain-Barré syndrome. Neurology 1999;52:635–638. 168. Bassile CC: Guillain-Barré syndrome and exercise guidelines. Neurol Rep 1996;20:31–36. 169. Ropper AH: Intensive care of Guillain-Barré syndrome. Can J Neurol Sci 1994;21:S23–S27. 170. Fulgham JR, Wijdicks EFM: Guillain-Barré syndrome. Crit Care Clin 1997;13:1–15. 171. Lawn ND, Fletcher DD, Henderson RD, et al: Anticipating mechanical ventilation in Guillain-Barré syndrome. Arch Neurol 2001;58: 893–898. 172. Singh NK, Jaiswal AK, Misra S, et al: Assessment of autonomic dysfunction in Guillain-Barré syndrome and its prognostic implications. Acta Neurol Scand 1987;75:101–105. 173. Thomashefsky AJ, Horwitz SJ, Feingold MH: Acute autonomic neuropathy. Neurology 1972;22:251–255. 174. Ng KK, Howard RS, Fish DR, et al: Management and outcome of severe Guillain-Barré syndrome. Q J Med 1995;88:243–250. 175. Lichtenfeld P: Autonomic dysfunction in the Guillain-Barré syndrome. Am J Med 1971;50:772–780. 176. McQuillan JJ, Bullock RE: Extreme labile blood pressure in GuillainBarré syndrome. Lancet 1988;2:172–173. 177. Mitchell PL, Meilman E: The mechanism of hypertension in GuillainBarré syndrome. Am J Med 1967;42:986–995. 178. Calleja MA: Autonomic dysfunction and Guillain-Barré syndrome. The use of esmolol in its management. Anesthesia 1990;45:726–727. 179. Narayan D, Huang MT, Mathew PK: Bradycardia and asystole requiring permanent pacemaker in Guillain-Barré syndrome. Am Heart J 1984;108:426–428. 180. Emmons PR, Blume WT, DuShaw JW: Cardiac monitoring and demand pacemaker in Guillain-Barré syndrome. Arch Neurol 1975;32:59–61. 181. Favre H, Foex P, Guggisberg M: Use of demand pacemaker in a case of Guillain-Barré syndrome. Lancet 1970;16:1062–1063. 182. Greenland P, Griggs RC: Arrhythmic complications in the GuillainBarré syndrome. Arch Intern Med 1980;140:1053–1055. 183. Palferman TG, Wright I, Doyle DV, et al: Electrocardiographic abnormalities and autonomic dysfunction in Guillain-Barré syndrome. BMJ 1982;284:1231–1232. 184. Winer JB, Hughes RA: Identification of patients at risk for arrhythmia in the Guillain-Barré syndrome. Q J Med 1988;68:735–739. 185. Raphael JC, Chevret S, Hughes RA, et al: Plasma exchange for Guillain-Barré syndrome. Cochrane Database Syst Rev 2001;(2): CD001798. 186. Hughes RA, Raphael JC, Swan AV, et al: Intravenous immunoglobulin for Guillain-Barré syndrome. Cochrane Database Syst Rev 2001; (2):CD002063. 187. Hughes RA, van Der Meche FG: Corticosteroids for treating GuillainBarré syndrome. Cochrane Database Syst Rev 2001;(3):CD001446. 188. Irani DN, Cornblath DR, Chaudhry V, et al: Relapse in Guillain-Barré syndrome after treatment with human immune globulin. Neurology 1993;43:872–875. 189. Castro LHM, Roper AH: Human immune globulin infusion in Guillain-Barré syndrome: worsening during and after treatment. Neurology 1993;43:1034–1036. 190. Misbah SA, Chapel HM: Adverse effects of intravenous immunoglobulin. Drug Safety 1993;9:254–262. 191. Kleyweg RP, van der Meche FGA: Treatment fluctuations in Guillain-Barré syndrome after high dose immunoglobulins or plasma exchange. J Neurol Neurosurg Psychiatry 1991;54:957–960.
192. Deiner H–C, Haupt WF, Kloss TM, et al: A preliminary, randomized multicenter study comparing intravenous immunoglobulin, plasma exchange, and immune adsorption in Guillain-Barré syndrome. Eur Neurol 2001;46:107–109. 193. Barohn RJ, Kissel JT, Warmolts R, et al: Chronic inflammatory demyelinating polyradiculoneuropathy: Cinical characteristics, course and recommendations for diagnostic criteria. Arch Neurol 1989;46:878–884. 194. Kurtzke JF, Kurland LT: The epidemiology of neurologic disease. In Joynt RJ (ed): Clinical Neurology. Rev. ed. Vol 4. Philadelphia, JB Lippincott, 1992, pp 80–88. 195. Grob D, Arsura EL, Brunner NG, et al: The course of myasthenia gravis and therapies affecting outcome. Ann NY Acad Sci 1987; 505:472–499. 196. Dushay KM, Zibrak JD, Jensen WA: Myasthenia gravis presenting as isolated respiratory failure. Chest. 1990;97:232–234. 197. Daroff RB: The office tensilon test for ocular myasthenia gravis. Arch Neurol 1986;43:843–844. 198. Keesey JC: Electrodiagnostic approach to defects of neuromuscular transmission. Muscle Nerve 1989;12:613–626. 199. Stalberg E, Trontelj J: Single Fibre Electromyography. Old Working, England, Mirvalle Press, 1979. Also, Oh SJ, Kim Se, Kuruoglu R, et al: Diagnostic sensitivity of the laboratory tests in myasthenia gravis. Muscle Nerve 1992;15:720–724. 200. Lindstrom JM, Seyboid ME, Lennon VA, et al: Antibody to acetylcholine receptor in myasthenia gravis: Prevalence, clinical correlates, and diagnostic value. Neurology 1976;26:1054–1059. 201. Linstrom JM: An assay for antibodies to human acetylcholine receptor in serum from patients with myasthenia gravis. Clin Immunol Immunopathol 1977;7:36–43. 202. Vincent A, Newsom-Davis J: Acetylcholine receptor antibody as a diagnostic test for myasthenia gravis: results in 153 validated cases and 2967 diagnostic assays. J Neurol Neurosurg Psychiatry 1985;48: 1246–1252. 203. Weissel M, Mayr N, Zeitholfer J: Clinical significant of autoimmune thyroid disease in myasthenia gravis. Exp Clin Endocrinol Diabetes 2000;108:63–65. 204. Batra P, Herrmann C Jr, Mulder D: Mediastinal imaging in myasthenia gravis: Correlation of chest radiography, CT, MR and surgical findings. AJR 1987;148:515–519. 205. Rowland LP: General discussion on therapy in myasthenia gravis. Ann NY Acad Sci 1987;505:607–609. 206. Drachman DB: Myasthenia gravis. N Engl J Med 1994;330:1797–1810. 207. Kuroda Y, Oda K, Neshige R, et al: Exacerbation of myasthenia gravis after removal of a thymoma having a membrane phenotype of suppressor T cells. Ann Neurol 1984;15:400–402. 208. O’Neill JH, Murray NMF, Newsom-Davis J: The Lambert–Eaton syndrome: A review of 50 cases. Brain 1988;111:557–596. 209. Lang B, Johnston I, Leys K, et al: Autoantibody specificities in Lambert-Eaton myasthenic syndrome. Ann NY Acad Sci 1993;681: 382–393. 210. Penn AS, Richman DP, Ruff RL, et al: Myasthenia gravis and related disorders: Experimental and clinical aspects. Ann NY Acad Sci 1993;681:425–514. 211. Moraes CT, DiMauro S, Zeviani M, et al: Mitochondrial DNA deletions in progressive external ophthalmoplegia and Kearns-Sayre syndrome. N Engl J Med 1989;320:1293–1299. 212. Moorthy G, Behrens MM, Drachman DB, et al: Ocular pseudomyasthenia or ocular myasthenia ‘plus’: A warning to clinicians. Neurology 1989;39:1150–1154. 213. Bucknall RC, Dixon A St J, Glick EN, et al: Myasthenia gravis associated with penicillamine treatment for rheumatoid arthritis. BMJ 1975;1:600–602. 214. Kuncl RW, Pestronk A, Drachman DB, Rechthand E: The pathophysiology of penicillamine-induced myasthenia gravis. Ann Neurol 1986;20:740–744.
W9418-13.qxd 11/3/03 2:51 PM Page 435
Chapter 13 215. Howard JF Jr: Adverse drug effects on neuromuscular transmission. Semin Neurol 1990;10:89–102. 216. Emeryk B, Rowinska K, Nowak-Michalska T: Do true remissions in myasthenia really exist? An electrophysiological study. J Neurol 1985;231:331–335. 217. Vincken W, Elleker M, Cosio M: Determinants of respiratory muscle weakness in stable chronic neuromuscular disorders. Am J Med 1987;82:53–64. 218. Rieder P, Louis M, Jolliet P, et al: The repeated measurement of vital capacity is a poor predictor of the need for mechanical ventilation in myasthenia gravis. Intensive Care Med 1995;21:663–668. 219. Collop NA: Management of respiratory problems in myasthenia gravis. Pulmon Crit Care Update 1992;7:2–6. 220. Swash M, Schwartz M: Neuromuscular disease: A practical approach to diagnosis and management. London, Springer-Verlag, 1988, p 233. 221. Howard JF: Treatment of myasthenia gravis with plasma exchange. Semin Neurol 1982;2:273–279. 222. Howard J: Intravenous immunoglobulin for the treatment of acquired myasthenia gravis. Neurology 1998;51:S30–S36. 223. Arsura EL, Bick A, Brunner NG, et al: High-dose intravenous immunoglobulin in the management of myasthenia gravis. Arch Intern Med 1986;146:1365–1368. 224. Berrouschot J, Baumann I, Kalischewski P, et al: Therapy of myasthenic crisis. Crit Care Med 1997;25:1228–1235. 225. Saltis LM, Martin BR, Traeger SM, et al: Continuous infusion of pyridostigmine in the management of myasthenic crisis. Crit Care Med 1993;21:938–940. 226. Matell G: Immunosuppressive drugs: Azothioprine in the treatment of myasthenia gravis. Ann NY Acad Sci 1987;505:588–594. 227. Tindall RSA, Phillips JT, Rollins JA, et al: A clinical therapeutic trial of cyclosporine in myasthenia gravis. Ann NY Acad Sci 1993;681: 539–551. 228. Younger DS, Raksadawan N: Medical therapies in myasthenia gravis. Chest Surg Clin North Am 2001;11:329–336. 229. Lindberg C, Andersen O, Larsson S, et al: Remission rate after thymectomy in myasthenia gravis when the bias of immunosuppressive therapy is eliminated. Acta Neurol Scand 1992;86:323–328. 230. Cooper JD, Al-Jilaihawa AN, Pearson FG, et al: An improved technique to facilitate transcervical thymectomy for myasthenia gravis. Ann Thorac Surg 1988;45:242–247. 231. Kirshner PA: Reoperation on the thymus. A critique. Chest Surg Clin North Am 2001;11:439–445. 232. d’Empaire G, Hoaglin DC, Perlo VP, et al: Effect of prethymectomy plasma exchange on postoperative respiratory function in myasthenia gravis. J Thorac Cardiovasc Surg 1985;89:592–596. 233. Mozaffar T: Critical illness myopathy. Muscle Nerve 2001;24:973–974. 234. Bolton CF, Young GB, Zochodne DW: The neurologic complications of sepsis. Ann Neurol 1993;33:94–100. 235. Bolton CF: Muscle weakness and difficulty in weaning from the ventilator in the critical care unit. Chest 1994;106:1–2. 236. Berek K, Margreiter J, Willeit J, et al: Polyneuropathies in critically ill patients: A prospective evaluation. Intens Care Med 1996;22: 849–855. 237. Witt NJ, Zochodne DW, Bolton CF, et al: Peripheral nerve function in sepsis and multiple organ failure. Chest 1991;99:176–184. 238. Hund E: Neurological complications of sepsis: Critical illness polyneuropathy and myopathy. J Neurol 2001;248:929–934. 239. Garnacho-Montero J, Madroazo-Osuna J, Garcia-Garmendia JL, et al: Critical illness polyneuropathy: risk factors and clinical consequences. A cohort study in septic patients. Intens Care Med 2001;27:1288–1296. 240. de Letter M-ACJ, Schmitz PIM, Visser LH, et al: Risk factors for the development of polyneuropathy and myopathy in critically ill patients. Crit Care Med 2001;29:2281–2286. 241. Coakley JH, Nagendran K, Ormerod IEC, et al: Prolonged neurogenic weakness in patients requiring mechanical ventilation for acute airflow limitation. Chest 1992;101:1413–1416.
Acute Stroke and Other Neurologic Emergencies
435
242. Bolton CF, Laverty DA, Brown JD, et al: Critically ill polyneuropathy: Electrophysiological studies and differentiation from Guillain-Barré syndrome. J Neurol Neurosurg Psychiatry 1986;49:563–573. 243. Witt NJ, Zochodne DW, Bolton CF, et al: Peripheral nerve function in sepsis and multiple organ failure. Chest 1991;99:176–184. 244. Zochodne DW, Bolton CF, Wells GA, et al: Critical illness polyneuropathy. A complication of sepsis and multiple organ failure. Brain 1987;110:819–842. 245. Hughes M, Grant IS, Biccard B, et al: Suxamethonium and critical illness polyneuropathy. Anaesth Intens Care 1999;27:636–638. 246. Zifko UA: Long-term outcome of critical illness polyneuropathy. Muscle Nerve 2000;9:S49–S52. 247. Leijten FS, Harinck-de Weerd JE, Poortvliet DC, et al: The role of polyneuropathy in motor convalescence after prolonged mechanical ventilation. JAMA 1995;274:1221–1225. 248. de Lemos JM, Carr RR, Shalnasky KF, et al: Paralysis in the critically ill: Intermittent bolus pancuronium compared with continuous infusion. Crit Care Med 1999;27:2648–2655. 249. de Letter MACJ, van Doorn PA, Savelkoul HFJ, et al: Critical illness polyneuropathy and myopathy (CIPNM): Evidence for local immune activation by cytokine expression in the muscle tissue. J Neuroimmunol 2000;106:206–213. 250. Trojaborg W, Weimer LH, Hays AP: Electrophysiological studies in critical illness associated weakness: Myopathy or neuropathy—a reappraisal. Clin Neurophys 2001;112:1586–1593. 251. Chad DA, Lacomis D: Critically ill patients with newly acquired weakness: The clinicopathological spectrum. Ann Neurol 1994;35:257–259. 252. Campellone JV, Lacomis D, Kramer DJ, et al: Acute myopathy after liver transplantation. Neurology 1998;50:46–53. 253. Lacomis D, Giuliani MJ, Van Cott A, et al: Acute myopathy of intensive care: Clinical, EMG, and pathological aspects. Ann Neurol 1996;40:645–654. 254. Faragher MW, Day BJ, Dennett X: Critical care myopathy: An electrophysiological and histological study. Muscle Nerve 1996;19:516–518. 255. Mozaffar T: Critical illness myopathy. Muscle Nerve 2001;24:973–974. 256. Tennilä A, Salmi T, Pettilä V, et al: Early signs of critical illness polyneuropathy in ICU patients with systemic inflammatory response syndrome or sepsis. Intens Care Med 2000;26:1360–1363. 257. Warren JD, Blumberg PC, Thompson PD: Rhabdomyolysis: A review. Muscle Nerve 2002;25:332–347. 258. Adams RD, Victor M, Mancall EL: Central pontine myelinolysis: A hitherto undescribed disease occurring in alcoholic and malnourished patients. Arch Neurol Psychiatry 1959;81:154. 259. Goebel HH, Zur PH: Central pontine myelinolysis: A clinical and pathological study of 10 cases. Brain 1972;95:495–504. 260. McKee Winkelman M, Banjer B: CPM in severely burned patients. Relationship to serum hyperosmolality. Neurology 1988;38:1211. 261. Karp BI, Laureno R: Pontine and extra-pontine myelinolysis: A neurological disorder following rapid correction of hyponatremia. Medicine 1993;72:359–373. 262. Riggs JE, Schochert SS: Osmotic stress, osmotic myelinolysis and oligodendrocyte topography. Arch Pathol Lab Med 1989;113: 1386–1388. 263. Wright DG, Laureno R, Victor M: Pontine and extrapontine myelinolysis. Brain 1979;102:361–385. 264. Gocht A, Colmant HJ: Central pontine and extrapontine myelinolysis: A report of 58 cases. Clin Neuropathol 1987;6:262–270. 265. Lampl C, Yazdi K: Central pontine myelinolysis. Eur Neurol 2002; 47:3–10. 266. Konno S, Nakagawa T, Hayashibe Y, et al: A case report of myelinolysis associate with serum hyperosmolality after open heart surgery. Kyobu Geka 1993;46:150–154. 267. Chemaly R, Halaby G, Mohasseb G, et al: Extrapontine myelinolysis: Treatment with TRH. Rev Neurol 1998;154:163–165. 268. Dickoff DJ, Raps M, Yahr MD: Striatal syndrome following hypernatremia and its rapid correction. Arch Neurol 1988;45:112–114.
W9418-13.qxd 11/3/03 2:51 PM Page 436
436
Section I
Introduction
269. Delay J, Pichot P, Lemperier MT, et al: Un neuroleptique majeur non phenothiazinique et non reserpinique, l’ haldol, dans le traitment des psychoses. Ann Med Physiol (Paris) 1960;118:145–152. 270. Gelenberg AJ, Bellinghausen B, Wojcik JD, et al: A prospective survey of neuroleptic malignant syndrome in a short-term psychiatric hospital. Am J Psychiatry 1988;145:517–518. 271. Keck PE, Sebastianelli J, Pope HG, et al: Frequency and presentation of neuroleptic malignant syndrome in a state psychiatric hospital. J Clin Psychiatry 1989;50:352–355. 272. Guzé BH, Baxter LR: Neuroleptic malignant syndrome. N Engl J Med 1985;313:163–166. 273. Lenler-Petersen P, Hansen BD, Hasselstrom L: A rapidly progressing lethal case of neuroleptic malignant syndrome. Intens Care Med 1990;16:267–268. 274. Keck PE, Caroff SN, McElroy SL: Neuroleptic malignant syndrome and malignant hyperthermia: End of a controversy. J Neuropsychiatry Clin Neurosci 1995;7:134–155. 275. May DC, Morris SW, Stewart RM: Neuroleptic malignant syndrome: response to dantrolene sodium. Ann Intern Med 1983;98:183–184. 276. McDonough CM, Swift G, Sheehan MB: Neuroleptic malignant syndrome: A diagnosis easily missed. Ir Med J 2000;93:152–154. 277. Pope HG, Keck JP, Mcelroy SL: Frequency and presentation of neuroleptic malignant syndrome in a large psychiatric hospital. Am J Psychiatry 1986;143:1227–1233. 278. Kargianis JL, Phillips LC, Hogan KP, et al: Clozapine-associated neuroleptic malignant syndrome: Two new cases and a review of the literature. Ann Pharmacother 1999;33:623–630. 279. Hasan S, Buckley P: Novel antipsychotics and the neuroleptic malignant syndrome: A review and critique. Am J Psychiatry 1998;155: 1113–1116. 280. Friedman LS, Weinrauch LA, D’Elia JA: Metoclopramide-induced neuroleptic malignant syndrome. Arch Intern Med 1987;147: 1495–1497. 281. Madakasira S: Amoxapine-induced neuroleptic malignant syndrome. Drug Intel Clin Pharm 1989;23:50–51. 282. Catterson ML, Martin RL: Anticholinergic toxicity masquerading as neuroleptic malignant syndrome: A case report and review. Ann Clin Psychiatry 1994;6:4267–4269. 283. Mills KC: Serotonin syndrome. Am Fam Physician 1995;52: 1475–1482.
284. Daras M, Kakkouras L, Tuchman AJ, et al: Rhabdomyolysis and hyperthermia after cocaine abuse: A variant of neuroleptic malignant syndrome. Acta Neurol Scand 1995;92:2161–2165. 285. Buckley PF, Hutchinson M: Neuroleptic malignant syndrome. J Neurol Neurosurg Psychiatry 1995;58:271–273. 286. Smego RA, Durack DT: The neuroleptic malignant syndrome. Arch Intern Med 1982;142:1183–1185. 287. Caroff SN, Mann SC: Neuroleptic malignant syndrome. Med Clin North Am 1993;77:185–202. 288. White DAC, Robins AH: Catatonia: Harbinger of the neuroleptic malignant syndrome. Br J Psychiatry 1991;158:419–421. 289. Rosebush P, Stewart T: A prospective analysis of 24 episodes of neuroleptic malignant syndrome. Am J Psychiatry 1989;146:717–725. 290. Keck PE, Pope HG, Cohen BM, et al: Risk factors for neuroleptic malignant syndrome. Arch Gen Psychiatry 1989;46:914–918. 291. Lee JWY: Serum iron in catatonia and neuroleptic malignant syndrome. Biol Psychiatry 1998;44:499–507. 292. Araki M, Takagi A, Higuchi I: Neuroleptic malignant syndrome: Caffeine contracture of single muscle fibers and muscle pathology. Neurology 1988;38:297–301. 293. Behen WM, Madigan M, Clark BJ, et al: Muscle changes in the neuroleptic malignant syndrome. J Clin Pathol 2000;53:223–227. 294. Pelonero AL, Levenson JL, Pandurangi AK: Neuroleptic malignant syndrome: A review. Psych Serv 1998;49:1163–1172. 295. Addonizio G, Susman VL, Roth SD: Neuroleptic malignant syndrome: Review and analysis of 115 cases. Biol Psychiatry 1987;22:1004–1020. 296. Lazarus A, Mann SC, Caroff SN: The Neuroleptic Malignant Syndrome and Related Conditions. Washington DC, American Psychiatric Press, 1989. 297. Davis JM, Janicak PG, Sakkas P, et al: Electroconvulsive therapy in the treatment of neuroleptic malignant syndrome. Convulsive Ther 1991;7:111–120. 298. Addonizio G, Susman VL: ECT as a treatment alternative for patients with symptoms of neuroleptic malignant syndrome. J Clin Psych 1987;48:102–105. 299. Gregorakos L, Thomaides T, Stratouli S, et al: The use of clonidine in the management of autonomic overactivity in neuroleptic malignant syndrome. Clin Auton Res 2000;10:193–196.
W9418-14.qxd 11/3/03 2:49 PM Page 439
Chapter 14 Prehospital Care of the Patient with Neurologic Injury Andrea Gabrielli, MD, Ahamed H. Idris, MD, A. Joseph Layon, MD
Introduction Optimal neurologic survival after cardiac arrest can be achieved only if all the links of chain of survival are performed efficiently and rapidly. Prehospital care in the setting of cardiac arrest ranges from recognition of early signs of respiratory, neurologic or cardiovascular distress, to basic cardiopulmonary resuscitation (CPR), to the potential for the use of automatic external defibrillation by the lay rescuer; from prompt response of the emergency medical services (EMS) system with institution of advanced cardiac life support (ACLS), to safe transport to an appropriate facility. While neurologic emergencies require specialized and multidisciplinary critical care, the quality of the initial management is the primary determinant for improving survival and neurologic outcome. Therefore, many principles applicable to prehospital care in general can be used in the approach to the neurologically injured patient, regardless if the injury is secondary to trauma or a cerebrovascular event.
Organization of Prehospital Care in North America In North America, the initial response to an out-of-hospital cardiac arrest or central nervous system (CNS) trauma is usually via the police or fire departments. These responders are proficient in providing basic life support (BLS) including CPR, opening the airway, bag-valve-mask (BVM) ventilation, and manual tamponade of external bleeding sites.
Emergency medical technicians-basic (EMT-B) and emergency medical technicians–paramedics (EMT-P) are trained in vehicular extrication, immobilization in transport, and in the case of paramedics, advanced life support.1 Transportation of neurotrauma patients to the emergency department (ED) is usually provided with destination priority given to a certified trauma center. However, distance, geography, logistics, and patient condition may determine the choice and site of transport destination. For example, it may be in the patient’s best interests to have hemodynamic stabilization initiated in a smaller hospital if transportation to a major center is not readily available.2 Even if the destination hospital is not a designated trauma center, outcome is enhanced with the use of the acute CNS injury clinical pathway, in which an emergency physician, neuroradiologist, neurosurgeon, and intensivist are promptly available as the victim reaches the ED.3,4 Generally, ambulance transport is used for transport within a 50-mile radius, assuming that roadway traffic is not a limiting factor in transport time. A rotary wing aircraft is often used for distances between 51 and approximately 150 miles, while a fixed wing aircraft is used for longer distances, when available.5 EMS systems vary considerably throughout the world. In Europe, for example, physicians are often used in the ambulance. Improved resuscitation and survival to discharge when cardiac arrest is present has been reported. In a Danish study of cardiac arrest patients, the presence of a physician in the ambulance increased the hospital discharge rate after cardiac arrest from 1% to 13%.6 The presence of a physician in the 439
W9418-14.qxd 11/3/03 2:49 PM Page 440
440
Section II
Neurointensive Care: General Issues
field seems of less importance for major trauma, regardless of association with CNS injury. Recently, a large European study of patients with major trauma (injury severity score [ISS] > 16) and severe head injury attempted to resolve the controversy between the advantage of prolonged ACLS in the field, managed by physicians, and a short BLS resuscitation with immediate transport to a hospital via helicopter with EMS personnel on board. No differences in outcome were demonstrated in this study.7 In a British study, in the presence of a similarly severe neurologic injury not associated with severe multiple trauma, a favorable trend in survival and better outcome were found when the patient was immediately transported by helicopter to an ED where necessary invasive procedures could be performed safely and effectively,8 even if a physician was involved in ACLS provision in the field. These studies suggest that the advantage of a physician in the field is limited to patients in cardiac arrest, and that health resources for trauma patients be focused on providing rapid transport to the ED. Related to costs and availability, and in the face of such evidence-based data, the health care system of the United States will continue to rely on EMT and paramedic first responders.
Overview and Big-Picture Changes: Advanced Cardiac Life Support for Experienced Providers Course The scope of emergency cardiovascular care has been expanded to include prearrest interventions, arrest prevention, and postarrest stabilization. All new guidelines have been rigorously reviewed, adhering to the principles of evidence-based medicine. The class I, IIa, IIb, III, and indeterminate designations after many guidelines have been used to indicate the strength of scientific evidence. More emphasis has been placed on first aid, CPR, and defibrillation in the workplace. A unifying approach to ACLS assessment and management for both lay rescuer and EMS has been proposed. Primary Survey: Airway (open), Breathing (two breaths), Circulation (chest compressions), Defibrillation (use automatic external defibrillation [AED]). Secondary Survey: Airway (advanced airway techniques), Breathing (placement confirmation, check effectiveness), Circulation (access the circulation; administer drugs as indicated), and Differential diagnosis. Education, Training, and Examination Changes
Cerebral Resuscitation and the 2000 American Heart Association Cardiopulmonary Resuscitation Guidelines The pathophysiology and rationale of cerebral resuscitation after cardiac arrest are extensively reviewed in Chapter 15. The Guidelines of the American Heart Association (AHA) 2000 Conference are the result of the first international application of a rigorous, evidence-based approach to produce resuscitation parameters, including neurological outcome.9 The new guidelines also include a reappraisal of old recommendations that have been reaffirmed, assigned for review, or removed. Several issues, pertinent to cerebral outcome after resuscitation during both BLS and ACLS, will be discussed here. These include 1. Overview and big-picture changes from the previous guidelines 2. Education, training, and examination 3. Ethical concerns in resuscitation 4. BLS 5. Airway and ventilation (BLS and ACLS) 6. ACLS 7. Pediatric advanced life support (PALS) 8. Neonatal resuscitation 9. Circulatory adjuncts approved for clinical use 10. Future research on cerebral resuscitation in prehospital setting.
Commitment to education and training based on core learning objectives has been increased. These include education and training innovations such as videotape-mediated training, and auditory and/or visual prompts. The educational value of ACLS, PALS, and BLS has been increased. Ethical Concerns in Resuscitation Family presence during resuscitation has been encouraged and is considered valuable. Evaluating and honoring “Do Not Attempt Resuscitation” (DNAR) status in field, and Certification of Death has been re-evaluated and is encouraged when appropriate. Basic Life Support: Adult and Pediatric The importance of early defibrillation has been further emphasized. In fact, the value of reducing the interval between adult sudden cardiac arrest and first defibrillatory shock by 1 to 2 minutes has been shown to do more to improve the probability of survival for an individual patient than all the medications, airway interventions, and newly designed defibrillation waveforms combined. Special situations that modify the phone first (call 911 first) versus phone fast (resuscitate immediately, then call 911) guidelines have been clarified. BLS recognizes clinical exceptions to the phone fast guideline (applies to children younger than 8 years of age) and the phone first guideline (applies to children older than 8 years of age). The major exception of the phone fast rule is in those children (younger
W9418-14.qxd 11/3/03 2:49 PM Page 441
Chapter 14
than 8 years of age) known to be at risk for ventricular fibrillation/ventricular tachycardia who experience sudden witnessed collapse. The new BLS guidelines for the following situations in adults have been altered to phone fast, that is, provide 1 minute of CPR before phoning the EMS system: 1. 2. 3. 4.
Submersion/near-drowning Poisoning/drug overdose Trauma Respiratory arrest.
Smaller Tidal Volumes During Adult Rescue Breathing More emphasis has been placed on proper BVM ventilation as a skill that all BLS health care providers must master to increase resuscitation efficacy and reduce complications. Rescuers have been given recommendations to deliver smaller tidal volumes during ventilation with BVM ventilation or when supplementary oxygen is available. Recommendations for rescue breaths delivered by mouth-to-mouth or mouth-to-barrier device have been set at an average of 700 to 1000 mL, delivered over 1.5 to 2 seconds. If supplementary oxygen is available, the skilled rescuer should attempt to provide smaller tidal volumes during mouth-tomouth and BVM ventilation, approximately 400 to 600 mL, with an inspiratory time of 1 to 2 seconds. If the smallest tidal volumes are used, the chest should rise visibly and the oxygen saturation should be maintained. Mouth-to-nose breathing has been accepted as an alternative to mouth-tonose-and-mouth or mouth-to-mouth rescue breathing for an infant. Cardiopulmonary Resuscitation The pulse check has been removed from the lay rescuer’s checklist. In its place are signs of circulation, which include normal breathing, coughing, or movement. If no signs of circulation are detected, the rescuer should begin chest compressions and attach an AED, if available. A new chest compression rate for adults has been recommended, irrespective of whether used in one- or two-rescuer CPR, by lay rescuers, or by health care professionals. This chest compression rate is approximately 100 compressions per minute. The adult compression-ventilation ratio has also been changed. For adult victims, two rescuers should no longer use a compression-ventilation ratio of 5 : 1. Instead, they should use a compression-ventilation ratio of 15 : 2 until the airway is secured. The 5 : 1 ratio should be used in pediatric arrest by professional responders regardless of whether one or two rescuers are involved. The two thumb-encircling hands chest compression technique has been recommended
Prehospital Care of the Patient with Neurologic Injury
441
over the two-finger compression technique for two-rescuer infant CPR, when performed by health care providers. The use of AEDs has been encouraged for victims of cardiac arrest older than approximately 8 years of age, or heavier than approximately 25 kg (55 lbs). Although data regarding the use of AEDs in pediatric cardiac arrest victims is limited, the guidelines suggest the use of defibrillators with adjustable energy dose for in-hospital use in areas that routinely care for infants and children. CPR performed without mouth-to-mouth ventilation has been reviewed and is encouraged if the rescuer is unwilling to perform rescue breathing for an adult victim. The rescuer should access the EMS system, open the airway, and perform chest compressions at the rate of approximately 100 compressions per minute CPR without ventilation; good quality CPR in the presence of an open airway will allow for acceptable gas exchange. Management of the Airway and Ventilation The guideline sections on airway management and ventilation contain the greatest number of new recommendations. Most of these recommendations apply to health care providers at both the BLS and ALS level. Airway devices (BVM, laryngeal mask airway [LMA], and Combitube) have been introduced as valid alternatives to the endotracheal tube. Recommendations for secondary confirmation of proper tracheal tube placement have been emphasized. Any EMS system that authorizes endotracheal intubation must ensure proper initial training, monitoring of skill retention, and ongoing monitoring of safety and effectiveness. Providers should assess endotracheal tube placement by primary confirmation using physical examination techniques, plus one or more secondary confirmation techniques, including qualitative end-tidal CO2 detectors, quantitative and continuous CO2 measurement (capnometry versus capnography), as well as devices that specifically detect tubes located in the esophagus. Specific approaches to prevent tube dislodgment have been encouraged, such as use of commercially manufactured tube holders. This is especially true in the prehospital setting, in which patient and transportation movements greatly increase the risk of dislodgment. Airway issues are discussed in detail in Chapter 16. Advanced Cardiac Life Support A comprehensive cardiac arrest algorithm has been introduced. This algorithm also presents the primary and secondary “ABCD Surveys” as a manner to organize the rescuers’ thoughts, action sequences, and anticipations. In ventricular fibrillation (VF)/pulseless ventricular tachycardia (VT), after attempting electrical conversion with a rapid sequence of three voltage administrations, intravascular epinephrine 1 mg or vasopressin 40 U for refractory
W9418-14.qxd 11/3/03 2:49 PM Page 442
442
Section II
Neurointensive Care: General Issues
VF/pulseless VT has been recommended. The fourth shock should be given 30 to 60 seconds after either epinephrine or vasopressin, to allow circulation and distribution of the drugs. The most effective adult dose of epinephrine remains 1 mg every 3 to 5 minutes. Higher doses of epinephrine are suspected of causing harm and, while acceptable, are no longer a recommended option. Vasopressin has been introduced as an equivalent to epinephrine for refractory VF/pulseless VT. As a vasoconstrictor, vasopressin appears as effective as epinephrine, with fewer negative cardiac effects. A unique advantage of vasopressin has been described in that it may be superior to epinephrine in increasing diastolic blood pressure,10 a fundamental component of cerebral blood flow (CPP), where CPP = {([systolic blood pressure - diastolic blood pressure] ∏ 3) + diastolic blood pressure} - ICP where ICP is intracranial pressure. Thus, when compared with epinephrine, it appears that vasopressin has the theoretical advantage during resuscitation of increasing both cerebral blood flow and cerebral oxygen delivery.11-13 While this advantage can be demonstrated in animal models of early and late CPR, it has not been confirmed in a recent, small, prospective, randomized human study.14 At this time, data from a similar, but larger, multicenter study in Europe are being analyzed for survival and neurologic outcome. Because vasopressin lasts much longer (10 to 20 minutes) than epinephrine, only one dose is recommended. By consensus, epinephrine 1 mg every 3 to 5 minutes may be resumed after 5 to 10 minutes if there has been no response to the vasopressin. Amiodarone has been added to the list of antiarrhythmic agents to consider for shock-refractory VF/pulseless VT as better choice than lidocaine. Amiodarone requires several time-consuming steps before administration: a glass ampule must be opened, the drug aspirated into a syringe with a large-gauge needle and diluted with 0.9% saline solution, the needle changed, and then the drug administered slowly through the intravenous (IV) line. Bretylium has been eliminated in the ACLS list of recommended antiarrhythmics due to its unavailability. Lidocaine remains acceptable for the treatment of shock-refractory VF, but the level of supportive evidence is weak. Magnesium has demonstrated effectiveness for treatment of VF/pulseless VT in two clinical situations: Torsades de Pointes and other arrhythmias associated with known hypomagnesemia. The algorithm for pulseless electrical activity and asystole has essentially remained the same for more than a decade, with epinephrine, atropine, and pacing the only three interventions available. Pediatric Advanced Life Support The intraosseous (IO) route has been recommended when no intravenous access is promptly available in arrest victims
six years of age and older; there is no upper age limit for the use of an IO device. PALS uses the 90-second, suggested “reasonable” limit, for establishment of vascular access in cardiac arrest, but offers a little more flexibility if the patient is stable. Vagal maneuvers have been added to the treatment algorithm for supraventricular tachycardia. Vagal maneuvers are recommended for the treatment of supraventricular tachycardia (SVT) provided these maneuvers do not delay cardioversion or the administration of adenosine for the child with poor systemic perfusion. Ice water applied to the face is a most effective vagal-enhancing maneuver in infants and young children. The administration of amiodarone is now recommended for pediatric SVT and VF/VT. Amiodarone can be used for both supraventricular and ventricular arrhythmias; in particular, amiodarone may be considered for refractory VF that persists despite three shocks. As in adult arrests, the use of high dose epinephrine has been de-emphasized. The recommended initial resuscitation dose of epinephrine for pediatric cardiac arrest is 0.01 mg/ kg, given IV or IO, or 0.1 mg/kg by the tracheal route. Repeated doses are recommended every 3 to 5 minutes for ongoing arrest. The same dose of epinephrine is recommended for second and subsequent doses for unresponsive asystolic and pulseless arrest, but higher doses of epinephrine (0.1 to 0.2 mg/kg) by any intravascular route may be considered. Attempting to defibrillate VF/pulseless VT detected by an AED is acceptable and recommended for children 8 years of age or older. Neonatal Resuscitation Although most new recommendations are related to in-hospital treatment of the newborn, significant prehospital care is still provided in rural areas. The most important recommendations are: 1. The importance of ventilation in the newly born infant with heart rate less than 100 beats per minute has been reaffirmed. 2. The indications for chest compressions have been simplified. Chest compressions are initiated if the heart rate is absent or if the heart rate remains less than 60 beats per minute after 30 seconds of adequate assisted ventilation. Chest compressions must be coordinated with ventilation at a ratio of 3 : 1, with a rate of 120 events per minute, comprised of 90 chest compressions and 30 ventilations. 3. Chest compression with two thumbs and encircling fingers is the preferred method for two-rescuer health care providers. Compressions should be delivered to the lower one third of the sternum. Other acceptable, although not preferred, chest compression techniques include the two thumb-encircling hand technique and the two-finger compression technique; of these, the two
W9418-14.qxd 11/3/03 2:49 PM Page 443
Chapter 14
thumb-encircling hands technique is preferred as the secondary technique. 4. The LMA is now available and acceptable for use in fullborn infants. The use of this device, while relatively intuitive, is considered appropriate only by properly trained prehospital providers. 5. As with the adult arrest victim, secondary confirmation of endotracheal tube placement is recommended, as are specific approaches to prevent tube dislodgment, such as the use of commercially manufactured tube holders. This is particularly true for the prehospital setting in which there is greater risk of dislodgment. 6. The use of isotonic crystalloid solutions for initial volume resuscitation instead of albumin-containing solutions has been confirmed. The fluid of choice for volume expansion in the prehospital phase is an isotonic crystalloid solution such as 0.9% saline or Ringer’s lactate solutions.
Prehospital Care of the Patient with Neurologic Injury
443
cardiac arrest is still within the defined window of opportunity ( 45 years History of seizures or epilepsy absent Symptom duration < 24 hours At baseline, patient is not wheelchair bound or bedridden Blood glucose between 60 and 400 Obvious asymmetry (right vs left) in any of the following 3 categories (must be unilateral)
Facial smile/grimace Grip Arm strength From Kidwell CS, et al: Stroke 2000;31:71–76, with permission.
Figure 14-1. A review of prehospital care of patients with stroke.
Yes
Unknown
No
[] [] [] [] [] []
[] [] [] [] [] []
[] [] [] [] [] []
Equal
R Weak
L Weak
[] Droop [] Weak grip [] No grip [] Drifts down [] Falls rapidly
[] Droop [] Weak grip [] No grip [] Drifts down [] Falls rapidly
[] [] [] [] []
W9418-14.qxd 11/3/03 2:49 PM Page 452
452
Section II
Neurointensive Care: General Issues
Figure 14-2. The “seven Ds” of stroke management.
activator (r-tPA) within the first 3 hours of onset of symptoms. The use of r-tPA and pro-urokinase has also been found to improve neurologic outcome in patients treated within 3 to 6 hours of onset of stroke. However, in this last category, the evidence-based medicine classification is that of anecdotal data. We have been discussing stroke as if it were always thrombotic in nature. While this is often the case, stroke may also be hemorrhagic, due either to intracerebral or subarachnoid hemorrhage (SAH). An SAH (grade 3 or 4 of the Hunt and Hess scale; Table 14-5) is a form of cerebrovascular accident
Table 14-5 Hunt and Hess Scale for Subarachnoid Hemorrhage Grade 1 2 3 4 5
Neurologic Status Asymptomatic Severe headache or nuchal rigidity; no neurologic deficit Drowsy; minimal neurologic deficit Stuporous; moderate-to-severe hemiparesis Deep coma; decerebrate posturing
W9418-14.qxd 11/3/03 2:49 PM Page 453
Chapter 14
that may be followed by cardiac arrest. In fact, aneurysmal SAH etiologically accounts for approximately 20% of the cardiac arrests reported in the prehospital arena or ED.114,115 Often, the patient cannot be resuscitated or remains devastated neurologically. Occasionally, however, aggressive CPR and rapid transport to an acute care facility where neurointensive care, including ICP monitoring, is available can be life-saving.116 In general, long-term survival in SAH patients in whom prehospital cardiac arrest occurred is quite poor, reported as less than 2%. Patients with SAH who present only with respiratory arrest have a better functional longterm survival than do those with full-blown cardiopulmonary arrest.118 The etiology of cardiac arrest in patients with SAH is complex, including sudden increase in intracranial pressure, brainstem compression with herniation, and respiratory arrest. Typically, a massive sympathetic discharge may result in a lethal cardiac arrhythmia. Common features characterize patients with SAH and cardiac arrest who survive, including bystander CPR that is initiated immediately, and return
Prehospital Care of the Patient with Neurologic Injury
453
of spontaneous circulation or respiration shortly after CPR is initiated.117 If SAH is suspected after resuscitation of a patient from cardiac arrest, hypertension (as discussed previously) should be withheld until the diagnosis is ruled out with a CT scan of the head. Based on our clinical experience, acquired in patients with SAH and vasospasm before surgical ablation, maintaining systolic blood pressure between 120 and 140 mm Hg is usually safe. In conclusion, recent advances in cardiopulmonary resuscitation, EMS, and trauma organization have had a dramatic impact on survival and quality of life in the neurologically injured patient. Despite numerous ongoing animal and human trials of agents thought to provide neuroprotection, few specific acute interventions have been recommended and these are noted primarily in victims of thrombotic stroke. In severe head and cervical spine trauma, emphasis is placed on the immediate correction of factors, such as hypoxia and hypotension, that will result in secondary injury, rapid transport to adequate hospital facilities, and prompt communication with the ED.118
P earls 1. While neurologic emergencies require specialized and multidisciplinary critical care, the quality of the initial management is the primary determinant for improving survival and neurologic outcome. 2. The presence of a physician in the field seems of less importance for major trauma, regardless of association with CNS injury. 3. For adult patients, two rescuers should no longer use a compression-ventilation ratio of 5 : 1. Instead, they should use a compression-ventilation ratio of 15 : 2 until the airway is secured. 4. The intraosseous (IO) route has been recommended when no intravenous access is promptly available in arrest victims 6 years of age and older; there is no upper age limit for the use of an IO device. 5. Postresuscitation interventions that may improve neurologic outcomes are of particular interest. These have been reviewed and include: a. Maintenaning normal ventilation without hyperventilation b. Monitoring temperature and preventing/treating hyperthermia c. Allowing mild hypothermia (to 34°C to 35°C) without aggressive rewarming d. Managaging postischemic myocardial dysfunction e. Maintaining glucose levels between about 80 mg/dL and 120 mg/dL. 6. An extensive MEDLINE review for the years 1966 through 1998, using head injury and prehospital care
7.
8.
9. 10.
11.
as keywords confirmed that hypotension (systolic BP < 90 mm Hg) and hypoxia (oxygen saturation < 90% or PaO2 < 60 mm Hg) were directly related to a poor outcome. Alcohol intoxication plays a major role in hospital outcome after acute brain injury. It has been shown that an alcohol level greater than 100 mg/dL is a risk factor for respiratory failure in the field, requiring endotracheal intubation. While alcohol clearance is associated with a “miraculous” awakening from head trauma coma, this group of patients still presents a high hospital morbidity and length to stay secondary mostly to gramnegative pneumonia. Approximately 5% to 10% of all patients with head injury have associated spinal injury. In the United States, there are more than 10,000 spinal injuries annually, with males predominating in this pathophysiology (81.6%). The average age of patients is 32.1 years, and most of the injuries are related to motor vehicle crashes (38.5%), followed by acts of violence (24.5%), and falls (21.8%). Although the decision as to the need for intubation is individually taken based on concurrent trauma and mental status, a patient with an FVC less than 15 mL/kg, or a negative inspiratory force (NIF) greater than -30 cm H2O will likely require intubation and ventilatory support. Continued
W9418-14.qxd 11/3/03 2:49 PM Page 454
454
Section II
Neurointensive Care: General Issues
12. The use of a plastic oral airway is not recommended if the patient is not intubated because it can contribute, in the presence of a strong gag reflex, to emesis with sudden neck movement and potential aspiration of gastric contents. 13. When trauma has occurred in a body of water, the patient is slowly floated to a long backboard while still in the water. Again, immobilization of the neck
References 1. Ramzy AI, Parry JM, Greenberg J: Head and spinal injury: Pre-hospital care. In Greenberg J (ed): Handbook of Head and Spine Trauma. New York, Marcel Dekker, 1993, pp 29–44. 2. Pre-hospital Trauma Life Support Committee of the National Association of Emergency Medical Technicians in Cooperation with the Committee on Trauma of the American College of Surgeons. PHTLS: Basic and Advanced Pre-hospital Trauma Life Support. 4th ed. St. Louis, Mosby, 1999. 3. Kish DL: Pre-hospital management of spinal trauma: An evolution. Crit Care Nurs Q 1999;22(2):36–43. 4. Pre-hospital Trauma Life Support Committee of the National Association of Emergency Medical Technicians in Cooperation with the Committee on Trauma of the American College of Surgeons. PHTLS: Basic and Advanced Pre-hospital Trauma Life Support. 4th ed. St. Louis, Mosby, 1999. 5. Crosby LA, Lewallen DG (eds): Emergency Care and Transportation of the Sick and Injured, 6th ed. Rosemont, IL, American Academy of Orthopaedic Surgeons, 1995. 6. Frandsen F, Nielsen JR, Gram L, et al: Evaluation of intensified prehospital treatment in out-of-hospital cardiac arrest: Survival and cerebral prognosis—The Odense Ambulance Study. Cardiology 1991;79:256–264. 7. Bartolomeo SD, Sanson G, Nardi G, et al: Effects of 2 patterns of prehospital care on the outcome of patients with severe head injury. Arch Surg 2001;136(11):1–15. 8. Nicholl JP, Brazier JE, Snooks HA: Effects of London helicopter emergency medical service on survival after trauma. BMJ 1995;311:217– 222. 9. Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. International Consensus on Science. Supplement to Circulation 2000;102(8). 10. Linder KH, Prengel AW, Brinkmann A, et al: Vasopressin administration in refractory cardiac arrest. Ann Intern Med 1996;124:1061–1064. 11. Wenzel, Lindner KH, Krismer AC, et al: Repeated administration of vasopressin but not epinephrine maintains coronary perfusion pressure after early and late administration during prolonged cardiopulmonary bypass. Circulation 1999;99:1379–1384. 12. Prengel AW, Lindner KH, Keller A: Cerebral oxygenation during cardiopulmonary resuscitation with epinephrine and vasopressin in pigs. Stroke 1996;27:1241–1248. 13. Wenzel V, Linder KH, Augenstein S, Prengel AW, Strohmenger HU: Vasopressin combined with epinephrine decreases cerebral perfusion compared with vasopressin alone during cardiopulmonary resuscitation in pigs. Stroke 1998;29:1462–1468. 14. Stiell IG, Hebert PC, Wells GA, et al: Vasopressin versus epinephrine for in hospital cardiac arrest: A randomized controlled trial. Lancet 2001;358(9276):105–109. 15. Safar P, Xiao F, Radovsky A, et al: Improved cerebral resuscitation from cardiac arrest in dogs with mild hypothermia plus blood flow promotion. Stroke 1996;27:105–113.
in the water should be done gently, grasping the head and moving it into a neutral, in-line or “eyes forward” position. 14. At this time, the best therapeutic option for fibrinolysis is intravenous recombinant tissue plasminogen activator (r-tPA) within the first 3 hours of onset of symptoms.
16. Bernard SA, Gray TW, Buist MD, et al: Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346(8):557–563. 17. Mild therapeutic hypothermia to improve the neurological outcome after cardiac arrest. The hypothermia after cardiac arrest study group. N Engl J Med 2002;346(8):549–561. 18. Callaway CW, Tadler SC, Katz LM, et al: Feasibility of external cranial cooling during out-of-hospital cardiac arrest. Resuscitation 2002; 52:159–165. 19. Safar PJ, Kochanek PM: Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346(8): 612–613. 20. Berek K, Schinnert A, Traweger C, Lechleitner P, Baubin M, Aichner F. The prognostic significance of coma-rating, duration of anoxia and cardiopulmonary resuscitation in out-of-hospital cardiac arrest. J Neurol (1997) 244:556–561. 21. Jorgensen EO, Holm S: The natural course of neurological recovery following cardiopulmonary resuscitation. Resuscitation 1998;36:111– 122. 22. Roine RO, Kajaste S, Kaste M: Neuropsychological sequelae of cardiac arrest. JAMA 1993;269:237–242. 23. Madl C, Grimm G, Kramer L, et al: Early prediction of individual outcome after cardiopulmonary resuscitation. Lancet 1993;341:855– 858. 24. Berek K, Lechleitner P, Luef G, et al: Early determination of neurological outcome after pre-hospital cardiopulmonary resuscitation. Stroke 1995;26:543–549. 25. Tirschwell DL, Lonstreth WT Jr., Rauch-Matthews ME, et al: Cerebrospinal fluid creatine kinase BB isoenzyme activity and neurologic prognosis after cardiac arrest. Neurology 1997;48:352–357. 26. Mullner M, Sterz F, Domanovits H, et al: The association between blood lactate concentration on admission, duration of cardiac arrest, and functional neurological recovery in patients resuscitated from ventricular fibrillation. Intensive Care Med 1997;23:1138– 1143. 27. Koller-Strametz J, Fritzer M, Gwechenberger M, et al: Elevation of prostate-specific markers after cardiopulmonary resuscitation. Circulation 2000;102:290–293. 28. Oropello JM: Predicting cerebral outcome after cardiopulmonary resuscitation: Are serine protease inhibitors soothsayers? Crit Care Med 2001;29:1842–1844. 29. Stelzl T, von Bose MJ, Hogl B, et al: A comparison of the prognostic value of neuron-specific enolase serum levels and somatosensory evoked potentials in 13 reanimated patients. Eur J. Emerg Med 1995;2:24–27. 30. Rosen H, Rosengren L, Herlitz J, et al: Increased serum levels of the S-100 protein are associated with hypoxic brain damage after cardiac arrest. Stroke 1998;29:473–477. 31. Sterz F, Leonov Y, Safar P, et al: Hypertension with or without hemodilution after cardiac arrest in dogs. Stroke 1990;21:1178–1184. 32. Behringer W, Safar P, Kentner R, et al: Antioxidant tempol enhances hypothermic cerebral preservation during prolonged cardiac arrest in dogs. J Cereb Blood Flow Metab 2002;22(1):105–117.
W9418-14.qxd 11/3/03 2:49 PM Page 455
Chapter 14 33. Idris AH, Gabrielli A, Caruso LJ, et al: Oxidant injury occurs rapidly after cardiac arrest, cardiopulmonary resuscitation (CPR) and reperfusion, abstracted. Circulation 2001;102(18):571. 34. Ward JD: Pre-hospital care—management of head injury. Neurosurg Clin North Am 1991;2(2):251–255. 35. Regel G, Stalp M, Lehmann U, Seekamp A: Pre-hospital care: Importance of early intervention on outcome. ACTA Anaesthesiol Scand 1997;71–76. 36. Colohan ART, Alves WM, Gross CR, et al: Head injury mortality in two centers with different emergency medical services and intensive care. J Neurosurg 1989;71:202–207. 37. Nicholl JP, Brazier JE, Snooks HA: Effects of London helicopter emergency medical service on survival after trauma. BMJ 1995;311:217– 222. 38. Baxt WG, Moody P: The impact of a rotocraft aeromedical emergency care service on trauma mortality. JAMA 1983;249:3047–3051. 39. Krausz MM, Bar-Ziv M, Rabinovici R, Gross D: “Scoop and run” or stabilize hemorrhage shock with normal saline or small-volume hypertonic saline? J Trauma 1992;33(1):6–10. 40. Hill DA, Delaney LM, Duflou J, Path (Forens) MM: A populationbased study of outcome after injury to car occupants and to pedestrians. J Trauma Injury Infect Crit Care 1996;40:351–355. 41. Siegel JH: The effect of associated injuries, blood loss, and oxygen debt on death and disability in blunt traumatic brain injury. The need for early physiologic predictors of severity. J Neurotrauma 1995; 12(4):579–590. 42. Baxt WG, Moody P: The impact of advanced pre-hospital emergency care on the mortality of severely brain-injured patients. J Trauma 1987;27(4):365–369. 43. Pitts LH: Neurotrauma and trauma systems. New Horizons 1995;3(3)546–548. 44. American College of Surgeons Committee on Trauma: Resources for Optimal Care of the Injured Patient. Chicago, IL, American College of Surgeons, 1993. 45. Thompson CT, Bickell WH, Siemens RA, et al: Community hospital level II trauma center outcome. J Trauma 1992;32:336–443. 46. Pitts LH, Ojemann RG, Quest DO: Neurotrauma care and the neurosurgeons: A statement from the Joint Section on Trauma of the AANS and CNS. J Neurosurg 1987;67:783–785. 47. Maciver IN, Lassman LP, Thomson CW, et al: Treatment of severe head injuries. Lancet 1958;544–550. 48. Seelig JM, Becker DP, Miller JD, et al: Traumatic acute subdural hematoma. Major mortality reduction in comatose patients treated within four hours. N Engl J Med 1981;304:1511–1512. 49. Colohan AR, Alves WM, Grooss CR, et al: Head injury mortality in two centers with different emergency medical services and intensive care. J Neurosurg 1989;71:202–207. 50. Waters J, Wells C: The effects of a modern emergency medical care system in reducing automobile crash death. J Trauma 1973;13:645– 647. 51. Klauber MR, Marshall LF, Toole BM, et al: Cause of decline in head injury mortality rate in San Diego County California. J Neurosurg 1985;62:528–531. 52. The Brain Trauma Foundation: Resuscitation of blood pressure and oxygenation. J Neurotrauma 2000;17:471–478. 53. Rosner MJ, Newsome HH, Becker DP: Mechanical brain injury: The sympatho-adrenal response. J Neurosurg 1984;61:76–86. 54. Brown GW, Brown ML: Cardiovascular responses to experimental cerebral concussion in the rhesus monkey. Arch Neurol 1954;71– 77. 55. Miller JD, Sweet RC, Narayan RK, et al: Early insults to the injured brain. JAMA 1978;240:439–442. 56. Eisenberg HM, Cayard C, Papanicoloaou AC, et al: The effect of three potentially preventable complications on outcome after severe closed head injury. In Ishii S, Nagai H, Brock M (eds): Intracranial Pressure V. Tokyo, Springer-Verlag, 1983, pp 549–553.
Prehospital Care of the Patient with Neurologic Injury
455
57. Frost EAM, Aracibia CU, Schulman K: Pulmonary shunt as a prognostic indicator in head injury. J Neurosurg 1979;50:768– 772. 58. Chesnut RM, Marshall LF, Klauber MR, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 1993(II);34:216–222. 59. Stocchetti N, Furlan A, Volta F: Hypoxemia and arterial hypotension at the accident scene in head injury. J Trauma 1996;40:764–767. 60. Scalea TM, Maltz S, Yelon J, et al: Resuscitation of multiple trauma and head injury: role of crystalloid fluids and inotropes. Crit Care Med 1994;22:1610–1615. 61. Mattox KL, Maningas PA, Moore EE, et al: Pre-hospital hypertonic saline/dextran infusion for post-traumatic hypotension. Ann Surg 1991;213:482–491. 62. Wade CE, Grady JJ, Kramer CG, et al: Individual patient cohort analysis of the efficacy of hypertonic saline/dextran in patients with traumatic brain injury and hypotension. J Trauma 1997;42:561– 565. 63. Vassar MJ, Perry CA, Holcroft JW: Pre-hospital resuscitation of hypotensive trauma patients with 7.5% NaCl versus 7.5% NaCl with added dextran: a controlled trial. J Trauma 1993;34:622–632. 64. Sayre MR, Daily SW, Stern SA, et al: Out-of-hospital administration of mannitol to head-injured patients does not change systolic blood pressure. Acad Emer Med 1996;3:840–848. 65. Hsiano AK, Michaelson SP, Hedges JR: Emergent intubation and CT scan pathology of blunt trauma patients with Glasgow Coma Scale scores of 3-13. Pre-hospital Disaster Med 1998;32:26–32. 66. White JC, Brooks JR, Goldthwait JC, et al: Changes in brain volume and blood content after experimental concussion. Ann Surg 1943; 118:619–634. 67. Lyeth BG, Dixon CE, Hamm RF, et al: Effects of anticholinergic treatment on transient behavioral suppression and physiological responses following concussive brain injury to the rat. Brain Res 1988;448:88– 97. 68. Rosomoff HL, Kochanek PM, Clark R, et al: Resuscitation from severe brain trauma. Crit Care Med 1996;24(2)(Suppl):S48– S56. 69. Stocchetti N, Penny, KI, Dearden M, et al: Intensive care management of head–injured patients in Europe: A survey from the European Brain Injury Consortium. Intensive Care Med 2001;27:400– 406. 70. The Brain Trauma Foundation: Glasgow coma scale score. J Neurotrauma 2000;17:563–571. 71. The Brain Trauma Foundation: Neurotrauma Guidelines. J Neurotrauma, 2000;17:457–627. 72. Gurney JG, Rivara FP, Mueller BA, et al: The effects of alcohol intoxication on the initial treatment and hospital course of patients with acute brain injury. J Trauma 1992;33(5):709–713. 73. Servadei F, Nasi MT, Cremonini AM, et al: Importance of a reliable admission Glasgow Coma Scale Score for determining the need for evacuation of posttraumatic subdural hematomas. J Trauma Injury, Infect Crit Care 1998;44:868–873. 74. Ivan LP: Spinal reflexes in cerebral death. Neurology 1973;23:650– 652. 75. Bullock R, Chesnut RM, Clifton G, et al: Guidelines for the Management of Severe Head Injury. New York, Brain Trauma Foundation, 1995. 76. Muizelaar JP, Marmarou A, Ward JD, et al: Adverse effects of prolonged hyperventilation in patients with severe head injury: A randomized clinical trial. J Neurosurg 1991;75:731–739. 77. Thomas SH, Orf J, Wedel SK, Conn AK: Hyperventilation in traumatic brain injury patients: Inconsistency between consensus guidelines and clinical practice. J Trauma 2002;52:47–53. 78. Clifton GL, Miller ER, Choi SC, et al: Lack of effect of induction of hypothermia after acute brain injury. N Engl J Med 2001;344(8):556– 563.
W9418-14.qxd 11/3/03 2:49 PM Page 456
456
Section II
Neurointensive Care: General Issues
79. Kornberger E, Schwarz B, Lindner KH, Mair P: Forced air surface rewarming in patients with severe accidental hypothermia. Resuscitation 1999;41:105–111. 80. Safar P: On the future of reanimatology. Acad Emerg Med 2000;7:75– 89. 81. Kochanek PM, Clark RSB, Ruppel RA, Dixon CE: Cerebral resuscitation after traumatic brain injury and cardiopulmonary arrest in infants and children in the new millennium. Pediatr Clin North Am 2001;48(3):661–681. 82. Lucas DR, Newhouse JP: The toxic effect of sodium 1-glutamate on the inner layers of the retina. Arch Ophthalmol 1957;58:193–201. 83. Bouma GJ, Muizelaar JP, Stringer WA, et al: Ultra-early evaluation of regional cerebral blood flow in severely head-injured patients using xenon-enhanced computerized tomography. J Neurosurg 1992;77: 360–368. 84. Morris GF, Bullock R, Marshall SB, et al: Failure of the competitive N-methyl-d-aspartate antagonist Selfotel (CGS 19755) in the treatment of severe head injury: Results of two phase III clinical trials. J Neurosurg 1999;91:737–743. 85. Cherian L, Chacko G, Goodman JC, et al: Cerebral hemodynamic effects of phenylephrine and l–arginine after cortical impact injury. Crit Care Med 1999;27:2512–2517. 86. Sasser HC, Safar P, Kelsey SF, et al: Arterial hypertension after cardiac arrest is associated with good cerebral outcome in patients [abstract]. Crit Care Med 1999;27:A29. 87. Eckstein M: The pre-hospital and emergency department management of the penetrating head injuries. Neurosurg Clin North Am 1995;6(4):741–751. 88. Smith JP, Bodai BI, Hill AS, et al: Pre-hospital stabilization of critically injured patients. A failed concept. J Trauma 1985;25:65. 89. Walls RM: Airway management. Emerg Med Clin North Am 1993;11:55. 90. American College of Surgeons: Advanced Trauma Life Support Manual. Chicago, American College of Surgeons, 1993. 91. Wilberger JE: Emergency care and initial evaluation. In Cooper PR (ed): Head Injury. Baltimore, Williams & Wilkins, 1993, pp 27– 41. 92. Sellick BA: Cricoid pressure to control regurgitation of stomach contents during induction of anesthesia. Lancet 1961;2:404. 93. Steinhause JE, Gaskin L: A study of intravenous lidocaine as a suppressant of cough reflex. Anesthesiology 1963;24:285–290. 94. Benzel EC, Day WT, Kesterson L, et al: Civilian cranio-cerebral gunshot wounds. Neurosurgery 1991;29:67–72. 95. Spinal Cord injury Information Network: Spinal cord injury: Facts and figures at a glance [University of Alabama at Birmingham Web site]. May 2001. Available at: http://www.spinalcord.uab.edu/show.asp?durki = 21446. Accessed Jan 23, 2002. 96. Miglietta MA, Levins T, Robb TV: Evaluation of spine injury in blunt trauma. J Am Osteopath Assoc 2002;102(2):87–91. 97. Ledsome GR, Sharp JM: Pulmonary functions in acute cervical cord injury. Am Rev Resp Dis 1981;124:41–44. 98. Cohen M: Initial resuscitation of the patient with spinal cord injury. Trauma Q 1993;9:38–43. 99. Keller C, Brimacombe J, Keller K: Pressures exerted against the cervical vertebrae by the standard and intubating laryngeal mask airways: A randomized, controlled, cross-over study in fresh cadavers. Anesth Analg 1999;89(5):1296–1300.
100. Vaccaro AR, An HS, Betz RR, Cotler JM, Balderston RA: The management of acute spinal trauma: pre-hospital and in-hospital emergency care. In Greenberg J (ed): Handbook of Head and Spine Trauma. New York, Marcel Dekker, 1993, pp 113–125. 101. Ramzy AI, Parry JM, Greenberg J: Head and spinal injury: Pre-hospital care. In Greenberg J (ed): Handbook of Head and Spine Trauma. New York, Marcel Dekker, 1993, pp 29–44. 102. Swenson TM, Lauerman WC, Blank RO, et al: Cervical spine alignment in immobilized football players: Photographic analysis before and after helmet removal. Am J Sports Med 1997;20:226–260. 103. Pre-hospital Trauma Life Support Committee of the National Association of Emergency Medical Technicians in Cooperation with the Committee on Trauma of the American College of Surgeons. PHTLS: Basic and Advanced Pre-hospital Trauma Life Support, 4th ed. St. Louis, Mosby, 1999. 104. Bracken MB, Shepard MJ, Collins WF, et al: A randomized controlled trial of methylprednisolone or naloxone in the treatment of acute spinal-cord injury: Results of the Second National Acute Spinal Cord Injury Study. N Engl J Med 1990;322:1405–1411. 105. Lyden PD, RappK, Babcock T, Rothrock J: Ultra-rapid identification, triage, and enrollment of stroke patients into clinical trials. J Stroke Cerebrovasc Dis 1994;4:106–107. 106. Barsan WG, Brott TG, Broderick JP, et al: Time of hospital presentation in patients with acute stroke. Arch Intern Med 1993;2558–2561. 107. 1997 Heart and Stroke Statistical Update. Dallas, TX: American Heart Association, 1996. 108. Easton JD, Hart RG, Sherman DG, Kaste M: Diagnosis and management of ischemic stroke, I: Threatened stroke and its management. Curr Probl Cardiol 1983;8:1–76. 109. Kothari R, Pancioli A, Liu T, et al: Cincinnati prehospital stroke scale: reproducibility and validity. Ann Emerg Med 1999;33:373–378. 110. Kidwell CS, Saver JL, Schubert GB, Eckstein M, Starkman S: Design and retrospective analysis of the Los Angeles prehospital stroke screen (LAPSS). Prehosp Emerg Care 1998;2:267–273. 111. Kidwell CS, Starkman S, Eckstein M, Weems K, Saver JL: Identifying stroke in the field: Prospective validation of the Los Angeles. Prehospital stroke screen (LAPSS). Stroke 2000;31:71–76. 112. Prasad K, Menon GR: Comparison of the three strategies of verbal scoring of the Glasgow coma scale in patients with stroke. Cerebrovasc Dis 1998;8:79–85. 113. Scott PA, Temovsky CJ, Lawrence K, Gudaitis E, Lowell MJ: Analysis of Canadian population with potential geographic access to intravenous thrombolysis for acute ischemic stroke. Stroke 1998;29:2304– 2310. 114. Kitahara T, Masuda T, Soma S: The etiology of sudden cardiopulmonary arrest in subarachnoid hemorrhage. No Shinkei Geka 1996; 21:781–786. 115. Schievink WI, Wijdicks EF, Parisi JE, Piepgras DG, Wisnant JP: Sudden death from subarachnoid hemorrhage. Neurology 1995;45:871–874. 116. Inamasu J, Saito R, Nakamura Y, et al: Survival of a subarachnoid hemorrhage patient who presented with pre-hospital cardiopulmonary arrest: Case report and review of the literature. Resuscitation 2001;51:207–211. 117. Shapiro S: Management of subarachnoid hemorrhage patients who presented with respiratory arrest resuscitated with bystander CPR. Stroke 1996;27:1780–1782. 118. Biros MH, Heegaard W: Pre-hospital and resuscitative care of the head-injured patient. Curr Opin Crit Care 2001;7:444–449.
W9418-15.qxd 11/3/03 2:46 PM Page 457
Chapter 15 Cerebral Resuscitation from Cardiac Arrest* Peter Safar, MD and Wilhelm Behringer, MD†
Introduction Definitions Cerebral protection (pretreatment) and preservation (intrainsult treatment) before and during (anticipated) cerebral ischemia are important in the management of patients undergoing elective cerebral or cardiac anesthesia and surgery (Fig. 15-1). Cerebral resuscitation is treatment to reverse the insult and support recovery.1,2 This chapter concerns cerebral resuscitation from the temporary, complete global brain ischemia (GBI) of cardiac arrest (CA). This topic is relevant for emergency medical services (EMS),3–5 which should deliver not merely cardiopulmonary resuscitation (CPR),6 but rather cardiopulmonary-cerebral resuscitation (CPCR).1,2,7 CPCR leads to intensive care with focus on brain and heart. The fate of the brain for many patients who, after a variety of cerebral insults, are brought to the intensive care unit (ICU), from either outside or inside the hospital, has already been determined—during and immediately after the insult. Nevertheless, intensivists should know about novel opportunities for cerebral resuscitation because they are often consulted on such cases outside the ICU, and because brainfocused prolonged life support can mitigate brain damage. Temporary hypotension with mean arterial pressure (MAP) of about 30 to 60 mm Hg, can be tolerated by the normal brain, but even mild hypotension can cause perma*Sections of this chapter are based on Safar P: Resuscitation of the ischemic brain. In Albin MA (ed): Textbook of Neuroanesthesia. New York, McGraw-Hill, 1997. † On August 3, 2003, Dr. Peter Safar died. Dr. Wilhelm Behringer finalized this chapter in Dr. Safar’s honor.
nent brain damage when it occurs in a state of severe hypoxemia, after brain trauma, or in the presence of atherosclerotic cerebral arteries that fail to go into autoregulatory vasodilation. Much of what applies to cerebral resuscitation from the GBI of CA is also relevant for special operations on the brain that require temporary circulatory arrest. One must differentiate between the temporary, complete GBI of CA (see Fig. 15-1); the permanent, complete GBI of panorganic death without resuscitation; the temporary, incomplete GBI of shock states; the temporary or permanent focal brain ischemia of stroke (e.g., cerebral embolism); traumatic brain injury with unifocal, multifocal, or global ischemic components; and a variety of toxic, inflammatory, or degenerative cerebral insults. A treatment that is effective for one of the above conditions may not be effective for another; one effective for protection-preservation during an insult may not be effective for resuscitation; and one effective during incomplete ischemia may not be effective after complete ischemia. One must further differentiate between ischemic tissue hypoxia or anoxia that is caused by reduced cerebral blood flow (CBF); hypoxic hypoxia that is low arterial PO2 (PaO2); and anemic hypoxia that is caused by very low hemoglobin levels (low hematocrit [Hct] or carbon monoxide [CO] poisoning). One must also differentiate between process variables, such as electroencephalographic activity (EEG); CBF; or cerebral metabolic rates for oxygen, glucose, or lactate (CMRO2, CMRG, CMRL) during and early after the insult, and the much more important outcome in terms of cerebral functional and morphologic changes, after maturing of the postischemic encephalopathy over at least 3 days, perhaps weeks. 457
?
?
Pulmonary failure
Shock
Circulatory Arrest 10
EEG isoelectric
EEG abnormal
Neurologic deficit
Neurologically normal
Unconscious
Brain death
Unconscious Conscious or stuporous
Conscious
Apnea
20
Vegetative state
Spontaneous breathing
15
Spontaneous breathing
Restoration of Circulation
5
Spontaneous breathing
Approximate time, min
Clinical Death
Figure 15-1. Diagram of different causes of cardiac arrest and their reversibility, with present standard normothermic CPR. Flow chart illustrates diagrammatically the development of total circulatory arrest, as it happens suddenly (terminal states no. 1, as in VF, or no. 2, as in a primary asystole); over minutes (terminal states no. 3–5); or protracted (terminal states no. 6–8). “Clinical death” is defined as “total circulatory arrest with potential reversibility to survival without brain damage.11 Longest duration of clinical death that is reversible depends on terminal state, temperature, resuscitation methods, CPR (low-flow) time, and the postresuscitation disease. After restoration of circulation, there are various possible outcomes. (From Safar P: The pathophysiology of dying and reanimation. In Schwartz G, Safar P, Stone JH, et al (eds): Principles and Practice of Emergency Medicine. Philadelphia, WB Saunders, 1985, pp 2–41.)
Brain failure
?
5–12 min
(4) Asphyxia, Airway obstruction, Apnea
Exsanguination
2–3 min
0 min
Alveolar anoxia
Primary asystole
Primary ventricular fibrillation
Terminal States
Panorganic Death
?
W9418-15.qxd 11/3/03 2:46 PM Page 458
W9418-15.qxd 11/3/03 2:46 PM Page 459
Chapter 15
In ischemic insults such as CA (e.g., sudden cardiac death [SCD]), one must further differentiate between the many different mechanisms leading to CA, such as asphyxia, exsanguination, ventricular fibrillation (VF), electromechanical dissociation (EMD = pulseless electrical activity [PEA]) with mechanical asystole, or electric asystole; and slow, secondary CA (see Fig. 15-1). One must determine, or at least estimate, arrest time (no-flow), CPR time (low-flow), and temperature—the most important variables determining cerebral outcome. Reversal of CA calls for CPCR in three phases, that is, basic, advanced, and prolonged life support (BLS-ALSPLS).1,2 When initiated outside the hospital, the steps of resuscitation are to be delivered throughout the EMS system—from scene via transportation to the most appropriate hospital’s emergency department (ED), operating room (OR), and ICU.3–5 The combination of CPCR and EMS3 is now called the “chain of survival,”4,5 and is only as effective as the weakest step of CPCR in the weakest link of EMS. When talking about hypothermia—presently the most effective cerebral preservation and resuscitation strategy— one must differentiate between controlled (therapeutic) and uncontrolled (accidental) hypothermia; between temperature levels, such as normothermia (37°C to 38°C), and mild (33°C to 36°C), moderate (28°C to 32°C), deep (16°C to 27°C), profound (5°C to 15°C), or ultraprofound (b b1 > b2 = a
0.01–0.03 mg/kg/min 0.04–0.15 mg/kg/min 0.01–0.15 mg/kg/min 2–5 mg/kg/min 5–10 mg/kg/min 10 mg/kg/min 5–20 mg/kg/min 2.5–10 mg/kg/min 1–4 mg/min Loading dose: 0.75 mg/kg over 10 min Continuous infusion: 5–10 mg/kg/min Loading dose: 50 mg/kg over 10–20 min Continuous infusion: 0.375–0.50 mg/kg/min
b2 and dopaminergic effects b1 = b2 effects
CBF, cerebral blood flow; ICP, intracranial pressure; SAH, subarachnoid hemorrhage.
Stenosis at the level of the aortic valve results in a pressure gradient from the left ventricle (LV) to the aorta. In adults with AS, this obstruction usually increases gradually over a prolonged period, and left ventricular output is maintained by the presence of left ventricular hypertrophy. However, the LV chamber size is essentially unchanged, and wall tension remains essentially normal in well-compensated patients with AS. With long-standing AS, myocardial contractility progressively deteriorates and further compromises left ventricular function. Myocardial ischemia can occur with AS due to the combination of increased oxygen needs by the hypertrophied myocardium and reduction of oxygen delivery secondary to the excessive compression of coronary vessels. In
addition, approximately 50% of patients have concomitant CAD.132 Hemodynamic Goals. Prevention of significant decreases in systemic vascular resistance is of paramount importance in management of patients with AS. Vasodilatation will not reduce the work of the left ventricle but will reduce coronary perfusion pressure leading, to life-threatening ischemia. Avoidance of bradycardia, maintenance of sinus rhythm, and normal intravascular volume are also important. The hypertrophied and noncompliant left ventricle requires a coordinated atrial contraction to provide adequate left ventricular filling; otherwise, a sudden decrease in cardiac
W9418-17.qxd 10/31/03 3:46 PM Page 549
Chapter 17
output may cause severe hypotension. Likewise, atrial fibrillation with rapid ventricular response may cause ischemia and should be treated aggressively. Bradycardia is also poorly tolerated because of the inability of the left ventricle to generate a compensatory response to maintain cardiac output by increasing stroke volume. Aortic Regurgitation Left ventricular volume overload is the characteristic feature of aortic insufficiency. As a result of increased end-diastolic wall tension, serial replication of sarcomeres occurs, producing the pattern of eccentric ventricular hypertrophy. As the disease progresses, recruitment of preload reserve and compensatory hypertrophy allow the ventricle to maintain normal ejection performance despite the elevated afterload. Left ventricular systolic dysfunction is initially reversible. However, as the amount of aortic regurgitation exceeds more than 60% of the stroke volume, progressive LV dilatation and hypertrophy occur, leading to irreversible myocardial damage.132–134 Hemodynamic Goals. Maintenance of normal intravascular
volume and normal or decreased systemic vascular resistance is crucial because the primary mechanism for the regurgitant flow is an increase in stroke volume. However, overzealous fluid replacement can result in pulmonary edema, and excessive decreases in afterload may result in myocardial ischemia because the aortic diastolic pressure is already low. Avoidance of bradycardia is beneficial because the increased diastolic time will exacerbate regurgitation. Mitral Stenosis Mitral stenosis (MS) is typically rheumatic in origin and primarily affects women. Only 25% of patients have isolated MS, while 40% have MS and mitral regurgitation (MR).135,136 The normal mitral valve area is 4.0 to 6.0 cm2. Narrowing of the valve area to less than 2.5 cm2 must occur before the development of symptoms.135,136 With a valve area less than 1 cm2, a significant left atrioventricular pressure gradient is required to maintain a normal cardiac output at rest. In turn, the elevated left atrial pressure increases pulmonary capillary pressures, resulting in exertional dyspnea.137 The pathophysiology of MS stems from increased left atrial pressure and reduced cardiac output, primarily caused by impaired filling of the left ventricle from the obstructed left atrium. In patients with a markedly increased pulmonary vascular resistance, right ventricular function is often impaired. Pulmonary hypertension increases the risk associated with surgery.136,137 Hemodynamic Goals. Maintenance of sinus rhythm and
avoidance of tachycardia is extremely important. In patients with MS, the gradient across the mitral valve is critically dependent on heart rate. In the presence of tachycardia, left atrial pressure will increase and is rapidly transmitted back
Cardiac Care in Neurosurgery
549
to the pulmonary capillaries, which may result in pulmonary edema and right ventricular failure. Moreover, the inability to maintain left ventricular filling may result in hypotension. Systemic vascular resistance must be maintained because stroke volume is unlikely to increase with afterload reduction. On the other hand, vasodilatation may induce right ventricular ischemia by decreasing the aortic diastolic pressure necessary for adequate perfusion of the right ventricle. Factors that increase pulmonary hypertension such as hypoxemia, hypercarbia, acidosis, and hypothermia should be avoided. Otherwise, the already compromised right ventricle can acutely fail. Preload must be maintained or slightly increased to provide adequate flow across the mitral valve for left ventricular filling. However, excessive fluid administration may result in pulmonary edema. Mitral Regurgitation MR occurs from defects of any of the structures of the mitral valve apparatus, which include the valve leaflets per se, the mitral annulus, the chordae tendineae, and the papillary muscles.133,138 In acute MR, the volume overload increases LV end-diastolic pressure. Although the measured ejection fraction increases, the forward stroke volume is reduced because part of the stroke volume is regurgitated into the left atrium. Severe MR causes a sudden increase in left atrial pressure due to the relatively noncompliant left atrium, leading to severe pulmonary edema and symptoms of acute left ventricular failure. Unless such patients are treated aggressively, a fatal outcome is almost certain. In contrast, chronic MR is compensated for by the development of ventricular dilatation, which allows for an increase in both the ejection fraction and the forward stroke volume. Accordingly, enlargement of the left atrium allows the volume overload to generate lower filling pressures. In this phase of compensated MR, the patient may be entirely asymptomatic, even during vigorous exercise. The transition to chronic decompensated MR is characterized by LV dysfunction. In this phase, end-systolic volume is increased, and forward stroke volume is decreased, resulting in increased LV filling pressure and pulmonary congestion.138 Hemodynamic Goals. Decreased systemic vascular resistance
is the basis for the care of patients with MR. Increases in afterload worsen the regurgitant fraction. In addition, faster than normal heart rates are beneficial because bradycardia can increase the regurgitant volume by increasing left ventricular end-diastolic volume and mitral annular distension. Tachycardia should be avoided if the MR is secondary to myocardial ischemia.139,140 Preload reduction may be beneficial, as long as it is not excessive, because adequate volume to maintain adequate forward stroke volume is essential. Similarly, excessive volume expansion can also worsen the regurgitation by increasing the radius of the left ventricle.
W9418-17.qxd 10/31/03 3:46 PM Page 550
550
Section II
Neurointensive Care: General Issues
P earls 1. Increasing data suggest that the insula is a modulator of cardiac autonomic tone and, as such, an important arrhythmogenic center. ECG changes are usually transient, resolving within weeks; QTc prolongation and U waves can be permanent. 2. Arrhythmias occur immediately after blood is introduced into the subarachnoid space and are associated with acute increases in intracranial pressure. 3. The ECG patterns most predictive of left ventricular dysfunction are T wave inversions and prolonged QTc.25 Not all patients with these ECG patterns or other abnormalities will have underlying myocardial injury. The echocardiographic patterns of myocardial injury often do not correlate with a specific coronary artery distribution. 4. In patients with SAH, there seems to be an association between female gender, poor neurologic grade, and the development of myocardial dysfunction. 5. An ECG should be performed in a patient of any age with documented intracranial pathology. If the ECG is abnormal or arrhythmias are present, a transthoracic or transesophageal ECHO should be considered. 6. Milrinone not only improves myocardial contractility but has both arterial and venodilation properties. However, its cerebral vasodilating properties could be detrimental in patients with abnormal intracranial elastance. 7. Postoperative factors that increase the risk for myocardial ischemia include hypothermia, anemia, tachycardia, inadequate analgesia, and extreme changes in blood pressure. 8. The differential diagnosis of ST-T depressions includes left ventricular hypertrophy, digitalis effect, left bundle branch block, and electrolyte abnormalities (hypokalemia, hypomagnesemia). 9. The increase of CK isoenzymes from the myocardium (MB fraction) is rather sensitive. CK-MB activity
References 1. Talman WT, Kelkar P: Neural control of the heart. Central and peripheral. Neurol Clin 1993;11:239. 2. Crick SJ, Sheppard MN, Anderson RH: In Ter Horst GJ (ed): The Nervous System and the Heart. Totowa, NJ, Humana Press, 2000, pp 3–54. 3. Di Pasquale G, Pinelli G, Andreoli A, et al: Holter detection of cardiac arrhythmias in intracranial subarachnoid hemorrhage. Am J Cardiol 1987;59:596. 4. Szabo MD, Crosby G, Hurford WE, et al: Myocardial perfusion following acute subarachnoid hemorrhage in patients with an abnormal electrocardiogram. Anesth Analg 1993;76:253.
10.
11.
12.
13.
14.
15.
16.
17.
increases within 6 to 8 hours and remains elevated for 48 hours. Troponin T increases at 4 hours and remains elevated for 10 to 14 days; levels exceeding 0.1 ng/mL are indicative of myocardial damage. Troponin I is the most specific serum marker, because it exists only in the myocardium. Serum levels also begin to increase around 4 hours and remain elevated for 7 to 10 days. Levels greater than 1 ng/mL are indicators of myocardial injury. The incidence of postoperative MI appears to peak within 48 to 72 hours of surgery (range, 24 hours to 5 days). The evidence thus far provides a compelling case for the use of b1-selective blockers; nonspecific beta blockers such as propranolol have been associated with hypotension and increased bronchospasm and should not be used. In the cerebral circulation, nitroglycerin dilates conductance vessels, and tends to increase cerebral blood flow and cerebral blood volume. Neurologic deterioration may occur due to increases in ICP. It is estimated that approximately four million people have CHF, with 400,000 new cases diagnosed every year. The most reliable signs that indicate chronic decompensated CHF include the presence of jugular venous distention and a third heart sound.117-119 Other signs, such as pulmonary crackles and peripheral edema, are less specific. In the absence of a pulmonary artery catheter, clinical surrogates of perfusion (e.g., urine output, pulse pressure and systemic blood pressure) and pulmonary vascular congestion such as dyspnea, S3 gallop, and chest radiography findings may provide the diagnosis of CHF; however, their predictive value in an intensive care unit is poor and may lead to inappropriate therapy.
5. Pollick C, Cujec B, Parker S, et al: Left ventricular wall motion abnormalities in subarachnoid hemorrhage: an echocardiographic study. J Am Coll Cardiol 1988;12:600. 6. Manninen PH, Gelb AW, Lam AM, et al: Perioperative monitoring of the electrocardiogram during cerebral aneurysm surgery. J Neurosurg Anesthesiol 1990;2:16. 7. Hersch C: Electrocardiographic changes in subarachnoid haemorrhage, meningitis, and intracranial space-occupying lesions. Br Heart J 1964;26:785. 8. Hersch C, Med D: Electrocardiographic changes in head injuries. Circulation 1961;23:853. 9. Goldstein DS: The electrocardiogram in stroke: Relationship to pathophysiological type and comparison with prior tracings. Stroke 1979;10:253.
W9418-17.qxd 10/31/03 3:46 PM Page 551
Chapter 17 10. Dujardin KS, McCully RB, Wijdicks EFM, et al: Myocardial dysfunction associated with brain death: Clinical, echocardiographic, and pathologic features. J Heart Lung Transplant 2001;20:350. 11. Byer E, Ashman R, Toth LA: Electrocardiograms with large, upright T waves and long Q-T intervals. Am Heart J 1947;33:796. 12. Oppenheimer SM, Cechetto DF, Hachinski VC: Cerebrogenic cardiac arrhythmias. Cerebral electrocardiographic influences and their role in sudden death. Arch Neurol 1990;47:513. 13. Cheung RTF, Hachinski V: The insula and cerebrogenic sudden death. Arch Neurol 2000;57:1685. 14. Eckardt M, Gerlach L, Welter FL: Prolongation of the frequency-corrected QT dispersion following cerebral strokes with involvement of the insula of Reil. Eur Neurol 1999;42:190. 15. Umali F, Gould L, Gomprecht RF: Electrocardiographic changes in intracranial lesions. Angiology 1971;22:616. 16. Zaroff JG, Rordorf GA, Newell JB, et al: Cardiac outcome in patients with subarachnoid hemorrhage and electrocardiographic abnormalities. Neurosurgery 1999;44:34. 17. Andreoli A, di Pasquale G, Pinelli G, et al: Subarachnoid hemorrhage: Frequency and severity of cardiac arrhythmias. A survey of 70 cases studied in the acute phase. Stroke 1987;18:558. 18. Tokgozoglu SL, Batur MK, Topcuoglu MA, et al: Effects of stroke localization on cardiac autonomic balance and sudden death. Stroke 1999;30:1307. 19. Estanol BV, Loyo MV, Mateos JH, et al: Cardiac arrhythmias in experimental subarachnoid hemorrhage. Stroke 1977;8:440. 20. Horowitz MB, Willet D, Keffer J: The use of cardiac troponin-I (cTnI) to determine the incidence of myocardial ischemia and injury in patients with aneurysmal and presumed aneurysmal subarachnoid hemorrhage. Acta Neurochir (Wien) 1998;140:87. 21. Parekh N, Venkatesh B, Cross D, et al: Cardiac troponin I predicts myocardial dysfunction in aneurysmal subarachnoid hemorrhage. J Am Coll Cardiol 2000;36:1328. 22. Mayer SA, Lin J, Homma S, et al: Myocardial injury and left ventricular performance after subarachnoid hemorrhage. Stroke 1999;30:780. 23. Kono T, Morita H, Kuroiwa T, et al: Left ventricular wall motion abnormalities in patients with subarachnoid hemorrhage: Neurogenic stunned myocardium. J Am Coll Cardiol 1994;24:636. 24. Zaroff JG, Rordorf GA, Ogilvy CS, et al: Regional patterns of left ventricular systolic dysfunction after subarachnoid hemorrhage: Evidence for neurally mediated cardiac injury. J Am Soc Echocardiogr 2000;13:774. 25. Mayer SA, LiMandri G, Sherman D, et al: Electrocardiographic markers of abnormal left ventricular wall motion in acute subarachnoid hemorrhage. J Neurosurg 1995;83:889. 26. Doshi R, Neil-Dwyer G: A clinicopathological study of patients following a subarachnoid hemorrhage. J Neurosurg 1980;52:295. 27. Connor RC: Heart damage associated with intracranial lesions. BMJ 1968;3:29. 28. Yanowitz F, Preston JB, Abildskov JA: Functional distribution of right and left stellate innervation to the ventricles: Production of neurogenic electrocardiographic changes by unilateral alteration of sympathetic tone. Circulation Res 1966;18:416. 29. Klouda MA, Brynjolfsson G: Cardiotoxic effects of electrical stimulation of the stellate ganglia. Ann NY Acad Sci 1969;156:271. 30. Burch GE, Sun SC, Colcolough HL, et al: Acute myocardial lesions. Following experimentally induced intracranial hemorrhage in mice: A histological and histochemical study. Arch Pathol 1967;84:517. 31. Van Vliet PD, Burchell HB, Titus JL: Focal myocarditis associated with pheochromocytoma. N Engl J Med 1966;274:1102. 32. Melville KI, Blum B, Shister HE, et al: Cardiac ischemic changes and arrhythmias induced by hypothalamic stimulation. Am J Cardiol 1963;12:781. 33. Offerhaus L, van Gool J: Electrocardiographic changes and tissue catecholamines in experimental subarachnoid haemorrhage. Cardiovasc Res 1969;3:433.
Cardiac Care in Neurosurgery
551
34. Neil-Dwyer G, Walter P, Cruickshank JM, et al: Effect of propranolol and phentolamine on myocardial necrosis after subarachnoid haemorrhage. BMJ 1978;2:990. 35. Svigelj V, Grad A, Tekavcic I, et al: Cardiac arrhythmia associated with reversible damage to insula in a patient with subarachnoid hemorrhage. Stroke 1994;25:1053. 36. Oppenheimer SM, Gelb A, Girvin JP, et al: Cardiovascular effects of human insular cortex stimulation. Neurology 1992;42:1727. 37. Haley EC Jr, Kassell NF, Torner JC: The International Cooperative Study on the Timing of Aneurysm Surgery. The North American experience. Stroke 1992;23:205. 38. Archer DP, Shaw DA, Leblanc RL, et al: Haemodynamic considerations in the management of patients with subarachnoid haemorrhage. Can J Anaesth 1991;38:454. 39. Cummings RO (ed): ACLS Provider Manual 2001, American Heart Association, pp 1–252. 40. Sulek CA, Blas ML, Lobato EB: Milrinone increases middle cerebral artery blood flow velocity after cardiopulmonary bypass. J Cardiothorac Vasc Anesth 2002;16:64. 41. Apostolides PJ, Greene KA, Zabramski JM, et al: Intraaortic balloon pump counterpulsation in the management of concomitant cerebral vasospasm and cardiac failure after subarachnoid hemorrhage: Technical case report. Neurosurgery 1996;38:1056. 42. Eagle KA, Berger PB, Calkins H, et al: ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery-Executive Summary. A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Anesth Analg 2002;94:1052. 43. Davies KR, Gelb AW, Manninen PH, et al: Cardiac function in aneurysmal subarachnoid haemorrhage: A study of electrocardiographic and echocardiographic abnormalities. Br J Anaesth 1991; 67:58. 44. Mayer SA, Lin J, Homma S, et al: Myocardial injury and left ventricular performance after subarachnoid hemorrhage. Stroke 1999;30: 780. 45. Britton M, de Fairu, Helmers C, et al: Arrhythmias in patients with acute cerebrovascular disease. Acta Med Scand 1979;205:425. 46. Badner NH, Knill RL, Brown JE, et al: Myocardial infarction after noncardiac surgery. Anesthesiology 1998;88:572. 47. Ashton CM, Petersen NJ, Wray NP, et al: The incidence of perioperative myocardial infarction in men undergoing noncardiac surgery. Ann Intern Med 1993;18:504. 48. Tarhan S, Moffit EA, Taylor WF, et al: Myocardial infarction after general anesthesia. JAMA 1972;220:1451. 49. Mangano DT, Goldman L: Perioperative assessment of patients with known or suspected coronary disease. N Engl J Med 1995;333: 1750. 50. Goldman L, Caldera DL, Nussbaum SR, et al: Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med 1997; 845:1977. 51. Lee TH, Marcantonio ER, Mangione CM, et al: Derivation and prospective evaluation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999;100:1043. 52. Mangano DT, Browner WS, Hollenberg M, et al: Association of perioperative myocardial ischemia with cardiac morbidity and mortality in men undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. N Engl J Med 1990;323:1781. 53. Mangano DT, Wong MG, London MJ, et al: Perioperative myocardial ischemia in patients undergoing noncardiac surgery–II: Incidence and severity during the 1st week after surgery. The Study of Perioperative Ischemia (SPI) Research Group. J Am Coll Cardiol 1991;17:851. 54. Landesberg G, Luria MH, Cotev S, et al: Importance of long-duration postoperative ST-T depression in cardiac morbidity after vascular surgery. Lancet 1993;341:715.
W9418-17.qxd 10/31/03 3:46 PM Page 552
552
Section II
Neurointensive Care: General Issues
55. Mangano DT, Layug EL, Wallace A, et al: Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. Multicenter Study of Perioperative Ischemia Research Group. N Engl J Med 1996;335:1713. 56. Sprung J, Abdelmalak B, Gottlieb A, et al: Analysis of risk factors for myocardial infarction and cardiac mortality after major vascular surgery. Anesthesiology 2000;93:129. 57. Ellis J: Perioperative myocardial ischemia and infarction in noncardiac surgery. Curr Rev Clin Anesth 2002;22:233. 58. Mangano DT, Browner WS, Hollenberg M, et al: Long-term cardiac prognosis following noncardiac surgery. The Study of Perioperative Ischemia Research Group. JAMA 1992;268:233. 59. Skidmore KL, London MJ: Myocardial ischemia. Monitoring for myocardial ischemia: how do I monitor therapy? Anesthesiol Clin North Am 2001;19:651. 60. Massie BM, Botvinick EH, Brundage BH, et al: Relationship of regional myocardial perfusion to segmental wall motion. A physiological basis for understanding the presence of reversibility of asynergy. Circulation 1978;58:1154. 61. Shanewise JS, Cheung AT, Aronson S, et al: ASE/SCA guidelines for performing a comprehensive intraoperative multiplane transesophageal examination: Recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999;89:870. 62. Mayer SA, Lin J, Homma S, et al: Myocardial injury and left ventricular performance after subarachnoid hemorrhage. Stroke 1999; 30:780. 63. Parekh N, Venkatesh B, Cross D, et al: Cardiac troponin I predicts myocardial dysfunction in aneurysmal subarachnoid hemorrhage. J Am Coll Cardiol 2000;36:1328. 64. Charlson ME, MacKenzie CR, Gold J, et al: The postoperative electrocardiogram and creatine kinase: Implications for diagnosis of myocardial infarction after non-cardiac surgery. J Clin Epidemiol 1989;42:25. 65. Weitz HH: Perioperative cardiac complications. Med Clin North Am 2001;85:1151, vi. 66. Lenke LG, Bridwell KH, Jaffe AS: Increase in creatine kinase MB isoenzyme levels after spinal surgery. J Spinal Disord 1994;7:70. 67. Healy JH, Kagen LJ, Velis KP, et al: Creatine kinase MB in skeletal muscle and serum of spine-fusion patients. Clin Orthop 1985;195:282. 68. Mair J, Artner-Dworzak E, Lechleitner P, et al: Cardiac troponin T in diagnosis of acute myocardial infarction. Clin Chem 1991;37:845. 69. Adams JE III, Sicard GA, Allen BT, et al: Diagnosis of perioperative myocardial infarction with measurement of cardiac troponin I. N Engl J Med 1994;330:670. 70. Haggart PC, Adam DJ, Ludman PF, et al: Comparison of cardiac troponin I and creatine kinase ratios in the detection of myocardial injury after aortic surgery. Br J Surg 2001;88:1196. 71. Mangano DT, Browner WS, Hollenberg M, et al: Association of perioperative myocardial ischemia with cardiac morbidity and mortality in men undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. N Engl J Med 1990;323:1781. 72. Boersma E, Poldermans D, Bax JJ, et al: Predictors of cardiac events after major vascular surgery: Role of clinical characteristics, dobutamine echocardiography, and beta-blocker therapy. JAMA 2001; 285:1865. 73. Fleisher LA, Eagle KA: Lowering cardiac risk in noncardiac surgery. N Engl J Med 2001;345:1677. 74. Orlowski JP, Shiesley D, Vidt DG, et al: Labetalol to control blood pressure after cerebrovascular surgery. Crit Care Med 1988;16:765. 75. Poldermans D, Boersma E, Bax JJ, et al: The effect of bisoprolol on perioperative mortality and myocardial infarction in high-risk patients undergoing vascular surgery. Dutch Echocardiographic
76. 77.
78. 79.
80.
81.
82.
83.
84.
85.
86.
87. 88.
89.
90.
91.
92.
93. 94.
95. 96.
97.
Cardiac Risk Evaluation Applying Stress Echocardiography Study Group. N Engl J Med 1999;341:1789. Auerbach AD, Goldman L: Beta-blockers and reduction of cardiac events in noncardiac surgery: Scientific review. JAMA 2002;287:1435. Bayliff CD, Massel DR, Inculet RI, et al. Propranolol for the prevention of postoperative arrhythmias in general thoracic surgery. Ann Thorac Surg 1999;67:182. Fallowfield JM, Marolw HF: Propranolol is contraindicated in asthma. BMJ 1996;313:1486. Shammash JB, Trost JC, Gold JM, et al: Perioperative beta-blocker withdrawal and mortality in vascular surgical patients. Am Heart J 2001;141:148. Miller RR, Olson HG, Amsterdam EA, Mason DT: Propranolol-withdrawal rebound phenomenon: Exacerbation of coronary events after abrupt cessation of antianginal therapy. N Engl J Med 1975;293:416. Gabrielli A, Gallagher TJ, Caruso LJ, et al: Diltiazem to treat sinus tachycardia in critically ill patients: A four-year experience. Crit Care Med 2001;29:1874. Caramella JP, Goursot G, Carcone B, et al: Prevention of pre- and postoperative myocardial ischemia in non-cardiac surgery by intravenous diltiazem. Ann Fr Anesth Reanim 1988;7:245. Bedford RF, Dacey R, Winn HR, et al: Adverse impact of a calcium entry-blocker (verapamil) on intracranial pressure in patients with brain tumors. J Neurosurg 1983;59:800. A multicenter trial of the efficacy of nimodipine on outcome after severe head injury. European Study Group on Nimodipine in Severe Head Injury. J Neurosurg 1994;80:797. Feigin VL, Rinkel GL, Algra A, et al: Calcium antagonists in patients with aneurysmal subarachnoid hemorrhage: a systematic review. Neurology 1998;50:876. Karinen P, Kouvikangas P, Ohinmaa A, et al: Cost-effectiveness analysis of nimodipine treatment after aneurysmal subarachnoid hemorrhage and surgery. Neurosurgery 1999;45:780. Sleight P: Calcium antagonists during and after myocardial infarction. Drugs 1996;51:216. Talke P, Li J, Jain U, et al: Effects of perioperative dexmedetomidine infusion in patients undergoing vascular surgery. The Study of Perioperative Ischemia Group. Anesthesiology 1995;82:620. Quintin L, Boullioc X, Butin E, et al: Clonidine for major vascular surgery in hypertensive patients: A double-blind, placebo controlled, randomized study. Anesth Analg 1996;83:687. Oliver MF, Goldman L, Julian DG, et al: Effect of mivazerol on perioperative cardiac complications during non-cardiac surgery in patients with coronary heart disease: the European Mivazerol Trial (EMIT). Anesthesiology 1999;91:951. Landesberg G, Erel J, Anner H, et al: Perioperative myocardial ischemia in carotid endarterectomy under cervical plexus block and prophylactic nitroglycerin infusion. J Cardiothorac Vasc Anesth 1993;7:259. Dodds TM, Stone JG, Coromilas J, et al: Prophylactic nitroglycerin infusion during noncardiac surgery does not reduce perioperative ischemia. Anesth Analg 1993;76:705. Ohar JM, Fowler AA, Selhorst JB, et al: Intravenous nitroglycerininduced intracranial hypertension. Crit Care Med 1985;13:867. Dahl A, Russell D, Nyberg-Hansen R, et al: effect of nitroglycerin on cerebral circulation measured by transcranial Doppler and SPECT. Stroke 1989;20:1733. Lagerkranser M: Effects of nitroglycerin on intracranial pressure and cerebral blood flow. Acta Anaesthesiol Scand Suppl 1992;97:34. Yamamoto S, Nishizawa S, Yokoyama T, et al: Subarachnoid hemorrhage impairs cerebral blood flow response to nitric oxide but not to cyclic GMP in large cerebral arteries. Brain Res 1997; 757:1. Frank SM, Fleisher LA, Breslow M, et al: Perioperative maintenance of normothermia reduces the incidence of morbid cardiac events. JAMA 1997;277:1127.
W9418-17.qxd 10/31/03 3:46 PM Page 553
Chapter 17 98. Bealer SL: Acute hypertensive and natriuretic responses following preoptic hypothalamic lesions. Am J Med Sci 1988;295:346. 99. Clifton GL, Robertson CS, Kyper K, et al: Cardiovascular response to severe head injury. J Neurosurg 1983;59:447. 100. Robertson CS, Clifton GL, Taylor AA: Treatment of hypertension associated with head injury. J Neurosurg 1983;59:455. 101. Cutler NR, Sramek JJ, Luna A, et al: Effect of the ACE inhibitor ceronapril on cerebral blood flow in hypertensive patients. Ann Pharmacother 1996;30:578. 102. Demolis P, Chalon S, Annane D, et al: Effects of an angiotensin converting enzyme, ramipril, on intracranial circulation in healthy volunteers. Br J Clin Pharmacol 1992;34:224. 103. Jover BF, McGrath BP: Beneficial effects of fenoldopam on systemic and regional hemodynamics in rabbits with congestive heart failure. J Cardiovasc Pharmacol 1988;11:483. 104. Olesen J: The effect of intracarotid epinephrine, norepinephrine, and angiotensin on the regional cerebral blood flow in man. Neurology 1972;22:978. 105. Prielipp RC, Wall MH, Groban L, et al: Reduced regional and global cerebral blood flow during fenoldopam-induced hypotension in volunteers. Anesth Analg 2001;93:45. 106. Ryan TJ, Antman EM, Brooks NH, et al: 1999 update: ACC/AHA Guidelines for the Management of Patients with Acute Myocardial Infarction. Executive Summary and Recommendations: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). Circulation 1999;100:1016. 107. Hop JW, Rinkel GJ, Algra A, et al: Randomized pilot trial of postoperative aspirin in subarachnoid hemorrhage. Neurology 2000;54:872. 108. Calvin JE, Klein LW: The use of antiplatelet agents in acute cardiac care. Crit Care Clin 2001;17:365. 109. Konstantopoulos K, Mousa SA: Antiplatelet therapies: Platelet GPIIb/IIIa antagonists and beyond. Curr Opin Investig Drugs 2001;2:1086. 110. Pineo GF, Hull RD: Unfractionated and low-molecular-weight heparin. Comparisons and current recommendations. Med Clin North Am 1998;82:587. 111. Schaible KL, Smith LJ, Fessler RG, et al: Evaluation of the risks of anticoagulation therapy following experimental craniotomy in the rat. J Neurosurg 1985;63:959. 112. Lahoaprasit V, Mayberg MR: Risks of anticoagulation therapy after experimental corticectomy in the rat. Neurosurgery 1993;32:625. 113. Kaul S, Shah PK: Low molecular weight heparin in acute coronary syndromes: Evidence for superior or equivalent efficacy compared with unfractionated heparin? J Am Coll Cardiol 2000;35:1699. 114. Adams K Jr, Zannad F: Clinical definition and epidemiology of advanced heart failure. Am Heart J 1998;135:S204. 115. Pedersen T, Kelbaek H, Munck O: Cardiopulmonary complications in high-risk surgical patients: The value of preoperative radionuclide cardiography. Acta Anaesthesiol Scand 1990;34:183. 116. Halm EA, Browner WS, Tubau JF, et al: Echocardiography for assessing cardiac risk in patients having noncardiac surgery. Study of Perioperative Ischemia Research Group. Ann Intern Med 1996;125:433. 117. Hunt SA, Baker DW, Chin MH, et al: ACC/AHA Guidelines for the Evaluation and Management of Chronic Heart Failure in the Adult. Executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to revise the 1995 Guidelines for the Evaluation and Management of Heart Failure). J Am Coll Cardiol 2001;38:2101. 118. Cheng V, Kazanagra R, Garcia A, et al: A rapid bedside test for B-type peptide predicts treatment outcomes in patients admitted for decompensated heart failure: A pilot study. J Am Coll Cardiol 2001;37:386.
Cardiac Care in Neurosurgery
553
119. Lobato EB: Perioperative care of the patient with congestive heart failure. Curr Rev Clin Anesth 2000;21:25. 120. Ommen SR, Nishimura RA, Appleton CP, et al: Clinical utility of Doppler echocardiography and tissue Doppler imaging in the estimation of left ventricular filling pressures, A comparative simultaneous Doppler-catheterization study. Circulation 2000;102:1788. 121. Nohriah A, Tsang S, Dries DL, et al: Bedside assessment of hemodynamic profiles identifies prognostic groups in patients admitted with heart failure. J Card Fail 2000;6:64. 122. Nohriah A, Lewis E, Stevenson LW: Medical management of advanced heart failure. JAMA 2002;287:628. 123. Felker GM, O’Connor CM: Inotropic therapy for heart failure: An evidence-based approach. Am Heart J 2001;142:393. 124. Rettig GF, Schieffer HJ: Acute effects of intravenous milrinone in heart failure. Eur Heart J 1989;10:39. 125. Arakawa Y, Kikuta K, Hojo M, et al: Milrinone for the treatment of cerebral vasospasm after subarachnoid hemorrhage: Report of seven cases. Neurosurgery 2001;48:723. 126. Drexler H, Hoing S, Faude F, et al: Central and regional vascular hemodynamics following intravenous milrinone in the conscious rat: Comparison with dobutamine. J Cardiovasc Pharmacol 1987;9: 563. 127. Royster RL, Butterworth JF 4th, Prielipp RC, et al: Combined inotropic effects of amrinone and epinephrine after cardiopulmonary bypass in humans. Anesth Analg 1993;77:662. 128. Montessuit M, Chevalley C, King J, et al: The use of intraaortic counterpulsation balloon for the treatment of cerebral vasospasm and edema. Surgery 2000;127:230. 129. Carabello BA, Crawford FA Jr: Valvular heart disease. N Engl J Med 1997;337:32. 130. Raymer K, Yang H: Patients with aortic stenosis: Cardiac complications in non-cardiac surgery. Can J Anaesth 1998;45:855. 131. Torsher LC, Shub C, Rettke SR, et al: Risk of patients with severe aortic stenosis undergoing noncardiac surgery. Am J Cardiol 1998;81: 448. 132. Guidelines for the management of patients with valvular heart disease. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Committee on Management of Patients with Valvular Heart Disease. Circulation 1998;98:1949. 133. Carabello BA: Progress in mitral and aortic regurgitation. Prog Cardiovasc Dis 2001;43:457. 134. Tornons MP, Olona P, Permanyer-Miralda G, et al: Clinical outcome of severe asymptomatic chronic aortic regurgitation: A long-term prospective follow-up study. Am Heart J 1995;130:333. 135. Barrington WW, Boudoulas H, Bashore T, et al: Mitral stenosis: Mitral dome excursion at M1 and the mitral opening snap—the concept of reciprocal heart sounds. Am Heart J 1988;115:1280. 136. Delahaye F, Delaye J, Ecochard R, et al: Influence of associated valvular lesions on long-term prognosis of mitral stenosis: A 20-year follow-up of 202 patients. Eur Heart J 1991;12:77. 137. Ward C, Hancock BW: Extreme pulmonary hypertension caused by mitral valve disease: Natural history and results of surgery. Br Heart J 1975;37:74. 138. Carabello BA: Mitral regurgitation: Basic pathophysiologic principles. Mod Concepts Cardiovasc Dis 1988;57:53. 139. Boltwood CM, Tei C, Wong M, et al: Quantitative echocardiography of the mitral complex in dilated cardiomyopathy: The mechanism of functional mitral regurgitation. Circulation 1983;68:498. 140. Schreiber TL, Fisher J, Mangla A, et al: Severe “silent” mitral regurgitation: A potentially reversible cause of refractory heart failure. Chest 1989;96:242.
W9418-18.qxd 10/31/03 3:46 PM Page 555
Chapter 18 Body Water and Electrolytes Alejandro A. Rabinstein, MD and Eelco F.M. Wijdicks, MD
The controlling mechanisms of homeostasis may become disturbed by acute injury to the brain, craniotomy, or a systemic complication. Perturbations of volume and electrolyte balance may develop rapidly and, if not adequately corrected, they may become a grave clinical problem that is difficult to solve. Most of the time, critical neurologic or neurosurgical patients are particularly prone to develop abnormalities in conservation of sodium and water. Other electrolyte abnormalities are germane to intensive care unit (ICU) stay and may be the result of certain necessary actions (e.g., intravenous fluid administration and drugs). When unrecognized and persistently severe, the potential consequences may be serious, even life threatening, when cardiac arrhythmias occur. Throughout this chapter, we will present general principles that apply to the care of any ICU patient regarding the diagnosis and management of fluid and electrolyte disorders. Additionally, we will address in detail the physiologic concepts that are most necessary for an in-depth understanding of frequently encountered clinical problems in the neurointensive care unit (neuro-ICU). Each section will start with a brief description of the basic physiologic elements required to maintain homeostasis, followed by a discussion of derangements (excess or deficiency) and their proper management illustrated with clinical examples. When appropriate, the consequences of mismanagement will be discussed. Finally, the chapter closes with a summary of essential clinical messages.
Disorders of Sodium and Water Homeostasis Before embarking on the discussion of the disorders and their clinical consequences, we will summarize the principles of homeostasis of body water. To begin with, two thirds of the total fluid content of the human body resides inside cells, while the remaining one third resides in the extracellular space. This extracellular space is further divided into two compartments: the vascular compartment, containing the plasma fluid that represents one-third of the total extracellular fluid volume, and the interstitial compartment, with the other two-thirds of the extracellular volume (Fig. 18-1). Osmotic forces exerted across the cellular membrane regulate the movement of water between the intracellular and the extracellular fluid compartments. These forces (osmotic activity) are determined by the concentration of solute particles in the plasma. The osmotic activity of the solutes in our body fluids is usually expressed in relation to the volume of water in which they are present, measured as serum osmolality and expressed as mOsm/kg H2O. In normal persons, serum osmolality can be estimated using the following formula: Serum osmolality (in mOsm kg H2 O) = 2(Na) + BUN 2.8 + glucose 18 where BUN represents blood urea nitrogen and both BUN and glucose are expressed in milligrams per deciliter. 555
W9418-18.qxd 10/31/03 3:46 PM Page 556
556
Section II
Neurointensive Care: General Issues
Figure 18-1. Distribution of total body fluid in the human organism. Two-thirds of total body fluid is intracellular. Of the remaining 1/3 (extracellular), 1/3 is intravascular and 2/3 is interstitial.
Not all solutes contribute to the movement of water across the cellular membrane to the same degree. Obviously, only solutes that are unable to freely cross the membrane are capable of producing an osmotic gradient and influence the distribution of water in the two compartments. This relative
osmotic activity in two solutions (in this case the intracellular and extracellular fluids) separated by a semipermeable membrane (the cellular membrane) is called effective osmolality or tonicity.1 This concept becomes clear when comparing sodium and urea (Fig. 18-2). Sodium is a solute that cannot move freely across the cellular membrane, thus increasing tonicity in the compartment where it is added. Urea, instead, can diffuse freely across the membrane and equilibrate its concentrations in both compartments, creating no net change in the tonicity of the solutions. Hence, both hypernatremia and azotemia are hyperosmotic conditions, but only hypernatremia represents a hypertonic state and leads to a change in the distribution of body water. The regulation of total body water is based primarily on a sensing system designed to respond to plasma osmolality and keep it relatively constant. As alluded earlier, plasma osmolality is determined by the total solute content in the plasma and the total plasma water volume. In normal conditions, the primary determinant of plasma osmolality (and most importantly, plasma effective osmolality or tonicity) is the plasma sodium concentration. And since effective osmolality determines the tendency of water to move across the cellular membrane, plasma sodium concentration is the principal determinant of the relative volumes of the intracellular and extracellular fluids.2 While the mechanisms of sodium conservation aim at maintaining a relatively constant effective circulating volume, the goal of those mediating (free) water balance is to maintain a relatively constant plasma tonicity. Baroreceptors sense intravascular volume status and regulate sodium conservation and excretion, while osmoreceptors sense plasma tonicity and regulate water intake and loss. Hypervolemia and hypovolemia are disorders of sodium balance; in contrast, hypotonic and hypertonic states are disorders of water balance (Fig. 18-3). Accurate assessment of total body sodium relies on a careful physical examination (with the extreme situations being hypovolemic shock and hypervolemic edematous states) and does not require any laboratory measurement. Meanwhile, water balance is best represented by serum sodium concentration (hyponatremia or hypernatremia). Thus, it is important to recognize that serum sodium concentration depends on water balance and not on total body sodium content.
Mechanisms of Water Balance (Regulation of Plasma Tonicity) Figure 18-2. Differences in contribution of sodium and urea to extracellular tonicity. The cellular membrane is relatively impermeable to sodium, thus, addition of sodium to the extracellular compartment will increase extracellular tonicity. Conversely, the cellular membrane allows free passage of urea resulting in rapid equilibration of urea concentrations between the intracellular and extracellular compartments and no net change of extracellular tonicity.
Plasma tonicity is regulated by thirst and the ability of the kidney tubules to conserve or excrete water. In addition, there are normal obligate water losses through the skin and lungs (so called “insensible losses”) and measurable minor losses through the gastrointestinal tract (when added to renal excretion they account for the “sensible losses”). A 70kg person loses a minimum of 1.5 L of water every day,
W9418-18.qxd 10/31/03 3:46 PM Page 557
Chapter 18
Figure 18-3. Differences between total body sodium and total body water. Changes in total body sodium lead to alterations in extracellular fluid volume (including blood volume) and are sensed by baroreceptors, whereas changes in total body water result in alterations of extracellular tonicity and are sensed by osmoreceptors.
and this loss needs to be offset by an intake of at least the same volume of water to avoid dehydration and plasma hypertonicity. Body tonicity is tightly regulated with a normal set point between 280 and 290 mOsm/kg. This tight control is accomplished by the action of osmoreceptors present in the brain that regulate the mechanism of thirst and modulate the secretion of antidiuretic hormone (ADH), both functions of the hypothalamus. ADH-responsive cells are present along the collecting duct. When ADH binds to its receptor, the permeability of the cells to water increases by the activation of water channel proteins known as aquaporin 2. Cell volume depends on the tonicity of the extracellular compartment. Extracellular hypotonicity leads to cell swelling, and cells must eliminate low-molecular-weight solutes (known as osmolytes) to recover their normal volume. Conversely, extracellular hypertonicity produces cell shrinking and subsequent compensatory accumulation of osmolytes to achieve restoration of volume. There are two types of osmolytes: inorganic (such as sodium, potassium and chloride) and organic (including amino acids, methylamines, myoinositol, and sorbitol, among others). Inorganic osmolytes can be shifted in and out of the cells rapidly in response to changes in tonicity gradients. But changes in the concentration of organic osmolytes take longer because the process implies genetic upregulation or downregulation of their synthesis and uptake.
Mechanisms of Total Body Sodium Balance (Regulation of Effective Circulating Volume) Multiple mechanisms contribute to regulate total body sodium balance. The occurrence of either true or perceived hypovolemia, caused by low extracellular fluid volume
Body Water and Electrolytes
557
or low tissue perfusion states such as cardiac failure, respectively, triggers a response by a series of vasopressor hormones. These include ADH ([antidiuretic hormone] arginine-vasopressin), renin, and norepinephrine. ADH increases water and secondarily sodium reabsorption by renal collecting ducts, and levels of ADH are elevated in hypovolemic states regardless of plasma tonicity. Renin leads to the activation of angiotensin II, and this hormone increases sodium by directly enhancing its reabsorption by the proximal tubule cells, stimulating the secretion of aldosterone (that, in turn, increases reabsorption of sodium by the distal tubule and collecting duct), and reducing the glomerular filtration rate thereby decreasing sodium delivery to the tubules. Norepinephrine, and epinephrine released by the adrenal medulla, also decrease glomerular filtration rate and stimulate sodium reabsorption at the proximal tubule level. (Fig. 18-4). In contrast, the release of these vasopressor hormones is suppressed in patients with hypervolemic states. In those situations, activation of a family of natriuretic peptides promotes sodium secretion. There are four recognized members of this family: atrial natriuretic peptide, brain natriuretic peptide, C-type natriuretic peptide, and the very recently discovered Dendraspis natriuretic peptide (DNP).3,4 Atrial natriuretic peptide is produced primarily by the heart atria, while brain natriuretic peptide predominates in the ventricles. In addition, all three peptides are produced in the brain, particularly the C-type, and have central as well as peripheral actions. In essence, these substances exert potent natriuretic, diuretic, and vasorelaxant activities. Currently available data suggest they do so by directly acting on renal tubules, increasing glomerular filtration rate, antagonizing the renal effects of ADH, suppressing the renin-angiotensin II-aldosterone axis, reducing sympathetic tone and the peripheral
Figure 18-4. Schematic summary of the physiologic response to hypovolemia. Increased release of “pressor hormones” including antidiuretic hormone (ADH), the members of the renin-angiotensin II-aldosterone (R/AGII/Ald) axis, and norepinephrine (NE) result in water and sodium conservation.
W9418-18.qxd 10/31/03 3:46 PM Page 558
558
Section II
Neurointensive Care: General Issues
Figure 18-5. Schematic summary of the physiologic response to hypervolemia. Activation of natriuretic peptides and inhibition of “pressor hormones” combine with peritubular changes to promote natriuresis.
release of catecholamines, and centrally inhibiting salt appetite and thirst. Other effects of hypervolemia that contribute to augment sodium excretion include a fall in plasma colloid osmotic pressure and an increase in interstitial hydraulic pressure; both favoring a lower uptake of sodium and water by the peritubular capillaries. This eventually results in increased interstitial hydrostatic pressures, increased backflow of fluid into the proximal tubule cells, and decreased reabsorption of water and sodium by those cells (Fig. 18-5). The contribution of a different type of natriuretic substance known as ouabain-like factor (OLF), a glycoside acting by inhibition of the Na+-K+ ATPase, is less well established, and interest in this poorly characterized substance has waned after the discovery of the family of natriuretic peptides.
Hypernatremia Hypernatremia is defined as serum sodium concentration exceeding 145 mmol/L. It always represents a deficit of water in relation to the total body sodium stores. This can occur as a consequence of net water loss or a hypertonic sodium gain. The presence of hypernatremia does not signify that there is an excess of sodium content in the body. Indeed, sodium stores may be either normal (in cases of pure water loss) or decreased (in cases of hypotonic fluid loss, such as vomiting or diarrhea).5 Sustained hypernatremia can only be perpetuated when thirst mechanisms are impaired or there is an inability to access water. Critical care patients frequently meet these conditions, especially those that are elderly or have altered states of consciousness. Patients presenting with altered consciousness may have been unable to access water for a pro-
longed period of time, which may be unknown at the time of first evaluation. Hypodipsia may be part of a variety of neurologic and neurosurgical entities. Excessive diaphoresis from agitation or fever can further exacerbate hypotonic fluid losses. The effects of hypernatremia on brain tissue activate a series of compensatory responses aimed at restoring brain volume. These involve the generation of osmotically active substances known as idiogenic osmoles or organic brain osmoles.6,7 Normalization of brain volume ensues, but this new equilibrium counts on the persistence of hyperosmolality. Rapid correction of hyperosmolality (especially when chronic) can render the brain incapable of eliminating the accumulated osmoles promptly enough, resulting in cerebral edema, generalized tonic-clonic seizures, coma, and death (Fig. 18-6).8 Careful assessment of the extracellular fluid volume should be the initial step in the evaluation of hypernatremia. Attention should be focused on recent changes in body weight, presence of peripheral edema, pulse rate, orthostatic changes in blood pressure, hourly urine output, and presence of gallop or jugular venous distention. A random (spot) urine sample can be used to measure sodium concentration (decreased extracellular fluid volume is suggested by a concentration lower than 10 mmol/L), and urine specific gravity (a contracted extracellular volume can be suspected when the urine is very concentrated). A fractional excretion of sodium lower than 1% is suggestive of volume depletion in nonedematous patients. Hypernatremia with low extracellular fluid volume implies a loss of hypotonic fluids. Excessive diuresis, vomiting, and diarrhea are the usual suspects in those situations. Hypernatremia with normal extracellular fluid volume indicates a loss of free water. Diabetes insipidus should be excluded and the appropriateness of the tonicity of fluids used for ongoing replacement reassessed. Hypernatremia with high extracellular fluid volume results from a gain of hypertonic fluids. It is usually iatrogenic, secondary to the use of hypertonic NaCl or NaHCO3 solutions, or hypertonic feeding; but it can also occur in cases of mineralocorticoid excess (Table 18-1). Clinical Manifestations of Hypernatremia Hypernatremia usually becomes symptomatic only when serum sodium concentration exceeds 160 mmol/L9,10 but the rapidity of the elevation of serum sodium levels is equally important; with rapid elevation, stupor emerges more frequently. The signs and symptoms of hypernatremia mostly reflect central nervous system dysfunction caused by the reduction of intracellular fluid volume in the brain. The clinical manifestations may go undetected in patients with other more obvious causes for the neurologic symptoms, such as extensive and prolonged brain injury. Decreased level of consciousness and confusion are the most
W9418-18.qxd 10/31/03 3:46 PM Page 559
Chapter 18
Body Water and Electrolytes
559
Figure 18-6. Effects of hypernatremia on the brain, adaptive responses, and the potential for iatrogenic damage. Within minutes after development of hypertonicity, loss of brain water results in a reduction in brain volume. Restitution of brain volume commences within a few hours with the entry and accumulation of electrolytes into brain cells (rapid adaptation). Volume normalization is then completed within several days by intracellular accumulation of organic osmolytes (slow adaptation). At that point, brain volume is normal despite persistent extracellular hypertonicity. Slow correction of the hypernatremia permits elimination of accumulated electrolytes and osmolytes from brain cells, thus uneventfully reestablishing the normal equilibrium. In contrast, rapid correction (exceeding 10 mmol/day) renders the brain incapable of eliminating the accumulated osmotically active substances. This leads to water uptake and may result in cerebral edema. (Reproduced with permission of the Massachusetts Medical Society from Adrogué HJ, Madias NE: Hypernatremia. N Engl J Med 2000;342:1493.)
Table 18-1 Common Causes of Hypernatremia in the Neurointensive Care Unit With hypovolemia Gastrointestinal losses (diarrhea, vomiting, nasogastric suctioning) Excessive insensible losses (tachypnea) Diabetes insipidus with insufficient fluid replacement Excessive diuresis (mannitol, loop diuretics) With normovolemia Diabetes insipidus with sufficient fluid replacement (reassess tonicity of replacement solution) With hypervolemia Iatrogenic (hypertonic sodium solutions, inadequate feeding preparations) Corticosteroid excess
common manifestations, and the level of consciousness tends to correlate with the severity of hypernatremia.11,12 Generalized tonic-clonic seizures can occur,13 but rapid progression to deep levels of coma is rare, and should prompt the search for other potential causes using brain imaging and additional laboratory tests. The hypothesis that brain shrinkage caused by the effects of hypernatremia could result in intracranial hemorrhages in pediatric patients has been proposed but never systematically studied,14 nor ever published in adults. Intense thirst may be present initially, but tends to gradually disappear with progression of the disorder. Absence of thirst in a patient with pronounced hypernatremia and preserved level of consciousness should raise suspicion for hypothalamic dysfunction. Finally, neuromuscular manifestations are uncommon, but rhabdomyolysis has been reported in a series of anecdotal cases.15,16
W9418-18.qxd 10/31/03 3:46 PM Page 560
560
Section II
Neurointensive Care: General Issues
Common Causes of Hypernatremia in the Neurointensive Care Unit Hypernatremia in the neuro-ICU is most commonly due to net free water loss, often related to a defective central release of ADH, or induced osmotic diuresis. Acute severe hypernatremia is most often iatrogenic. It may be produced by the intravenous administration of hypertonic saline solutions (e.g., 1.5% or 3% NaCl given as treatment of increased intracranial pressure, or as hypervolemic therapy in patients with subarachnoid hemorrhage and symptomatic vasospasm), or sodium bicarbonate (for correction of lactic acidosis). Less commonly, acute hypernatremia can be caused by nutritional formulas or repeated use of hypertonic saline enemas. Induced Osmotic Diuresis Although to a lesser degree than hypertonic saline infusion, mannitol is expected to cause hypernatremia from free water loss. This is a predicted and calculated risk that is acceptable as long as the indication for the use of osmotic agents is clear. Serum osmolality must be monitored and should not exceed 320 mOsm/kg because this dehydrated state correlates with a risk of renal failure.17 Tromethamine (THAM), an alkalinizing agent, can represent a useful therapeutic alternative to mannitol in patients with increased intracranial pressure and hypernatremia.18 Diabetes Insipidus Theoretically, patients with head injury, primary or metastatic brain neoplasms, global anoxic brain damage, meningitis or encephalitis, massive cerebral edema, or rapid diencephalic herniation are at increased risk for the development of diabetes insipidus.19 However, diabetes insipidus is mostly encountered after pituitary or diencephalic surgery (e.g., removal of a hypothalamic hamartoma).20 It is almost an obligatory feature in patients fulfilling brain death criteria. Diabetes insipidus is diagnosed in patients with hypotonic urine (urine osmolality 13 Hz), alpha (8 to 13 Hz), theta (4 to 8 Hz), and delta (50% decrease 4+: >100% increase 1-: incompatible with monitoring D, Dose-related; EMG, electromyography.
10 mseconds of the stimulus presentation (see Fig. 21-13). Midlatency and cognitive auditory-evoked potentials are recorded predominantly over the temporal cortex. In an ideal situation, recording electrodes should be placed on the skin overlying each earlobe, the seventh cervical vertebra, and the vertex. A typical montage would be • • • •
Channel 1: Cz—Ipsilateral ear Channel 2: Cz—Contralateral ear Channel 3: Cz—Seventh cervical vertebra Channel 4: Ipsilateral ear—contralateral ear
The channel likely to provide the most information, showing all waves, is Channel 1 (Cz—ipsilateral ear). The Cz—contralateral ear shows the responses from the auditory pathway that has crossed to the opposite side of the brainstem and may help identify wave V. Channel 3 will often also empha-
size wave V of the response and may be used when wave V is only poorly seen in Channel 1. Occasionally, wave I will not be well visualized in Channel 1 and may be better seen in Channel 4. Wave I must be visualized adequately to make certain that a stimulus is actually reaching the cochlea. If wave I is seen clearly, changes in the auditory evoked potential are unlikely to be caused by failure of adequate stimulation. Auditory Pathway. Initially, investigators thought that each component of the BAEP represented the activity of a distinct structure along the brainstem auditory pathway. The sites of origin of each wave were thought to be: wave I, eighth cranial nerve; wave II, cochlear nucleus; wave III, superior olivary complex; wave IV, lateral lemniscus; wave V, inferior colliculus; wave VI, medial geniculate; and wave VII, primary audi-
Figure 21-13. BAEP waveform and associated structures that are thought to generate each wave.
W9418-21.qxd 10/31/03 3:49 PM Page 633
Chapter 21
tory cortex. This schema seemed to hold up rather well when abnormalities of the potentials were compared with locations of clinical lesions.52,53 For example, lesions of the midbrain were reported to correlate well with abnormalities of waves IV and V; lesions in the pons altered wave III. Lesions of the eighth cranial nerve itself were associated with loss of the entire potential or prolongation of the latency of all components after wave I. Unfortunately, the types of lesions encountered in clinical situations usually involve more than one auditory structure and may also have remote effects via pressure or edema. Limitations in clinical correlation between anatomic lesions and BAEP abnormalities have led to both human and animal experimentation. Animal experimentation where recordings were made directly from brainstem structures suggests that, during most of the evoked potential, high-amplitude fields occur throughout the brainstem. This observation led to the conclusion that most of the BAEP waves are generated in multiple auditory structures. Many different discrete brainstem lesions in animals decrease the amplitude of but have no effect on the latency of the BAEP waveforms. Only with lesions of the eighth cranial nerve were all BAEP components suppressed. Large lesions elsewhere generally led to attenuation, but not loss, of these components. The BAEP waveforms beyond wave I thus likely represent a highly complex interaction between multiple structures (see Fig. 21-13).54–58 Several studies have involved recording from human brainstem structures during posterior fossa explorations. The compound eighth cranial nerve action potential latency recorded from the eighth cranial nerve just proximal to the internal auditory meatus occurred midway between surfacerecorded waves II and III.59 This finding indicated that the eighth cranial nerve plays a major role in the generation of both waves I and II. Potentials recorded near the cochlear nucleus have corresponded to the surface-recorded wave III.60 Although the precise generators of the various components of the BAEPs remain controversial, most investigators agree that longer latency components are generated in structures located progressively more rostrally along the brainstem auditory pathway. This schema continues to have great clinical utility. BAEPs assume an adult configuration in early childhood.61 Many factors will affect their latencies and amplitudes, including head size, body temperature, anesthetic drugs, and, especially important in the neuro-ICU, auditory acuity. Interpeak latencies (usually I to III and I to V), which tend to be less affected by these factors than are absolute latencies, are therefore the preferred measures of normality. Compared to somatosensory- or visual-evoked potentials, systemic factors are relatively unlikely to seriously alter the BAEP. Intravenous anesthetic agents, even in very high doses, have no significant effects on the BAEP. Middle latency- and cognitive auditory-evoked responses on the other hand are exquisitely sensitive to systemic factors and anesthetic drugs. In fact, dose-dependent changes in middle
Neurologic Monitoring in the Neurointensive Care Unit
633
latency auditory-evoked potentials are being used as a measure of anesthetic depth. Visual-Evoked Potentials Visual-evoked potentials (VEPs) reflect the function of the visual pathway, which extends from the optic nerve, through the chiasm, to the visual cortex. The VEP is generated primarily by the visual cortex. The stimulus may consist of a repetitive bright flash of light or of a reversing black-andwhite checkerboard pattern. It is applied by goggles through closed eyelids or by contact lenses containing light-emitting diodes applied directly to the cornea. Recordings are from scalp electrodes placed over the calcarine cortex. The timing of peak I of the VEP coincides with the electroretinogram and confirms that the sensory input was processed in the eye. The electroretinogram or peak I, however, does not confirm input of the signal into the CNS, because conduction of the stimulus along an intact optic nerve is still required. Motor-Evoked Potentials Interest in the use of MEPs stems from the importance of motor function for the patient’s ultimate functional status and from the theoretical limitations of SSEPs in monitoring motor function. Although MEPs are not technically difficult to produce in the conscious patient, the lack of widespread use of MEPs in the OR and neuro-ICU suggests that their use may not be as simple as it seems. Theoretical Basis. The same type of system that is used to
monitor SEPs may also be used to monitor the motor system. The stimulus to the motor system is either electrical or magnetic in nature. Both of these types of stimuli have been used at several different stimulus sites: the motor cortex, the spinal cord (any level), or the peripheral nerve. MEPs recorded from the peripheral nerves and spinal cord are called neurogenic MEPs. Their application is limited to intraoperative use. These potentials are relatively small and usually require some signal averaging to be readily seen. MEPs recorded from muscle are called myogenic MEPs. These potentials are much larger and frequently do not require averaging. The reproducibility of MEPs may be increased by techniques such as facilitation by peripheral sensory stimulation and temporal summation of responses by short interval sequential stimulation.62 The use of transcranial magnetic stimulation to elicit myogenic MEPs was recently approved by the U.S. Food and Drug Administration.63 Use of MEPs to assess motor function may be difficult for several reasons. First, regardless of whether the stimulus is applied transcranially to the motor cortex or directly to the spinal cord, there is an uncertainty as to which structures are being activated by the stimulus and which structures are producing the recorded waveforms of neurogenic MEPs.64–71 While part of the stimulus is conducted via the corticospinal tract, other descending tracts are also involved.65,72 MEPs are
W9418-21.qxd 10/31/03 3:49 PM Page 634
634
Section II
Neurointensive Care: General Issues
exquisitely sensitive to many anesthetic and sedative drugs. Barbiturates, benzodiazepines, and volatile agents abolish MEPs, whereas low doses of propofol and narcotics appear to interfere less. Muscle relaxants may be used in amounts that will suppress most gross motor movement associated with stimulation provided quantitative neuromuscular blockade monitoring is used. Simple peripheral nerve stimulators are not suitable for MEP monitoring because they are unable to accurately quantify either the amount of stimulus given to the nerve or the amount of muscular response produced.
Applications of Evoked Potentials in the Neurointensive Care Unit Assessment of Motor and Somatosensory Function One obvious area of utility for EP testing is the assessment of the integrity of a pathway by the appropriate EP modality. This application represents a simple extension of the diagnostic use of EPs, but may be very helpful in the neuroICU to clarify a clinical picture at the bedside or at a time when imaging findings are still equivocal. In this respect, MEPs have been used successfully in patients with strokes or trauma to the brainstem to predict ultimate motor function.73–75 Cognitive VEPs may be useful to differentiate the de-efferented state of a locked-in syndrome from psychogenic unresponsiveness or isolated brainstem lesions.76 Absence of cortical SSEPs may signify a lesion high in the cervical spinal cord or at the craniocervical junction and thus explain the lack of reactivity to stimuli applied in the territory of the peripheral nervous system. BAEP wave I may be used to test for the presence of hearing. Absence of wave I, on the other hand, may be difficult to interpret outside of the clinical context as it may be due to such diverse etiologies as a preexisting hearing loss, a transverse petrosal fracture, the effect of ototoxic drugs, isolated damage to the auditory pathways in the brainstem, or brain death. Continuous EP monitoring expands the diagnostic possibilities further. If the patient’s condition or pathology puts a pathway that is amenable to EP monitoring at risk, continuous EP monitoring represents a unique modality to receive early warning and positively affect outcome. Likewise, continuous EPs can document the neurologic consequences of increased intracranial pressure. Coma Evaluation SSEPs combined with BAEPs are helpful in the treatment of comatose patients.77–85 In general, if both BAEPs and cortical SSEPs are intact at presentation and remain intact, ultimate outcome is good. A relatively good outcome may occur in this case even if all clinical signs indicate a very poor prognosis.78 If the cortical SSEPs are absent at presentation and
the BAEPs are present, the best outcome expected is a chronic vegetative state. If both cortical SSEPs and all BAEP waves beyond wave I are absent, brain death is very likely. It is important to note that drug overdose will not eliminate either the BAEP or the early and intermediate latency components of the SSEP. While the EEG may be entirely absent in the case of drug overdose or therapy such as in barbiturate coma, the BAEP and SSEP should be present if the patient has brain function. Slight differences in sensitivity, specificity, and predictive value of EPs exist that depend on the etiology of the comatose state. In anoxic-ischemic coma, absence of cortical SSEPs 24 hours after the precipitating event was found to be the best method to predict poor outcome.5 Similarly, bilateral loss or absence of cortical SSEPs is always associated with a poor outcome in comatose patients whose EEG reveal alpha, theta, or alpha-theta coma.76 Conversely, presence of cortical SSEPs is associated with a favorable outcome. AEPs are less useful in anoxic-ischemic coma. Brainstem responses may initially be absent, due to cochlear ischemia, but are otherwise only affected very late in the course of anoxia or ischemia. Presence of midlatency or late auditory potentials is predictive of a good outcome, but is subject to the same modulating influences as the EEG. In coma due to head trauma, both the cerebral hemispheres and the brainstem may be involved in a pattern of lesions that reflects more the mechanism of injury and less the intrinsic tolerance to hypoxia/ischemia of a given brain structure. Presence of cortical EPs such as the N20 of the median SSEP or midlatency AEPs is still associated with favorable outcomes even if the latency of the peaks is increased.86 Conversely, absence of cortical peaks and progressive rostro-caudal deterioration of BAEPs, as occurs with transtentorial herniation, leads to brain death.87 The decreased predictive power of absent cortical responses in post-traumatic coma is demonstrated by the fact that all cases of good clinical outcomes despite absent cortical SSEPs stem from such trauma patients.88,89 Confirmation of Brain Death The declaration of brain death is usually based on clinical criteria or radiographic evidence of absence of cerebral blood flow. In special clinical circumstances, neurophysiologic testing can be supportive of this diagnosis.83,84 The principal advantage of EP testing lies in the fact that EP signals from the brainstem are virtually unaffected by CNS depressant drugs. For BAEPs to support the diagnosis of brain death, wave I must be evident. Unfortunately, wave I is frequently absent in brain death.90 Although it is generated in the peripheral portion of the auditory nerve, the blood supply to the nerve and to the cochlea itself often has an intracranial origin. Thus, increased intracranial pressure, with resultant decreases in blood flow, can lead to cochlear damage and loss of wave I. If no components of the BAEP are evident, it is consistent with but not diagnostic of brain death. SSEP testing is faced with the opposite problem.
W9418-21.qxd 10/31/03 3:49 PM Page 635
Chapter 21
Although documentation of the peripheral input to the brainstem is easily achieved, the evoked response from the cervico-medullary junction persists in the majority of brain dead patients, because it originates at the watershed between intracranial and extracranial blood supply.91
Monitoring Cerebral Blood Flow All of the previously discussed monitors are monitors of nervous system function. When function fails, they do not necessarily give information about the mechanism of nervous system damage. One of the most common mechanisms of CNS damage is inadequate blood flow. The remainder of this chapter will examine those methods that are available to monitor the adequacy of CBF. These monitors (Table 21-7) provide information that is complementary to the functional assessment discussed above, because function only becomes altered when CBF decreases by more than half. The most common clinical measure aimed at assuring adequate CBF is to maintain the cerebral perfusion pressure above the lower limit of cerebral autoregulation. While a markedly decreased or elevated cerebral perfusion pressure may lead to ischemia or spontaneous hemorrhage, respectively, a normal cerebral perfusion pressure by no means
Neurologic Monitoring in the Neurointensive Care Unit
635
assures normal CBF. For example, increased cerebrovascular resistance because of carotid stenosis, cerebral vasospasm, or microcirculatory compromise may cause ischemia, despite normal cerebral perfusion pressure. Similarly, normal cerebral perfusion pressure may coexist with abnormally increased CBF in settings such as post-traumatic vasoparalysis or normal perfusion pressure breakthrough after resection of an arteriovenous malformation. Direct Cerebral Blood Flow Measurement The ideal clinical method for CBF measurement in the neuro-ICU should be a noninvasive inexpensive bedside procedure that is continuous, or at least frequently repeatable, and provides good spatial resolution for superficial and deep structures of all vascular territories.92 No currently available method comes close to having all these characteristics. Nonetheless, determinations of CBF served to validate other techniques of assessing cerebral perfusion and provided important insights into the pathophysiologic events in head injury or stroke. Direct measurement of CBF is possible by determining kinetics of either wash-in or wash-out of an inert tracer compound, in a variation of the method of originally described by Kety and Schmidt.93 The most widely used meas-
Table 21-7 Techniques for Measuring Cerebral Blood Flow Resolution Category
Technique
Temporal
Spatial
Indirect
Neurologic examination Electroencephalogram/ evoked potentials Cerebral perfusion pressure Kety-Schmidt 133 Xe wash-out AVDO2, Jugular venous oxygen saturation (SjvO2) Double indicator dilution
>3 min 1–3 min SNP).81 Co-existing diseases may dictate the use of a particular agent. For example, use of NTG for the patient with ischemic cardiac disease may improve the endocardial-epicardial flow ratio without inducing coronary steal, and thus may improve
W9418-22.qxd 11/1/03 12:06 PM Page 675
Chapter 22
Pharmacotherapy in the Neurosurgical Intensive Care Unit
myocardial ischemia. Infusions generally have less marked effects on ICP than boluses and may be preferable. Because both agents are widely used in the neurosurgical intensive care unit setting, they will be reviewed in detail. While positive inotropes are useful in improving the contractile state of the failing ventricle to improve cardiac output and arterial blood pressure, vasodilators can improve ventricular function by optimizing loading conditions. Severe heart failure is characterized by both poor left ventricular function (decreased cardiac output with elevated filling pressure) and elevated systemic vascular resistance, and added therapeutic benefit may be obtained by treating both abnormalities. Thus, the effects of inotropes and vasodilators are additive in augmenting cardiac output and decreasing filling pressures in the treatment of severe left ventricular failure. Sodium nitroprusside and nitroglycerin are the most useful vasodilators because of their potency, rapid onset and short duration of action. These properties make them ideal for rapid titration of arterial pressure. Other vasodilators that are used occasionally but which have slower onsets and longer durations of action include hydralazine and phentolamine (an a1-adrenergic antagonist). Sodium nitroprusside is a potent arterial vasodilator as well as venodilator, while nitroglycerin is a potent venodilator but is less potent as an arterial vasodilator. This is clinically evident in the greater efficacy of sodium nitroprusside in reducing arterial pressure in normovolemic patients, while both agents are effective in reducing ventricular filling pressures. Nitroglycerin, and to a lesser extent sodium nitroprusside, are also effective pulmonary vasodilators and inhibitors of hypoxic pulmonary vasoconstriction (hence, the increased pulmonary shunt frequently observed with their use). Other pulmonary vasodilators useful for the acute control if right ventricular afterload include prostaglandin E1, prostacyclin, and inhaled nitric oxide. Nitroglycerin also exhibits utility in treating myocardial ischemia. In the perioperative period, sodium nitroprusside (0.5 to 10 mg/kg/minute) and nitroglycerin (0.5 to 4 mg/kg/minute) are useful in treating ventricular dysfunction with elevated filling pressures, pulmonary hypertension, and systemic hypertension. Sodium nitroprusside appears to be particularly effective in treating postoperative hypertension in patients with hyperdynamic circulation and in reducing afterload and preload in patients with poor ventricular function and acceptable arterial blood pressure. Concurrent volume infusion may be required to maintain adequate filling pressures and achieve increased stroke volume. Both sodium nitroprusside and nitroglycerin have also been used to offset the vasoconstriction produced by inotropes with intrinsic a-adrenergic activity. The use of sodium nitroprusside is complicated by the potential for cyanide toxicity at higher doses, while nitroglycerin has the advantage of being relatively nontoxic.
675
Sodium Nitroprusside (Nitropress) Sodium nitroprusside (Table 22-18) is an intravenous vasodilator effective in the acute management of hypertensive crisis as well as in congestive heart failure. It is an extremely potent vasodilator, with rapid onset and a short duration of action, used primarily to manage hypertensive emergencies; it is also useful when immediate reduction of preload or afterload is needed. The peripheral vasodilatory effects of nitroprusside are due to a direct action of the drug on arterial and venous smooth muscle; other tissues containing smooth muscle are not affected by the drug. The hypotensive effects of SNP are enhanced by other hypotensive agents. Sympathomimetics that exert a direct stimulatory effect (e.g., epinephrine) are the only class of drugs that effectively increase blood pressure during nitroprusside therapy. As a direct vasodilator, SNP increases cerebral blood flow and volume. Nitroprusside administered to normocarbic subjects can produce significant increases in intracranial pressure, and cause neurologic dysfunction with only slight decreases in blood pressure.82 Increases in ICP produced by SNP are maximal during modest decreases (80 mg/dL • D5 w 0.9% saline if 70 mm Hg • RAP 5–10 mm Hg • SpO2 >95% • PaCO2 35 mm Hg (ET CO2 35–40 mm Hg) • Temp 36 ∞ to 37 ∞C • Osmolality 2 drainage required within 10 mins Action Phase lllb
W9418-25.qxd 11/21/03 2:58 PM Page 728
Phase lllb Persistent ICP >20 mm Hg despite previous measures
Is a head CT required to exclude a surgical lesion?
Yes • Regain control of cerebral perfusion Ø ICP (CSF drain, 3% saline, mannitol, hyperventilation) • CT scan
Surgical lesion
No Is the underlying cause vascular or nonvascular?
No surgical lesion
Operating Room
Vascular—Hyperemia, Vasospasm • Confirm Dx by perfusion indices Nonvascular—Cerebral Edema (cellular, BBB breakdown, etc.) • If hyperemia (high PbtO2, SjvO2, TCD) Augment blood pressure (phenylephrine) until: Is sedation adequate? CPP >70 mm Hg Is CPP appropriate? Perfusion indices (PbtO2, SjvO2, TCD) normalize Titrate CO2 against perfusion indices • If vasospasm (low Pbto2, hi SjvO2, TCD) Achievable Ensure RAP >10 mm Hg Accept ICP level to 30 mm Hg* Augment BP to MAP >100 mm Hg Unachievable (or if ICP >30 mm Hg) ICP remains elevated and perfusion • Drain CSF inadequate • 3% saline • Drain CSF • Mannitol 0.5 gm/kg every 6 h (keep osmolarity