Textbook of Stereotactic and Functional Neurosurgery (2nd Edition, Vol.s 1 & 2)

Textbook of Stereotactic and Functional Neurosurgery Andres M. Lozano, Philip L. Gildenberg, Ronald R. Tasker (Eds.) ...

Author: Andres M. Lozano | Philip L. Gildenberg | Ronald R. Tasker

533 downloads 1644 Views 67MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Textbook of Stereotactic and Functional Neurosurgery

Andres M. Lozano, Philip L. Gildenberg, Ronald R. Tasker (Eds.)

Textbook of Stereotactic and Functional Neurosurgery With 1088 Figures and 232 Tables

Editors: Andres M. Lozano Professor of Surgery and RR Tasker Chair in Functional Neurosurgery University of Toronto Senior Scientist, Toronto Western Research Institute President, World Society for Stereotactic and Functional Neurosurgery Canadian Research Chair in Neuroscience (Tier 1) 399 Bathurst St. WW 4-447 Toronto, Ontario M5T 2S8 Canada [email protected] Philip L. Gildenberg Houston Stereotactic Concepts, Inc. 2260 West Holcombe Boulevard Suite 309 Houston, Texas 77030 USA [email protected] Ronald R. Tasker University of Toronto 399 Bathurst St. WW 4-447 Toronto, Ontario M5T 2S8 Canada A C.I.P. Catalog record for this book is available from the Library of Congress ISBN: 978-3-540-69959-0 This publication is available also as: Electronic publication under ISBN 978-3-540-69960-6 and Print and electronic bundle under ISBN 978-3-540-70779-0 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publication or parts thereof is only permitted under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer‐Verlag. Violations are liable for prosecution under the German Copyright Law. ß Springer‐Verlag Berlin Heidelberg 2009 The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product liability: The publishers cannot guarantee the accuracy of any information about the application of operative techniques and medications contained in this book. In every individual case the user must check such information by consulting the relevant literature. Springer is part of Springer Science+Business Media springer.com Publishing Editor: Gabriele Schro¨der MRW Editor: Sandra Fabiani Printed on acid‐free paper

SPIN: 11534778

2109 – 5 4 3 2 1 0

Andres Lozano dedicates this work To Marie, Alexander and Christopher

Philip L. Gildenberg, MD, PhD, dedicates this ‘‘Textbook of Stereotactic and Functional Neurosurgery’’ to Patricia Franklin Gildenberg, who has been a significant collaborator in everything he has written since 1973. Her contributions and dedication to both the American and World Societies for Stereotactic and Functional Neurosurgery have been important in maintaining the integrity of those organizations throughout the years.

I would like to dedicate my contribution to this book to my late wife, Mary Morris Craig Tasker, who made my career possible. Ronald R. Tasker OC, MD, FRCSC

Preface

This second edition of the Textbook of Stereotactic and Functional Neurosurgery appears just 11 years after the first. The first edition of the Textbook marked the 50th anniversary of the birth of human stereotactic neurosurgery. After an initial period of rapid growth, the field became quiescent in the 1960s when treatment of Parkinson’s disease migrated from stereotactic surgery to medication, primarily L-dopa. When it became apparent that L-dopa was not the final answer to the management of Parkinson’S disease, human stereotaxis had a rebirth, and the first edition followed a few years later. Many patients returned once again to surgical management in the early 1990s, so many neurosurgeons returned to the fields of stereotactic and functional neurosurgery. In the meantime, computer science and imaging techniques had advanced to foster the development of frameless image guided surgery, as well as stereotactic radiosurgery. Consequently, many neurosurgeons were becoming involved with stereotactic surgery, either returning to the field or being introduced to it for the first time. Both groups needed a comprehensive review of the then-current state of the field which was provided by the first edition of the Textbook of Stereotactic and Functional Neurosurgery. During the past 11 years there have been additional significant advances in this field. Computers have advanced to the point where exotic complex surgical programs have become commonplace and computers have become commonly used in the operating room, both for image guidance and physiological verification of targets. Image guided surgery is being used by most neurosurgeons, many of whom may not have had a prior background in stereotactic surgery. Deep brain stimulation for the treatment of motor disorders has become technically feasible and is commonly practiced at many stereotactic centers, using some techniques that were still under development a decade ago and technically advanced stimulating devices. Radiosurgical techniques involve more complex dosimetry than was available earlier, again because of advances in computer technology. Many new possibilities are being opened for both neurosurgical guidance and functional surgery involving targets that had not previously been recognized. These advances have proceeded so rapidly and broadly across several scientific fields that we have a need for a new edition of the Textbook of Stereotactic and Functional Neurosurgery. It is meant to provide a broad background for neurosurgeons, neurologists, and radiation oncologists, radiation physicists, and other scientists in the ever expanding fields of stereotactic neurosurgery, functional neurosurgery, and stereotactic radiosurgery. These fields have become so broad that no other publication covers the information necessary for all those specialists in a single reference source. This book will provide an update on the significant advances, information, and knowledge that have developed during this past decade has and have been scattered through the literature of these disparate specialties. The first edition was edited by Philip L. Gildenberg and Ronald Tasker. Both those stereotactic neurosurgeons retired from active clinical practice in the past 10 years, so they recruited Andres Lozano to be the primary editor. They felt that he brings to this position a unique combination of clinical #

Springer-Verlag Berlin/Heidelberg 2009

xii

Preface

expertise, extensive knowledge of background laboratory information, a record of providing innovation to functional neurosurgery, and has provided leadership for a new generation of stereotactic neurosurgeons. The editors wish to thank McGraw-Hill, who published the first edition, for relinquishing the copyright to Dr. Gildenberg, which opened the door for Springer to publish the second edition. The difference in the editorial process between the first edition and the second edition in some ways reflects the development of computers during that time, and consequently the advances in the fields of stereotactic and functional neurosurgery. The authors who contributed to the first edition sent hard copy printed manuscripts to the editors who would then read for content and copy edit, a timeconsuming and inefficient process. Manuscripts were then mailed back to the authors for final approval and/or correction and sent back to the Editors, possibly several times, who relayed them to the publisher when they were perfected. All manuscript preparation and submission for the second edition were done electronically, so no paper changed hands, which sometimes introduced new complexities, but ultimately proved more efficient, not to mention friendly to the environment. It is noteworthy that between the time of the first edition and the second edition, the operating room also became digitized, with consequent improvement in the display of diagnostic studies, processing information from those studies, and integrating that information into the surgical plan. The editors thank Garbiele Schro¨der and Stephanie Benko of Springer, who made this Textbook possible. The editors especially wish to thank Andrew Spencer and his staff at Springer, who served as the publisher’s representative to this project and maintained the flow of digital information between authors, editors and publisher. The editors also wish to thank Joan Richardson, who served as a liaison between the authors, editors, and publisher’s staff, and provided an accurate update on the progress and problems of each chapter and author. She managed the organization of 195 manuscripts with over 250 authors, with occasional changes in authors or topics, to facilitate communication between the editors and authors.

Table of Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix

Section 1 History of Stereotactic Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 History of Stereotactic Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 P. L. Gildenberg . J. K. Krauss

2 History of the Stereotactic Societies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 P. L. Gildenberg . J. K. Krauss

3 History of Stereotactic Surgery in US . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 P. L. Gildenberg

4 History of Stereotactic Surgery in Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 J. K. Krauss

5 History of Stereotactic Surgery in Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 C. Ohye

6 History of Stereotactic Neurosurgery in the Nordic Countries . . . . . . . . . . . . . . . . . . 65 B. A. Meyerson . B. Linderoth

7 A Brief History of Stereotactic Neurosurgery in Switzerland . . . . . . . . . . . . . . . . . . . 73 E. Taub

8 History of Stereotactic Surgery in Great Britain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 E. A. C. Pereira . A. L. Green . D. Nandi . T. Z. Aziz

9 History of Stereotactic Surgery in France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 A. L. Benabid . S. Chabardes . E. Seigneuret

10 History of Stereotactic and Functional Neurosurgery in Canada . . . . . . . . . . . . . . . 113 A. G. Parrent

xiv

Table of contents

11 History of Stereotactic Surgery in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 F.-C. Lee . B. Sun . J. Zhang . K. Zhang . F.-G. Meng

12 History of Stereotactic Surgery in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 P. K. Doshi

13 History of Stereotactic Surgery in Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 S. S. Chung

14 History of Stereotactic Surgery in Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 J. Guridi . M. Manrique

15 History of Stereotactic Neurosurgery in Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 A. Franzini . V. A. Sironi . G. Broggi

16 History of Stereotactic and Functional Neurosurgery in Brazil . . . . . . . . . . . . . . . . 197 O. Vilela Filho

Section 2 Imaging in Stereotactic Surgery . . . . . . . . . . . . . . . . . . . . . . . 249 17 General Imaging Modalities: Basic Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 A. A. Gorgulho . W. Ishida . A. A. F. De Salles

18 CT/MRI Technology: Basic Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 M. I. Hariz . L. Zrinzo

19 CT/MRI Safety in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 M. Schulder . A. Oubre´

20 Functional MRI in Image Guided Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 T. Sankar . G. R. Cosgrove

21 Angiography, MRA in Image Guided Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . 299 P. Jabbour . S. Tjoumakaris . R. Rosenwasser

22 Diagnostic PET in Image Guided Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 B. Ballanger . T. van Eimeren . A. P. Strafella

23 Neurophysiologic Mapping for Glioma Surgery: Preservation of Functional Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325 R. M. Richardson . M. S. Berger

Table of contents

24 Image Reconstruction and Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 B. A. Kall

Section 3 Stereotactic Targeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 25 Printed Stereotactic Atlases, Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347 R. J. Coffey

26 Electronic Stereotactic Atlases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 J. Yelnik . E. Bardinet . D. Dormont

27 Anatomical and Probabilistic Functional Atlases in Stereotactic and Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 W. L. Nowinski

28 Accuracy in Stereotactic and Image Guidance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 A. Hartov . D. W. Roberts

29 Development of a Classic: The Todd-Wells Apparatus, the BRW, and the CRW Stereotactic Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 J. Arle

30 Leksell Stereotactic Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 L. D. Lunsford . D. Kondziolka . D. Leksell

31 The Riechert/Mundinger Stereotactic Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 J. K. Krauss

32 The Talairach Stereotactic System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495 A. L. Benabid . S. Chabardes . E. Seigneuret . D. Hoffmann . J. F. LeBas

33 Laitinen Stereotactic Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 M. I. Hariz . L. V. Laitinen

34 Miniframe Stereotactic Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521 M. A. Madera . W. D. Tobler

Section 4 Image Guided Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . 533 35 Engineering Aspects of Electromagnetic Localization in Image Guided Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 E. C. Parker . P. J. Kelly

xv

xvi

Table of contents

36 The History, Current Status, and Future of the StealthStation Treatment Guidance System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 R. Bucholz . L. McDurmont

37 BrainLab Image Guided System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567 J. F. Fraser . T. H. Schwartz . M. G. Kaplitt

38 Robotic Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 P. L. Gildenberg

39 MRI in Image Guided Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599 M. Schulder . L. Jarchin

40 CT in Image Guided Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619 D. Kondziolka . L. D. Lunsford

41 Impedance Recording in Central Nervous System Surgery . . . . . . . . . . . . . . . . . . . 631 R. J. Andrews . J. Li . S. A. Kuhn . J. Walter . R. Reichart

42 Stereotactic and Image-Guided Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 J. B. Elder . A. P. Amar . M. L. J. Apuzzo

43 Pathology Techniques in Stereotactic and Image Guided Biopsy . . . . . . . . . . . . . . 663 P. T. Chandrasoma . N. E. Klipfel

44 Stereotactic and Image Guided Craniotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 E. C. Parker . P. J. Kelly

45 Image Guided Craniotomy for Brain Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699 I. E. McCutcheon

46 Virtual Reality in the Operating Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 P. L. Gildenberg

47 Comprehensive Brain Tumor Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735 M. Tamber . M. Bernstein

48 Novel Therapies for Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 G. Al-Shamy . R. Sawaya

49 Image-Guided Management of Brain Abscess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769 E. Taub . A. M. Lozano

Table of contents

50 Image-Guided Management of Brain Stem Lesions . . . . . . . . . . . . . . . . . . . . . . . . . 779 M. Levivier

51 Stereotactic Approaches to the Brain Stem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 L. U. Zrinzo . D. G. T. Thomas

52 Image Guided Management of Intracerebral Hematoma . . . . . . . . . . . . . . . . . . . . . 797 A. Losiniecki . G. Mandybur

53 Technical Aspects of Image-Guided Neuroendoscopy . . . . . . . . . . . . . . . . . . . . . . . 807 J. D. Caird . J. M. Drake

54 Intraoperative Image Guidance in Skull Base Tumors . . . . . . . . . . . . . . . . . . . . . . . 815 D. Omahen . F. Doglietto . D. Mukherjee . F. Gentili

55 Image Guided Management of Cerebral Metastases . . . . . . . . . . . . . . . . . . . . . . . . 831 P. Kongkham . M. Bernstein

Section 5 Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851 56 Radiobiology of Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853 D. C. Shrieve . J. S. Loeffler

57 Overview of Radiosurgery Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 M. Schulder

58 Gamma Knife: Technical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897 D. J. Schlesinger . C. P. Yen . C. Lindquist . L. Steiner

59 Linac Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 929 W. A. Friedman . F. J. Bova

60 CyberKnife: Technical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 J. R. Adler . D. W. Schaal . A. Muacevic

61 Proton Beam Radiotherapy: Technical and Clinical Aspects . . . . . . . . . . . . . . . . . . 957 S. Y. Woo

62 IMRT: Technical and Clinical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965 M. P. Carol

63 What Every Neurosurgeon Should Know About Stereotactic Radiosurgery . . . . . . 977 P. M. Black . F. Tariq

xvii

xviii

Table of contents

64 Radiosensitizers in Neurooncology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987 D. Khuntia . A. Chakravarti . H. I. Robins . K. Palanichamy . M. P. Mehta

65 Gamma Knife: Clinical Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007 A. Niranjan . L. D. Lunsford . J. C. Flickinger . J. Novotny . J. Bhatnagar . D. Kondziolka

66 Gamma Knife: Clinical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037 L. Steiner . C. P. Yen . J. Jagannathan . D. Schlesinger . M. Steiner

67 Linac Radiosurgery: Technical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087 F. J. Bova . W. A. Friedman

68 Cyberknife: Clinical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1111 F. C. Henderson Sr . W. Jean . N. Nasr . G. Gagnon

69 Proton Beam Radiosurgery: Clinical Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131 H. A. Shih . P. H. Chapman . J. S. Loeffler

70 Radiosurgery for Metastases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139 M. Maarouf . C. Bu¨hrle . M. Kocher . V. Sturm

71 Focused and Conventional Radiation for Acoustic Nerve Tumors . . . . . . . . . . . . . 1151 R. Den . S. H. Paek . D. W. Andrews

72 Radiosurgery for Pituitary Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1171 J. P. Sheehan . J. Jagannathan . W. J. Elias . E. R. Laws

73 Radiosurgery for Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1191 D. Kondziolka

74 Whole Body and Spinal Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1203 P. C. Gerszten

75 Gamma Knife Radiosurgery: Technical Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1225 D. Kondziolka . A. Niranjan . J. Novotny . J. Bhatanagar . L. D. Lunsford

Section 6 Functional Neurosurgery – Technical Aspects . . . . . . . . . 1237 76 Image Guided Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239 S. Khan . N. K. Patel . E. White . P. Plaha . S. Ashton . S. S. Gill

77 Evoked Potentials in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255 J. L. Shils . J. E. Arle

Table of contents

78 Microelectrode Recording in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . 1283 W. D. Hutchison . J. O. Dostrovsky . M. Hodaie . K. D. Davis . A. M. Lozano . R. R. Tasker

79 Impedance Recording in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . 1325 L. Zrinzo . M. I. Hariz

80 Anesthesia for Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1331 P. H. Manninen . N. Apichatibutra

81 Lesions Versus Implanted Stimulators in Functional Neurosurgery . . . . . . . . . . . 1349 W. S. Anderson . R. E. Clatterbuck . K. Kobayashi . J.-H. Kim . F. A. Lenz

82 Radiofrequency Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359 E. R. Cosman Sr. . E. R. Cosman Jr.

83 Stimulation Physiology in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . 1383 A. W. Laxton . J. O. Dostrovsky . A. M. Lozano

84 Stimulation Technology in Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . 1401 B. H. Kopell . A. Machado . C. Butson

85 Therapeutic Lesions Through Chronically Implanted Deep Brain Stimulation Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1427 S. Raoul . D. Leduc . C. Deligny . Y. Lajat

Section 7 Functional Neurosurgery for Movement and Motor Disorders – Clinical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . 1443 86 Surgery for Movement Disorders: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 1445 K. M. Prakash . A. E. Lang

87 History of Surgery for Movement Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1467 A. G. Parrent

88 Psychiatric Considerations in Management of Movement Disorders . . . . . . . . . . 1487 M. Zurowski . V. Voon . V. Valerie

89 Pathophysiology of Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1497 M. R. DeLong . T. Wichmann

90 Medical Management of Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1507 E. V. Encarnacion . R. A. Hauser

xix

xx

Table of contents

91 Patient Selection for Surgery for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . 1529 E. K. Tan . J. Jankovic

92 Pallidotomy for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1539 M. I. Hariz

93 Selective Thalamotomy and Gamma Thalamotomy for Parkinson Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549 C. Ohye

94 Subthalamotomy for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1569 J. A. Obeso . L. Alvarez . R. Macias . N. Pavon . G. Lopez . R. Rodriguez-Rojas . M. C. Rodriguez-Oroz . J. Guridi

95 Globus Pallidus Stimulation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . 1577 M. Deogaonkar . J. L. Vitek

96 Subthalamic Nucleus Stimulation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . 1603 A. L. Benabid . J. Mitrofanis . S. Chabardes . E. Seigneuret . N. Torres . B. Piallat . A. Benazzouz . V. Fraix . P. Krack . P. Pollak . S. Grand . J. F. LeBas

97 Thalamic Stimulation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1631 R. E. Wharen . R. J. Uitti . J. A. Lucas

98 PPN Stimulation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649 S. Stone . C. Hamani . A. M. Lozano

99 Other Targets to Treat Parkinson’s Disease (Posterior Subthalamic Targets and Motor Cortex) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665 F. Velasco . S. Palfi . F. Jimeńez . J. D. Carrillo-Ruiz . G. Castro . Y. Keravel

100 Motor Cortex Stimulation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . 1679 M. Meglio . B. Cioni

101 Tissue Transplantation for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691 K. Mukhida . M. Hong . I. Mendez

102 Gene Transfer for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1719 P. A. Starr . K. S. Bankiewicz

103 Intraparenchymal Drug Delivery for Parkinson’s Disease . . . . . . . . . . . . . . . . . . . 1731 R. D. Penn . A. A. Linninger

Table of contents

104 Management of Essential Tremor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1743 J. M. Nazzaro . K. E. Lyons . R. Pahwa

105 Management of Tremors other than Essential Tremor and Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1757 J. P. Nguyen . S. Raoul . C. Deligny . V. Roualdes . Y. Keravel

106 Diagnosis and Medical Management of Dystonia . . . . . . . . . . . . . . . . . . . . . . . . . . 1767 S. Fahn

107 Pathophysiology of Dystonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779 J. A. Bajwa . M. D. Johnson . J. L. Vitek

108 Central Procedures for Primary Dystonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1801 X. A. Vasques . L. Cif . B. Biolsi . P. Coubes

109 Functional Stereotactic Procedures for Treatment of Secondary Dystonia . . . . . 1835 H-H. Capelle . J. K. Krauss

110 Diagnosis and Medical Management of Cervical Dystonia . . . . . . . . . . . . . . . . . . . 1857 R. Bhidayasiri . D. Tarsy

111 Central Procedures for Cervical Dystonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1871 J. Q. Oropilla . Z. H. T. Kiss

112 Peripheral Procedures for Cervical Dystonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1885 T. Taira

113 Microvascular Decompression for Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . 1911 J. R. Pagura

114 History and Current Neurosurgical Management of Spasticity . . . . . . . . . . . . . . . 1925 R. D. Penn

115 Destructive Neurosurgical Procedures for Spasticity . . . . . . . . . . . . . . . . . . . . . . . 1935 M. Sindou . P. Mertens

116 Surgery in the Dorsal Root Entry Zone for Spasticity . . . . . . . . . . . . . . . . . . . . . . . 1959 M. P Sindou . P. Mertens

117 Intrathecal Drugs for Spasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1973 R. D. Penn

xxi

xxii

Table of contents

Section 8 Functional Neurosurgery for Pain . . . . . . . . . . . . . . . . . . . . 1983 118 Anatomy and Physiology of Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985 W. D. Willis Jr. . K. N. Westlund

119 Neuroimaging and Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2019 R. Peyron

120 What have PET Studies Taught us about Cerebral Mechanisms Involved in Analgesic Effect of DBS? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2031 R. Kupers . J. Gybels

121 History of DBS for Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2049 D. Richardson

122 Comprehensive Management of Cancer Pain Including Surgery . . . . . . . . . . . . . . 2061 P. S. Kalanithi . J. M. Henderson

123 The Central Lateral Thalamotomy for Neuropathic Pain . . . . . . . . . . . . . . . . . . . . 2081 D. Jeanmonod . A. Morel

124 Technique of Trigeminal Nucleotractotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2097 M. J. Teixeira . E. T. Fonoff

125 Bulbar DREZ Procedures for Facial Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2125 J. P. Gorecki

126 Percutaneous Cordotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2137 R. R. Tasker

127 CT-Guided Percutaneous Cervical Cordotomy for Cancer Pain . . . . . . . . . . . . . . . 2149 Y. Kanpolat

128 Ablative Spinal Cord Procedures for Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . 2159 P. L. Gildenberg

129 Intrathecal Opiates for Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2171 J. C. Sol . J. C. Verdie . Y. Lazorthes

130 Management of Pain of Benign Versus Cancer Origin . . . . . . . . . . . . . . . . . . . . . . 2197 P. L. Gildenberg . R. A. DeVaul

Table of contents

131 DBS for Persistent Non-Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2227 C. Hamani . D. Fontaine . A. Lozano

132 Motor Cortex Stimulation for Persistent Non-Cancer Pain . . . . . . . . . . . . . . . . . . . 2239 A. G. Machado . A. Y. Mogilner . A. R. Rezai

133 Radiofrequency Dorsal Root Entry Zone Lesions for Pain . . . . . . . . . . . . . . . . . . . 2251 P. Konrad . F. Caputi . A. O. El-Naggar

134 Surgical Dorsal Root Entry Zone Lesions for Pain . . . . . . . . . . . . . . . . . . . . . . . . . . 2269 M. P. Sindou

135 Facet Denervation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2291 R. R. Tasker . Wen Ching Tzaan

136 Sympathectomy for Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2297 C. R. Telles-Ribeiro . L. F. de Oliveira

137 Spinal Cord Stimulation. Techniques, Indications and Outcome . . . . . . . . . . . . . . 2305 B. Linderoth . B. A. Meyerson

138 Mechanisms of Action of Spinal Cord Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . 2331 B. Linderoth . R. D. Foreman . B. A. Meyerson

139 Peripheral Nerve Stimulation for Neuropathic Pain . . . . . . . . . . . . . . . . . . . . . . . . 2349 A. G. Shetter

140 The Pathophysiology of Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2359 R. W. Hurt

141 Radiofrequency Rhizotomy for Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . . . . 2421 E. Taub

142 Retrogasserian Glycerol Injection for Trigeminal Neuralgia . . . . . . . . . . . . . . . . . 2429 B. Linderoth . G. Lind

143 Balloon Compression for Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2457 J. A. Brown . J. G. Pilitsis

144 Microvascular Decompression for Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . 2465 S. Sup Chung

145 Gamma Knife Surgery for Trigeminal Neuralgia and Facial Pain . . . . . . . . . . . . . . 2475 A. C. J. de Lotbinie`re

xxiii

xxiv

Table of contents

146 Treatment of Headache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2483 N. T. Mathew

147 Occipital Neuralgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2507 D. B. Cohen . M. Y. Oh . D. M. Whiting

148 Hypothalamic Stimulation for Cluster Headache . . . . . . . . . . . . . . . . . . . . . . . . . . 2517 A. Franzini . M. Leone . G. Messina . R. Cordella . C. Marras . G. Bussone . G. Broggi

149 Surgical Treatment of Chronic Cluster Headache . . . . . . . . . . . . . . . . . . . . . . . . . . 2525 J. M. Castilla

150 Mesencephalotomy for Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2533 P. L. Gildenberg

Section 9 Management of Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2541 151 Indications for Surgical Management of Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . 2543 H. G. Wieser . D. Zumsteg

152 Classification of Epileptic Seizures and Epilepsies . . . . . . . . . . . . . . . . . . . . . . . . . 2561 H. O. Lu¨ders . S. Noachtar

153 EEG in Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2575 M. Hoppe . R. Wennberg . P. Tai . B. Pohlmann-Eden

154 The Wada Test-60th Year Anniversary Update-In Epilepsy Surgery . . . . . . . . . . . 2587 J. A. Wada . B. Kosaka

155 Imaging Evaluation of Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2617 D. Madhavan . R. Kuzniecky

156 Image Guided Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2633 Y. G. Comair . R. B. Chamoun

157 Intraoperative Monitoring in Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2651 G. Ojemann

158 MEG in Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2661 A. Fujimoto . T. Akiyama . H. Otsubo

159 Medical Management of Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2669 M. E. Newmark

Table of contents

160 Selective Amygdalo-Hypocampectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2677 T. A. Valiante

161 Subpial Transection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2715 Z. S. Tovar-Spinoza . J. T. Rutka

162 Corpus Callosotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2723 R. E. Maxwell

163 Hemispherectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2741 J.- G. Villemure . R. T. Daniel

164 Radiosurgery in Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2761 I. Yang . N. M. Barbaro

165 Centromedian Thalamic Stimulation for Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . 2777 F. Velasco . A. L. Velasco . M. Velasco . F. Jimeńez . J. D. Carrillo-Ruiz . G. Castro

166 Anterior Nucleus DBS in Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2793 M. Hodaie . C. Hamani . R. Wennberg . W. Hutchison . J. Dostrovky . A. M. Lozano

167 Vagal Nerve Stimulation for Seizures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2801 A. P. Amar . J. B. Elder . M. L. J. Apuzzo

168 Cerebellar Stimulation for Seizure Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2823 R. Davis

169 Stimulation of the Hippocampus and the Seizure Focus . . . . . . . . . . . . . . . . . . . . 2839 A. L. Velasco . F. Velasco . M. Velasco . G. Castro . J. D. Carrillo-Ruiz . J. M. Nuń˜ez D. Trejo

Section 10 Psychiatric Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2853 170 Ethical Considerations in Psychiatric Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2855 B. S. Appleby . P. V. Rabins

171 Psychosurgery – A Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2867 C. R. Bjarkam . J. C. Sørensen

172 Cingulotomy for Depression and OCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2887 G. R. Cosgrove

xxv

xxvi

Table of contents

173 DBS for OCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2897 L. Gabrie¨ls . P. Cosyns . K. van Kuyck . B. Nuttin

174 Medical Management and Indications for Surgery in Depression . . . . . . . . . . . . . 2925 P. Giacobbe . S. Kennedy

175 Ablative Procedures for Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2943 V. A. Coenen . C. R. Honey

176 Deep Brain Stimulation for Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2953 C. Hamani . B. Snyder . A. Laxton . A. Lozano

177 Surgical Procedures for Tourette’s Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2963 V. Visser-Vandewalle

178 Treatment of Aggressive Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2971 G. Broggi . A. Franzini

Section 11 Special and Emerging Applications . . . . . . . . . . . . . . . . . . 2979 179 DBS Disorders of Consciousness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2981 N. D. Schiff

180 Apnea: Phrenic Nerve Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2991 S. Rehncrona . G. Sedin . H. Fodstad

181 DBS for Bladder Dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2999 R. Almusa . M. M. Hassouna

182 Impaired Vision: Visual Prosthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3009 J. P. Girvin . A. G. Martins

183 Impaired Hearing: Auditory Prosthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3021 J. K. Niparko . A. Marlowe . H. W. Francis

184 Impaired Motor Function: Functional Electrical Stimulation . . . . . . . . . . . . . . . . . 3047 R. B. Stein . A. Prochazka

185 Gene Therapy for Neurological Disorders (Except Oncology) . . . . . . . . . . . . . . . . 3061 M. G. Kaplitt

Table of contents

186 Gene Therapy for Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3083 M. L. M. Lamfers . E. A. Chiocca

187 Microdialysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3117 M. J. Smith . M. G. Kaplitt

Section 12 The Future of Stereotactic and Functional Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3129 188 The Future of Computers and Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3131 B. A. Kall

189 The Future of Neuronavigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3137 D. W. Roberts

190 The Future of Radiosurgery and Radiotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3143 L. Ma . P. K. Sneed

191 The Future of Infusion Systems in Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . 3155 R. D. Penn

192 The Future of Cell Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3161 M. B. Newman . R. A. E. Bakay

193 The Future of Neural Interface Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3185 M. C. Park . M. A. Goldman . T. W. Belknap . G. M. Friehs

194 The Future of Molecular Neuro-Oncology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3201 J. A. J. King . M. D. Taylor

195 Future Ethical Challenges in Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3229 N. Lipsman . M. Bernstein

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3239

xxvii

List of Contributors

John R. Adler, Jr. Department of Neurosurgery, Stanford University, Stanford, CA, USA

Women’s Hospital, 75 Francis Street, CA 138F, Boston, MA 02115, USA Email: [email protected]

Tomoyuki Akiyama Division of Neurology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada

David W. Andrews Department of Neurosurgery, Jefferson Medical College, 834 Walnut St. Suite 650, PA 19107, Philadelphia, USA Email: [email protected]

Riyad Almusa Toronto Western Hospital, 399 Bathurst St, M5T 2S8, Toronto, Ontario, Canada George Al-Shamy Department of Neurosurgery - 442, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA L. Alvarez Movement Disorders, Functional Neurosurgery and Neurophysiology Units, Centro Internacional de Restauracion Neurologica (CIREN), Havana, Cuba Arun Paul Amar Departments of Neurosurgery, University of California San Francisco and the Permanente Medical Group 2025 Morse Avenue, Sacramento, CA 95825, USA Email: [email protected] William S. Anderson Instructor in Neurosurgery, Harvard Medical School, Department of Neurological Surgery, Brigham and

Russell J. Andrews Ames Associate (Smart Systems and Nanotechnology), NASA Ames Research Center, CA 94035, Moffett Field, USA Email: [email protected] Narisa Apichatibutra Department of Anesthesia, Toronto Western Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada Brian S. Appleby John Hopkins Hospital, Meyer 279, 600 N. Wolfe St., 21287–7279, MD, Baltimore, USA Michael L. J. Apuzzo Edwin M. Todd/Trent H. Wells, Jr., Professor of Neurological Surgery and Professor of Radiation Oncology, Biology, and Physics Keck School of Medicine, University of Southern California, 1200 North State Street, Suite 5046, Los Angeles, CA 90033, USA Email: [email protected]

xxx

List of contributors

Jeffrey E. Arle Department of Neurosurgery, Lahey Clinic, 41 Mall Road, 01805, Burlington, MA, USA

Eric Bardinet INSERM U679, Hoˆpital de la Salpeˆtrie`re, 47, Bd de l’Hoˆpital, 75013, Paris, France

Sharon Ashton Renishaw plc, New Mills, Wotton-under-Edge, Gloucestershire GL12 8JR, UK

Thomas Belknap Department of Clinical Neurosciences, Program in Neurosurgery, The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI 02903, USA

Tipu Z. Aziz Oxford Functional Neurosurgery, Nuffield Department of Surgery, University of Oxford, and Department of Neurological Surgery, The West Wing, The John Radcliffe Hospital, Oxford. Imperial College London and Charing Cross Hospital, London Jawad A. Bajwa Capistrant Parkinson and Movement Disorder Center, Bethesda Hospital and Nasseff Neuroscience Center, United Hospital, Saint Paul, MN, USA Neurological Associates of Saint Paul, Maplewood, MN, USA Roy A. E. Bakay Department of Neurosurgery, Rush University Medical Center, Chicago IL, USA Email: [email protected] B. Ballanger PET Imaging Centre, Center of Addiction Mental Health, University of Toronto, Toronto, Ontario, Canada Krystof S. Bankiewicz Department of Neurological Surgery, University of California, San Francisco, USA Nicholas M. Barbaro Department of Neurological Surgery, University of California, San Francisco, USA

Alim-Louis Benabid Joseph Fourier University, Pavillon B - Grenoble University Hospital, Grenoble, France Email: [email protected] Abdelhamid Benazzouz University of Bordeaux, France Mitchel S. Berger Department of Neurological Surgery, University of California, 505 Panassus Ave, Box 0112, San Francisco, CA, USA Email: [email protected] Mark Bernstein Division of Neurosurgery, Toronto Western Hospital and University of Toronto, Toronto, Canada Email: [email protected] Jagdish Bhatnagar The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA Roongroj Bhidayasiri Chulalongkorn Comprehensive Movement Disorders Center, Chulalongkorn University Hospital, 1873 Rama 4 Road, Bangkok 10330, Thailand Email: [email protected]


B. Biolsi CHRU Montpellier, Service de Neurochirurgie, Universite´ Montpellier I, Montpellier, France

Chris Butson Medical School of Wisconsin, 8701 Watertown Plank Road, WI 53226, Milwaukee, USA

Carsten Reidies Bjarkam Department of Neurobiology, Institute of Anatomy, University of Aarhus, DK–8000 Aarhus C, Denmark Email: [email protected]

John D. Caird Division of Neurosurgery, Department of Surgery, The Hospital for Sick Children, University of Toronto, 555 University Avenue, M5G 1X8, Toronto, Ontario, Canada

Peter M. Black Department of Neurosurgery, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115, USA Email: [email protected] Frank J. Bova Department of Neurological Surgery, University of Florida, PO Box 100265, Gainesville, FL 32610, USA Giovanni Broggi Department of Neurosurgery, Fondazione Istituto Neurologico “C. Besta”, Milan, Italy Email: [email protected] Jeffrey A. Brown 600 Northern Boulevard #118, Great Neck, NY 11021, USA Email: [email protected] Richard Bucholz Division of Neurosurgery, Department of Surgery, Saint Louis University School of Medicine, Saint Louis, Missouri, USA Email: [email protected] Christian Bu¨hrle Associate Professor, Department of Stereotactic and Functional Neurosurgery, University Hospital, 50924 Cologne, Germany G. Bussone Department of Neurosurgery, Neurological Institute Foundation “C. Besta”, Milan, Italy

Hans-Holger Capelle Department of Neurosurgery, Medical School Hannover, MHH, Hannover, Germany Franco Caputi Department of Neurosurgery, Rome, Italy Email: [email protected] Mark P. Carol 744 Lexington Way, Burlingame, CA 94010, USA Email: [email protected] Jose´ Damiań Carrillo Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico Jose´ Manuel Castilla Servicio de Neurocirugıá, Hospital “General Yaguë”, Avenida de Cid 96, 09005, Burgos, Spain Email: [email protected] Guillermo Castro Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico Stephan Chabardes Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Arnab Chakravarti Massachusetts General Hospital, Department of

xxxi

xxxii


Radiation Oncology, Cox 3, 100 Blossom Street, Boston, MA, USA Email: [email protected] Roukoz B. Chamoun Department of Neurosurgery, Baylor College of Medicine, Houston, Texas, USA Parakrama T. Chandrasoma Department of Pathology, Keck School of Medicine University of Southern California, GNH 2900, 1200 North State St., CA, USA Paul H. Chapman Departments of Radiation Oncology and Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA Email: [email protected] E. Antonio Chiocca James Cancer Hospital/Solove Research Institute, The Ohio State University Medical Center, Columbus, OH 43210–1240, USA Email: [email protected] Sang Sup Chung Bundang CHA General Hospital, 351 Yatap-dong, 463–712, Bundang-gu Sungnam-se, Kyong-doSungnam, South Korea Email: [email protected] L. Cif CHRU Montpellier, Service de Neurochirurgie, Universite´ Montpellier I, Montpellier, France Beatrice Cioni Neurochirurgia Funzionale e Spinale, Universita` Cattolica, Roma, Italy Richard E. Clatterbuck Hattiesburg Clinic, 4155 28th Avenue, Hattiesburg,

MS 39401, USA Email: [email protected] Volker A. Coenen Surgical Center for Movement Disorders, Division of Neurosurgery, University of British Columbia, Vancouver, BC, Canada Robert J. Coffey Medtronic Neurological, 4000 Lexington Avenue North, MN, USA Email: [email protected] David B. Cohen Drexel University College of Medicine Department of Neurosurgery, Allegheny General Hospital, 420 East North Ave, PA 15212, Pittsburgh, USA Youssef G. Comair Department of Neurosurgery, Baylor College of Medicine, 1709 Dryden, TX 77030, Houston, Texas, USA Email: [email protected] R. Cordella Department of Neurosurgery, Neurological Institute Foundation “C. Besta”, Milan, Italy G. Rees Cosgrove Department of Neurosurgery, Lahey Clinic Medical Center and Tufts University School of Medicine, Boston, Massachusetts, USA Email: [email protected] Eric R. Cosman, Jr. Cosman Medical, 76 Cambridge Street, MA, Burlington, USA Email: [email protected] Eric R. Cosman, Sr. Cosman Medical, 76 Cambridge Street, MA, Burlington, USA Email: [email protected]


Paul Cosyns Department of Psychiatry, University Hospital Antwerp, Antwerp, Belgium

Woodruff Memorial Research Building, 101 Woodruff Circle, Atlanta, GA 30322, USA Email: [email protected]

Philippe Coubes Unite´ de Recherche sur les Mouvements Anormaux, Hoˆpital Gui de Chauliac, Service de Neurochirurgie, 80 Avenue Augustin Fliche, 34295 Montpellier cedex 05, France Email: [email protected]

Robert Den Department of Neurosurgery, Jefferson Medical College, 834 Walnut St. Suite 650, PA 19107, Philadelphia, USA

Karen D. Davis Department of Physiology, University of Toronto, 1 King’s College Circle, M5S 1S8, Toronto, Canada

Milind Deogaonkar Functional Neuroscience Research Center, Cleveland Clinic Lerner Research Center, 9500 Euclid Avenue, OH 44195, Cleveland, USA

Ross Davis Neural Engineering Clinic, 330 Hammock Shore Dr., Melbourne Beach, FL, 32951, USA Email: [email protected]

Richard A. DeVaul Prof. Psychiatry, College of Medicine, Texas A & M System Health Science Center, College Station, Texas, USA

Alain C. J. de Lotbinie`re FACS Brain & Spine Surgeons of New York, 244 Westchester Avenue, NY 10604, White Plains, USA Email: [email protected]

John P. Donoghue Department of Neuroscience and Brain Science Program, Brown University, Providence, RI 02912, USA

Luiz F. de Oliveira Neurodor - Neurosurgery and Pain Clinic, Rua Visconde de Piraja 351/509, Cep: 22471–002, Ipanema, Rio de Janeiro, RJ, Brazil Antonio A. F. De Salles Department of Neurosurgery, David Geffen School of Medicine, University of California Los Angeles, 10495 Le Conte Avenue, Suite 2120, Los Angeles, California 90095, USA Email: [email protected] Ce´line Deligny Department of Neurology, CHU de Nantes, Nantes, France Mahlon R. DeLong Department of Neurology, Emory University

Paresh K. Doshi Department of Neuroscience, Jaslok Hospital and Research Centre, 15 - Dr. Deshmukh Marg, Pedder Road, Mumbai 400 026, Maharashtra, India Email: [email protected] Jonathan O. Dostrovsky Department of Physiology, University of Toronto, 1 King’s College Circle, M5S 1S8, Toronto, Canada Email: [email protected] James M. Drake Division of Neurosurgery, Department of Surgery, The Hospital for Sick Children, University of Toronto, 555 University Avenue, M5G 1X8, Toronto, Ontario, Canada Email: [email protected]

xxxiii

xxxiv


James B. Elder Department of Neurological Surgery, Keck School of Medicine, University of Southern California, 1200 North State Street, Suite 5046, Los Angeles, CA 90033, USA Email: [email protected]

Saõ Paulo Medical School, Rua Mena Barreto 765 - Itaim Bibi, SP- 014033–010, Saõ Paulo, Brazil

Jeff Elias Department of Neurosurgery, University of Virginia, Box 800212, HSC, VA, 22908–0212, Charlottesville, USA Email: [email protected]

Robert D. Foreman Department of Physiology, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA

Amr O. El-Naggar Lake Cumberland Neurosurgery, Somerset, KY, USA Email: [email protected] Elmyra V. Encarnacion Plummmer Movement Disorders Center, Texas A & M Health Sciences Center/Scott & White, Temple, Texas, USA Stanley Fahn Department of Neurology, Columbia University College of Physicians & Surgeons, New York, NY 10032, USA Email: [email protected] Osvaldo Vilela Filho Goiaˆnia Neurological Institute, Goiaˆnia, GO, Brazil Email: [email protected] John C. Flickinger The Departments of Neurological Surgery, and Radiation Oncology The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA Harald Fodstad University of Uppsala, Uppsala, Sweden Erich Talamoni Fonoff Division of Functional Neurosurgery, University of

Denys Fontaine Service de Neurochirurgie, Hoˆpital Pasteur, Centre Hospitalier Universitaire de Nice, Nice, France

Valerie Fraix Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Doglietto Francesco Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5T 2S8, Toronto, Canada Howard W. Francis Department of Otolaryngology, Head & Neck Surgery, The Johns Hopkins School of Medicine, Baltimore, USA Angelo Franzini Department of Neurosurgery, Fondazione Istituto Neurologico “C. Besta”, Milan, Italy Justin F. Fraser Department of Neurological Surgery, Weill Medical College of Cornell University, New YorkPresbyterian Hospital, New York, NY, USA William A. Friedman Department of Neurological Surgery, University of Florida, PO Box 100265, Gainesville, FL 32610, USA Email: [email protected] Gerhard M. Friehs Department of Clinical Neurosciences, Program in Neurosurgery, The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI 02903, USA


Ayataka Fujimoto Division of Neurology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada

John P. Girvin 1078 The Parkway, ON, N6A 2W0, London, Canada Email: [email protected]

Loes Gabrie¨ls Department of Psychiatry, University Hospital Gasthuisberg, Leuven, Belgium

Marc A. Goldman Department of Clinical Neurosciences, Program in Neurosurgery, The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI 02903, USA

Gregory Gagnon Department of Neurosurgery and Radiation Medicine, Georgetown University Hospital, 3800 Reservoir Rd, 20007, Washington DC, USA

John P. Gorecki Department of Neurosurgery, Wichita Surgical Specialists, P.A. The Heritage Plaza, 818 N. Emporia, Suite 200, 67214–3788, Wichita, Kansas, USA Email: [email protected]

Fred Gentili Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5T 2S8, Toronto, Canada Email: [email protected] Peter C. Gerszten Department of Neurological Surgery, University of Pittsburgh, UPMC Presbyterian B–400, 200 Lothrop Street, PA 15213, Pittsburgh, USA Email: [email protected]

Alessandra A. Gorgulho Department of Neurosurgery, David Geffen School of Medicine, University of California Los Angeles, 10495 Le Conte Avenue, Suite 2120, Los Angeles, California 90095, USA Sylvie Grand Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France

Peter Giacobbe University Health Network, University of Toronto, 200 Elizabeth Street, M5G 2C4, Toronto, Ontario, Canada

Alexander L. Green Oxford Functional Neurosurgery, Nuffield Department of Surgery, University of Oxford and Department of Neurological Surgery, The West Wing, The John Radcliffe Hospital, Oxford, UK

Philip L. Gildenberg Houston Stereotactic Concepts, 3776 Darcus St., 77005, Houston, Texas, USA Email: [email protected]

Jorge Guridi Division of Neurosurgery, Clinica Universitaria, Universidad de Navarra, Centro de Investigacioń Medica Aplicada. CIMA, Pamplona, Spain

Steven S. Gill Department of Neurosurgery, Frenchay Hospital, Bristol BS16 1LE, UK Email: [email protected]

Jan Gybels Department of Neurosurgery, Gasthuisberg University Hospital, University of Leuven, Leuven, Belgium

xxxv

xxxvi


Sun Ha Paek Department of Neurosurgery, Jefferson Medical College, 834 Walnut St. Suite 650, PA 19107, Philadelphia, USA Clement Hamani Division of Neurosurgery, Department of Surgery, Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada Marwan I. Hariz Edmond J. Safra Chair of Functional Neurosurgery, Unit of Functional Neurosurgery, Institute of Neurology, London, UK Email: [email protected] Alex Hartov Thayer School of Engineering, Dartmouth College, Hannover, NH 03755, USA Magdy M. Hassouna Toronto Western Hospital, 399 Bathurst St, M5T 2S8, Toronto, Ontario, Canada Email: [email protected] Robert A. Hauser Parkinson’s Disease and Movement Disorders Center, University of South Florida, 4 Columbia Drive, Suite 410, Tampa, Florida 33606, USA Email: [email protected] Jaimie M. Henderson Stanford University Medical Center, 300 Pasteur Dr, CA 94305, Stanford, USA Email: [email protected] Fraser Cummins Henderson, Sr. Georgetown University Medical Center, 3800 Reservoir Rd, 20007, Washington DC, USA Email: [email protected] Leigh R. Hochberg Department of Neuroscience and Brain Science

Program, Brown University, Providence, RI 02912, USA Department of Neurology, Massachusetts General Hospital, Brigham and Women’s Hospital, Spaulding Rehabilitation Hospital, Harvard Medical School, Boston, MA, USA Center for Restorative and Regenerative Medicine, Rehabilitation Research and Development Service, Department of Veterans Affairs, Veterans Health Administration, Providence, RI 02908, USA Mojgan Hodaie Department of Physiology, University of Toronto, 1 King’s College Circle, M5S 1S8, Toronto, Canada Dominique Hoffmann Joseph Fourier University, Pavillon B - Grenoble University Hospital, Grenoble, France Christopher R. Honey Surgical Center for Movement Disorders - Division of Neurosurgery – University of British Columbia Vancouver, BC, Canada Email: [email protected] Murray Hong Departments of Anatomy & Neurobiology and Surgery (Neurosurgery), Dalhousie University, Halifax, Nova Scotia, Canada Matthias Hoppe Bethel Epilepsy Centre, Bielefeld, Germany R. Wayne Hurt Chief of Neurological Surgery, St. Joseph Medical Center and Assistant Clinical Professor of Neurological Surgery, Baylor College of Medicine and University of Texas Medical Branch, Houston, Texas, USA William Hutchison Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5S 2T8, Toronto, Ontario, Canada


Warren Ishida Department of Neurosurgery, David Geffen School of Medicine, University of California Los Angeles, 10495 Le Conte Avenue, Suite 2120, Los Angeles, California 90095, USA Pascal Jabbour Department of Neurosurgery, Philadelphia, PA, USA E-mail: [email protected] Jay Jagannathan Lars Leksell Center for Gamma Surgery, Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA Joseph Jankovic Parkinson’s Disease Center and Movement Disorders Clinic, Department of Neurology, Baylor College of Medicine, 6550 Fannin, Suite 1801, Houston, Texas 77030, USA Email: [email protected]

Matthew D. Johnson Department of Biomedical Engineering, Cleveland Clinic Foundation, Cleveland, Ohio, USA Paul Kalanithi Stanford University Medical Center, 300 Pasteur Dr, CA 94305, Stanford, USA Bruce A. Kall Departments of Neurologic Surgery and Information Technology, Mayo Clinic, Rochester, MN, USA Email: [email protected] Yu¨cel Kanpolat Inkilap Sokak No: 24/2, 06640, Kizilay, Turkey Email: [email protected] Michael G. Kaplitt Department of Neurological Surgery, Weill Cornell Medical College, New York, NY, USA Email: [email protected]

Lauren Jarchin Department of Neurosurgery, North Shore LIJ, Manhasset, NY 11030, USA

Patrick J. Kelly Department of Neurological Surgery, NYU School of Medicine, 530 First Ave, Suite 8R, NY, USA Email: [email protected]

Walter Jean Department of Neurosurgery and Radiation Medicine, Georgetown University Hospital, 3800 Reservoir Rd, 20007, Washington DC, USA Email: [email protected]

Sidney Kennedy University Health Network, University of Toronto, 200 Elizabeth Street, M5G 2C4, Toronto, Ontario, Canada Email: [email protected]

Daniel Jeanmonod Department of Functional Neurosurgery, University Hospital Zu¨rich, Zu¨rich, Switzerland Email: [email protected]

Yves Keravel Service de Neurochirurgie, Hoˆpital Henri Mondor, Cre´teil, France

Fiacro Jimeńez Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico

Sadaquate Khan Institute of Neurosciences, Frenchay Hospital, Bristol BS16 1LE, UK Deepak Khuntia Department of Human Oncology, University of

xxxvii

xxxviii


Wisconsin, 600 Highland Ave, K4-B100, Madison, WI, USA Email: [email protected]

Paul Kongkham Division of Neurosurgery, Toronto Western Hospital and University of Toronto, Toronto, Canada

Jong-Hyun Kim Fellow in Neurosurgery, Department of Neurosurgery, 600 North Wolfe Street, Meyer 8–181, Baltimore, MD 21287, USA Email: [email protected]

Peter Konrad Director of Functional Neurosurgery, Vanderbilt University Medical Center, Department of Neurological Surgery, Rm T–4224; MCN, 37232–2380, Nashville, Tennessee, USA Email: [email protected]

James A. J. King Sick kids Hospital, 555 University Avenue, M5G 1X8, Toronto, Ontario, Canada Zelma H. T. Kiss Associate Professor, Neurosurgery University of Calgary, Calgary, Alberta T2N 4N1, Canada Email: [email protected] Nancy E. Klipfel Department of Pathology, Keck School of Medicine University of Southern California, GNH 2900, 1200 North State St., CA, USA Email: [email protected] Kazu Kobayashi Fellow in Neurosurgery, Department of Neurosurgery, 600 North Wolfe Street, Meyer 8–181, Baltimore, MD 21287, USA Martin Kocher Professor, Department of Radiation Oncology, University Hospital, 50924 Cologne, Germany Douglas Kondziolka The Departments of Neurological Surgery, and Radiation Oncology, The University of Pittsburgh, PA, USA The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA E-mail: [email protected]

Brian Harris Kopell Medical School of Wisconsin, 8701 Watertown Plank Road, WI 53226, Milwaukee, USA Email: [email protected] Brenda Kosaka Division of Neurosciences, University of British Columbia, 2329 West Mall, V6T 1Z4, Vancouver, Canada Paul Krack Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Joachim K. Krauss Chairman and Director, Department of Neurosurgery, Medical University, MHH CarlNeuberg-Str. 1, 30625 Hannover, Germany Email [email protected] S. A. Kuhn Ames Associate (Smart Systems and Nanotechnology), NASA Ames Research Center, CA 94035, Moffett Field, USA Ron Kupers PET Unit & Department of Surgical Pathophysiology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark Email: [email protected]


Ruben Kuzniecky Professor, NYU Comprehensive Epilepsy Center, Department of Neurology, New York University Medical Center, NY, USA Email: [email protected] Lauri V. Laitinen Unit of Functional Neurosurgery, Institute of Neurology, Box 146, Queen Square, WC1N 3BG, London, UK Martine L. M. Lamfers Department of Neurosurgery, Erasmus Medical Center, Rotterdam, The Netherlands Anthony E. Lang Division of Patient Based Clinical Research, Toronto Western Hospital McLaughlin Pavilion, 7th Floor Rm 7–403, 399 Bathurst Street, M5T 2S8, Toronto, Ontario, Canada Email: [email protected] Edward R. Laws Department of Neurosurgery, University of Virginia, Box 800212, HSC, VA, 22908–0212, Charlottesville, USA Email: [email protected] Adrian Laxton Division of Neurosurgery, Toronto Western Hospital, University of Toronto, UHN, Toronto, Ontario, Canada Yves Lazorthes Service de Neurochirurgie, Hospital LarreyRangueil, Avenue Jean Poulhes, 31403, Toulouse, France Email: [email protected] Jean François LeBas Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France

Dominique Leduc IREENA, Faculte´ des Sciences de Nantes, Nantes, France Foo-Chiang Lee Consultant Neurosurgeon, Division of Neuroscience Sunway Medical Center, No. 5 Jln Lagoon Selatan, Bandar Sunway, 46150, Selangor, Subang Jaya, Malaysia Email: [email protected] Dan Leksell Lars Leksell Professor of Neurological Surgery, University of Pittsburgh, B400 UPMC, Pittsburgh, PA 15213, USA Frederick A. Lenz Professor of Neurosurgery, The Johns Hopkins University School of Medicine, Department of Neurosurgery, 600 North Wolfe Street, Meyer 8–181, Baltimore, MD 21287, USA Email: [email protected] M. Leone Department of Neurosurgery, Neurological Institute Foundation “C. Besta”, Milan, Italy Marc Levivier Department of Neurosurgery, CHUV, and Centre Universitaire Romand de Neurochirurgie, Lausanne, Switzerland Email: [email protected] J. Li Ames Associate (Smart Systems and Nanotechnology), NASA Ames Research Center, CA 94035, Moffett Field, USA Go¨ran Lind Department of Neurosurgery, Karolinska Institutet and Karolinska University Hospital Stockholm, Sweden

xxxix

xl


Bengt Linderoth Department of Neurosurgery, Karolinska University Hospital, SE 171–76 Stockholm, Sweden Email: [email protected]

John A. Lucas Associate Professor of Psychology, Mayo Clinic College of Medicine, 4500 San Pablo Rd, FL 32224, Jacksonville, USA

Christer Lindquist Director, Gamma Knife Centre, Cromwell Hospital, London, England

L. Dade Lunsford Department of Neurological Surgery, Suit B–400, University of Pittsburgh Medical Center, 200 Lothrop Street, Pittsburgh, PA 15213, USA Email: [email protected]

Andreas A. Linninger Associate Professor of Chemical Engineering and Bioengineering, University of Illinois at Chicago, Laboratory for Product and Process Design, M/C 063, 851 S. Morgan St. - 218 SEO, Chicago, Illinois 60607–7000, USA Email: [email protected]

Hans O. Lu¨ders Department of Neurology, Cleveland Clinic, 9500 Euclid Ave, OH 44195, Cleveland, USA Email: [email protected]

Nir Lipsman Division of Neurosurgery, Toronto Western Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada

Kelly E. Lyons University of Kansas Medical Center, Department of Neurology, 3599 Rainbow Blvd, Mailstop 2012, Kansas City, KS 66160, USA Email: [email protected]

Jay S. Loeffler Department of Radiation Oncology, Massachusetts General Hospital, 55 Fruit Street, MA 02114, Boston, USA Email: [email protected]

Lijun Ma Department of Radiation Oncology, University of California-San Francisco, 505 Parnassus Avenue, CA 94143–0226, San Francisco, USA

G. Lopez Movement Disorders, Functional Neurosurgery and Neurophysiology Units, Centro Internacional de Restauracion Neurologica (CIREN), Havana, Cuba

Mohammad Maarouf Department of Stereotactic and Functional Neurosurgery, University Hospital, 50924 Cologne, Germany

Andrew Losiniecki Department of Neurosurgery, University of Cincinnati College of Medicine, Cincinnati, OH, USA

Andre G. Machado Center for Neurological Restoration, Department of Neurosurgery, Department of Biomedical Engineering, Cleveland Clinic, Cleveland, OH, USA

Andres M. Lozano Division of Neurosurgery, Department of Surgery, Toronto Western Hospital, University of Toronto, 399 Bathurst Street, Toronto, Ontario, Canada Email: [email protected]

R. Macias Movement Disorders, Functional Neurosurgery and Neurophysiology Units, Centro Internacional de Restauracion Neurologica (CIREN), Havana, Cuba


Marcella A. Madera University of Cincinnati, Department of Neurosurgery, Cincinnati, Ohio, USA

MMC 96, 420 E. Delaware St., MN 554455–0374, Minneapolis, USA Email: [email protected]

Deepak Madhavan Department of Neurological Sciences, University of Nebraska Medical Center, Omaha, NE, USA Email: [email protected]

Ian E. McCutcheon Department of Neurosurgery, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Box 44, Houston, Texas 77030–4009, USA Email: [email protected]

George Mandybur The Neuroscience Institute: Department of Neurosurgery, University of Cincinnati College of Medicine and Mayfield Clinic, Cincinnati, OH, USA Email: [email protected] Pirjo Manninen Department of Anesthesia, Toronto Western Hospital, 399 Bathurst Street, Toronto, Ontario, Canada Email: [email protected] Miguel Manrique Division of Neurosurgery, Clinica Universitaria, Universidad de Navarra, Centro de Investigacioń Medica Aplicada. CIMA, Pamplona, Spain Andrea Marlowe Department of Otolaryngology, Head & Neck Surgery, The Johns Hopkins School of Medicine, Baltimore, USA C. Marras Department of Neurosurgery, Neurological Institute Foundation “C. Besta”, Milan, Italy Antonio G. Martins Ninan T. Mathew Houston Headache Clinic, Houston, Texas, USA Email: [email protected] Robert Maxwell Department of Neurosurgery, Univ of Minnesota/S.E.

Lee McDurmont Division of Neurosurgery, Department of Surgery, Saint Louis University School of Medicine, Saint Louis, Missouri, USA Mario Meglio Istituto di Neurochirurgia, Policlinico Agostino Gemelli, Lg Gemelli, 00168, Roma, Italy Email: [email protected] Minesh P. Mehta Department of Human Oncology, University of Wisconsin, 600 Highland Ave, K4-B100, Madison, WI 53792, USA Email: [email protected] Ivar Mendez Departments of Anatomy & Neurobiology and Surgery (Neurosurgery), Dalhousie University, Halifax, Nova Scotia, Canada Fang-Gang Meng Center for Functional Neurosurgery (BMS), Shanghai Jiao Tong University Ruijin Hospital and Department of Functional Neurosurgery (JGZ,KZ, FGM), Beijing Tiantan Hospital, People’s Republic of China Patrick Mertens Department of Neurosurgery, Hopital Neurologique P. Wertheimer, University of Lyon, 59 Bd Pinel, F–69003 Lyon, France

xli

xlii


G. Messina Department of Neurosurgery, Neurological Institute Foundation “C. Besta”, Milan, Italy

Nadim Nasr Department of Neurosurgery and Radiation Medicine, Georgetown University Hospital, 3800 Reservoir Rd, 20007, Washington DC, USA

Bjo¨rn A. Meyerson Department of Neurosurgery, Karolinska Institute and Karolinska University Hospital Stockholm, Stockholm, Sweden Email: [email protected]

Jules M. Nazzaro University of Kansas Medical Center, Department of Neurology, 3599 Rainbow Blvd, Mailstop 2012, Kansas City, KS 66160, USA

John Mitrofanis University of New South Wales, Australia

Mary B. Newman Department of Neurosurgery, Rush University Medical Center, Chicago, IL, USA

Alon Y. Mogilner Section of Functional and Restorative Neurosurgery, North Shore-LIJ Health System, Manhasset, New York, USA Anne Morel Department of Functional Neurosurgery, University Hospital Zu¨rich, Zu¨rich, Switzerland Alexander Muacevic European CyberKnife™ Center Munich, Max Lebsche Platz 31, 81377 Munich, Germany Email: [email protected] Debabrada Mukherjee Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5T 2S8, Toronto, Canada Karim Mukhida Cell Restoration Laboratory, Room 12H1, Departments of Anatomy & Neurobiology and Surgery (Neurosurgery), Dalhousie University, 5850 College Street, Halifax, Nova Scotia, Canada B3H 1X5 Email: [email protected] Dipankar Nandi Imperial College London and Charing Cross Hospital, London, UK

Michael E. Newmark Kelsey-Seybold Clinic, 2727 W. Holcombe Blvd, 2nd floor Neurology, TX 77025, Houston, USA Email: [email protected] Jean Paul Nguyen Department of neurosciences, CHU Laennec, Nantes, France Email: [email protected] John K. Niparko Department of Otolaryngology, Head & Neck Surgery, The Johns Hopkins School of Medicine, Baltimore, USA Email: [email protected] Ajay Niranjan The Departments of Neurological Surgery, and Radiation Oncology, The University of Pittsburgh, PA, USA The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA Soheyl Noachtar Department of Neurology, Cleveland Clinic, 9500 Euclid Ave, OH 44195, Cleveland, USA Josef Novotny The Departments of Neurological Surgery, and


Radiation Oncology, The University of Pittsburgh, PA, USA The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA Wieslaw L. Nowinski Biomedical Imaging Lab, Agency for Science, Technology & Research (ASTAR), 30 Biopolis Street, 138671, The Matrix, Singapore Email: [email protected] Jose´ Marıá Nuń˜ez Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico Bart Nuttin Department of Neurosurgery, University Hospital Gasthuisberg and Katholieke Universiteit Leuven, Herestraat 49, B–3000, Leuven, Belgium Email: [email protected] Jose A. Obeso Department of Neurology and Neurosurgery, Clinica Universitaria and Medical School and Neuroscience Division, CIMA Centro de Investigacioń Biome´dica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), University of Navarra, Pamplona, Spain Email: [email protected] Michael Y. Oh Assistant Professor of Neurosurgery, Drexel University College of Medicine Co-Director, Division of Neuromodulation Department of Neurosurgery, Allegheny General Hospital, 420 East North Ave, PA 15212, Pittsburgh, USA Email: [email protected] Chihiro Ohye Functional & Gamma Knife Surgery Center, 886

Nakao-machi, 370–0001, Takasaki, Gunma, Japan Email: [email protected] George Ojemann Department of Neurological Surgery RI–20, University of Washington Medical Center, Campus Box 356470, 1959 N.E. Pacific Street, WA 98195, Seattle, USA Email: [email protected] David Omahen Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5T 2S8, Toronto, Canada Hiroshi Otsubo Division of Neurology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada Email: [email protected] Alondra Oubre´ Department of Neurosurgery, North Shore University Hospital, 9 Tower, Manhassett, NY 11030, USA Jorge Roberto Pagura Centro Integrado de Dor, Rua Baltazar da Veiga 490, S.P. 04510, Sao Paolo, Brazil Email: [email protected] Rajesh Pahwa University of Kansas Medical Center, Department of Neurology, 3599 Rainbow Blvd, Mailstop 2012, Kansas City, KS 66160, USA Kamalakannan Palanichamy Massachusetts General Hospital, Department of Radiation Oncology, Cox 3, 100 Blossom Street, Boston, MA 02114, USA

xliii

xliv


Stephan Palfi Service de Neurochirurgie, Hoˆpital Henri Mondor, Cre´teil, France Michael C. Park Department of Clinical Neurosciences, Program in Neurosurgery, The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI 02903, USA E-mail: [email protected] Erik C. Parker Department of Neurological Surgery, NYU School of Medicine, 530 First Ave, Suite 8R, NY, USA Andrew G. Parrent Department of Clinical Neurosciences, University Hospital, 339 Windemere Rd., N6A 5A5, London, Ontario, Canada Email: [email protected] Nikunj K. Patel Institute of Neurosciences, Frenchay Hospital, Bristol BS16 1LE, UK N. Pavon Movement Disorders, Functional Neurosurgery and Neurophysiology Units, Centro Internacional de Restauracion Neurologica (CIREN), Havana, Cuba Richard D. Penn Professor of Neurosurgery, The University of Chicago Medical Center Section of Neurosurgery, MC 3026 5841 South Maryland Avenue, Chicago, Illinois 60637, USA E-mail: [email protected] Erlick A. C. Pereira Oxford Functional Neurosurgery, Nuffield Department of Surgery, University of Oxford and Department of Neurological Surgery, The West Wing, The John Radcliffe Hospital, Oxford, OX3 9DU, UK Email: [email protected]

Roland Peyron CHU de Saint-Etienne & INSERM U879, University of St. Etienne, Lyon, France Email: [email protected] Brigitte Piallat Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Julie G. Pilitsis Assistant Professor, University of Massachusetts, School of Medicine, Worcester, MA, USA Email: [email protected] Puneet Plaha Institute of Neurosciences, Frenchay Hospital, Bristol BS16 1LE, UK Bernd Pohlmann-Eden Bethel Epilepsy Centre, Bielefeld, Germany Epilepsy Service, Division of Neurology, Dalhousie University, Halifax, Canada Pierre Pollak Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Kumar M. Prakash Division of Patient Based Clinical Research, Toronto Western Hospital McLaughlin Pavilion, 7th Floor Rm 7–403, 399 Bathurst Street, M5T 2S8, Toronto, Ontario, Canada Arthur Prochazka Department of Physiology and Centre for Neuroscience, University of Alberta, Edmonton AB T6G 2S2, Canada Jean Quint Oropilla Consultant in Neurosurgery, Makati Medical Center, Philippines Email: [email protected]


Peter Rabins John Hopkins Hospital, Meyer 279, 600 N. Wolfe St., 21287–7279, MD, Baltimore, USA Email: [email protected] Sylvie Raoul Service de Neurochirurgie, Hoˆpital G. et R. Lae¨nnec, Bd J. Monod, 44093 NANTES cedex, France Email: [email protected] Stig Rehncrona Lunds Universitet, Box 117, SE-221 00 Lund, Sweden Email: [email protected] R. Reichart Ames Associate (Smart Systems and Nanotechnology), NASA Ames Research Center, CA 94035, Moffett Field, USA Ali Rezai Director, Center for Neurological Restoration, Jane and Lee Seidman Chair in Functional Neurosurgery, Department of Neurosurgery, Cleveland Clinic, Cleveland, OH, USA Email: [email protected]

H. Ian Robins Department of Medicine and Human Oncology, University of Wisconsin, 600 Highland Ave, K4-B100, Madison, WI 53792, USA Email: [email protected] M. C. Rodriguez-Oroz Department of Neurology and Neurosurgery, Clinica Universitaria and Medical School and Neuroscience Division, CIMA Centro de Investigacioń Biome´dica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), University of Navarra, Pamplona, Spain R. Rodriguez-Rojas Movement Disorders, Functional Neurosurgery and Neurophysiology Units, Centro Internacional de Restauracion Neurologica (CIREN), Havana, Cuba Robert Rosenwasser Department of Neurosurgery, Thomas Jefferson University Hospital, Philadelphia, PA, USA V. Roualdes Department of neurosciences, CHU Laennec, Nantes, France

Donald Richardson Emeritus Professor Neurosurgery, Tulane University Health Science Center, New Orleans, Louisiana, USA Email: [email protected]

James T. Rutka Division of Neurosurgery, Suite 1503, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8 Email: [email protected]

R. Mark Richardson Department of Neurological Surgery, University of California, San Francisco, 505 Panassus Ave, Box 0112, CA, USA

Tejas Sankar Department of Neurosurgery, Lahey Clinic Medical Center, Tufts University School of Medicine, Boston, Massachusetts, USA

David W. Roberts Section of Neurosurgery, Dartmouth Medical School, Hannover, NH 03755, USA Email: [email protected]

Raymond Sawaya Department of Neurosurgery - 442, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA Email: [email protected]

xlv

xlvi


David W. Schaal Clinical Development, Accuray Incorporated, Sunnyvale, CA, USA Nicholas D. Schiff Department of Neurology and Neuroscience, Weill Medical College of Cornell University, 1300 York Avenue Room F610, New York, New York, 10021, USA Email: [email protected] David Schlesinger Lars Leksell Center for Gamma Surgery, Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA Michael Schulder Department of Neurosurgery, North Shore University Hospital, 9 Tower, Manhassett, NY 11030, USA Email: [email protected] Theodore H. Schwartz Department of Neurological Surgery, Weill Medical College of Cornell University, New YorkPresbyterian Hospital, New York, NY, USA Gunnar Sedin University of Uppsala, Uppsala, Sweden Email: [email protected]

Helen A. Shih Departments of Radiation Oncology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA Email: [email protected] Jay L. Shils Department of Neurosurgery, Lahey Clinic, 41 Mall Road, 01805, Burlington, MA, USA Email: [email protected] Dennis C. Shrieve Department of Radiation Oncology, Massachusetts General Hospital, 55 Fruit Street, MA 02114, Boston, USA Email: [email protected] Marc P. Sindou Chairman Department of Neurosurgery, University Claude-Bernard of Lyon Hoˆpital Neurologique Pierre Wertheimer, 59 Bd Pinel, 69677, BRON CEDEX, France Email: [email protected] V.A. Sironi Fondazione Istituto Neurologico, “C. Besta”, Milan, Italy

Eric Seigneuret Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France

Michelle J. Smith Department of Neurological Surgery, Weill Cornell Medical College, New York, NY, USA

Jason P. Sheehan Box 800–212, Department of Neurological Surgery, Health Sciences Center, Charlottesville, VA 22908, USA

Penny K. Sneed Department of Radiation Oncology, University of California-San Francisco, 505 Parnassus Avenue, CA 94143–0226, San Francisco, USA Email: [email protected]

Andrew G. Shetter Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona 85013, USA Email: [email protected]

Brian Snyder Division of Neurosurgery, Toronto Western Hospital, University of Toronto, UHN, Toronto, Ontario, Canada


J. C. Sol Multidisciplinary Pain Center Unit of Stereotactic and Functional Neurosurgery, Department of Neuroscience, University Hospital of Toulouse, Toulouse, France Jens Christian Sørensen Department of Neurosurgery, Aarhus University Hospital, DK–8000 Aarhus C, Denmark Philip A. Starr Department of Neurological Surgery, University of California, San Francisco, USA Email: [email protected] Richard B. Stein Department of Physiology and Centre for Neuroscience, University of Alberta, Edmonton AB T6G 2S2, Canada Email: [email protected] Ladislau Steiner Lars Leksell Center for Gamma Surgery, Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA Email: [email protected] Melita Steiner Lars Leksell Center for Gamma Surgery, Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA Scellig Stone Division of Neurosurgery, Department of Surgery, Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada A. P. Strafella PET Imaging Centre, Center of Addiction Mental Health, University of Toronto, Toronto, Ontario, Canada Movement Disorders Center, Toronto Western Hospital & Research Institute, University of Toronto,

Toronto, Ontario, Canada Email: [email protected] Volker Sturm Professor, Department of Stereotactic and Functional Neurosurgery, University Hospital, 50924 Cologne, Germany Email: [email protected] Bomin Sun Center for Functional Neurosurgery (BMS), Shanghai Jiao Tong University Ruijin Hospital and Department of Functional Neurosurgery (JGZ,KZ,FGM), Beijing Tiantan Hospital, People’s Republic of China Peter Tai Epilepsy Program, Toronto Western Hospital, University of Toronto, Toronto, Canada Takaomi Taira Director of Stereotactic and Functional Neurosurgery, Department of Neurosurgery, Tokyo Women’s Medical University, 8–1 Kawada, Shinjuku, Tokyo 1628666, Japan Email: [email protected] Mandeep Tamber Division of Neurosurgery, Toronto Western Hospital and University of Toronto, Toronto, Canada Eng-King Tan National Neuroscience Institute, Singapore General Hospital, Singapore Farzana Tariq Department of Neurosurgery, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115, USA Daniel Tarsy Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA

xlvii

xlviii


Ronald R. Tasker Department of Physiology, University of Toronto, 1 King’s College Circle, M5S 1S8, Toronto, Canada Ethan Taub Oberarzt, Department of Neurosurgery, Basel University Hospital, Spitalstrasse 21, CH–4031 Basel, Switzerland Email: [email protected] Michael D. Taylor Sick Kids Hospital, 555 University Avenue, M5G 1X8, Toronto, Ontario, Canada Email: [email protected] Manoel Jacobsen Teixeira Division of Functional Neurosurgery, University of Saõ Paulo Medical School, Rua Mena Barreto 765 - Itaim Bibi, SP- 014033–010, Saõ Paulo, Brazil Email: [email protected]

Ohio, USA Email: [email protected] Napoleon Torres Joseph Fourier University, Pavillon B - Grenoble University Hospital, 38700, Grenoble, France Zulma Tovar-Spinoza Division of Neurosurgery, Suite 1503, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8 David Trejo Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico Wen-Ching Tzaan Department of Neurosurgery, Chang Gung Medical College and Memorial Hospital, 5 Zu Shung St., Kweishan, Taoynan 333, Taiwan

Carlos R. Telles-Ribeiro Neurodor - Neurosurgery and Pain Clinic, Rua Visconde de Piraja 351/509, Cep: 22471–002, Ipanema, Rio de Janeiro, RJ, Brazil Email: [email protected]

Ryan J. Uitti Mayo Clinic College of Medicine, 4500 San Pablo Rd, FL 32224, Jacksonville, USA

David G. T. Thomas Division of Neurosurgery, The National Hospital for Neurology & Neurosurgery, Queen Square, Box 32, WC1N 3BG, London, UK

Taufik A. Valiante Univeristy of Toronto, Department of Surgery, CoDirector, Epilepsy Program, Krembil Neuroscience Center, UHN, 399 Bathurst Street 4W–436, Toronto, Ontario, Canada, M4S3H6 Email: [email protected]

Daniel Roy Thomas Department of Neurological Sciences, Christian Medical College, Vellore, Tamil Nadu 632004, India Stavropoula Tjoumakaris Department of Neurosurgery, Thomas Jefferson University Hospital, Philadelphia, PA, USA William D. Tobler Department of Neurosurgery, University of Cincinnati College of Medicine, ML 0515, Cincinnati,

T. van Eimeren PET Imaging Centre, Center of Addiction Mental Health, University of Toronto, Toronto, Ontario, Canada Kris van Kuyck Department of Neurosurgery, Laboratory of Experimental Neurosurgery and Neuroanatomy, Katholieke Universiteit Leuven, Belgium


X. A. Vasques CHRU Montpellier, Service de Neurochirurgie, Universite´ Montpellier I, Montpellier, France

Ana Luisa Velasco Cerrada Bosques de Moctezuma 55, La Herradura Huixquilucan, Estado de Me´xico 52780, Me´xico Email: [email protected] Francisco Velasco Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico Email: [email protected]

Jerrold L. Vitek Department of Neurosciences, Cleveland Clinic Foundation, 9500 Euclid Ave, NC30, Cleveland, OH 44195, USA Email: [email protected] Valerie Voon NINDS/NIH, 10 Center Dr, 20892–1428, Bethesda, Maryland, USA Email: [email protected] Juhn A. Wada Division of Neurosciences, University of British Columbia, 2329 West Mall, V6T 1Z4, Vancouver, Canada Email: [email protected]

Marcos Velasco Epilepsy Surgery Clinic; Unit for Stereotactic, Functional Neurosurgery and Radiosurgery, General Hospital of Mexico City, Mexico

J. Walter Ames Associate (Smart Systems and Nanotechnology), NASA Ames Research Center, CA 94035, Moffett Field, USA

J. C. Verdie Multidisciplinary Pain Center Unit of Stereotactic and Functional Neurosurgery, Department of Neuroscience, University Hospital of Toulouse, France

Richard Wennberg Department of Neurosurgery, Toronto Western Hospital, 399 Bathurst Street, M5S 2T8, Toronto, Ontario, Canada Email: [email protected]

Jean-Guy Villemure Director, Division of Neurosurgery, University of Montreal, Professor and Chairman, Neurosurgery Service, Centre hospitalier de l’universite´ de Montreál, 1560 Sherbrooke East, Montreal, Canada H2L 4M1 Email: [email protected]

Karin N. Westlund Marine Biomedical Institute, University of Texas Medical Branch, 2925 Beluche, TX 77551, Galveston, USA

Veerle Visser-Vandewalle Department of Neurosurgery, Academic Hospital Maastricht, P.O. Box 5800, 6202, AZ Maastricht, The Netherlands Email: [email protected]

Robert E. Wharen, Jr. Mayo Clinic College of Medicine, 4500 San Pablo Rd, FL 32224, Jacksonville, USA Email: [email protected] Edward White Institute of Neurosciences, Frenchay Hospital, Bristol BS16 1LE, UK

xlix

l


Donald M. Whiting Drexel University College of Medicine Director, Division of Neuromodulation Vice-Chairman, Department of Neurosurgery, Allegheny General Hospital, 420 East North Ave, PA 15212, Pittsburgh, USA Thomas Wichmann Department of Neurology, Emory University Woodruff Memorial Research Building 101 Woodruff Circle, Atlanta, GA 30322, USA Heinz Gregor Wieser Neurology Department, University Hospital Zurich, Frauenklinikstrasse 26, 8091 Zurich, Switzerland Email: [email protected] William D. Willis, Jr. Marine Biomedical Institute, University of Texas Medical Branch, 2925 Beluche, TX 77551, Galveston, USA Email: [email protected] Shiao Y. Woo The University of Texas MD Anderson Cancer Center, Proton Therapy Center- Houston, 1840 Old Spanish Trail, TX 77054, Houston, USA Email: [email protected] Isaac Yang Department of Neurological Surgery, University of California San Francisco, USA Email: [email protected] Je´roˆme Yelnik INSERM U679, Hoˆpital de la Salpeˆtrie`re, 47, Bd de

l’Hoˆpital, 75013, Paris, France Email: [email protected] Chun-Po Yen Lars Leksell Center for Gamma Surgery, Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA Jianguo Zhang Center for Functional Neurosurgery (BMS), Shanghai Jiao Tong University Ruijin Hospital and Department of Functional Neurosurgery (JGZ,KZ, FGM), Beijing Tiantan Hospital, People’s Republic of China Kai Zhang Center for Functional Neurosurgery (BMS), Shanghai Jiao Tong University Ruijin Hospital and Department of Functional Neurosurgery (JGZ,KZ, FGM), Beijing Tiantan Hospital, People’s Republic of China Ludvic U. Zrinzo Division of Neurosurgery, The National Hospital for Neurology & Neurosurgery, Queen Square, Box 32, WC1N 3BG, London, UK Email: [email protected] Dominik Zumsteg Neurology Department, University Hospital Zurich, Frauenklinikstrasse 26, 8091 Zurich, Switzerland

7 A Brief History of Stereotactic Neurosurgery in Switzerland E. Taub

Though he was not a neurosurgeon, the first great stereotactician in Switzerland was surely Walter Rudolf Hess (1881–1973), Professor of Physiology at the University of Zurich from 1917 to 1951. From the 1920s onward, he performed basic studies on the functional organization of the diencephalon, using electrical stimulation through stereotactically implanted depth electrodes in freely moving cats [1]. For this work, he was given the Nobel Prize in Physiology or Medicine in 1949, sharing the award with Antońio Egas Moniz (1874–1955), the Portuguese inventor of leukotomy and of cerebral angiography. Hugo Krayenbu¨hl (1902–1985), who had been trained as a neurosurgeon by Sir Hugh Cairns in London, became Professor of Neurosurgery in Zurich in 1948 – the first such chair in the country. He was active in functional neurosurgery from the beginning, publishing a study of prefrontal leukotomy and topectomy for the treatment of pain in 1950 [2] and a case series of temporal cortical excision and lobectomy for epilepsy in 1953 [3]. It is reported that he performed the first stereotactic procedures in Switzerland in 1958, in collaboration with Mehmet Gazi Yas¸argil, operating with a Riechert frame [4]. (It was, of course, Yas¸argil who was to succeed Krayenbu¨hl as chief of neurosurgery upon his retirement in 1973). In these early years of functional (lesional) stereotaxy, Krayenbu¨hl and Yas¸argil used equipment designed by their Zurich physiologist colleague O. A. M. Wyss [5] to perform both thalamotomies and #


pallidotomies for the treatment of Parkinson’s disease and other extrapyramidal disorders [6]. In the 1960s, functional and stereotactic neurosurgery in Krayenbu¨hl’s department became the domain of Jean Siegfried, who had developed his interests in the field as a Fellow under William H. Sweet at Massachusetts General Hospital in 1962. In the late 1960s, Siegfried wrote on percutaneous cordotomy and the surgical treatment of spasmodic torticollis and proposed stereotactic dentatotomy for the treatment of movement disorders [7,8,9]. Toward the end of the decade, with the introduction of L-DOPA, Siegfried took on the question of medical versus surgical treatment of Parkinson’s disease, writing extensively on the subject and gaining a very large clinical experience in stereotactic lesionmaking [10]. At the same time, he became Switzerland’s leading expert in Sweet’s technique of radiofrequency thermocoagulation of the Gasserian ganglion for the treatment of trigeminal neuralgia, eventually amassing a huge case series: a report on the first 500 cases, published in 1977, was followed by a report on 1,000 cases in 1981 [11,12]. Siegfried went on to become a pioneer of deep brain stimulation and one of the prime movers in the transformation of functional neurosurgery – now nearly, though not entirely, complete – from a ‘‘lesioning’’ to a ‘‘stimulating’’ discipline. He stimulated the sensory thalamus for the treatment of chronic pain in a series of 89 patients from 1978 to 1985 [13] and began to perform thalamic stimulation for tremor in the

74

7

A brief history of stereotactic neurosurgery in switzerland

late 1980s [14]. After the reintroduction of pallidotomy by Laitinen and colleagues in the early 1990s, Siegfried was the first to describe the achievement of comparable effects by pallidal stimulation [15]. After Siegfried’s move from the University across the Lake of Zurich to the Klinik Im Park (originally called AMI Klinik Im Park), a private clinic that had been opened in 1986, he also had the distinction of introducing Gamma Knife radiosurgery to Switzerland. He founded the Gamma Knife Center at the Klinik Im Park in 1994 and headed it until his retirement in 2001. Siegfried’s activities at the Klinik Im Park were continued by Thomas Mindermann (2000-present) and Ethan Taub (2001– 2007), a former fellow of Ronald Tasker and Andres Lozano at the University of Toronto. Epilepsy surgery had already been an interest of Krayenbu¨hl’s in the 1950s, as mentioned above, and was further developed in Zurich by both Siegfried and Yas¸argil. In 1970, Siegfried and the epileptologist Christoph Bernoulli brought the technique of stereoencephalography (SEEG) for the localization of epileptic discharges to Zurich, 5 years after its description by Bancaud and Talairach in Paris [16]. In the early 1980s, together with the epileptologist Heinz Gregor Wieser, Yas¸argil pioneered the technique of selective amygdalohippocampectomy for the treatment of mesiobasal limbic epilepsy [17]. At around the same time, SEEG gave way to semi-invasive localization via foramen ovale electrodes, which were first described by Siegfried, Wieser, and H. R. Stodieck [18]. From 1993 onward, the Zurich tradition of selective amygdalohippocampectomy was continued by Yas¸argil’s successor as chairman, Yasuhiro Yonekawa. Also at the University of Zurich, Daniel Jeanmonod has headed the Laboratory for Functional Neurosurgery for some years, continuing into the present. Known in the clinical sphere as one of the few current proponents of deep brain lesioning, as opposed to stimulation, he has pursued basic research on the pathophysiology of

movement disorders and other neurological conditions. His research group has published a detailed stereotactic atlas of the thalamus [19]. Off to the west, in the French-speaking part of Switzerland, both epilepsy surgery and functional stereotaxy became major areas of clinical activity in Lausanne after Jean-Guy Villemure moved there from the Montreal Neurological Institute to become chairman in 1997. Villemure’s department achieved high case numbers and became a renowned center of excellence in both areas. In particular, he and his group amassed a large series of patients who underwent subthalamic stimulation for Parkinson’s disease and published very important findings about both the usefulness of this procedure and its complications [20,21]. After Villemure’s retirement in 2006, his activities in the functional field were continued by his disciples Jocelyne Bloch and Claudio Pollo under his successor as chairman, Marc Levivier. Joachim-Kurt Krauss, currently chief of neurosurgery at the Medizinische Hochschule in Hanover (Germany), briefly headed the functional neurosurgery unit at the University of Berne in the 1990s. He was succeeded in this capacity (1999–2001) by Ethan Taub, who later brought deep brain stimulation procedures to Basel (2007present) in collaboration with Morten Wasner. Deep brain stimulation in Berne is currently the responsibility of Alexander Stibal. Heinz Fankhauser began a deep brain stimulation program at the private Clinique Cecil in Lausanne in 2002, while Ronald Bauer, a former fellow of Tipu Aziz at Oxford, has led the functional neurosurgery unit at the Cantonal Hospital in St. Gallen since 2006. The current state of functional neurosurgery in Switzerland can thus be briefly summarized: There are active clinical programs in deep brain stimulation at the University Hospitals of Lausanne, Berne, and Basel, as well as at the Cantonal Hospital in St. Gallen and at a private clinic in Lausanne. The Gamma Knife Center at the Klinik Im Park, long the only dedicated facility for stereotactic radiosurgery

A brief history of stereotactic neurosurgery in switzerland

in Switzerland, is still in operation as of this writing and is likely to be joined by other centers in this field in the near future. Across the country, scientific presentations for the medical public and articles in the lay press are gradually heightening awareness of the uses of functional neurosurgery among referring physicians and prospective patients, and activity in the field can be expected to expand and to flourish.

References 1. Hess WR. Die funktionelle Organisation des vegetativen Nervensystems. Basel: Schwabe; 1948. 2. Krayenbu¨hl H, Stoll W. [Prefrontal leucotomy and topectomy for the treatment of irreducible pain]. Rev Neurol (Paris) 1950;83(1):40-1. 3. Krayenbu¨hl H, Hess R, Weber G. [Electroencephalographic, corticographic, and surgical considerations on 21 cases of temporal epilepsy treated by cortical excision and lobectomy]. Rev Neurol (Paris) 1953;88(6):564-7. 4. Siegfried J, quoted in Nashold BS. The history of stereotactic neurosurgery. Stereotact Funct Neurosurg 1994;62:29-40. 5. Wyss OAM. Hochfrequenzkoagulationsgera¨t zur reizlosen Ausschaltung. Helv Physiol Pharmacol Acta 1945;3:437-43. 6. Krayenbu¨hl H, Wyss OAM, Yas¸argil MG. Bilateral thalamotomy and pallidotomy as treatment for bilateral Parkinsonism. J Neurosurg 1961;18:429-44. 7. Siegfried J. [Percutaneous cordotomy]. Schweiz Med Wochenschr 1967;97(40):1325-6. 8. Siegfried J. [Surgery for spasmodic torticollis]. Schweiz Med Wochenschr 1967;97(40):1325. 9. Siegfried J, Perret E. [Stereotaxic dentatotomy. New method of surgical treatment of hyperkinesis]. Rev Otoneuroophtalmol 1968;40:(7)341-3.

7

10. Krayenbu¨hl H, Siegfried J. [Treatment of Parkinson’s disease: L-dopa or stereotaxic technics?]. Neurochirurgie 1970;16(1):71-6. 11. Siegfried J. 500 Percutaneous thermocoagulations of the Gasserian ganglion for trigeminal pain. Surg Neurol 1977;8(2):126-31. 12. Siegfried J. Percutaneous controlled thermocoagulation of Gasserian ganglion in trigeminal neuralgia: experiences with 1000 cases. In: Samii M, Jannetta P, editor. The cranial nerves. Berlin: Springer; 1981. p. 322-30. 13. Siegfried J. Sensory thalamic neurostimulation for chronic pain. Pacing Clin Electrophysiol 1987;10(1 Pt 2):209-12. 14. Blond S, Siegfried J. Thalamic stimulation for the treatment of tremor and other movement disorders. Acta Neurochir Suppl (Wien) 1991;52:109-11. 15. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of ventroposterolateral pallidum: a new therapeutic approach for alleviating all parkinsonian symptoms. Neurosurgery 1994;35(6):1126-9. 16. Lu¨ders H, Comair YG, (eds). Epilepsy surgery. 2nd ed. Philadelphia, PA: Lippincott Williams and Wilkins; 2000. p. 48. 17. Wieser HG, Yas¸argil MG. Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy. Surg Neurol 1982;17(6):445-57. 18. Siegfried J, Wieser HG, Stodieck SR. Foramen ovale electrodes: a new technique enabling presurgical evaluation of patients with mesiobasal temporal lobe seizures. Appl Neurophysiol 1985;48(1–6):408-17. 19. Morel A, Magnin M, Jeanmonod D. Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387(4):588-630. 20. Vingerhoets FJ, Villemure JG, Temperli P, Pollo C, Pralong E, Ghika J. Subthalamic DBS replaces levodopa in Parkinson’s disease: two-year follow-up. Neurology 2002;58(3):396-401. 21. Burkhard PR, Vingerhoets FJ, Berney A, Bogousslavsky J, Villemure JG, Ghika J. Suicide after successful deep brain stimulation for movement disorders. Neurology 2004;63(11):2170-2.

75

16 History of Stereotactic and Functional Neurosurgery in Brazil O. Vilela Filho

When Andres Lozano invited me to write this chapter, although honored, I declined the invitation at first. After all, very little had been written about this issue. The neurosurgeon Sebastiaõ Gusmaõ, the person who has most studied the history of Brazilian neurosurgery, has already written a significant number of papers about this subject, but basically nothing about the history of the Brazilian stereotactic and functional neurosurgery. Therefore, being the president of the Brazilian Society for Stereotactic and Functional Neurosurgery, I felt impelled to accept the challenge and write this chapter, but not without some extra pressure from Lozano. It was a tough job. I called and emailed colleagues throughout this country innumerable times, at any time, and I suppose they must be tired of me, obviously not without reason. In the case of colleagues already dead, their relatives, friends and colleagues were my target. Not very infrequently, the same fact was presented with different versions, obliging me to dig even deeper. But here it comes at last, the first paper on this issue. I have done my very best and, I must say, it truly fascinated me. If any incongruity has occurred, it was not my fault, since most of the times I had to base myself on told stories instead of recorded information. Dear editors and friends, please accept my apologies for having missed the deadline so many times. The beginning of the history of Brazilian general and functional neurosurgery is closely intermingled, being humanly impossible to separate them, and that is how I decided to start this chapter. #


A Brief History of Brazilian Neurosurgery and the First Functional Procedures Performed in Brazil According to Horrax [1], the history of neurosurgery in modern times may be divided into three periods: pre-Lister (1710–1846), preHorsley (1846–1890), and modern neurosurgery (from 1890 until today). The first known neurosurgical procedure performed in Brazil was carried out by the surgeon Luis Gomes Ferreyra, in 1710, in the vicinities of Sabara´, a town located in the countryside of Minas Gerais State. The patient was a slave presenting with a compound depressed skull fracture caused by the natural fall of a tree branch. Ferreyra removed the bone fragments, performed hemostasis using silk threads from cobwebs soaked in egg white, and covered the bone defect with a piece of calabash shell until bone healing. The patient eventually recovered and returned to work, presenting as only sequel a very mild expression dysphasia [2]. The pre-Horsley period started with the clinical introduction of general anesthesia by Morton, in 1846, and of antisepsis by Lister, in 1867. In Brazil, differently from Europe, this period extended until 1928. During this phase, neurosurgical interventions were almost completely limited to the treatment of depressed skull fractures and drainage of extracerebral hematomas and abscesses [2,3].

198

16

History of stereotactic and functional neurosurgery in brazil

Curiously, many functional procedures were started in Brazil in the end of the pre-Horsley period. Following are the pioneers of some of these operations [3]. Paolo Josetti, in 1896, performed a gasserian ganglion resection for the treatment of trigeminal neuralgia. In 1903, Joaõ Burnier and Jose´ Hungria Jr performed periarterial sympathectomy for the treatment of causalgia. Paes Leme, in an imprecise date, but certainly before 1910, carried out the first surgical procedure for epilepsy in a patient presenting with Jacksonian seizures, starting in the right hemiface, caused by encephalomalacia of the central part of the left middle frontal gyrus, which was secondary to head injury and subsequent brain abscess. In 1913, Nascimento Gurgel employed peripheral neurectomies (Stoffel’s technique) to treat spasticity. From this period, worth mentioning was the contribution from Augusto Paulino de Souza and Ame´rico Vale´rio [3]. At that time, the prevalent idea was that the entire rolandic area (pre- and postcentral gyri) was related to motor control. In 1907, these surgeons operated on a victim from cranial missile wound presenting with contralateral hemiplegia. During the operation, being the brain widely exposed through a large craniectomy, they observed the complete integrity of the posterior rolandic area (postcentral gyrus). Based on this surgical finding, they hypothesized that the posterior rolandic area was not involved in motor control. This hypothesis, presented at the Fourth Latin American Medical Congress, held in Rio de Janeiro in 1909, and mentioned in the 1922 paper of the authors [3], was confirmed by Horsley, Campbell, and Dana 2 years later (1909). The above mentioned authors, among many others, are considered the predecessors of neurosurgery in our country [2]. Two surgeons are regarded as the precursors of Brazilian neurosurgery: Augusto Brandaõ Filho, known as the prince of the surgeons, and Alfredo Monteiro, another formidable general

surgeon, both from Rio de Janeiro, at that time the capital of the country [2]. Brandaõ Filho, in 1924, was the first to perform ventriculography in Brazil, and on 6 August 1928, directly supervised by Egas Moniz, he performed the first cerebral angiography in the Americas, in fact, only 14 months after Egas Moniz, assisted by Almeida Lima, invented this procedure (28 June 1927) [2]. Brandaõ Filho also performed some open cordotomies (according to Jose´ Portugal), two gasserian ganglion resections for trigeminal neuralgia (though the first patient died, an excellent result was achieved in the second), as well as brain tumor resection in seven patients, all of which died according to his publication of 1931 [2]. Noteworthy is the fact that the initial experimental researches performed by Egas Moniz and Almeida Lima that culminated with the development of human cerebral angiography were carried out at Instituto Rocha Cabral, in Lisbon. This institute was founded and maintained by the Portuguese Bento da Rocha Cabral, greatuncle of one of the most important Brazilian neurosurgeons, still very active, Guilherme Cabral, from Belo Horizonte, capital of Minas Gerais State [2]. In 1928, Antoˆnio Austrege´silo Rodrigues Lima, an eminent neurologist and the first Professor of Neurology in Brazil (1912, Federal University of Rio de Janeiro, as it is called today), visited many neurosurgical services in USA, including those headed by Cushing, Dandy, Adson, and Frazier, and was very impressed with what he saw. Back to Brazil, he urged the brilliant and very skillful general surgeon, Alfredo Monteiro, and his young and recently graduated (1927) assistant, Jose´ Ribe Portugal, to start neurosurgery in Brazil. To do so, he inaugurated the Neurosurgical Service of the Department of Neurology of Federal University of Rio de Janeiro (Universidade Federal do Rio de Janeiro – UFRJ). In 1932, this service became a department, and Alfredo Monteiro was


invested its head and the first Professor of Neurosurgery in Brazil. In 1935, Monteiro, under his own request, was transferred to the Department of Operative Techniques and Experimental Surgery, and Portugal, stimulated by his former boss, took over the chair of Neurosurgery [2]. In the early 1930s, both Brandaõ Filho and Alfredo Monteiro, probably disenchanted with their results, mainly in brain tumor operations, and realizing that surgery of the nervous system should be performed by people with adequate knowledge and formation in neurological sciences, including anatomy, pathophysiology, semiology, neurology, and neurosurgical techniques, that is, by true neurosurgeons, decided to abandon neurosurgery. They really never considered themselves as neurosurgeons, but general surgeons performing neurosurgery [2]. For the reasons aforementioned, Antoˆnio Austrege´silo Rodrigues Lima may be regarded as the godfather of Brazilian neurosurgery. The inauguration of modern neurosurgery in the world, in 1890, was led by the association of the following essential technological innovations: general anesthesia, antisepsis, Broca’s theory of cerebral localization (further corroborated and enlarged by the findings of Jackson, Fritsch & Hitzig, and Ferrier) and the studies on cranioencephalic topography carried out by Broca (still very important, but even more in a time when neuroimaging studies were not available) [2]. In the late 1920s, being neurosurgery well established in the world, neurology and general surgery well developed in Brazil, and considering the fact that state-of-the-art neuroimaging techniques (ventriculography and cerebral angiography) had already been introduced in the country, the moment was favorable for the birth of modern neurosurgery in Brazil [2]. Modern Brazilian neurosurgery was then inaugurated [2], having as pioneers Jose´ Ribe Portugal (1901–1992) and Elyseu Paglioli

16

(1898–1985), and in our opinion, also Antoˆnio Carlos Gama Rodrigues, or Carlos Gama (1904– 1963), as he was known. All of them started neurosurgery in their respective states, Rio de Janeiro, Rio Grande do Sul, and Saõ Paulo, approximately at the same time. Jose´ Portugal and Carlos Gama were, initially, self-taught, and just sometime later in their carriers they visited the services of the icons of world neurosurgery. Portugal, in 1945, spent 4 months with John Scarff, in New York, returning in 1947 for 6 more months. Gama, on the other hand, visited the service of Harvey Cushing. Differently, Elyseu Paglioli only started his neurosurgical practice after spending 8 months in Paris (1930) with De Martel, one of the pioneers of French neurosurgery [2]. Elyseu Paglioli and Jose´ Portugal founded two of the most important Brazilian neurosurgical schools. Of interest for this chapter, Paglioli formed Manoel Krimberg, one of the beginners of stereotactic and functional neurosurgery in Rio Grande do Sul State, and Djacir Figueiredo, the pioneer of stereotactic and functional neurosurgery in Ceara´ State. From Portugal’s school emerged Renato Barbosa, regarded as the pioneer of stereotactic and functional neurosurgery in Brazil, Gianni Temponi, another very well known Brazilian functional neurosurgeon, and Joffre Lima, the beginner of stereotactic and functional neurosurgery in Para´ State. Carlos Gama did not found a neurosurgical school, but he did form Rolando Tenuto, who headed the Neurosurgical Service at the University of Saõ Paulo (Universidade de Saõ Paulo – USP) Medical School teaching hospital (Clinic Hospital) from its inauguration, in 1945, until his retirement, in 1970, having formed, among many others, Jose´ Zaclis, one of the pioneers in stereotactic and functional neurosurgery in Saõ Paulo State, and Raul Marino Jr, the most widely known Brazilian functional neurosurgeon; this service became also one of the most important Brazilian neurosurgical schools [2].

199

200

16


The Brazilian Society of Neurosurgery (Sociedade Brasileira de Neurocirurgia – SBN) was founded by the twelve Brazilian neurosurgeons that attended the First International Congress of Neurological Surgeons in Brussels, Belgium, on 26 July 1957 [2], and became affiliated to the World Federation of Neurosurgical Societies in the same year. Jose´ Portugal and Elyseu Paglioli were among the twelve founding members, and were elected, respectively, the first and third president of the SBN. Both were later reelected for another term. According to the last census, performed in 2005, there are currently 2,251 neurosurgeons in Brazil, 1,735 of which are members of the SBN, making it the third in the world in number of members [2]. Interestingly, the three pioneers of neurosurgery in Brazil were in some way involved with functional neurosurgery, making relevant contributions to the field. Graduated from medical school in 1927, 2 years later Jose´ Portugal underwent examination for full professorship of the Department of Surgical Techniques and Experimental Surgery, Medical School, Federal University of Rio de Janeiro, occasion in which he presented the thesis entitled Contribuiçaõ a` Neurectomia Retrogasseriana (Contribution to Retrogasserian Neurectomy). This work was itself already a register of his main contributions to the field. His interest in the surgical treatment of trigeminal neuralgia, in fact, started during the last 3 years (total duration of medical school, including internship: 6 years) of his medical school, when he was a monitor of Anatomy (in the Brazilian university system, the monitor is a student who helps teaching a discipline he/she has already attended; to get this academic position, those interested should undergo an examination, when the best candidate is chosen), and remained his primary interest throughout his whole carrier. He was really a superb neurosurgeon [2]. According to Mario Brock, one of his disciples, Portugal used to perform a retrogasserian neurectomy in

roughly 20 min [4]. This procedure was so perfected by him that, in 1938, Leriche, the pioneer of pain surgery, came to Rio de Janeiro to observe one of his operations [2]. Regarding his contributions to the field of functional neurosurgery, the first one was registered when he was still a medical student and monitor of Anatomy. At that time, the prevailing idea, as shown in the first edition of Hovelaque’s book, was that the motor root of the trigeminal nerve crossed the gasserian ganglion though its outer aspect and joined the mandibular division of the trigeminal nerve externally. While making an anatomical preparation of the Fifth cranial nerve, however, Portugal noted that its motor root crossed the gasserian ganglion through its middle part and joined the mandibular division internally. Puzzled with this finding, which was in contradiction with the ‘‘bible’’ of Anatomy, he dissected many other specimens, confirming his initial observation. His findings were published in 1926, and again in 1929, being cited in Rouviere’s Anatomy [2]. His second contribution was in regard to the relative distribution of the three divisions of the trigeminal nerve in its sensory root. According to Stookey, the lateral, intermediate and medial parts of the trigeminal sensory root corresponded, respectively, to V3, V2 and V1. Portugal, however, showed that the entire external half of the sensory root corresponded to V3, and that the preservation of the medial 1/5 of the sensory root was enough to maintain corneal innervation and so prevent keratitis and consequent visual loss [2]. Finally, the third and most important contribution from Portugal to this field was the original clinical introduction of intradural retrogasserian neurectomy for the treatment of trigeminal neuralgia (1929) in a time when everybody else used the extradural approach [2]. Carlos Gama was also very fond of trigeminal procedures. In 1929 he improved the technique of gasserian ganglion alcoholization and published his first papers on this subject. Later, in 1938, Gama became Professor of the Department


of Neurology at the USP Medical School presenting his thesis entitled Neuralgias do Trigeˆmeo (Trigeminal Neuralgias) [2]. Elyseu Paglioli was a neurosurgeon of many publications, two of which are of particular interest to the field of functional neurosurgery. The first was the thesis Circulaçaõ Venosa dos Nućleos Centrais do Ce´rebro (Venous Circulation of the Basal Ganglia), which was presented in 1927 to secure the chair of Anatomy at the Federal University of Rio Grande do Sul, and published in 1929 [2]. The second was the thesis entitled Ventriculografia (Ventriculography), prefaced by his former chief, De Martel, and presented in 1938 to obtain the chair of Surgical Clinic at the Federal University of Rio Grande do Sul Medical School [2]. The apex of his stupendous academic carrier was reached in 1950, when he became the dean of the university, a position he occupied for twelve consecutive years. Besides his neurosurgical and academic carriers, he was also engaged in politics, being mayor of Porto Alegre, the capital of Rio Grande do Sul State, and Minister of Health during the administration of the Brazilian President Joaõ Goulart [2].

Stereotactic and Functional Neurosurgery in Brazil: From Early Days to Present Time Brazil is a large country, the fifth in the world in extension. It is divided into 26 states and the federal district, which are agrouped in five regions: North, Northeast, Midwest, Southeast, and South. For didatical reasons, we decided to describe the history of stereotactic and functional neurosurgery in Brazil considering its different regions.

Southeast Region The Southeast region is made up by the states of Rio de Janeiro, Saõ Paulo, Minas Gerais, and Espı´rito Santo.

16

It is hard to ascertain who the pioneer of stereotactic and functional neurosurgery in the Southeast and in Brazil was. It depends on a number of parameters not clearly defined, but certainly one of the following will fulfill the criteria: Renato Tavares Barbosa, Paulo Niemeyer, both from Rio de Janeiro, or Aloysio Mattos Pimenta, from Saõ Paulo.

Rio de Janeiro State Renato Tavares Barbosa (1912–2003) (> Figure 16-1) graduated from the Federal University of Minas Gerais Medical School. After that, he went to Rio de Janeiro, being the first disciple formed by Jose´ Portugal, pioneer of Brazilian neurosurgery. Initially, he worked as an assistant to Portugal [2]. In the 1950s, Barbosa went to Freiburg, Germany, where he got his training in stereotactic and functional neurosurgery with Riechert and Mundinger [5]. His first stereotactic operation was performed at Rio de Janeiro Neurological Institute, Federal University of Rio de Janeiro, . Figure 16-1 Renato Barbosa

201

202

16


in 1958, using the Riechert and Mundinger apparatus [2]. Throughout his life he accumulated a huge experience with stereotactic treatment of movement disorders, mainly Parkinson’s disease, and psychosurgery (open procedures such as prefrontal leucotomy and cingulectomy), initially at Rio de Janeiro Neurological Institute, and later at Lagoa Hospital, where he was head of the Neurosurgery Service for 27 years, and at Sorocaba Clinic, a private hospital. According to Luiz Fernando Martins, in 1971, at that time doing his internship at Lagoa Hospital, Barbosa was the first in Brazil to use a radiofrequency generator to perform thermocoagulation lesions. Until then, thermocoagulation lesions were made by using electrocautery: placing the stethoscope on the skull, one could ‘‘hear’’ the lesion being produced, which was stopped when tremor arrest was achieved or side effects started to appear. He was one of the founders of the Brazilian Society of Neurosurgery, being elected its president for the biennium 1970/1972 [2]. He was also one of the founders of Brazilian Society for Stereotactic and Functional Neurosurgery (1980), becoming its first president [6]. Renato Barbosa, one of the pioneers of stereotactic and functional neurosurgery in our country, was the first to dedicate most of his practice to this field, which he did during his whole career. For these reasons, he is regarded as the father of Brazilian stereotactic and functional neurosurgery [2]. Paulo Niemeyer (1914–2004) (> Figure 16-2), brother of the worldwide known architect, Oscar Niemeyer (a 100 years old and still very active, the planner of the most important buildings in Brası´lia, remarried his 60-year-old secretary when he was already 98) and father of Paulo Niemeyer Filho, at present one of the most renowned Brazilian neurosurgeons, was actually the first to perform stereotactic surgery in Brazil. Niemeyer graduated from the Federal University of Rio de Janeiro Medical School in 1936. His initial training was in general surgery, which he did under the supervision of Augusto Paulino,

. Figure 16-2 Paulo Niemeyer

Professor of Surgery and Head of the Surgical Clinic at Santa Casa de Miserico´rdia do Rio de Janeiro Hospital (at that time the teaching hospital of the Federal University of Rio de Janeiro Medical School), and Alfredo Monteiro, Professor of the Department of Operative Techniques and Experimental Surgery at the same institution and one of the precursors of Brazilian neurosurgery [2,7,8]. His interest in neurosurgery became apparent in 1939, when he started dedicating to the surgery of head injury, until then part of the general surgery practice [2,7]. At the beginning a self-taught neurosurgeon, just like Jose´ Portugal, later on he visited important and well-established neurosurgical centers of other countries [2]. Since 1943 he became interested in functional neurosurgery, particularly in the surgical treatment of movement disorders and epilepsy [2,7,8]. Before the stereotactic era, he treated many patients harboring Parkinson’s disease, athetosis


and other hyperkinesias employing pyramidotomy or cortical resection [7,9], and on 14 April 1954, he performed the first stereotactic surgery (left pallidotomy for PD in a 60-year-old patient) in South America, using the frame he was given by Riechert [7]. The report of his experience with stereotactic treatment of dyskinesias was the first publication of this kind in South America [10]. According to Niemeyer, when he was still an 8-year-old child, while alone with his beloved sister, Judith, she presented a seizure, the first of a series to follow. It was this event that later led him to Medicine, in general, and to epilepsy surgery, in particular [8]. In 1946 he was the first to perform electrocorticography in Brazil [7]. Initially, he used this technique in an attempt to demonstrate, in humans, the existence of cortical spreading depression, as previously reported in animals by Aristides Leaõ, a Brazilian scientist [7,8,11]. The paper derived from this research was widely recognized. Subsequently, he used not only electrocorticography, but also electrographic exploration with implanted electrodes, to determine epileptic foci [7,8]. In 1949, together with the neurologist Abraham Ackerman, he founded the Brazilian League Against Epilepsy [2,7,8]. Some meetings on epilepsy and electrocorticography were then organized. The French epileptologist Henri Gastaut was invited for two of these meetings (1954 and 1955). During his lectures, Gastaut showed evidence suggesting the involvement of amygdala and hippocampus in the genesis of temporal lobe epilepsy. At that time already experienced with Falconer’s en bloc temporal lobectomy technique (26 cases), Niemeyer, in 1956, inspired by Gastaut’s findings, was the first in the world to perform selective amygdalo-hippocampectomy for treating temporal lobe epilepsy. He approached the mesial temporal structures through a 2-cm incision placed in the second temporal gyrus, and named his technique transventricular amygdalohippocampectomy [2,7,8,12].

16

Recommended by Paul Bucy, he was invited by the Department of Health, Education, and Welfare of the American Government to participate in the Second International Colloquium on Temporal Lobe Epilepsy, in Washington, organized by the National Institute of Neurological Diseases and Blindnesses, in 1957, to show his technique, which was published in the following year [7,8,12]. Another important contribution from Niemeyer was his original observation that immediately after selective amygdalo-hippocampectomy, the abnormal cortical activity, as shown by pre-resection electrocorticography, becomes surprisingly worse, and that this abnormality subsides in a few days, as demonstrated through electrodes left in the subdural space [8,12]. This finding was reported in his 1958 paper [12]. More recently, two Brazilians, during their fellowship at Montreal Neurological Institute, the epileptologist Fernando Cendes and the epilepsy surgeon Arthur Cukiert, came to the same conclusion and called the aforementioned phenomenon ‘‘Niemeyer’s effect’’ [13]. In 1953, Niemeyer became head of the Department of Neurosurgery of Santa Casa de Miserico´rdia do Rio de Janeiro Hospital [7]. In 1957, he was among the 12 Brazilian neurosurgeons that founded the Brazilian Society of Neurosurgery [2]. In 1981, in Munich, he was elected second Vice-President of the World Federation of Neurosurgical Societies (WFNS), and in 1997, in Netherlands, he received the Honor Medal of the WFNS, which was awarded to five living neurosurgeons of the world that most contributed for the progress of neurosurgery [7]. The single thing he was most proud of, though, was the large number of neurosurgeons that he trained, now spread all over the country [7]. After 1958, unfortunately, his interest in functional neurosurgery progressively declined, and he became dedicated almost exclusively to general neurosurgery, in particular to microsurgery for

203

204

16


intracranial aneurisms and transesphenoidal approach for pituitary tumors [2,7]. On 10 March 2004, Paulo Niemeyer passed away [7,8]. Undoubtedly, a great loss to Brazilian and World neurosurgery. Gianni Maure´lio Temponi (1934–2007), another pioneer of stereotactic and functional neurosurgery in Brazil, graduated from the Federal University of Triaˆngulo Mineiro Medical School in 1959 [2,14]. His neurosurgical training (1961/1964) was carried out at Rio de Janeiro Neurological Institute under Jose´ Portugal. Renato Barbosa was the functional neurosurgeon in charge at the same institution, and it was from him that he got his training in stereotactic and functional neurosurgery. When Barbosa moved to Lagoa Hospital, Temponi was left in charge of functional neurosurgery. Posteriorly, he became Director of Rio de Janeiro Neurological Institute, Full Professor of Neurosurgery at Federal University of Rio de Janeiro Medical School (1972), and founded the Neurosurgical Service at Clementino Fraga Filho Hospital, the teaching hospital at Federal University of Rio de Janeiro. As head of the Service, he formed 45 neurosurgeons. Like his master in stereotactic and functional neurosurgery, Renato Barbosa, his main areas of interest were stereotactic treatment of movement disorders and psychosurgery. During his career, always using the Riechert and Mundinger apparatus, Temponi performed 478 ventriculographyguided procedures, mainly thalamotomy for Parkinson’s disease [15], dentatotomy for cerebral palsy (dystonia, chorea and spasticity), the theme of his PhD thesis defended in 1971 (‘‘Stereoencephalotomy for cerebral palsy in children’’), and psychosurgery (especially for heretic oligophrenia). He was so confident and satisfied with his results that, contrarily to almost everyone else, he never replaced ventriculography with computed tomography and/or magnetic resonance imaging as a tool for determining the stereotactic coordinates of functional targets. Temponi left four offspring, two of

them (Vicente and Gianni) neurosurgeons in Rio de Janeiro. Carlos Roberto Telles Ribeiro or simply Carlos Telles (> Figure 16-3), as he is known, is certainly one of the most respected and renowned functional neurosurgeons in the country nowadays. Telles graduated from the State University of Rio de Janeiro Medical Sciences School in 1969. In 1973 he finished his residence training at Brası´lia District Hospital, in Brası´lia, under Paulo Mello. From 1976 to June of 1979, he stayed as a fellow at University of Hannover, doing his PhD under Wolfhard Winkelmu¨ller. Invited by Mario Brock, in July 1979 he moved to the University of Berlin, where he inaugurated the Pain Clinic and, in 1980, defended his PhD thesis, supervised by Winkelmu¨ller, entitled ‘‘Treatment of chronic pain by electrical stimulation of the spinal cord.’’ After almost 1 year in Berlin, he went to Duke University, spending 1 month with Blaine Nashold Jr. Returning to Brazil in 1980, by invitation of Pedro Sampaio, Professor of Neurosurgery and Head of the Division of Neurosurgery at

. Figure 16-3 Carlos Telles


Pedro Ernesto Hospital, teaching hospital at the State University of Rio de Janeiro, Telles inaugurated the first multidisciplinary pain clinic in Brazil, which was probably one of his most important contributions to Brazilian stereotactic and functional neurosurgery. He was also the first to perform spinal cord stimulation for the treatment of pain in our country. Still in 1980, as will be mentioned later in this chapter, his role was fundamental for the foundation of the Brazilian Society for Stereotactic and Functional Neurosurgery. His main contribution to stereotactic and functional neurosurgery was the development of the technique he called combined solitary tractotomy (RF solitary tractotomy + RF lesioning of the subnucleus caudalis of the trigeminal nerve + C2 and C3 dorsal rhizotomy), which was performed through an open procedure, and designed to treat pain secondary to invasive cancer of the head and neck in the distribution of the V, IX and X cranial nerves and cervical region [16–18]. This technique was performed in more than 50 patients, providing excellent pain relief in 80% of them. Though never published, it was presented in three important meetings: 14th Brazilian Congress of Neurosurgery (Belo Horizonte, Brazil, 1982) [16], 20th Latin American Congress of Neurosurgery (Saõ Paulo, Brazil, 1983) [17], and Annual Meeting of the American Society for Stereotactic and Functional Neurosurgery (North Caroline, US, 1983) [18]. In 1991, he spent 6 months at the University of Freiburg, under Christopher Ostertag, doing his post-doctorate thesis (‘‘Stereotactic biopsy for the diagnosis of deep brain lesions’’). In 1992, after the retirement of Pedro Sampaio, he became Professor and Head of the Service of Neurosurgery at Pedro Ernesto Hospital, State University of Rio de Janeiro. Still very active, and having dedicated most of his life to the treatment of pain, he supervised the formation of more than 40 physicians in the area of pain, many of them illustrious pain specialists in our country. Other of his areas of expertise are

16

stereotactic biopsy of deep brain lesions, percutaneous cordotomy for cancer pain (> 800 cases), sympathetic procedures for the treatment of complex regional pain syndromes, DREZotomy, and trigeminal procedures (> 1,300 cases using RF retrogasserian rhizotomy, microvascular decompression, or balloon microcompression of the Gasserian ganglion). Carlos Telles was president of both the Brazilian Society for Stereotactic and Functional Neurosurgery (1988/1990) and Brazilian Society of Neurosurgery (1996/1998). Currently, there are some other colleagues in Rio de Janeiro with an impressive experience in stereotactic and functional neurosurgery, among them: Alexandre Amaral, Servidores Pu´blicos Estaduais Hospital (training with Manoel Teixeira); Ce´sar Fantezia Andrauss, Rio de Janeiro Neurological Institute (training with Gianni Temponi); Eduardo Carlos Barreto, Quinta D’Or Hospital (training with Carlos Telles and Mario Brock); Paulo L. C. Cruz, Santa Casa de Miserico´rdia do Rio de Janeiro Hospital (training with Manoel Teixeira); Ney Jose´ Monteiro, Lagoa Hospital (training with Renato Barbosa); Jose´ Augusto Nasser dos Santos, Estaćio de Sa´ University (training at Columbia University); and Marcello Reis da Silva, Clementino Fraga Filho Hospital (training with Jean Claude Peragut). Eduardo Barreto was one of the founders of the Brazilian Chapter of the International Neuromodulation Society – Brazilian Neuromodulation Society (May 2007), occasion in which he was elected its first president. Ce´sar Andrauss, in 1997, helped by an engineer, built the TCA stereotactic apparatus, a modification of the Leksell frame; until 2005, when the last unity was built, 28 apparatuses had been manufactured by hand.

Saõ Paulo State Two of the pioneers in stereotactic and functional neurosurgery in Brazil are from Saõ Paulo State: Aloysio Mattos Pimenta and Jose´ Zaclis.

205

206

16


Aloysio Mattos Pimenta (1912–1987) (> Figure 16-4) graduated from the University of Saõ Paulo (Universidade de Saõ Paulo – USP) Medical School in 1935. During his internship, in 1935, he divided his time between the Surgical Clinic, headed by Alves Lima; Neurological Clinic, headed by Enjolras Vampre´ (the pioneer of Neurology in Saõ Paulo), both at Santa Casa de Miserico´rdia de Saõ Paulo Hospital, at that time the teaching hospital at the USP Medical School; and Juqueri Hospital, a psychiatric institution. We assume that, since Carlos Gama, pioneer of neurosurgery in Saõ Paulo and Brazil, was the only neurosurgeon of the Neurological Clinic at the time, Mattos Pimenta accompanied him also. In 1936, he was part of the delegation of the USP Medical School that visited Argentina, spending most of his time with Manuel Balado, then one of the most prominent South American neurosurgeons. The combination of these different experiences, we believe, enabled Mattos Pimenta to start his neurosurgical practice [2,19]. In 1936, the same year Egas Moniz published the initial results of prefrontal leucotomy, Mattos . Figure 16-4 Aloysio Mattos Pimenta

Pimenta, oriented by Pacheco e Silva (Professor of Psychiatry at the University of Saõ Paulo Medical School), was one of the first in the world and the first in the Americas to perform the prefrontal leucotomy of Egas Moniz, which took place at Juqueri Hospital, in fact only a few days before Freeman and Watts performed their first psychosurgical procedure in the USA [2,19]. Until the late 1940s, assisted by the psychiatrist Mario Yahn and the neurosurgeon Afonso Sette Jr, Mattos Pimenta performed 279 operations of this type. The technique of Egas Moniz was used in the first 161 patients (before 1945), the technique proposed by Freeman and Watts in the 48 intermediate patients of their series, and the staged Freeman and Watts procedure (operation performed in three stages), as proposed by Mario Yahn, in the last 70 patients. The results obtained in the 279 patients were compiled in a book authored by Yahn, Mattos Pimenta and Sette Jr, and was prefaced by Egas Moniz and published in 1951 [20]. In August 1948, Egas Moniz chaired the First International Congress of Psychosurgery, which was held in Lisbon. During its closing session, the Brazilian delegation (Mattos Pimenta, Paulino Longo, Pacheco e Silva, Ma´rio Yahn, Anı´bal Silveira, E´lio Simo˜es, and Antoˆnio Carlos Barreto), led by Paulino Longo (head of the Neurological Division, Paulista Medical School, Federal University of Saõ Paulo) proposed the official nomination of Egas Moniz for the Nobel Prize considering his two great and universally accepted achievements, cerebral angiography and prefrontal leucotomy. The proposal was enthusiastically accepted and embodied by the members of the delegations of the 26 other countries that attended the meeting [2]. He had already been nominated for this award twice before, but in both occasions his nomination was declined. In 1949 Egas Moniz was at last awarded the Noble Prize in Physiology and Medicine for his invaluable contribution with the prefrontal leucotomy, no mention made, however, of his discovery of the cerebral angiography.


At the beginning, without formal training, Mattos Pimenta performed only simpler procedures. In order to polish his neurosurgical formation, he spent 2 years (1938 and 1939) in Europe, most of the time with To¨nnis, in Berlin, but also with Zu¨lch, in Berlin (2 months), Foërster, in Breslau (1 month), Olivercrona, in Stockholm (1 month), Busch, in Copenhagen, Leriche, in Strassbourg, and Clovis Vincent, in Paris. Due to the eclosion of World War II, however, he had to return to Brazil much earlier than he had planned. Always a restless creature, in January 1942 he went to the New York Neurological Institute, spending 4 months with Stookey and Putnam, and then moved on to the University of Michigan, where he stayed for 20 months under the supervision of Max Peet [2,19]. Back to Brazil, in 1944 he started working at the Neurological Clinic of Paulista Medical School, Federal University of Saõ Paulo, headed by Paulino Longo. In 1947, the Neurosurgical Division of this institution was officially inaugurated and Mattos Pimenta was invested its head, position that he held until his retirement (1982). In the 1960s he became Full Professor of Neurosurgery of Paulista Medical School, and in 1980 he started up the post-graduation course in neurosurgery (master degree and PhD), still the only one in the country [2,19]. Eventually he became involved with stereotactic surgery for Parkinson’s disease (PD). The date, though, is uncertain. We had the opportunity to analyze his curriculum vitae prepared in November, 1959 [19]. It is a very detailed CV, but no mention is made concerning any training in stereotactic surgery. Fernando Braga, who started his neurosurgical training under Mattos Pimenta in 1961, however, told us that his uncle, presenting with Parkinson’s disease, was operated on by his chief, who used the Riechert and Mundinger apparatus, in the same year he started his training. For this very reason, we inferred that Mattos Pimenta spent some time with Riechert and Mundinger in Freiburg in 1960, and after

16

returning to Brazil, started performing stereotactic surgery in 1961. Since then he operated on Parkinson’s disease with certain regularity for many years, but always as an appendage of his neurosurgical career. Neuro-oncology was always the apple of his eye. His associative activity was also intense. He was one of the twelve founders and the second president of Brazilian Society of Neurosurgery (1958/1959), presiding its second congress in 1959 (Campos do Jordaõ, Saõ Paulo State) [2]. In 1965 he was president of the 11th Latin-America Congress of Neurosurgery, held in Saõ Paulo, having as official topics only stereotactic surgery and urgencies in neurosurgery [21]. Maybe even more important were his activities in the World Federation of Neurosurgical Societies: First vice-president (1973/1977), honorary president (elected in 1977), and president of its congress held in Saõ Paulo in 1977. Two other neurosurgeons performed stereotactic and functional neurosurgery at Saõ Paulo Hospital, the teaching hospital at Federal University of Saõ Paulo: Fernando Menezes Braga and Fernando Antonio Patriani Ferraz. Fernando Braga, once finished his training at Paulista Medical School, spent 1 year (1965/ 1966) in Edinburgh with Gillinghan, doing mostly general neurosurgery, but also some stereotactic and functional neurosurgery. Before coming back to Brazil, he purchased the McCaul apparatus, and in 1967, performed his first thalamotomy in a parkinsonian patient at Saõ Paulo Hospital. Along the years, mainly until 1970, he accumulated a significant experience with stereotactic treatment of movement disorders. In 1992, as Division Head and Professor of Neurosurgery, Braga entrusted Fernando Ferraz the task of creating the Stereotactic and Functional Neurosurgery Service at the Division of Neurosurgery at Federal University of Saõ Paulo. Ferraz, an ex-resident of the institution, already had some training in stereotactic and functional neurosurgery, which he had acquired

207

208

16


from Mattos Pimenta and Braga. Before starting this new task, though, he went to UCLA School of Medicine and spent a short time with Antonio De Salles. A CT/MR-compatible stereotactic apparatus was acquired (Micromar stereotactic system), and then, still in 1992, the service, almost completely dedicated to the treatment of movement disorders, was finally inaugurated. Ferraz did a great job. The service went remarkably well until the beginning of the 2000s, when, due to a number of institutional problems, it faded away. It is very unfortunate, since Federal University of Saõ Paulo is considered one of the centers of excellence for neurosurgical training in Brazil. In this regard, it is probably wise mentioning that, just as in the case of Mattos Pimenta, stereotactic and functional neurosurgery was only an appendage of the neurosurgical practice of Braga and Ferraz. None of them was actually a functional neurosurgeon by ‘‘birth’’ and this can have made a real difference. The most important center of training in stereotactic and functional neurosurgery in Saõ Paulo is, in fact, the USP Medical School, which started by the hands of Jose´ Zaclis, another pioneer of the field. Graduated from the University of Saõ Paulo (USP) Medical School in 1944, Jose´ Zaclis (1917–1983) (> Figure 16-5a) was formed neurosurgeon by Rolando Tenuto, Professor of Neurosurgery and Head of the Division of Neurosurgery at the Clinic Hospital, teaching hospital at the USP Medical School [22]. The main assistant to Tenuto, in 1947 he was given the task of organizing the Neuroradiology Service of the institution [2,22]. From this time on, he dedicated almost exclusively to this area, performing ventriculography, pneumoencephalography, myelography, and cerebral angiography. He was the first to describe the technique of cerebral pan-angiography by means of injection of contrast medium into a single artery while simultaneously producing transient intrathoracic hypertension [23,24]. His service was the first

entirely dedicated to neuroradiology in Brazil. Since at that time stereotactic surgery relied basically on ventriculography, in which he was an expert, after a trip to Germany, Russia, and USA, places where he became acquainted with different stereotactic apparatuses, and with the help of the engineer Koralek, he built, in 1961, the HC stereotactic apparatus (HC stands for Hospital das Clıńicas) (> Figure 16-5b) [25]. The equipment was set in a room of the neuroradiology unit completely dedicated to stereotactic surgery, where he used part of his time to perform stereotactic procedures for movement disorders and pain. The operations were carried out with the patient in a sitting position. This was the second Brazilian made stereotactic frame, in fact the first really elaborated and arc-centered apparatus. Only two of these frames were produced: one for the USP Medical School, and another for the Rio de Janeiro Neurological Institute. From 1971 to 1974, he was Head of the Division of Neurosurgery, and Professor at the USP Medical School. After Zaclis came Raul Marino Jr and Manoel Teixeira. Raul Marino Jr (> Figure 16-6) graduated from the USP Medical School in 1960, and after finishing his neurosurgical training in the same institution in 1964, under Rolando Tenuto, he went to North America to brush up his formation. In 1965, while in Boston, he divided his time between the Lahey Clinic, accompanying James Poppen and Charles Fager, who introduced him to stereotaxis and surgery for movement disorders, and the laboratory of Walle Nauta at the Massachusetts Institute of Technology (MIT), where he was for the first time really exposed to the limbic system, basal ganglia, and functional neuroanatomy. In 1966, still in Boston, he moved to the Massachusetts General Hospital (MGH), and was extremely fortunate to have as simultaneous supervisors professionals like Thomas Ballantine Jr (psychiatric surgery), William Sweet (pain surgery), and Raymond Kjellberg (surgery for movement disorders, brachytherapy, and


16

. Figure 16-5 (a) Jose´ Zaclis (b) HC stereotactic apparatus, built by Zaclis and Koralek in 1961

. Figure 16-6 Raul Marino Jr

cyclotron radiosurgery for pituitary tumors and other endocrine diseases), all luminaries of the field of stereotactic and functional neurosurgery. Learning of Marino’s previous work on limbic system with Nauta at MIT, Ballantine Jr invited him to join his team and work on the cingulotomy project for mental illnesses and pain [26]. In 1967, this team published their pioneer paper on stereotactic anterior cingulotomy for psychiatric

disorders and pain [27]. According to Marino Jr (personal communication), in the 1965 publication of Ballantine’s group [28] on the same subject, the free-hand technique was used instead of the stereotactic technique and the number of cases was still too small. Marino spent the year of 1967 in Montreal. At Montreal Neurological Institute (MNI), McGill University, where he spent most of his time, he learned electroencephalography, electrocorticography, and the Wada test with Peter Gloor and epilepsy surgery with Theodore Rasmussen. There he had the opportunity to meet Wilder Penfield and Herbert Jasper. His free time was spent at the Notre Dame Hospital, University of Montreal, accompanying the stereotactic procedures of Claude Bertrand and the transsphenoidal surgeries of Jules Hardy, who later on became his brother-in-law. In 1968 he returned to the USA for an extra year at the National Institutes of Health (Bethesda, Maryland) supervised by Paul MacLean (lab work), the father of the limbic system, and Van Buren, at the time in charge of the stereotactic program [26]. After 4 years abroad, Marino finally returned to Brazil. In 1969, requested by Rolando Tenuto, his former chief, he resumed his work at the USP Medical School. Initially, he performed his stereotactic operations using the Zaclis’ HC apparatus. In 1971, after a substantial reform of

209

210

16


the operating rooms and infirmaries of the Psychiatric Institute at the USP Medical School, Raul Marino inaugurated the first integrated, multidisciplinary Stereotactic and Functional Neurosurgery Service in Brazil, a service with 51 beds completely devoted to stereotactic and functional neurosurgery [26]. Well prepared as he was, he performed the most varied functional procedures, although his main interests, from the beginning until his retirement, always remained the same, that is, surgical treatment of epilepsy and psychiatric disorders, from which areas derived his most important contributions. Aiming to improve the electrophysiological investigation of refractory epilepsy, in 1972 he spent 6 months in Paris with Jean Talairach and Gabor Szikla, masters of the stereotactic electroencephalography technique [26]. Back to Brazil, a Talairach frame was acquired by his institution, finally replacing the Zaclis apparatus. A member of the American Branch of the International Society for Research in Stereoencephalotomy since its foundation (1968), in 1973, during the Sixth Symposium of the International Society in Tokyo, he became its member and was elected local Chairman of the next meeting (1977), which would be held in Saõ Paulo, Brazil, as a satellite meeting of that of the World Federation of Neurosurgical Societies (WFNS) [26]. Still during the Tokyo meeting, when was proposed a change in the name of the International Society, Marino suggested adding the term functional, which was accepted and the new designation became World Society for Stereotactic and Functional Neurosurgery. In 1977, during the Seventh Meeting of the World Society for Stereotactic and Functional Neurosurgery, Marino was elected Vice President of the Eighth Meeting of the World Society for Stereotactic and Functional Neurosurgery (the President elected was Jean Siegfried) and appointed member of the editorial board of the official journal of the society, Applied Neurophysiology (previously – 1938/1975 – called Confinia

Neurologica), later (1988) renamed as Stereotactic and Functional Neurosurgery, position that he held until his retirement in 2007 [26]. As a result of the 1977 meeting, the book Functional Neurosurgery, edited by Theodore Rasmussen and Raul Marino Jr, having as contributors an impressive number of outstanding functional neurosurgeons, was published in 1978 [29]. In his introductory chapter, Marino explains the reasons that led him to propose the term functional neurosurgery for this specialty, making a parallel with the bees behavior: ‘‘General neurosurgery tends to concentrate on the lesion, rather than on the symptoms, while functional neurosurgery focus on the symptoms, that is, on the abnormal functions, which are often hyperfunctional states that appear at a distance as a consequence of the primary lesion. Many specialties deal with the human brain, as many insects alight on the prairie flowers. However, only the bees know how to extract the honey. The bees alone are able to do that job and leave the flowers intact, without hurting or making them lose their freshness, allowing them to remain exactly as they were before. This is the hope and aim of functional neurosurgery’’ [29]. In 1980, Marino also played a very important role in the foundation of the Brazilian Society for Stereotactic and Functional Neurosurgery, and it was his close relationship with Blaine Nashold, who was present during this event, that enabled the Brazilian Society for Stereotactic and Functional Neurosurgery to be regarded as the Brazilian branch of World Society for Stereotactic and Functional Neurosurgery since its creation [26]. Raul Marino was elected the second President of Brazilian Society for Stereotactic and Functional Neurosurgery. Undoubtedly due to his great fondness for and expertise in psychiatric surgery, Marino also became Vice President of the International Society of Psychosurgery. In the 1980s (Raul Marino, personal communication), Marino was the first and the only (to the best of our knowledge) Brazilian neurosurgeon


to use neural (dopaminergic cells) grafts in humans for the treatment of Parkinson’s disease. Autologous adrenal medullary grafts were attempted in three patients (Beneficeˆncia Portuguesa Hospital), but no benefit was achieved and this technique was abandoned. Homologous substantia nigra grafts derived from fetuses of legally approved abortions (pregnancy following rape) were tried in three other patients. However, soon afterwards these previously legal abortions became prohibited and the trials were terminated. In 1990, Marino became Full Professor of Neurosurgery and Chief of the Division of Neurosurgery at the USP Medical School, one of the most prestigious Brazilian schools of Medicine, position that he held until his retirement in 2007 [26]. During his years at the University (1971–2007) he helped to form a large number of neurosurgeons. In 1993, after spending some time at the Temple University (Philadelphia, USA), he inaugurated the second LINAC Radiosurgery Service in the country at Beneficeˆncia Portuguesa Hospital (Saõ Paulo). The first X-Knife service, although quite rudimentary, had already been installed by Jack Beraha at Oswaldo Cruz Hospital (Saõ Paulo) in 1983, having being closed in 1989. His most important contributions were in the areas of epilepsy surgery and psychosurgery [27,29–37]. As already mentioned, as part of Ballantine’s team, he played a very important role in the development of stereotactic anterior cingulotomy for the treatment of psychiatric disorders (major depression and obsessive-compulsive disorder) and pain, a technique he later spread in our country [27]. Concerning epilepsy surgery, his contributions were various, standing out: he was the first in Brazil to implement and diffuse the need of a multidisciplinary team to better evaluate the surgical candidates; his team was pioneer in establishing the electrophysiological parameter to determine the extension of the

16

callosotomy, that is, the disruption of bilateral synchrony of spike and wave discharges [30]; finally, Marino Jr was the first to propose and use the stereotactic callosotomy technique [31]. Raul Marino Jr was the first Brazilian neurosurgeon with formal training in stereotactic and functional neurosurgery and also the first to introduce the concept of a multidisciplinary stereotactic and functional neurosurgery team in our country. From the beginning, his career was almost completely devoted to the field of stereotactic and functional neurosurgery. Besides, he is by far the most renowned and prominent Brazilian functional neurosurgeon. For all these reasons, Marino Jr should be the one to be regarded as the real father of stereotactic and functional neurosurgery in Brazil. Graduated from the USP Medical School in 1972, Manoel Jacobsen Teixeira (> Figure 16-7a) finished his residence training in the same institution in 1976. During his training, he was for the first time exposed to stereotactic surgery by Jose´ Zaclis. Still in 1976, he spent 3 months at Goiaˆnia Neurological Institute under Luiz Fernando Martins, a pupil of Wilhelm Umbach, deepening his knowledge and practice in stereotactic and functional neurosurgery. In 1977 he went to Europe, where he stayed for more than 2 years. Initially, Teixeira spent 1 year and a half in Edinburgh with Edward Hitchcock, his most beloved chief and icon. Afterwards, he moved on to Zurich, spending 3 months with Jean Siegfried, and then, finally, to Freiburg, where he accompanied Mundinger during 4 months. During his outstanding career, he has visited a number of other functional neurosurgery services around the world. Teixeira started working at the Division of Neurosurgery of the Department of Neurology at the USP Medical School in 1984. In 1985, he obtained his master degree, and in 1990, his PhD. From 1997 to 2004, he was Technical Director of the Health Service of the Division of Functional Neurosurgery of the Institute of

211

212

16


. Figure 16-7 (a) Manoel Teixeira (b) Micromar stereotactic system, model TM-03B

Psychiatry at the USP Medical School, and since 2004 he became Technical Director of the Division of Functional Neurosurgery. In March 2007, after the retirement of Raul Marino, there was an examination for full professorship in neurosurgery at the USP Medical School, and Teixeira was approved in first place, becoming Full Professor of the discipline and Head of the Division of Neurosurgery. His academic career has been quite profitable. Teixeira and his group have published a large number of papers, book chapters and books, and presented innumerable oral abstracts and posters both in national and international meetings, the vast majority in the area of pain. The multidisciplinary Pain Clinic of USP Medical School, which he runs, is most probably the busiest pain center in the country. Along the years he has formed a large number of functional neurosurgeons and pain specialists not only from Brazil but also from many other South American countries. Manoel Teixeira is a very versatile functional neurosurgeon, being experienced in all areas of the field, including brachytherapy and radiosurgery; his expertise in epilepsy and psychiatric disorders surgery, however, is less impressive. But the apple of his eye is really pain, anything in this area, from pathophysiology to the most

refined surgical technique. In fact, he is regarded by many as a living encyclopedia in this field. Also, he is certainly one of the most experienced active pain surgeons in the world. Accordingly, his most important contributions were on the surgical treatment of pain. A disciple of Edward Hitchcock, he probably is the still active neurosurgeon with the largest experience in medullary stereotactic trigeminal nucleotractotomy [38,39], a technique developed by Hitchcock in 1970 [40]. Hitchcock and Teixeira, in 1987, described the technique of pontine stereotactic trigeminal nucleotractotomy, indicated in cases of deafferentation extending through a long distance in the trigeminal nuclear complex, when the pain can not be adequately alleviated by medullary tractotomy alone [38,41]. In 2005, his group showed in rats that the motor cortex presents an antinociceptive function even under physiological circumstances and that motor cortex stimulation-induced analgesia is mediated by the opioid system [42]; for this research, Erich Fonoff received the Young Investigator Award from the World Society for Stereotactic and Functional Neurosurgery [43]. Teixeira et al., in 2007, demonstrated that the deafferentation pain associated with actinic brachial plexopathy and trigeminal neuropathy, usually unresponsive to


various therapeutic approaches, both conservative and surgical, can be successfully relieved by, respectively, cervical DREZotomy and stereotactic trigeminal nucleotractotomy [44]. Also in 2007, using transcranial magnetic stimulation to map the motor cortex, Teixeira et al demonstrated that the response to motor cortex stimulation in patients harboring neuropathic pain secondary to brachial plexus avulsion could be predicted by the size of the field of representation of the deafferented limb in the motor cortex, being the result good or excellent in patients with larger fields and poor in those with smaller or undetectable fields; this research was awarded the NeuPSIG Poster Prize (category: clinical studies) during the Second International Congress on Neuropathic Pain (Berlin, June, 2007) [45]. Recently, his group proposed a technique for stereotactic disconnection of the hypothalamic hamartoma and so preventing gelastic and secondary seizures and aggressive behavior [46], and phrenic nerve stimulation for the treatment of refractory singultus [47]. Other important but unpublished contributions from his group were the use of computer-assisted stereotactic fenestration of aqueductal cysts for the treatment of hydrocephalus [48] and the stereotactic intracavitary instillation of amphotericin B through an indwelling catheter to treat deep intracerebral paracoccidioidomycosis cysts [49]. In our opinion, these are the most relevant contributions from Teixeira and his group, among others [50–58]. A great amount of the development of stereotactic and functional neurosurgery in our country was due to the integrated actions of Teixeira and the neurosurgical industry. In 1985, Teixeira and the engineer Antoˆnio Martos, CEO of Micromar, built the first commercially available Brazilian stereotactic apparatus, the Micromar stereotactic system, model TM-01B (Teixeira & Martos), a modification of that of Hitchcock’s. Continuous refinement led to the production of new versions of the first frame,

16

the models TM-02B, in 1991, and TM-03B, in 1996 (> Figure 16-7b). In 1998, Teixeira, Martos and the physicist Armando Alaminos produced the first linear accelerator- compatible radiosurgery system in Brazil, the Micromar radiosurgery system. Manoel Teixeira, a very well known and admired functional neurosurgeon in our country, was president of the Brazilian Society for Stereotactic and Functional Neurosurgery for two terms: 1990/1992 and 1998/2000. Other important names of stereotactic and functional neurosurgery in Saõ Paulo State are Jorge Roberto Pagura, Jose´ Oswaldo Oliveira Jr, Claúdio Fernandes Correâ, Arthur Cukiert, Jack Beraha, and Nilton Luis Latuf. Jorge Roberto Pagura (> Figure 16-8) graduated from the ABC Medical School in 1974. Once finished his neurosurgical residence (1975/1978) at Nove de Julho Hospital, in Saõ Paulo, under the supervision of Gilberto Machado de Almeida (also Professor of Neurosurgery and Head of the Division of Neurosurgery at the USP Medical School), he went to . Figure 16-8 Jorge Pagura

213

214

16


Germany, where he spent 15 months (1978/1979) with Wolfhard Winkelmu¨ller at the University of Hannover. In 1980 Pagura returned to Hannover and stayed with Majid Samii for 1 month; at that time, Peter Janetta was also there showing his technique of neurovascular decompression for trigeminal neuralgia. This procedure became the one he is most fond of. In 1983 he obtained his PhD with the thesis entitled ‘‘Percutaneous radiofrequency spinal rhizotomy’’ at the Paulista Medical School, Federal University of Saõ Paulo, which was later published [59]. In the same year, he went to Duke University where he spent 40 days with Blaine Nashold learning the DREZ procedure. In 1996 Pagura became Full Professor and Head of the Division of Neurosurgery at ABC Medical School. His main areas of expertise are trigeminal neurovascular decompression [60], percutaneous radiofrequency glossopharyngeal rhizotomy [61], DREZotomy for pain and spasticity, and neuronavigation. His most important contributions to the literature are exactly in the two first topics above mentioned. His paper on glossopharyngeal rhizotomy, a classic in this issue, is still frequently cited [61]. Jorge Pagura was president of the Brazilian Society for Stereotactic and Functional Neurosurgery in the biennium 2002/2004. Also worth mentioning was his role in the creation of the Brazilian Society for the Study of Pain (Sociedade Brasileira para o Estudo da Dor – SBED), the Brazilian branch of the International Association for the Study of Pain. The SBED was created by initiative of Pagura and the neurologist Moacir Schnapp. Pagura was elected the second president of this society. Jose´ Oswaldo Oliveira Jr (> Figure 16-9a) graduated from the USP Medical School in 1977, institution where he also did his residence training from 1978 to 1982. In 1984 he obtained his master degree, and in 1986, his PhD, both from the USP Medical School. His training in functional neurosurgery was done in Edinburgh,

under the supervision of Edward Hitchcock, in 1985 (7 months), and in Paris, supervised by Gabor Szikla, in 1986 (9 months). A very inventive person, in the 1980s, time in which the radiofrequency generators (RFG) in Brazil had to be imported, with technical assistance, he built his own equipment, not for commercial purposes, but for his personal use. Two other versions of this RFG were later developed. The main drawback of his equipment was the lack of temperature monitoring, which had to be inferred based on the amperage. Placement of the DBS electrode is, to say the least, a quite boring and time consuming maneuver, as everyone involved with this technique knows. In 2005, Oliveira Jr developed the DBS electrode placer (> Figure 16-9b), later commercialized by Micromar. This very simple equipment makes DBS electrode placement an easy and fast procedure. Besides, he also developed a head and electrode holder for percutaneous cordotomy, a kit for transsphenoidal chemical hypophysectomy, and a number of electrode kits. Oliveira Jr heads the Department of Functional Neurosurgery, Antalgic Therapy and Palliative Care at AC Camargo Hospital, the most important and antique institution for cancer treatment in Brazil, and the Functional Neurosurgery Service at the Division of Neurosurgery at Servidores Pu´blicos Estaduais Hospital, both in Saõ Paulo. Although also a very versatile functional neurosurgeon, his main area of interest and expertise has always been pain surgery. During his very active career, he has formed more than 20 functional neurosurgeons. Claúdio Fernandes Correâ (> Figure 16-10) graduated from the Federal University of Triaˆngulo Mineiro in 1978 and did his neurosurgical residence (1979/1982) at Beneficeˆncia Portuguesa Hospital, in Santos, Saõ Paulo State. Correâ obtained both his master degree (spinal cord stimulation for refractory pain, 1994) and PhD (brachytherapy for brain tumors, 1999) at the Paulista Medical School, Federal University


16

. Figure 16-9 (a) Jose´ Oswaldo Oliveira Jr (b) Micromar electrode placer

. Figure 16-10 Claúdio Correâ

of Saõ Paulo. His formation in stereotactic and functional neurosurgery was initially supervised by Carlos Telles (1986/1987), and later by Manoel Teixeira. At present he heads the Stereotactic and Functional Neurosurgery Service at Nove de Julho Hospital, in Saõ Paulo. His main areas of expertise are pain surgery and stereotactic neurooncology. So far Correâ has performed more than 1,300 microcompressions of the gasserian ganglion [62], which probably represents one of the largest series in the world. His experience in brachytherapy is also impressive [52], seconding only that of

Teixeira in our country. Correâ was president of the Brazilian Society for Stereotactic and Functional Neurosurgery in the biennium 1996/1998 and the founder of SIMBIDOR, a biennial multidisciplinary pain meeting, with already eight versions (first version: 1994), undoubtedly one of the most important Brazilian pain meetings. Arthur Cukiert graduated from the USP Medical School in 1986, completing his neurosurgical training at the same institution in 1990 (1987/1990). He then moved to Montreal Neurological Institute, where he trained in epilepsy and stereotactic surgery with Andre Olivier (1991/1992). Since he returned to Brazil Cukiert has headed the Division of Neurology and Neurosurgery at Brigadeiro Hospital, in Saõ Paulo. In 1996, he obtained his PhD at the USP. Almost all his practice has been dedicated to epilepsy and pituitary tumor surgery. So far he has performed more than 1,200 epilepsy surgeries. Cukiert has published extensively on all aspects of epilepsy [63–68]. His main interests in this area are the normal and abnormal function of the corpus callosum and the development of clinical and flowchart paradigms for the medical and surgical treatment of epilepsy. He is one of the authors of the first epidemiological paper on

215

216

16


epilepsy released in Brazil [63]. Cukiert was also the first in the country to perform deep brain stimulation for the treatment of refractory epilepsy (anterior thalamic nucleus). He is author of five books: three on epilepsy and two on neuroendocrinology. Jack Beraha graduated from the Pontifical Catholic University of Saõ Paulo Medical School in 1978. He did his neurosurgical training at Oswaldo Cruz Hospital (Saõ Paulo), supervised by Darcy Vellutini, going later to Geneva, where he accompanied the services of A. Werner and J. Berney. He then moved on to the University of Genebra, where he obtained his PhD. As he became interested in radiosurgery he went to the University of Valencia Medical School, staying with Barcia-Salorio in 1982/1983, and afterwards to Buenos Aires and Charlottesville, accompanying Ladislau Steiner for a period of 21=2 months. In 1987, Beraha became Full Professor at the Pontifical Catholic University of Saõ Paulo Medical School. Until 1983 radiosurgery was performed by using gamma-knife, telecobaltotherapy, proton beam, and betatron. During his stay with Barcia-Salorio, he learned about telecobaltotherapy radiosurgery. Using a principle similar to that of Barcia-Salorio, Beraha, helped by the physicists Luiz Scaff and Dirceu Vizeu, modified a 4 MeV linear accelerator (LINAC) at Oswaldo Cruz Hospital, in Saõ Paulo, to make it radiosurgery-compatible, and in July of 1983 Beraha et al performed the first LINAC radiosurgery in the world in a patient harboring a craniopharyngioma [69]. From then until 1989 he treated more than 100 patients presenting with intracranial tumors or arteriovenous malformations [69,70]. One of these patients, who had a vestibular schwannoma that apparently failed to respond to radiosurgery, was subsequently operated on by a very skillful and famous neurosurgeon. After the operation, this patient presented significant complications, and the neurosurgeon, seemingly in an attempt to get rid of the possible

implications of the poor result, told the patient’s family that the sequel was consequence of the prior radiosurgery. Beraha was sued by the patient’s family, but even having been proved innocent, as expected, the number of referrals declined drastically, and his radiosurgery service was finally closed in 1989. It must be remembered, however, that the world owes Jack Beraha the clinical introduction of LINAC radiosurgery. Also worth mentioning is the fact that this story was told us not by Beraha, but by a number of other colleagues. Nilton Luis Latuf (1937–2005) graduated from the USP Ribeiraõ Preto Medical School in 1963, where he also attended his residence in neurosurgery (1964/1966). In 1967 Latuf went to Paris and spent 1 year with Gerard Guiot at Foch Hospital [71]. Back to Brazil and to Ribeiraõ Preto, in Saõ Paulo State, he was the first functional neurosurgeon in Brazil to establish his practice in the countryside. His first functional procedure, a thalamotomy in a parkinsonian patient, was performed by using the Guiot frame, and it took place at Santa Casa de Miserico´rdia Hospital, in Ribeiraõ Preto, in October 1968. In 1969 Latuf inaugurated the Division of Neurology and Neurosurgery in the same institution, heading it for 25 years. He formed 48 residents, and performed more than eight thousand neurological surgeries throughout his life [71], a significant amount of these constituted by stereotactic and functional neurosurgery procedures. Latuf was a member of the executive council of the Brazilian Society of Neurosurgery for 21 years, as well as the South American Representative and Honorary President of the World Association of Lebanese Neurosurgeons for many years. A number of other colleagues in Saõ Paulo State have established a busy practice in the field of stereotactic and functional neurosurgery, among them: Erich Fonoff (Saõ Paulo), Nilton Lara (Saõ Paulo), Salomon Benabou (Saõ Paulo), Valter Cescato (Saõ Paulo), Antoˆnio de Almeida


(Saõ Paulo), Evandro de Souza (Saõ Paulo), Soraya Cecı´lio (Saõ Paulo), Jose´ Claúdio Marinho da No´brega (Saõ Paulo), Luis Augusto Rogano (Saõ Paulo), Edson Amaˆncio (Santos), Arthur Ungaretti Jr (Santo Andre´), Joaõ Alberto Assirati Jr (Ribeiraõ Preto), Marcus Colbachini (Ribeiraõ Preto), Sebastiaõ Carlos da Silva Jr (Saõ Jose´ do Rio Preto), Kleber Duarte (Saõ Jose´ do Rio Preto), Carlos Tadeu Parisi Oliveira (Bragança Paulista), Guilherme Cantatore Castro (Bragança Paulista), Jose´ Paulo Montemor (Campinas), and Edmur Piza Filho (Jundiaı´).

Minas Gerais State The pioneer of stereotactic and functional neurosurgery in Minas Gerais State was Jose´ de Arau´jo Barros. Jose´ de Arau´jo Barros (> Figure 16-11a) graduated from the Federal University of Minas Gerais Medical School and in 1953 he finished his initial training in neurosurgery with Francisco Rocha, a disciple of Herbert Olivecrona and one of the pioneers of neurosurgery in Minas Gerais State. In 1954 he moved on to the University of Illinois in Chicago, spending 2 years under Percival Bailey. During his fellowship, he went several times to New York to learn stereotactic surgery with Irving Cooper. Back to Belo Horizonte, capital of Minas Gerais State, Barros constructed his own frame (> Figure 16-11b), based on that of Cooper, and started performing stereotactic surgery in 1960. This was the first stereotactic apparatus built in Brazil. Along the years he performed 248 procedures in 186 patients harboring movement disorders, mainly Parkinson’s disease, but also dystonia, hemiballism, and choreoathetosis. In 1961 he became Full Professor of Neurology at the Minas Gerais Medical Sciences School, position that he occupied until his retirement in 1998. A very active neurosurgeon at Saõ Jose´ Hospital, the teaching hospital of the Medical

16

Sciences School, Barros formed 58 neurosurgeons. One of his ex-residents, Lourenço de Freitas Neto, replicated his frame and was the pioneer in the stereotactic field in Espı´rito Santo State. At present, five are the best known and active functional neurosurgeons in Minas Gerais: Jose´ Maurıćio Siqueira, Gerva´sio Teles Cardoso de Carvalho, Sebastiaõ Nataniel Silva Gusmaõ, Rodrigo de Mattos Labruna, and Gilberto de Almeida Fonseca Filho. Jose´ Maurıćio Siqueira graduated from the Federal University of Minas Gerais Medical School in 1976 and attended neurosurgical residence at Felıćio Roxo Hospital (1977/1979), in Belo Horizonte, previously one of the two teaching hospitals at the Federal University of Minas Gerais. In 1981 he went to the University of Toronto, spending 1 year as a fellow under Ronald Tasker. Consecutively, he accompanied Claude Bertrand at the Notre Dame Hospital for 2 months and Patrick Kelly, at the New York State University at Buffalo, for 1 month. Some years later he also spent a short period (1 month) with Erik Backlund at the Karolinska Institute in Stockholm. Siqueira made two major contributions to the field of stereotactic and functional neurosurgery. In 1985 Siqueira was the first in the literature to describe the technique of open bulbospinal trigeminal nucleotomy through radiofrequency lesioning for the treatment of refractory facial deafferentation pain [72]. This paper, in which he reported the 1-year follow-up results obtained in two patients operated on in 1983 (complete abolition of pain was achieved in both patients), had been previously presented as a poster at the IX Meeting of the World Society for Stereotactic and Functional Neurosurgery in Toronto, in 1985, chaired by Ronald Tasker. In 1986, Blaine Nashold et al published an abstract in Neurosurgery describing a technique identical to that of Siqueira, claiming that this ‘‘new surgery’’ had

217

218

16


. Figure 16-11 (a) Jose´ de Arau´jo Barros (b) Stereotactic frame built by Barros

been developed by his group [73]. Actually, Nashold has never mentioned Siqueira’s pioneer paper in any of his publications. In 1987, in a letter sent to the editor of Neurosurgery, Schvarcz contested the idea that Nashold was the original author of the technique and stated that Hitchcock and himself, in 1972, were the first to report that technique, using the stereotactic approach, and mentioned Siqueira’s poster, who had performed the procedure through an open operation [74]. Schvarcz, however, failed to mention that the very first to perform this procedure by employing the stereotactic technique was Edward Hitchcock alone in 1970 [40]. In 1997, during the 11th Congress of the World Federation of Neurosurgical Societies, held in Amsterdam, Tasker introduced Siqueira to John Gorecki, an ex-resident of Nashold’s, and told him about the pioneer paper of Siqueira. Afterwards, Siqueira sent Gorecki a reprint of his paper, as well as the certificate of his poster presentation of 1985 at the World Society for Stereotactic and Functional Neurosurgery Meeting in Toronto. In 2004 Gorecki, in his chapter on DREZ and brainstem ablative procedures for Youmans Neurological Surgery fifth edition, finally gave Siqueira the deserved credit, reputing his report in this issue as the first published in the literature [75].

During his fellowship at the University of Toronto, Siqueira helped to review Tasker’s experience on unilateral Vim thalamotomy for the treatment of severe drug-resistant parkinsonian tremor. The conclusions derived from this study, published in 1983, when very few was being said about the surgical treatment of Parkinson’s disease, were very instrumental for the renascence of interest in this modality of treatment [76]. Currently, Siqueira heads the Services of Functional and Epilepsy Surgery of the Division of Neurology and Neurosurgery at Felıćio Roxo Hospital. Gerva´sio Teles Cardoso de Carvalho graduated from the Minas Gerais Medical Sciences School in 1981. He finished his residence training at Santa Casa de Miserico´rdia de Belo Horizonte Hospital in 1985 (1982/1985), and spent 1 year (1986) with Carlos Telles at Pedro Ernesto Hospital, the teaching hospital at the State University of Rio de Janeiro. In 1987, following the lead of Siqueira, he went to the University of Toronto and spent 1 year under Ronald Tasker (1987/ 1988). In 1995 he obtained his master degree from the Federal University of Saõ Paulo. Carvalho has a vast experience in the surgical treatment of pain [77–79], psychiatric and movement disorders [80], and coordinates the


Stereotactic and Functional Neurosurgery Service at the Division of Neurosurgery at Santa Casa de Miserico´rdia de Belo Horizonte Hospital. During the time he was a fellow under Tasker, he was part of the team that demonstrated that, in patients suffering from cord injury pain, the intermittent neuralgic and evoked elements of neuropathic pain differed from the steady causalgic/dysesthetic element in their response to destructive and modulatory procedures (the latter, contrarily to the first two elements, which are frequently relieved by interruption of the pain pathways, responds better to stimulation of the spinal cord or the brain), suggesting the presence of distinct mechanisms underlying these different modalities of pain [77–79]. Sebastiaõ Nataniel Silva Gusmaõ graduated from the Minas Gerais Medical Sciences School in 1973 and did his neurosurgical residence in the same institution (1974/1976). Three years later, to polish his initial formation, he spent 1 year at the Department of Neurosurgery at the National Hospital for Nervous Diseases, in London, and a shorter period at the Louis Pasteur University Medical School, in Strasbourg. Gusmaõ obtained his master degree from the Federal University of Minas Gerais in 1984, and his PhD from the Federal University of Saõ Paulo in 1993. His training in stereotactic and functional neurosurgery was done under Carlos Telles, in Rio de Janeiro, and also, for a shorter period, at Stockholm University. Gusmaõ was president of the Brazilian Society for Stereotactic and Functional Neurosurgery in the biennium 1992/1994 and is currently Full Professor of Neurosurgery and Head of the Division of Neurosurgery at the Federal University of Minas Gerais. Pain surgery constitutes his main area of expertise in the field of stereotactic and functional neurosurgery, and from this area derived his most important contribution. Gusmaõ was the first in the literature to propose the use of intraoperative computed tomography to guide percutaneous

16

radiofrequency trigeminal rhizotomy, greatly facilitating the identification of the foramen ovale [81]. Two other very experienced colleagues from Minas Gerais are Rodrigo de Mattos Labruna, trained by Teixeira, and Gilberto de Almeida Fonseca Filho, trained by Teixeira, Dieckmann and Ostertag. Their main areas of expertise are pain and spasticity surgery. Labruna has also a significant experience in surgery for psychiatric disorders.

Espı´rito do Santo State The pioneer of stereotactic and functional neurosurgery in Espı´rito Santo State was Lourenço de Freitas Neto. Freitas Neto did his residence training in Belo Horizonte under Jose´ de Arau´jo Barros, who introduced him to stereotactic surgery. He then moved back to Vito´ria, the capital of Espı´rito Santo, in 1968, where, after replicating the frame developed by Barros, he started performing stereotactic surgery for movement disorders. Until 1984 he performed 32 procedures, basically for Parkinson’s disease, occasion in which he abandoned functional neurosurgery due to the lack of candidates for surgery. Se´rgio Otoni, after a 6-month training with Edward Hitchcock, returned to Vito´ria in 1978, but never really dedicated to stereotactic and functional neurosurgery. Pedro Menezes, trained by Teixeira, worked part of the time in Saõ Paulo and part in Vito´ria. Between 2000 and 2006, while in Vito´ria, he performed a significant number of functional procedures. Only recently (2003), though, Walter Fagundes-Pereyra for the first time established a really active and ongoing practice in functional neurosurgery in Espı´rito Santo (Vito´ria). Fagundes-Pereyra was first exposed to stereotactic and functional neurosurgery by Gerva´sio Carvalho, during his neurosurgical residence at Santa Casa de Miserico´rdia de Belo Horizonte Hospital. In 2001 he obtained his master degree from the same institution. He

219

220

16


then moved on to Lille, where, during 1 year and a half, he refined his training in stereotactic and functional neurosurgery with Serge Blond. In 2007 he obtained his PhD from the Federal University of Saõ Paulo.

Midwest Region The Midwest region is composed by the states of Goia´s, Mato Grosso, and Mato Grosso do Sul and the Federal District.

Goia´s State The pioneer of stereotactic and functional neurosurgery in the Midwest region was Luiz Fernando Martins (> Figure 16-12), undoubtedly one of the most renowned and important functional neurosurgeons in Brazil and South America. Martins graduated from the Federal University of Goia´s Medical School in 1971. His internship

. Figure 16-12 Luiz Fernando Martins

(1971), the last year (sixth) of Medical School in the Brazilian university system, was spent with Renato Barbosa, at Lagoa Hospital, in Rio de Janeiro. Barbosa was the one who first introduced him to stereotaxis. Since then, and certainly forever, Martins became a victim of a platonic love for this marvelous field of neurosurgery. In 1972 he went to the University of Berlin, where he stayed for a period of 4 years, doing his neurosurgical residence under Wilhelm Umbach, at the time one of the most important functional neurosurgeons in Germany. In 1975 Martins obtained his PhD from the University of Berlin, and in 1976, back to Brazil, he started his practice in Goiaˆnia, the capital of Goia´s, joining the team of the Goiaˆnia Neurological Institute, a private hospital founded the year before, then and now one of the most important neurosurgical institutions in the country. On 13 February 1976, using a Riechert and Mundinger apparatus, he performed his first stereotactic procedure, lesioning the zona incerta in a patient presenting with Parkinson’s disease. In 1982 he returned to Europe, spending more than 1 year: 3 months with Brock, in Berlin; 2 months with Dieckmann, in Go¨ttingen; 7 months with Siegfried, in Zurich; and 2 months with Mundinger, in Freiburg. A team player, Martins has always proclaimed the need of a multidisciplinary group to evaluate surgical candidates in all areas of functional neurosurgery. Though performed in an irregular basis since 1976, only in 1989 the epilepsy surgery program was inaugurated at Goiaˆnia Neurological Institute. Thanks to this government-supported program, for which implementation were instrumental Martins, the neurosurgeon Orlando Arruda, and the epileptologist Paulo Ragazzo (trained at the Montreal Neurological Institute), 1,782 patients, not only from the Midwest, but also from the North and Northeast of the country could be operated on in this institution. These operations should be credited to the following team of neurosurgeons: Luiz Fernando Martins, Orlando Arruda (already


dead), Valter da Costa, Joaõ Arruda, Joaquim da Costa, and more recently, Henrique Lobo, Marcelo Martins, and Osvaldo Vilela Filho. In 1999 Martins innovated again, creating the first radiosurgery service in the Midwest region, and one of the first in the country. This service is a joint project of Goiaˆnia Neurological Institute and CEBROM (Brazilian Center of Radiotherapy, Oncology and Mastology), a private outpatient clinic dedicated to oncology, radiotherapy and chemotherapy and owner of the LINAC used in the procedures. Since then, three hundred and 16 radiosurgeries have been performed. Both epilepsy surgery and radiosurgery programs are clear examples of a multidisciplinary team work. Martins is a highly experienced neurosurgeon in all areas of stereotactic and functional neurosurgery, but spasticity. Throughout his outstanding career he has performed, besides those aforementioned, 402 surgeries for movement disorders (mainly Parkinson’s disease); 714 for pain (mainly trigeminal neuralgia); 468 for psychiatric disorders (mainly aggressiveness and OCD), probably one of the largest series in the world at present; 186 stereotactic biopsies; and 36 stereotactic-guided craniotomies. Martins was one of the founders of the Brazilian Society for Stereotactic and Functional Neurosurgery, being its president for two terms (1984/1986 and 2000/2002), and of the recently created Brazilian Society of Radiosurgery (2007), which has as president the functional neurosurgeon Salomon Benabou, and is now president of the SLANFE (Latin American Society for Stereotactic and Functional Neurosurgery). Though author of just a few papers [82–88], he made two significant scientific contributions: the determination of the radiographic stereotactic coordinates of the foramen ovale [82], and the correlation between the position and size of medial thalamic lesions and the degree of pain relief obtained by patients presenting cancer pain (an autopsy study) [83,84]. Another important contribution, although not published, was his

16

proposal of combined amygdalotomy and anterior capsulotomy for the treatment of refractory aggressiveness (1985). But as he says, ‘‘I am not a writer, but a surgeon. I like to operate and also to lecture, this is how I have chosen to pass along my experience.’’ A superb lecturer, he played a very important role spreading the knowledge concerning surgical treatment of psychiatric disorders, movement disorders, and epilepsy across the country. Unfortunately, the lack of habit of writing papers is a characteristic shared by the vast majority of Brazilian neurosurgeons, which seems to be a cultural deficiency. How much more could someone like him have contributed to the field by writing down such a huge experience. There are six other functional neurosurgeons in Goia´s State, five of them from Goiaˆnia and one from Ana´polis (Nivaldo Evangelista Teles). Four of those from Goiaˆnia are part of the team of the Goiaˆnia Neurological Institute (Joaquim da Costa, Joaõ Batista Arruda, Vladimir Arruda Zaccariotti, and Osvaldo Vilela Filho), and the other one has his Service at Santa Helena Hospital (Hamilton Ayres da Silva). Hamilton da Silva, once finished his 5-year residence in neurosurgery (1987/1991) at Servidores Pu´blicos Estaduais Hospital (Saõ Paulo), decided to dedicate his career to stereotactic and functional neurosurgery and spent 4 years (1992/ 1995) with Teixeira at USP Medical School and 6 months (1996) with Ronald Tasker at University of Toronto. Eventually he returned to Goiaˆnia and established his practice at Santa Helena Hospital and Goiaˆnia General Hospital. Most of his practice in stereotactic and functional neurosurgery has been devoted to the surgical treatment of pain. Joaquim da Costa had his initial contact with stereotaxis as a resident under Jose´ de Arau´jo Barros, the pioneer of stereotactic and functional neurosurgery in Minas Gerais State, but his true master was Martins, with whom he works until now. In 1983 he spent a short period with Falk Oppel in Berlin learning epilepsy surgery.

221

222

16


Like Martins, he is a very experienced functional neurosurgeon. Joaõ Arruda and Vladimir Zaccariotti did their neurosurgical residence at Goiaˆnia Neurological Institute. Both had Martins as their first master in stereotactic and functional neurosurgery. Arruda completed his training with Blaine Nashold at Duke University (1 year), and Zaccariotti, with Falk Oppel (17 months), in Bielefeld, and Antonio De Salles, at UCLA (7 months). Arruda works both at Goiaˆnia Neurological Institute and Arau´jo Jorge Hospital (Cancer Hospital), where he runs the Neurosurgical Service, devoted mainly to neurooncology, treatment of pain and radiosurgery, areas in which he has a large experience. Besides, he is also a very experienced epilepsy surgeon. Zaccariotti works with him in both institutions. Osvaldo Vilela Filho (> Figure 16-13) graduated from the Federal University of Goia´s (Universidade Federal de Goia´s – UFG) Medical School in 1984. During his last 2 years of Medical School, he accompanied many functional procedures performed by Luiz Fernando Martins at Goiaˆnia Neurological Institute, one of his . Figure 16-13 Osvaldo Vilela Filho

personal icons. This simple fact would later influence his whole career. His neurosurgical residence (1985/1988) was done at the Brası´lia District Hospital, in Brası´lia, being chief resident of neurosurgery in 1988. Unfortunately, stereotactic and functional neurosurgery was not performed in this institution at that time. Worried about the difficulty to approach some lesions of difficult location, mainly those deeply placed, Vilela Filho (1987) developed a technique that allowed the transposition of the lesions seen on CT-scan to the scalp, greatly easing the surgical procedure [89]. Later, before his stereotactic training, he used the same technique to perform CT- or ultrasound-guided free-hand biopsy of intracranial lesions [90]. In February of 1991 he went to the Toronto Western Hospital, University of Toronto, to do a fellowship in stereotactic and functional neurosurgery under Ronald Tasker, staying until 1992. While at UT he reviewed all patients with neuropathic pain submitted to DBS by Ronald Tasker. This research led to a number of important publications [91–95]. Tasker became another of his icons, with whom he established a close friendship that remains until now, someone that he gladly reputes as his Canadian father. In 1993 he returned to Canada, this time to the University of Western Ontario, in London, to learn epilepsy surgery under the supervision of Andrew Parrent and John Girvin. Later he went to other centers of excellence in stereotactic and functional neurosurgery for shortperiod visits, like those of Roy Bakay, at the time at Emory University (2000); Andres Lozano, at University of Toronto (2000); and Ali Rezai, at Cleveland Clinic (2004). In 2006 he obtained his PhD from the Paulista Medical School, Federal University of Saõ Paulo. Back to his hometown, Vilela Filho inaugurated and headed the Goiaˆnia Brain Institute, devoted to neurosurgery and neurology, in general, but particularly to stereotactic and functional neurosurgery. Initially located at the Goiaˆnia Orthopedic Institute (1993/2001), it


was later moved to Lucio Rebelo Hospital (2001/ 2003). In 1994 he underwent a competitive examination for neurosurgeon of the UFG Medical School, achieving the highest score, and since 1995 he has worked as an Invited Professor of Neurosurgery in the Department of Surgery of the same institution. After 3 years of continuous insistence, the Clinic Hospital, teaching hospital of the UFG Medical School, finally acquired the necessary equipment and in 1998 he created the Stereotactic and Functional Neurosurgery Service at the Division of Neurosurgery at the Clinic Hospital, which he has chaired since then. This was the first university and public hospital in the Midwest region of the country to offer stereotactic and functional neurosurgery. In 2001 Vilela Filho inaugurated the Laboratory of Experimental Stereotaxis at the UFG Medical School, which he still runs. In 2002 he became a consultant in neurosurgery and neurology to the Council of Healthcare Advisors, which was presided by the Nobel Laureate James Watson (dead in 2004), and in 2003 he became part of the team at the Goiaˆnia Neurological Institute. Since 2006 he has worked as an Invited Professor of the Department of Medicine at the Catholic University of Goia´s, becoming Associate Professor of Neurosciences after a competitive examination in 2007, once more achieving the highest score. Vilela Filho has had a very intense associative activity. Still as a student, he became a member of the Society for the Defense of Natural Resources (1977), a society made up almost exclusively by professors and researchers of the local universities (Federal and Catholic Universities of Goia´s), serving as its first secretary (1980/1982) and president (1982/1984). In 1997 he founded the Young Neurosurgeons Committee of the Brazilian Society of Neurosurgery, becoming its coordinator for three consecutive terms (1997/2002), and in the same year he was the chairman of the Scientific Committee of the VII National Congress of the Brazilian Academy of Neurosurgery. Again in 1997 he was one of the

16

founders and the first president of the Goia´s Association for the Study of Pain, position that he held for three terms (1997/2002), period in which he presided two important international congresses on pain, the I and II INDOR (International Congress on Pain). In 1999 he was also one of the founders and the first president of the Goia´s Society of Neurosurgery (1999/2002), presiding in 2001 its first congress, the first joint meeting Brazil & Canada on neurosurgery. For two consecutive terms he served as member of the International Relations Committee of the Brazilian Society of Neurosurgery (2000/2004), and in 2002 he was elected member of the Executive Council of this society (2002/2008). In 2004 he chaired the Scientific Committee of the XXV Brazilian Congress of Neurosurgery, organized by the Brazilian Society of Neurosurgery. Vilela Filho has been a true soldier of the Brazilian Society for Stereotactic and Functional Neurosurgery, having served as its secretary (2000/2002), vice-president (2004/2006), and president (2006/2008). In 2007 he presided the Eighth Congress of the Brazilian Society for Stereotactic and Functional Neurosurgery & First International Joint Meeting on Stereotactic and Functional Neurosurgery, undoubtedly the most important meeting organized by the Brazilian society. During this event was founded the Brazilian Neuromodulation Society, the Brazilian branch of the International Neuromodulation Society, and Vilela Filho was elected its vicepresident. In 2008 he became secretary of the Latin American Society for Stereotactic and Functional Neurosurgery (SLANFE). He is also member of a number of other societies, including the World and American Societies for Stereotactic and Functional Neurosurgery, International Neuromodulation Society, Brazilian Society for the Study of Pain, and American Association of Neurological Surgeons (1995/2002). Vilela Filho is member of the editorial board of three international journals (Neuromodulation; Brazilian Contemporary Neurosurgery, the

223

224

16


Brazilian supplement of Surgical Neurology; and Neurotarget, the official journal of the Latin American Society for Stereotactic and Functional Neurosurgery) and six Brazilian journals (Neurocirurgia Contemporaˆnea Brasileira; Jornal Brasileiro de Neurocirurgia; Dor – 2002/2004; Simbidor; Jovem Me´dico; and Pra´tica Hospitalar), and was Associate Editor of the SBN Newsletter, the bulletin of the Brazilian Society of Neurosurgery (2002/2004). Throughout his career, Vilela Filho and his team have made numerous original scientific contributions, besides those aforementioned. Reviewing Tasker’s 16 patients with neuropathic pain of spinal cord origin submitted to DBS (thalamic ventrocaudal nucleus – VC and periventricular gray matter – PVG), he identified four patients with complete cord injury. PVGDBS was performed in three of these patients to treat the evoked and intermittent components of their pain, being unsuccessful in all them. VC-DBS, however, was successful in three out of four patients. Based on these results he contested previously proposed hypotheses, according which VC-DBS produces pain relief by inhibition of dorsal horn nociceptive neurons through a variety of pathways, which could not be at work in patients presenting complete cord transection [96,97]. Alternatively, he proposed that VC-DBS induces pain relief by inhibition of medial thalamus nociceptive neurons through a polysynaptic pathway, as follows: VC stimulation » somatosensory cortex excitation (» motor cortex activation) » anterior putamen excitation » peptidergic (substance P) activation of GPi/ SNR » medial thalamus inhibition [97]. Vilela Filho later proposed that this polysynaptic pathway would be operative even under physiological circumstances and could work as a modulatory pain center, which led him to name it as prosencephalo-mesencephalic modulatory circuit [98,99]. He also suggested that the interruption of this circuit or of its activating pathway, which he supposed to be the anterior spinothalamic

tract, by determining overactivation of the medial pain system, would be the pathophysiological substrate of the steady dysesthetic component of neuropathic pain [98,99]. According to his hypothesis, stimulation of various structures such as the peripheral nerves, spinal cord, medial lemniscus, VC, internal capsule and motor cortex induces pain relief by activation of this circuit [100,101]. Still in the field of pain surgery, Vilela Filho introduced the CT-guided percutaneous technique to perform punctuate midline myelotomy [102]; proposed the use of doppler-scan for localization of the occipital artery, which lies just lateral to the greater occipital nerve, greatly facilitating the realization of percutaneous radiofrequency occipital neurotomy and occipital nerve stimulation [103]; and was the first to propose the use of MRI for postoperative evaluation of percutaneous cordotomy [104]. It is well known that functional neuroimaging studies may frequently show abnormalities in patients harboring psychiatric disorders and that these abnormalities tend to disappear after successful surgical treatment. Vilela Filho, however, was the first to propose the use of functional studies to determine the best target to treat each patient. The idea was to surgically deactivate the areas shown to be hyperactive in the functional studies [105]. Some other studies performed by other authors, based on the principle suggested by Vilela Filho, led to the discovery of Brodmann area 25 as a target for depression [106] and of the posteromedial hypothalamus as a target for cluster headache [107]. Vilela Filho and Souza, in 1996 [108], and again in 1998 [109], originally proposed that Tourette syndrome is the clinical manifestation of the hyperactivity of the globus pallidus externus and prefrontal area. Based on this hypothesis, Vilela Filho et al, in 2004, were the first to perform bilateral GPe-DBS for the treatment of Tourette syndrome refractory to conservative management [110]. Four patients have been operated on so far. The excellent results obtained


lend support to the presented hypothesis [111– 113]. In 2001, Vilela Filho proposed a modification in the technique of stereotactic subcaudate tractotomy, using landmarks seen on CT-scan instead of those seen in the skull X-ray, as it was usually performed [114]. Vilela Filho and Silva current series of 54 patients harboring Parkinson’s disease submitted to unilateral subthalamotomy is one of the largest in the literature. They have shown that STN lesioning is a highly effective and safe procedure [115–117]. A possible complication of this procedure is hemiballism, which occurred in 11% of the patients. These authors demonstrated that damage to the dorsolateral territory of STN and concomitant sparing of the zona incerta seem to be essential for the development of hemiballism, and that the presence of intraoperative stimulation-induced dyskinesia and, possibly, levodopainduced dyskinesia apparently are significant risk factors for the development of this complication [117,118]. In 2001 Vilela Filho et al reported the case of a patient with unilateral essential tremor whose tremor completely disappeared after a stroke restricted to the contralateral posterior putamen, leading him to originally propose a role for the basal ganglia in the genesis of essential tremor, possibly through the hyperactivity of the posterior putamen [119]. To test this hypothesis, they used an animal model of essential tremor (harmaline-induced tremor) and showed that the harmaline-induced tremor could be greatly reduced by an electrolytic lesion placed in the ipsilateral posterior striatum of the rat, giving further support to the proposed hypothesis [120]. Vilela Filho et al., in 1997, introduced an apparently low-risk and accurate technique for CT-guided free-hand percutaneous biopsy of spinal cord tumors [121]. In 2001 Vilela Fiho and his colleagues developed a new product that did not increase the inherent MRI distortion and could be seen in the main MRI sequences used for stereotactic surgery (T1-weighted,

16

T2-weighted and inversion recovery). This product, however, was never patented or commercialized [122]. Vilela Filho and Carneiro Filho, in 2002, were the first to introduce the gamma probe in neurosurgery, which was initially used for intraoperative detection of brain tumors and to ascertain its complete removal [123]. In 2006 the same authors again pioneered using the gamma probe for intraoperative detection of epileptogenic areas during epilepsy surgery, a technique still in investigation that seems particularly helpful in non-mesial temporal lobe epilepsy [124]. Finally, Vilela Filho and Carneiro Jr developed an MRI-based frameless technique to determine the stereotactic coordinates. A very precise technique, it allows to perform sort of an image fusion (CT and MRI) without the necessity of any software other than those of the MRI and CT scanners [125].

Mato Grosso State The pioneer of stereotactic and functional neurosurgery in Mato Grosso State was Jony Soares Ramos. Ramos did his residence training at Lagoa Hospital, in Rio de Janeiro. The Head of the Neurosurgery Service, Ney Monteiro, who supervised his residence, was one of the disciples of Renato Barbosa, the pioneer of stereotactic and functional neurosurgery in Brazil. During the last year of his training (2001), he completed his formation in pain surgery accompanying Carlos Telles at Pedro Ernesto Hospital, the teaching hospital at the State University of Rio de Janeiro. In 2002 he started his practice in Cuiaba´, the capital of Mato Grosso, initially using an antique Riechert and Mundinger apparatus, which he had bought from Renato Barbosa, already retired for a long time, and later he acquired the Micromar stereotactic system. From the beginning to 2004, he worked with Bruno Silveira, who trained with Teixeira, but Silveira stopped functional neurosurgery in 2004, and Ramos carried on by

225

226

16


himself. Already performing stereotactic procedures for movement disorders, pain and biopsy for some years, in 2007, oriented by Martins, Ramos performed his first operation for the treatment of a psychiatric disorder. He is still the only active functional neurosurgeon in the state.

Mato Grosso do Sul State Two general neurosurgeons have recently started performing stereotactic and functional neurosurgery in Campo Grande, capital of Mato Grosso do Sul State. Cesar Nicolatte did his residence in neurosurgery at the Beneficeˆncia Portuguesa Hospital (1996/2000), in Saõ Paulo, where he learned stereotactic and functional neurosurgery with Paulo Dorsa, arriving in Campo Grande in 2001. Joaquim Oliveira Vieira Jr did his neurosurgical training at USP Medical School (1995/ 1999), accompanying the functional procedures performed by Teixeira, Cescato, and Oliveira Jr. After his residency, he did his PhD at the Brigadeiro Hospital, in Saõ Paulo, oriented by Arthur Cukiert, and started his practice in Campo Grande in 2003. Nicolatte and Vieira Jr, working together, have recently started to perform stereotactic and functional neurosurgery procedures, so far restricted to percutaneous procedures for pain, stereotactic biopsy and stereotactic-guided craniotomy.

Federal District The Federal District, located in Goia´s State, houses Brası´lia, the capital of the country. It seems unbelievable that stereotactic and functional neurosurgery was inaugurated in Brası´lia only in the 1990s, but there are some explanations for that. First, Brası´lia is a very young city, having been founded only in 1960, when it became the country’s capital (until then Rio de Janeiro was the Brazilian capital). Second, Paulo Mello, the

pioneer of neurosurgery in Brası´lia (1961), went to England a few years after its inauguration. Mello spent 4 years (1964/1967) at the Newcastle Upon Tyne Hospitals, becoming acquainted with stereotactic surgery by the hands of John Hankinson. Returning to Brası´lia, he was appointed Head of the Division of Neurosurgery at Brası´lia District Hospital (1971). In 1973 he went to Freiburg and spent 1 month with Mundinger. Still in the seventies the hospital acquired a Riechert and Mundinger apparatus and a radiofrequency generator, but at the time few stereotactic operations were being performed everywhere and Mello, already involved with vascular neurosurgery, felt his interest in stereotactic and functional neurosurgery almost completely vanish. Besides, the surgical room equipped with the system for orthogonal teleradiography has never been built. Third, Carlos Telles, who did his neurosurgical residence at Brası´lia District Hospital, was incited by Paulo Mello to dedicate himself to stereotactic and functional neurosurgery. Accepting the suggestion, he went to Germany to attend a fellowship in this field, where he spent almost 5 years. During his last year there (1980), Telles was visited by Pedro Sampaio, Professor of Neurosurgery and Head of the Division of Neurosurgery at Pedro Ernesto Hospital, the teaching hospital at Rio de Janeiro State University, who invited him to join his team. Telles accepted his invitation and established his practice in Rio de Janeiro in 1980. Brası´lia, then, remained without its so needed functional neurosurgeon. In 1994 the Division of Neurosurgery at ´ Brasılia District Hospital was headed by Carlos Silve´rio de Almeida, who had attended a fellowship in England. Almeida was strongly determined to unearth the old Riechert and Mundinger apparatus and put it at work, and so he did. Orienting his 4th year neurosurgical resident Luiz Claúdio Modesto Pereira, they transposed the lesion seen on the CT-scan to the stereotactic skull radiography and finally performed the first stereotactic


procedure in Brası´lia, the drainage of a deeply placed brain abscess. Luiz Claúdio Modesto Pereira, excited with this first step, decided to dedicate to the field and became the pioneer in stereotactic and functional neurosurgery in the Federal District. In 1995 he visited the services of Christopher Ostertag, in Freiburg (2 months); Robert Goodman, in New York (1 month); and Manoel Teixeira, in Saõ Paulo (1 month). In 1996 he spent 1 month with Rees Cosgrove, in Boston, and 15 days with Ronald Tasker, in Toronto. From 1995 to 1998 his practice in stereotactic and functional neurosurgery was constituted by sporadic procedures and restricted to anesthetic blockades for pain, stereotactic biopsy, stereotactic-guided surgery and spinal cord stimulation. Even though, Pereira was apparently the first in Brazil to perform stereotactic fibrinolysis with rTPA to treat intracerebral hematomas (1995). In 1999 he finally went to Toronto, spending 1 year under Andres Lozano at the University of Toronto, and since his return to Brası´lia (2000) he has established a full and busy practice in stereotactic and functional neurosurgery [126,127]. The second and last functional neurosurgeon in the Federal District so far is Tiago Freitas, who, like Pereira, was a resident at the Brası´lia District Hospital (2002/2005). Graduated from the Federal University of Goia´s Medical School (2001), during his last year of neurosurgical training he intermittently accompanied Vilela Filho at Goiaˆnia Neurological Institute. In 2006 Freitas went to USP Medical School, accompanying Manoel Teixeira for 3 months, and then to the Cleveland Clinic for a 7-month fellowship under Ali Rezai. Still in the same year he started his practice in Brası´lia.

South Region The South region is made up by the states of Parana´, Rio Grande do Sul and Santa Catarina.

16

The pioneer of stereotactic and functional neurosurgery in the South was Renato de Muggiati, from Parana´.

Parana´ State Renato de Muggiati (1932–1996) (> Figure 16-14) graduated from the Federal University of Parana´ in 1955. During his last 2 years at Medical School, he had the opportunity to accompany the neurosurgeon Jose´ Portugal Pinto, nephew of Jose´ Ribe Portugal, who had recently arrived in Curitiba (state capital) after 6 years at Cleveland Clinic under Gardner, inaugurating neurosurgery in this state. As an assistant to Pinto, he helped him to perform a large number of prefrontal lobotomies at Adauto Botelho Hospital. Finished his graduation, he continued at that institution until 1959, occasion in which he went to Freiburg, spending 2 years under Riechert. A Riechert and Mundinger apparatus was the most important piece of his luggage on his trip back to Curitiba. In 1961 he performed the first stereotactic procedure in Parana´ State, a

. Figure 16-14 Renato de Muggiati

227

228

16


thalamotomy in a parkinsonian patient [128]. At the beginning, he built his own electrodes using the spokes from bike wheels. Still in 1961 he was invited to create and head the Division of Neurosurgery at the Clinic Hospital, the teaching hospital of the Federal University of Parana´, in fact the first university neurosurgical service in Parana´. In 1962 he inaugurated the Neurosurgery Service at Santa Casa de Miserico´rdia de Curitiba Hospital, which later became one of the teaching hospitals of the Pontifical Catholic University of Parana´ [128]. Despite being a general neurosurgeon, stereotactic surgery always remained as one of the most important activities in his outstanding career, along which he performed 304 stereotactic procedures, mainly for Parkinson’s disease [129]. The oldest functional neurosurgeon in Parana´ today is Alceu Correia, who trained with Wolfhard Winkelmu¨ller at the University of Hannover for 5 years (1971/1975). Correia has a vast experience in the surgical treatment of pain and was president of the Brazilian Society for Stereotactic and Functional Neurosurgery in the biennium 1994/1996. At present there are eleven other colleagues in Parana´ performing functional neurosurgery. However, stereotactic and functional neurosurgery constitutes the main area of activity of only four of them: Alexandre Novicki Francisco, Murilo Sousa Meneses, Danel Benzecry de Almeida, and Helve´rcio Fernando Polsaque Alves. Alexandre Novicki Francisco graduated from the Federal University of Parana´ in 1992. After finishing his neurosurgical residence at Santa Casa de Miserico´rdia de Curitiba Hospital in 1996, he went to the USP Medical School, spending 6 months with Teixeira (1997). The next step was the University of Toronto, where he did a 1-year fellowship under Ronald Tasker and Andres Lozano (1997/1998). Since his return to Curitiba in 1998 he has coordinated the Stereotactic and Functional Service at the Division of Neurosurgery at Cajuru Hospital, a

teaching hospital at the Pontifical Catholic University of Parana´. His main areas of expertise are the surgical treatment of pain [95,130], spasticity and movement disorders. Francisco was secretary of the Brazilian Society for Stereotactic and Functional Neurosurgery (2006/2008), and is now vice-president (president-elect) of the same society and secretary of the Brazilian Neuromodulation Society, the Brazilian branch of the International Neuromodulation Society (INS). Murilo Sousa Meneses graduated from the Federal University of Parana´ in 1980. He then went to France for a 6-year residence in neurosurgery. During his last year (1987), at Lariboisiere Hospital (Paris) under Claude Thurel, Meneses got most of his training in stereotactic and functional neurosurgery. In 1987 he obtained his master degree, and in 1990 his PhD, both from the University of Picardie, in Amiens. In 1988, back to Curitiba, he became Assistant Professor of Anatomy (neuroanatomy) at the Federal University of Parana´; he is now Associate Professor of the same discipline. His initial formation in stereotactic and functional neurosurgery was later complemented with short-period (1 month) visits to other centers of excellence in the field, such as Mayo Clinic, with Patrick Kelly (1991); Saint Anne Hospital (Paris), with Chudkiewicz (1993); University of South California, with Michael Appuzzo (1994); and in Bordeaux, with Alain Rougier (1998). The areas in which he is most experienced are epilepsy [131–133] and movement disorders [134] surgery. Meneses is the author of three books, one in Parkinson’s disease (two editions), and two in neuroanatomy (one of them with two editions). Daniel Benzecry de Almeida graduated from the Federal University of Para´ in 1990. His neurosurgical residence was done at Paulista Medical School, Federal University of Saõ Paulo (1991/ 1994). In 1995 he started his training in stereotactic and functional neurosurgery, spending 1 year with Teixeira at USP Medical School (1995/1996), and after that, another year with


Oliveira Jr at AC Camargo Hospital (1996/1997). In 1998 he went to Lyon for a 3-month fellowship under Marc Sindou. Later he visited Richard North at Johns Hopkins for a short period (2 weeks). Almeida dedicates most of his practice to pain surgery [135,136]. In 2007 he obtained his master degree from the Federal University of Saõ Paulo. Helve´rcio Fernando Polsaque Alves graduated from the State University of Maringa´ in 1997. Once finished his residency at Santa Casa de Miserico´rdia de Ribeiraõ Preto Hospital (2001), where he was first exposed to stereotactic surgery by Nilton Latuf and Marcus Colbachini, he spent 15 months at AC Camargo Hospital under Oliveira Jr. In 2003 he went to Maringa´, a beautiful and relatively small town in the interior of Parana´, where he established his busy practice in stereotactic and functional neurosurgery, consisting predominantly of surgical treatment of pain and movement disorders. The other colleagues performing stereotactic and functional neurosurgery in Parana´ to a lesser extent are: Ce´sar Vinicius Grande, Sonival Caˆndido Hunhevicz, Sı´lvio Machado, and Samir Ale Bark, in Curitiba; Stenio H. de Souza, in Cascavel; and Pedro Garcia Lopes and Wander Miguel Tamburus, in Londrina.

Rio Grande do Sul State The pioneer of stereotactic and functional neurosurgery in Rio Grande do Sul was Paris Ferreira Souza. Paris Ferreira Souza (> Figure 16-15) graduated from the Federal University of Rio Grande do Sul in 1952. His neurosurgical training was done at USP Medical School, under Rolando Tenuto, from 1954 to 1958. Willing to polish his neurosurgical technique, Souza went to Mainz, accompanying Schu¨rmann for 2 years (1961/1963). Before leaving Germany, he decided to go to Freiburg, where he spent 2 months with Riechert and Mundinger to learn stereotactic

16

. Figure 16-15 Paris Ferreira Souza

surgery and purchased a Riechert and Mundinger apparatus. Back to Brazil, he decided to return to Porto Alegre, capital of Rio Grande do Sul, and start his practice. Still in 1963 Souza performed his first stereotactic procedure to treat a patient with Parkinson’s disease at the Beneficeˆncia Portuguesa Hospital, in Porto Alegre, becoming the pioneer of stereotactic and functional neurosurgery in this state. Due to some problems of political nature, however, he had to leave Porto Alegre in 1964. In 1965 he restarted his stereotactic practice in Saõ Paulo, at Oswaldo Cruz Hospital, a private institution. A joint project of Oswaldo Cruz Hospital and Servidores Pu´blicos Estaduais Hospital, a public institution, allowed him to operate on patients of the public hospital in the private hospital. Until the beginning of the 1970s he operated on a lot of parkinsonian patients, occasion in which he abandoned functional neurosurgery due to the sharp decline in surgical indications and went on as a general neurosurgeon. Paris Souza is now 84 years old, the oldest Brazilian functional neurosurgeon alive. He spends his lifetime

229

230

16


between Saõ Paulo and his ranch in Rio Grande do Sul, and still flies his ultralight plane like a young man (previously, he used to fly a Cessna 182, the skyline), despite the number of forced landings he has endured. The void left with the sudden exit of Paris Souza was filled by his prior classmate from the Medical School in Porto Alegre, Manoel Krimberg. Krimberg was formed neurosurgeon by Elyseu Paglioli at Saõ Jose´ Hospital. After 2 years in France with Talairach and Guiot (1963/1964), he returned to Porto Alegre, and in 1965 performed his first stereotactic procedure, a thalamotomy in a parkinsonian patient, using the Talairach apparatus. His stereotactic practice was going remarkably well until he associated with the anatomist Paolo Contu, an expert in glia. Apparently incited by Contu, as we were told (Ney Azambuja, personal communication), Krimberg started performing autologous sciatic nerve grafting to treat spinal cord injury patients, proclaiming spectacular results. Paglioli, his chief, became upset and in 1969 Krimberg left Saõ Jose´ Hospital and started working in two other places: Brigada Militar Hospital, in Porto Alegre, and Nossa Senhora das Graças Hospital, in Canoas, in the countryside. At this point, we lost his track. We were recently told about his death years ago (uncertain date). Once again practically orphan of stereotactic and functional neurosurgery, the next in the row to restart the specialty in Porto Alegre was Telmo Tonetto Reis. Reis graduated from the Federal University of Rio Grande do Sul in 1964 and finished his neurosurgical training at Maia Filho Hospital, in Porto Alegre, in 1966. In 1967, unable to afford an air trip, and with one thousand dollars in his pocket, he endured a 15-day ship trip to London, where he spent 1 year at the National Hospital Queen Square under Valentine Logue. He then moved on to Edinburgh, where he accompanied John Gillingham and Edward Hitchcock for another year (1968). Part of this period (1 month and a

half) was spent in France with Guiot. Returning to Porto Alegre, he performed his first stereotactic procedure (thalamotomy) to treat a parkinsonian patient in 1969, by using the Guiot and Gillingham apparatus, at Moinho de Ventos Hospital. Still an active functional neurosurgeon, his main area of expertise is movement disorders. Graduated from the Federal University of Parana´ in 1974, Jose´ Vitor Pinto did his neurosurgical training at Maia Filho Hospital and Moinho de Ventos Hospital (1975/1976), in Porto Alegre, under Telmo Reis, who introduced him to stereotaxis. Enchanted with stereotactic and functional neurosurgery, Pinto decided to follow the steps of his chief. In 1977 he went to Edinburgh, spending 2 years under Edward Hitchcock and John Gillingham. In 1979, when Hitchcock moved to Birmingham, Pinto did the same, staying for another year. Returning to Porto Alegre with a Hitchcock apparatus and a Radionics radiofrequency generator in his luggage, he started his practice working with Telmo Reis at Moinho de Ventos Hospital, a partnership that still remains. In the 1980s, Pinto was probably the first to perform pallidotomy (Leksell’s target) for Parkinson’s disease in our country. His initial series of 30 cases was first presented in the XVIII Congress of the Brazilian Society of Neurosurgery in 1990. Pinto was president of the Brazilian Society for Stereotactic and Functional Neurosurgery in the biennium 1988/1990. Two other superb functional neurosurgeons started their practice in Porto Alegre before Pinto: Paulo Petry Oppitz and Ney Arthur Vilamil de Castro Azambuja. Paulo Oppitz graduated from the Catholic University of Pelotas in 1969. Some time after his 2-year neurosurgical training at Saõ Jose´ Hospital with Nelson Ferreira, in Porto Alegre (1970/ 1971), Oppitz went to Buenos Aires, spending 1 year with Jorge Schvarcz, and then to Edinburgh, accompanying Hitchcock for 6 months. Soon after his arrival in Edinburgh, he purchased the fifth Hitchcock apparatus sold in the world,


which costed him five thousand pounds. In fact, Hitchcock even used Oppitz recently acquired frame in some of his operations. Returning to Porto Alegre, Oppitz started his practice at Cristo Redentor Hospital in 1976, and like almost all functional neurosurgeons of his state, his main area of expertise is the surgical treatment of movement disorders. Ney Azambuja graduated from the Catholic Medical School of Porto Alegre (now a federal foundation) in 1974, and like Oppitz, did his 3-year neurosurgical residence at Saõ Jose´ Hospital (1975/1977). In 1978 Azambuja went to Zurich, spending 1 year under Jean Siegfried. Before returning to Brazil, he visited Hitchcock for 1 month. Back to Porto Alegre, he started his practice at Saõ Jose´ Hospital in 1980. Later, Schvarcz introduced him to Oppitz, and since then they have developed an associated stereotactic practice, first at Cristo Redentor and Saõ Jose´ Hospitals, and later at Moinho de Ventos and Saõ Lucas Hospitals. Azambuja, in the beginning of the 1980s, was probably the first in Brazil to perform periaqueductal gray matter DBS for the treatment of pain. In 1985 he obtained his master degree from the Federal University of Rio Grande do Sul. As he became interested in epilepsy surgery, Azambuja moved from Saõ Jose´ Hospital to Saõ Lucas Hospital, the teaching hospital at the Pontifical Catholic University of Rio Grande do Sul, and undoubtedly one of the most renowned centers of epilepsy surgery in the country. Azambuja is now Associate Professor and regent of the discipline of neurosurgery at the Pontifical Catholic University of Rio Grande do Sul, as well as researcher of the Institute of Biomedical Research at the same institution. His main areas of expertise are the surgical treatment for epilepsy [137–139], movement disorders and pain [140], in particular trigeminal neuralgia. Other colleagues devoting part of their practice to stereotactic and functional neurosurgery in Rio Grande do Sul State are: Elyseu Paglioli Neto

16

(Porto Alegre, epilepsy surgery only), Jorge Bizzi (Porto Alegre), and Leonardo Frighetto (Passo Fundo, mainly radiosurgery).

Santa Catarina State Marcelo Neves Linhares was the pioneer of stereotactic and functional neurosurgery in Santa Catarina State, starting his practice in Floriano´polis, the state capital, in 1999. Curiously, he is still the only functional neurosurgeon in his state. Linhares graduated from the Federal University of Santa Catarina in 1992. After finishing his neurosurgical residence (1994/1997), he stayed at the USP Medical School under Teixeira for 5 months (1998), and then moved on to Toronto, doing a 1-year fellowship under Ronald Tasker and Andres Lozano (1998/1999) [141]. Linhares obtained his PhD from the USP Medical School in Ribeiraõ Preto, Saõ Paulo State, in 2005, and in 2007 he became Associate Professor of Neurosurgery at the Federal University of Santa Catarina.

Northeast Region The Northeast region is composed by the following states: Pernambuco, Ceara´, Bahia, Paraı´ba, Alagoas, Sergipe, Rio Grande do Norte, Maranhaõ, and Piauı´. There is a certain degree of uncertainty regarding the place, neurosurgeon and date that stereotactic and functional neurosurgery was inaugurated in the Northeast. We will try to clear this up along the next paragraphs. Manoel Caetano de Barros, the oldest Brazilian neurosurgeon still alive (sadly, Caetano de Barros died on 31 October 2008, sometime after we finished writing this chapter), was the pioneer of neurosurgery in the Northeast region of the country, starting his career in Recife, capital of Pernambuco, in 1947, initially as a self-taught neurosurgeon. Still in 1947 he went to Paris, spending 2 years with Clovis Vincent at Salpetrie`re

231

232

16


Hospital. In 1950 he returned to Europe, this time to London, spending 1 year under McKissock at the National Hospital Queen Square. In 1958 Caetano de Barros became Full Professor of Neurosurgery and Head of the Department of Neurology and Neurosurgery at the Federal University of Pernambuco. In his service at Pedro II Hospital, at that time the teaching hospital at the Federal University of Pernambuco, there was a Talairach apparatus acquired in the 1960s. For this reason, it has been assumed that he was the pioneer also of stereotactic and functional neurosurgery in the Northeast region of the country. We tried to contact him, but he is already very old (94 years old) and unfortunately unable to answer our questions. His son, Alex Caetano de Barros, a very active neurosurgeon in Recife, was unaware if his father had any involvement with stereotactic and functional neurosurgery in the early years of his career. Many other colleagues were consulted, but none had the answer to our question. At last, after a tough search, we were able to come up with the name of the neurologist and clinical neurophysiologist Salustiano Gomes Lins, a colleague and friend of Caetano de Barros from the old days. Salustiano Lins did his training in clinical neurophysiology in Montreal, where he had the opportunity to observe a number of stereotactic procedures. Back to Recife, he incited Caetano de Barros to purchase a stereotactic frame and start performing functional neurosurgery. The stereotactic apparatus was acquired, but, according to Lins, Caetano de Barros never really started his stereotactic practice. Once cleared up the aforementioned issue, it seems to us that the pioneer of stereotactic and functional neurosurgery in the Northeast was Djacir Figueiredo, from Fortaleza, capital of Ceara´. Graduated from the Federal University of Ceara´ in 1955, Figueiredo was formed neurosurgeon by Elyseu Paglioli, in Porto Alegre (Rio Grande do Sul State), at Saõ Jose´ Hospital, one of the two first Brazilian neurosurgical schools. In 1964 he went to New York, spending 1 year under

Lawrence Pool at the Neurological Institute of Columbia University. During his time in the USA, Figueiredo visited Earl Walker at Johns Hopkins for 2 months, where he first became involved with stereotactic and functional neurosurgery. At that time Walker used the McKinney stereotactic instrument. Before returning to Fortaleza, Figueiredo purchased the improved version of the McKinney apparatus, and even not being a genuine functional neurosurgeon, he was the first to perform a stereotactic procedure in the Northeast. From 1965 to 1968 he operated on five parkinsonian patients, and then quit his stereotactic practice. With the vacuum left by Figueiredo, someone else wishing to dedicate to stereotactic and functional neurosurgery was needed. This space was occupied by Vicente de Paula Lobo, who worked in the team coordinated by Djacir Figueiredo. Vicente de Paula Lobo (1930–2007) graduated from the Federal University of Ceara´ in 1960 and finished his neurosurgical training at Rio de Janeiro Neurological Institute, under Jose´ Ribe Portugal, in 1963. During his residency he was first exposed to stereotactic and functional neurosurgery by Renato Barbosa. In 1972 Lobo returned to Rio de Janeiro to update his training in stereotactic and functional neurosurgery, again under Renato Barbosa, but at this time at the Lagoa Hospital, since Barbosa had already left the Neurological Institute. Returning to Fortaleza, after the acquisition of a Riechert and Mundinger apparatus, Lobo performed his first procedure for Parkinson’s disease at Fortaleza General Hospital in May 1973. Approximately in 1980, due to the pronounced declining in the surgical indication for Parkinson’s disease, Lobo also abandoned the field of stereotaxis. In 1985, Fla´vio Belmino Barbosa Evangelista, who did his residence training at the Rio Janeiro Neurological Institute under Gianni Temponi, one of the pioneers of stereotactic and functional neurosurgery in Rio de Janeiro and Brazil, started performing stereotactic surgery in Fortaleza, mainly for Parkinson’s disease and only in private clinic.


The pioneer of stereotactic and functional neurosurgery in Pernambuco was Glaúcio Veras. Veras trained with G. Dieckmann, in Go¨ttingen, from 1979 to 1985. Back to Brazil he stayed in Foz do Iguaçu, Parana´, until 1989 and then moved on to Recife, capital of Pernambuco. In 1991 he inaugurated the Pain Clinic at the Cancer Hospital, and in 1993 he started performing stereotactic surgery in the same institution. His practice has been dedicated almost exclusively to pain surgery [142] and brachytherapy, having performed only a few procedures for other functional disorders. The most active functional neurosurgeon in this state, however, is Paulo Thadeu Brainer-Lima. Brainer-Lima graduated from the University of Pernambuco Medical School in 1989 and did his residence training at Restauraçaõ Hospital, in Recife, from 1991 to 1994. In 1995 he went to the USP Medical School, spending 1 year and a half under Marino Jr. During this period he also had the privilege to accompany Teixeira, Cukiert, and Oliveira Jr. He then went to Oxford University (1998), staying with Tipu Aziz for 9 months. Returning to Recife, he started performing stereotactic and functional neurosurgery in 1999, and since then he has coordinated the Stereotactic and Functional Neurosurgery Service of the Division of Neurosurgery at Restauraçaõ Hospital, the most important neurosurgical school in the Northeastern region. Brainer-Lima obtained his master degree in 1999, and his PhD in 2003, both from the same institution, the Federal University of Pernambuco. In September 2008, he became president of the Brazilian Society for Stereotactic and Functional Neurosurgery. The main foci of his career have been surgical treatment of pain, movement disorders and epilepsy [143–145]. Maria da Glo´ria S. Pabst was the pioneer of stereotactic and functional neurosurgery in Bahia State. Trained by Teixeira at the USP Medical School in 1991, she started her practice in Salvador (state capital) in 1992, performing her first procedure for Parkinson’s disease (helped by

16

Teixeira) in the same year. Due to a number of difficulties, however, her activities in the field have been restricted to stereotactic biopsy and stereotactic guided surgery since then. Stereotactic and functional neurosurgery in Bahia only flourished very recently (2006), though, with the arrival of Yuri Andrade-Souza in Salvador. Andrade-Souza did his neurosurgical training (1999/2002) at Saõ Jose´ Hospital, in Porto Alegre, under Nelson Ferreira (Ferreira, though a general neurosurgeon, was the first in Brazil to perform trigeminal retrogasserian rhizothomy by using radiofrequency thermocoagulation). In 2003 he went to the University of Toronto, spending 1 year under Andres Lozano at the Toronto Western Hospital, and 1 year and a half under Michael Schwartz at Sunnybrook Hospital. He then moved on to the University Hospital, University of Western Ontario, in London (Canada), staying 1 year under Andrew Parrent to learn epilepsy surgery. During his fellowships he published a number of significant papers [146–153]. Back to his hometown, Salvador, in 2006, Andrade-Souza has established a growing and active practice devoted to functional neurosurgery. The pioneer of stereotactic and functional neurosurgery in Paraı´ba was Valdir Delmiro Neves. Neves started his practice in the field in Joaõ Pessoa, state capital, in 1997, after being trained by Teixeira. Some years later (2000) the colleague Ussaˆnio Mororo´ Meira [154], trained by Oliveira Jr, arrived in Joaõ Pessoa. Both have a busy practice in the field of stereotactic and functional neurosurgery. Heider Lopes de Souza was the first to perform stereotactic and functional neurosurgery in Rio Grande do Norte. Souza did his neurosurgical training at Paulista Medical School, Federal University of Saõ Paulo, period in which he intermittently accompanied Teixeira at USP Medical School. He obtained his master degree in 1993, and his PhD (trigeminal neuralgia) in 2000, both from the Federal University of

233

234

16


Saõ Paulo. In 1996 he started his practice in Natal, capital of the state. Se´rgio Adrian Fernandes Dantas, in 2002, was though the first neurosurgeon with formal training in stereotactic and functional neurosurgery to establish his practice in Natal. Dantas initially spent 1 year with Oliveira Jr at AC Camargo Hospital, in Saõ Paulo. He then went to Lille and spent two more years under Serge Blond [155]. The first neurosurgeon to perform a stereotactic procedure in Maranhaõ, more precisely in its capital, Saõ Luı´s, was Osmir de Ca´ssia Sampaio. Sampaio, in 1995, after a short period (2 months) with Teixeira, started performing stereotactic biopsy, but he never went beyond this. He is truly a general neurosurgeon, and just wanted to widen his surgical possibilities. The first ‘‘authentic’’ functional neurosurgeon in Maranhaõ was in fact Jose´ Roberto Pereira Guimara˜es, who started his practice in 1999 after a formal training with Oliveira Jr in Saõ Paulo. Reynaldo Mendes de Carvalho Jr is the pioneer and still the only functional neurosurgeon in Piauı´, having established his practice in Teresina, state capital, in 1998, after a regular training with Teixeira. Likewise, Paulo Roberto Santos Mendonça, from Aracaju, the capital of Sergipe State, is the only functional neurosurgeon in his state. He trained with Oliveira Jr for 2 years and only recently (2006) started his practice. According to Abynada´ de Siqueira Lyro, the oldest neurosurgeon from Alagoas, there is no one in his state dedicated to stereotactic and functional neurosurgery; some colleagues, though, do perform stereotactic biopsy, like Ricardo Camelo and Aldo Calaça. Other colleagues from the Northeast performing stereotactic and functional neurosurgery to a lesser extent are: Antoˆnio Marcos de Albuquerque (Pernambuco), Ju´lio Augusto Lustosa Nogueira (Pernambuco), Joaõ Carlos Soares de Souza Jr (Maranhaõ), Ro´dio Luiz Brandaõ Caˆmara (Rio Grande do Norte), Ricardo Rodrigues de Carvalho (Paraı´ba), and Orlando Espinheiro Freire de Carvalho Filho (Bahia).

North Region Though a vast region, the largest in the country, Para´, to the best of our knowledge, is the only state of the North region (made up by the states of Amazonas, Para´, Roraima, Rondoˆnia, Acre, Amapa´, and Tocantins) in which stereotactic and functional neurosurgery has been implanted. The first neurosurgeon to perform stereotactic and functional neurosurgery in the North region was Joffre Moreira Lima. Joffre Moreira Lima (1918–1990) graduated from the Federal University of Para´ Medical School in 1955. After finishing his neurosurgical training under Jose´ Ribe Portugal at Rio de Janeiro Neurological Institute (1956 and 1957), he returned to Bele´m, capital of Para´, becoming together with Eloy Simo˜es Bona one of the pioneers of neurosurgery in this state [2]. In the next 2 years (1958 and 1959), he returned frequently to Rio de Janeiro Neurological Institute, keeping in close touch with Portugal and Renato Barbosa, the father of stereotactic and functional neurosurgery in Brazil. While doing the research for this chapter, we were told by Ce´sar Neves, one of the most active and renowned neurosurgeons of Para´, that Joffre Lima had performed stereotactic surgery for Parkinson’s disease a long time before. The physical therapist Lila Janahu, Lima’s daughter, under our request consulted very old operative recordings of Santa Casa de Miserico´rdia de Bele´m Hospital, where her father worked, and discovered that Joffre Lima operated on a patient with epilepsy on 22 September 1958, and a patient with Parkinson’s disease on 22 July 1959. Therefore, it seems that Lima was also the pioneer of stereotactic and functional neurosurgery in Para´. Unfortunately, we could not determine for how long he kept his stereotactic practice or any other details in this regard. More recently, stereotactic and functional neurosurgery was restarted in Para´ by Scylla Lage Silva Neto. Silva Neto, during the last year of his residency in Saõ Paulo (1988), accompanied


Teixeira part of the time at USP Medical School. He then moved on to Freiburg, spending 2 years (1989/1990) with Christopher Ostertag. Returning to Bele´m, Silva Neto performed his first functional procedure in March, 1991. Albedy Moreira Bastos, also trained by Teixeira, started his practice in stereotactic and functional neurosurgery in 1994. Though very experienced in the areas of pain surgery and stereotactic biopsy, and despite their adequate training, little have they done in the other areas of stereotactic and functional neurosurgery. Kleber Duarte, another disciple of Teixeira, stayed in Bele´m from 1995 to 1999. During this period he performed a variety of functional procedures, including 32 operations for Parkinson’s disease, but then he moved on to Rio de Janeiro and later to Saõ Jose´ do Rio Preto (Saõ Paulo State), where he keeps a busy practice. More recently, two other colleagues, Mauro Almeida and Jose´ Claúdio Rodrigues, both trained by Oliveira Jr, initiated their stereotactic and functional neurosurgery practice in Bele´m.

The Brazilian Functional Neurosurgeons Abroad At present there are five Brazilian functional neurosurgeons working abroad (USA, 3; Canada, 1; and Germany, 1), three of them already well established. The first to follow this trajectory was Antonio De Salles, 25 years ago. Antonio Afonso Ferreira De Salles (Antonio De Salles) (> Figure 16-16) graduated from the Federal University of Goia´s Medical School in 1978. After his 4-year residency at Goiaˆnia Neurological Institute (1979/1982), he went to the Division of Neurosurgery at the Medical College of Virginia for a research fellowship in head injury (1983/1985), and in 1986 he got his PhD at the Virginia Commonwealth University. Both activities were supervised by Donald Becker. His primary interest was on head injury, and he

16

. Figure 16-16 Antonio Afonso De Salles

certainly made very important contributions to this field, which are beyond the scope of this chapter. After that, his interest turned to stereotactic and functional neurosurgery and he moved to Massachusetts General Hospital, Harvard Medical School, for a clinical and research fellowship under Raymond Kjellberg (1986/1988). He then spent 1.5 year at Umea University with Lauri Laitinen (1988/1989). Back to the USA, he was invited by his former boss, Donald Becker, to the UCLA School of Medicine, where he was appointed Assistant Professor, Head of the Stereotactic Surgery Section, and co-director of the radiosurgery program, positions that he still holds. In 1999 he became Full Professor at the Division of Neurosurgery and Department of Radiation Oncology at UCLA School of Medicine. His contributions to stereotactic and functional neurosurgery have been substantial. He was the first to demonstrate that low-frequency/ low-voltage electrical stimulation of the pontine parabrachial region in cats produces a nonopiate-mediated pain suppression, as well as the connections of this region with the frontal fields

235

236

16


[156–158]. De Salles demonstrated for the first time that MRI and CT scans could be matched in the stereotactic space, which prompted the development of the image fusion technique, widely used nowadays [159]. Over the years, his team did several firsts: MR-guided radiofrequency ablation of brain tumors [160]; performance of functional neurosurgery in the MRI operating room, showing the feasibility of electrophysiological studies in such an environment [161,162]; LINAC radiosurgery for the treatment of trigeminal neuralgia [163]; demonstration of the changes in the neurotransmitters environment following low-dose LINAC radiosurgery in non-human primates, which may be implicated in the response to radiosurgery [164]; the use of shaped beam radiosurgery in a multitude of applications [165]; and the impact of ventromedial hypothalamus DBS on food intake in freely moving non-human primates [166]. Once graduated from the Paulista Medical School, Federal University of Saõ Paulo, in 1993, Clement Hamani did his PhD in Neurosciences in the same institution (1994/1996), under Luiz Mello, studying the reorganization of hippocampal circuitry in a pilocarpine-induced rat model of epilepsy [167,168]. During this period, he spent 8 months at the University of Illinois in Chicago as a research fellow, working with microdialysis in the brain [169]. From 1997 to 2002, he did his residence training at the USP Medical School, under Raul Marino Jr, and after a 6-month fellowship in functional neurosurgery with Manoel Teixeira, in the same institution, he went to the University of Toronto, where, under the supervision of Andres Lozano, he did his postdoctorate fellowship. In 2005 he became Invited Professor (neurophysiology) of the Federal University of Saõ Paulo. In 2007, he was offered the position of Associate Researcher of the Center of Addiction and Mental Health of the University of Toronto, and in July, 2008, became Assistant Professor, Division of Neurosurgery, Toronto Western Hospital, of this great university. As a member of Andres Lozano’s team, he has been part of a

number of very important researches, both clinical and laboratorial, such as subgenual cingulum DBS for depression [106], anterior thalamic nucleus DBS for epilepsy [170,171], hypothalamic/ fornix DBS for memory enhancement [172], and pedunculopontine nucleus DBS for progressive supranuclear palsy and Parkinson’s disease [173], among others [147,148,174–180]. After finishing his residency (1998/2002) and preceptorship (2003/2004) at the University of Saõ Paulo Medical School, under Raul Marino Jr, Andre´ Machado did a 2-year fellowship at the Center of Neurological Restoration, Department of Neurosurgery, Cleveland Clinic, under the supervision of Ali Rezai, and since then he became a staff member of the same institution. He runs a lab at the Cleveland Clinic Lemer Research Institute focused on motor rehabilitation in animal models of stroke [181]. As part of Ali Rezai’s team, he has greatly contributed to clinical researches in motor cortex [182] and gasserian ganglion [183] stimulation for pain, DBS for movement disorders [184,185] and psychiatric illnesses [186,187], and thalamic stimulation for minimally conscious state [188]. The two other colleagues working abroad are Alessandra Gorgulho and Guilherme Lepski. Gorgulho attended a 1-year fellowship with Teixeira, in Saõ Paulo, and a 3-year research fellowship with the Salles, at UCLA. At present, she is an assistant researcher at the Stereotactic Surgery Section at UCLA, but is still not working as a functional neurosurgeon. Lepski, of German origin, an ex-fellow and assistant to Teixeira, in Saõ Paulo, has just recently (October 2007) moved to Tu¨bingen, and is now working at the University of Eberhard-Karls under Marcos Tatagiba.

The Brazilian Society for Stereotactic and Functional Neurosurgery Despite its constant growth, the Brazilian Society of Neurosurgery (Sociedade Brasileira de


Neurocirurgia – SBN) started to be departmentalized into its various subspecialties only in 1992, during the presidency of Carlos Batista Alves de Souza. The history of Brazilian stereotactic and functional neurosurgery, however, is somewhat different from the other neurosurgical subspecialties. It started as an independent society, although connected to the SBN. After the departmentalization of SBN, it also became its Department of Functional Surgery and Pain. During the 13th Brazilian Congress of Neurosurgery, chaired by Laelio Lucas and held in Guarapari, Espı´rito Santo State, by initiative of Carlos Telles, who had recently arrived from Germany after a 4-year fellowship, the Brazilian Society for Stereotactic and Functional Neurosurgery was founded on 16 September 1980. Before the conference, Telles had written to colleagues around the country who were dealing with functional neurosurgery and invited them for a meeting dedicated to the foundation of the Brazilian Society for Stereotactic and Functional Neurosurgery. The following colleagues attended the meeting, becoming founding members of the Brazilian Society for Stereotactic and Functional Neurosurgery: Renato Barbosa, Jose´ Barros, Alceu Correia, Bernardo Couto, Djacir Figueiredo, Lourenço de Freitas Neto, Nilton Latuf, Raul Marino Jr, Luiz Fernando Martins, Pedro Motta, Delfim Nunes Neto, Sergio Ottoni, Jorge Pagura, Otoı´de Pinheiro, Jose´ Vitor Pinto, Telmo Reis, Manoel Teixeira, Carlos Telles, and Gianni Temponi. Blaine Nashold Jr and Edward Hitchcock, invited speakers to the Brazilian Congress, were also present, and thankful to the invaluable support lent by Nashold, the Brazilian Society for Stereotactic and Functional Neurosurgery became affiliated to the World Society for Stereotactic and Functional Neurosurgery since its birth. Renato Barbosa, Raul Marino Jr, Carlos Telles, Otoı´de Pinheiro, and Jose´ Barros were elected, respectively, as president, vice-president, secretary, assistant secretary, and treasurer of the first board of directors of our Society. Although there

16

is some controversy about to whom should be attributed the initiative to create Brazilian Society for Stereotactic and Functional Neurosurgery, the data here reported were collected from the blotter of Brazilian Society for Stereotactic and Functional Neurosurgery. The leadership of the Brazilian Society for Stereotactic and Functional Neurosurgery is renewed every 2 years, and the following were the presidents from its birth up to now: 1980/ 1982, Renato Barbosa; 1982/1984, Raul Marino Jr; 1984/1986, Luiz Fernando Martins; 1986/ 1988, Carlos Telles; 1988/1990, Jose´ Vitor Pinto; 1990/1992, Manoel Teixeira; 1992/1994, Sebastiaõ Gusmaõ; 1994/1996, Alceu Correia; 1996/ 1998, Claúdio Correâ; 1998/2000, Manoel Teixeira; 2000/2002, Luiz Fernando Martins; 2002/ 2004, Jorge Pagura; 2004/2006, Edvaldo Cardoso; 2006/2008, Osvaldo Vilela Filho; and 2008/2010, Paulo Brainer-Lima. The main focus of Brazilian Society for Stereotactic and Functional Neurosurgery has always been educational. Approximately three or four meetings are organized every year, usually as a parallel activity within another greater meeting such as those of the Brazilian Society of Neurosurgery, Brazilian Academy of Neurosurgery, SIMBIDOR – Brazilian Symposium and International Meeting on Pain, and CINDOR – Interdisciplinary Congress on Pain. Besides, every 2 years, in Saõ Paulo, there is a meeting co-organized by Brazilian Society for Stereotactic and Functional Neurosurgery and Micromar. The first time the Brazilian Society for Stereotactic and Functional Neurosurgery organized a completely independent meeting, was on 16–19 May 2007, the 1st International Joint Meeting on Stereotactic and Functional Neurosurgery & 8th Congress of the Brazilian Society for Stereotactic and Functional Neurosurgery: The Future Today!, held at Rio Quente Resorts, Rio Quente, Goia´s State, and chaired by Osvaldo Vilela Filho, president of the Brazilian Society for Stereotactic and Functional Neurosurgery. The conference venue, with its hot springs, was magnificent, the

237

238

16


scientific level was the highest possible, and the social activities were superb. The faculty was made up by 42 speakers, 18 from Brazil and 24 from other countries (US, Canada, Mexico, France, Italy, Sweden, China, Argentina, Chile, Venezuela, and Colombia), all well-known functional neurosurgeons. The Brazilian Society for Stereotactic and Functional Neurosurgery did not fund any speaker. Everyone graciously accepted to come on their own, in a gesture of great consideration and friendship never to be forgotten. The official clothing was casual wear, being suits and ties officially prohibited, which was instrumental in establishing a very informal atmosphere. The great innovation of the meeting was the ‘‘spicy-session,’’ in which controversial topics were fully debated by experts in an attempt to reach a consensus. One hundred and forty three participants attended the meeting, half from Brazil and half from various other countries. The Brazilian Society for Stereotactic and Functional Neurosurgery paid homage to two friends and colleagues, Ronald Tasker, International Honored President, and Luiz Fernando Martins, Brazilian Honored President, for their invaluable contributions to, respectively, world and Brazilian stereotactic and functional neurosurgery. Finally, by initiative of Eduardo Barreto and Osvaldo Vilela Filho, there was a session during the meeting, coordinated by Elliot Krames, president of the International Neuromodulation Society (INS), destined to the foundation of the Brazilian Neuromodulation Society, the Brazilian chapter of the INS, which happened on 18 May 2007. Its first board of directors was then elected: Eduardo Barreto, president; Osvaldo Vilela Filho, vice-president; and Alexandre Francisco, secretary. The Brazilian Society for Stereotactic and Functional Neurosurgery has 98 active members (85 are adequately formed functional neurosurgeons, nine of which are either currently inactive or never performed functional neurosurgery, and 13 are non-functional neurosurgeons or

neurologists) and five honorary members, who were chosen based on their great contribution to the field and relationship with Brazilian functional neurosurgeons: Ronald Tasker, Antoˆnio de Salles, Edward Hitchcock, Blaine Nashold Jr, and Wolfhard Winkelmu¨ller. Besides, there are 61 other neurosurgeons performing at least some of the stereotactic and/or functional procedures, like epilepsy alone or pain surgery and stereotactic biopsy, which gives a total of 137 (98–9–13 = 76 + 61 = 137) neurosurgeons at least partially involved with functional neurosurgery (stereotactic biopsy alone was not considered), as shown in > Figure 16-17. However, only 60 Brazilian neurosurgeons (41 members and 19 non-members) are truly functional neurosurgeons, that is, most of their practice is dedicated to stereotactic and functional neurosurgery. Finally, also worth mentioning is the creation of the Brazilian Society for the Study of Pain, which happened on 29 August 1983, by initiative of the functional neurosurgeon Jorge Pagura and the neurologist Moacir Schnapp. One year later it was affiliated to the International Association for the Study of Pain. Pagura became its second president, in fact the only functional neurosurgeon elected for president of this society.

Non-Medical Contributors for the Development of Stereotactic and Functional Neurosurgery in Brazil Until 1985, with the exception of the stereotactic frames built by the neurosurgeons Jose´ Barros and Jose´ Zaclis for their own use, all other stereotactic apparatuses in the country were imported, at a very high cost, being those of Riechert-Mundinger and Hitchcock the most commonly found. In 1985, the engineer Antonio Martos (CEO of Micromar), with the supervision of Manoel Teixeira, built the first commercially available Brazilian stereotactic apparatus,


16

. Figure 16-17 Distribution of Brazilian neurosurgeons performing functional procedures per regions and states

. Figure 16-18 Micromar radiofrequency generator

the Micromar stereotactic system, a modification of that of Hitchcock’s. Since then, this frame has been continuously refined, achieving a very high quality. Micromar is still the only Brazilian company dedicated to the stereotactic field.

It was also very instrumental to this field manufacturing electrode kits for the different neurosurgical procedures, radiofrequency generators (the first version, released in 1996, was basically analogical and received a series of refinements along the years; in 2008, an entirely digital version was released) (> Figure 16-18), and the radiosurgery system for linear accelerator (its production was again supervised by Manoel Teixeira, being released for clinical use in 1998). Another very important person in this regard has been the physicist Armando Alaminos Bouza, with a master degree in neurosciences. He was the one responsible for the development of the software (MNPS) used in association with the Micromar stereotactic system and Micromar radiosurgery system, allowing the determination of stereotactic coordinates, detailed planning of the electrode trajectory, overlaying CT and MR

239

240

16


. Figure 16-19 Neuroaugmentative procedures performed in Brazil from 2002 to 2008. SCS = spinal cord stimulation; DBS = deep brain stimulation; MCS = motor cortex stimulation; VNS = vagal nerve stimulation

images with digitized maps of stereotactic atlas (Schaltenbrand & Wahren and Talairach), shrunk or stretched as necessary to match the patient’s intercomissural distance, multimodality image fusion, and planning for radiosurgery and brachytherapy. Also important for the development of the stereotactic field in our country were Sandra Ferraz and Victor Dabah, from Dabasons Importaçaõ, Exportaçaõ e Come´rcio Ltda, local representative of Medtronic, and Joaquim Cordeiro and Ma´rcio Sossai, from JM Come´rcio e Importaçaõ Ltda, local representative of ANS, for their role in directly or indirectly (through neurosurgeons) spreading the reasoning for the use of neuroaugmentative techniques across the country and sponsoring the stereotactic and functional neurosurgery meetings. Sandra Ferraz should also be accounted for tirelessly backing the functional neurosurgeons in their fight against the health insurance companies in an attempt to make them pay for the procedures and necessary hardware. The results of their combined efforts can be better appreciated in > Figure 16-19, which shows a significant increase in the number of neuromodulatory procedures in our country.

Recent developments will most probably change forever this panorama, strengthening the tendency shown in > Figure 16-19. Until recently, neuromodulatory surgeries were not officially recognized by the National Health Agency (Ageˆncia Nacional de Sau´de – ANS). On 2 April 2008, though, according to the normative resolution number 167 of ANS, published in the Official Diary of the Union, neuroaugmentative surgeries became part of the roll of procedures recognized by ANS, which means that, from now on, private health insurance companies will be obliged to authorize and pay for these procedures. The Brazilian functional neurosurgical community will always be in great debt to all these people.

Acknowledgments Besides those already mentioned in the introductory part of the chapter, the author would like to express his deep gratitude to Fernanda Vilela for proofreading this manuscript.


References 1. Horrax G. Neurosurgery: a historical sketch. Springfield, IL: Thomas; 1952. p. 10-15. 2. Gusmaõ SS, de Souza JG. Histo´ria da Neurocirurgia no Brasil [History of Brazilian neurosurgery]. Joinville: Letra Me´dica; 2000. 3. Souza APS, Vale´rio AG. A cirurgia nervosa no Brasil [Neurological surgery in Brazil]. Anais do Segundo Congresso Brasileiro de Neurologia-Psychiatria e Medicina Legal. Arch Neurol Psiq e Med Legal 1922;6:30-67. 4. Mario Brock. Jose´ Ribeiro Portugal. J Bras Neurocir 2008;19(1):38-41. 5. Coutinho MF. Obitua´rio: Renato Tavares Barbosa. J Bras Neurocir 2003;14(1):29. 6. Livro de atas da Sociedade Brasileira de Neurocirurgia Funcional e Estereota´xica (Blotter of the Brazilian Society for Stereotactic and Functional Neurosurgery). 7. Filho PN. Paulo Niemeyer. Arq Bras Neurocir 2004;23 (1):8-11. 8. Yacubian EM, Cavalheiro EA. In memoriam: Paulo Niemeyer Soares. Epilepsia 2004;45(9):1166-8. 9. Niemeyer P. Tratamento ciru´rgico do tremor parkinsoniano, da atetose e outras hipercinesias. Med Cirurg Farm 1945;116:1-16. 10. Niemeyer P. Tratamento ciru´rgico das discinesias. Rev Bras Cir 1955;30(4):211-48. 11. Niemeyer P. Supressaõ da resposta motora e da atividade cortical ele´trica no homem. Arq Neuropsiquiatr 1946;4:109-17. 12. Niemeyer P. The transventricular amygdala-hippocampectomy in temporal lobe epilepsy. In: Baldwin MBP, editor. Temporal lobe epilepsy. Springfield, IL: Thomas; 1958. p. 461-82. 13. Cendes F, Dubeau F, Olivier A, Cukiert A, Andermann E, Quesney LF, et al. Increased neocortical spiking and surgical outcome after selective amygdalo-hippocampectomy. Epilepsy Res 1993;16(3):195-206. 14. Lynch JC. Gianni Maure´lio Temponi (1934–2007). Boletim da SBN (SBN Newsletter). Outubro; 2007. p. 22. 15. Temponi G. Estudo sobre 100 intervenço˜es estereota´xicas cerebrais no tratamento do parkinsonismo (Abstr.). In: Mattos Pimenta A, editor. XI Congresso Latino-Americano de Neurocirurgia (Relato´rios Oficiais e Comunicaço˜es Livres), Saõ Paulo; 1965. p. 272. 16. Telles CR. Nova tećnica para o tratamento das algias cervicofaciais neopla´sicas. Presented at the 14th Brazilian Congress of Neurosurgery, Belo Horizonte, Sept 1982. 17. Telles CR. Tratamento ciru´rgico das dores cervicofaciais de origem neopla´sica. Presented at the 20th Latin American Congress of Neurosurgery, Saõ Paulo, Mar 1983. 18. Telles CR. Facial pain due to cancer of head and neck. Presented at the Annual Meeting of the American Society for Stereotactic and Functional Neurosurgery, North Caroline, Apr 1983. 19. Aloysio Mattos Pimenta. Curriculum Vitae, 1959.

16

20. Yahn M, Mattos Pimenta A, Sette A, Jr. Tratamento ciru´rgico das mole´stias mentais (Leucotomia), Saõ Paulo; 1951. 21. Mattos Pimenta A. XI Congresso Latino-Americano de Neurocirurgia (Relato´rios Oficiais e Comunicaço˜es Livres). Saõ Paulo; 1965. 22. Barros NG. In memoriam: Jose´ Zaclis (1917–1983). Arq Neuropsiquiatr 1984;42:84-5. 23. Zaclis J. [X-ray visualization of the entire cerebrovascular system by means of contrast medium injection into one single artery; cerebral pan-angiography.] Arq Neuropsiquiatr 1959;17(1):1-22. (Portuguese). 24. Zaclis J, Longo PH, Cruz OR. Intrathoracic hyperpressure in neuroradiology; results in (1) cerebral panangiography and (2) arteriography of the thoracocervical region. J Neurosurg 1964;21:1087-94. 25. Zaclis J, Koralek JM, Longo PH, Almeida GM. Novo aparelho para neurocirurgia estereota´xica (Abstr.). In: Mattos Pimenta A (Ed). XI Congresso Latino-Americano de Neurocirurgia Saõ Paulo; (Relato´rios Oficiais e Comunicaço˜es Livres). 1965. p. 272. 26. Marino R. Jr. Reflections on a stereotactic and functional journey: the blessing and inspiration of working among giants. Neurosurgery 2005;56(1):172-7. 27. Ballantine HT, Jr, Cassidy WL, Flanagan NB, Marino R, Jr. Stereotaxic anterior cingulotomy for neuropsychiatric illness and intractable pain. J Neurosurg 1967;26 (5):488-95. 28. Cassidy WL, Ballantine HT, Jr, Flanagan NB. Frontal cingulotomy for affective disorders. Recent Adv Biol Psychiatry 1965;8:269-82. 29. Rasmussen T, Marino R, Jr. Functional neurosurgery. New York: Raven Press; 1978. 30. Ragazzo PC, Manzano GM, Marino R, Jr. Functional microsurgical partial callosotomy in patients with secondary generalized epilepsies. I. Disruption of bilateral synchrony of spike and wave discharges. Appl Neurophysiol 1988;51(6):297-306. 31. Marino Juńior R, Gronich G. Corpus callosum stimulation and stereotactic callosotomy in the management of refractory generalized epilepsy. Arq Neuropsiquiat 1989;47(3):320-5. 32. Marino Juńior R. Surgery for epilepsy. Selective partial microsurgical callosotomy for intractable multiform seizures: criteria for clinical selection and results. Appl Neurophysiol 1985;48(1–6):404-7. 33. Cukiert A, Puglia P, Scapolan HB, Vilela MM, Marino Juńior R. Congruence of the topography of intracranial calcifications and epileptic foci. Arq Neuropsiquiatr 1994;52(3):289-94. 34. Marino Juńior R, Cosgrove GR. Neurosurgical treatment of neuropsychiatric illness. Psychiatr Clin North Am 1997;20(4):933-43. 35. Wen HT, Rhoton AL, Jr, Marino R, Jr. Anatomical landmarks for hemispherotomy and their clinical application. J Neurosurg 2004;101(5):747-55.

241

242

16


36. de Almeida AN, Marino R, Jr. The early years of hemispherectomy. Pediatr Neurosurg 2005;41(3): 137-40. 37. Wen HT, Rhoton AL, Jr, Marino R, Jr. Gray matter overlying anterior basal temporal sulci as an intraoperative landmark for locating the temporal horn in amygdalohippocampectomies. Neurosurgery 2006;59 (4 Suppl 2):ONS221-7. 38. Teixeira MJ. Various functional procedures for pain. Part II: Facial pain. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. p. 1389-402. 39. Teixeira MJ, Lepski G, Aguiar PH, Cescato VA, Rogano L, Alaminos AB. Bulbar trigeminal stereotactic nucleotractotomy for treatment of facial pain. Stereotact Funct Neurosurg 2003;81(1–4):37-42. 40. Hitchcock E. Stereotactic trigeminal tractotomy. Farmatsiia 1970;19(3):131-5. 41. Hitchcock E, Teixeira MJ. Pontine stereotactic surgery and facial nociception. Neurol Res 1987;9(2):113-7. 42. Fonoff ET, Dale CS, Pagano RL, Paccola CC, Ballester G, Teixeira MJ, et al. Antinociception induced by epidural motor cortex stimulation in naive conscious rats is mediated by the opioid system. Behav Brain Res (2008). doi:10.1016/j.bbr.2008.07.027. 43. Fonoff ET, Dale CS, Pagano RL, Teixeira MJ, Ballester G, Giorgi R. Antinociceptive effect induced by sensorimotor cortex stimulation in normal rats. Young Investigator Award. Presented at the 14th Meeting of the World Society for Stereotactic and Functional Neurosurgery, Rome, Italy, June 2005. 44. Teixeira MJ, Fonoff ET, Montenegro MC. Dorsal root entry zone lesions for treatment of pain-related to radiation-induced plexopathy. Spine 2007;32(10): E316-9. 45. Teixeira MJ, Fonoff ET, Macri F, Picarelli H, Lin TY, Marcolin MA. Prediction of Motor Cortex Simulation in Treatment of Brachial Plexus Avulsion Pain by Transcranial Magnetic Stimulation. NeuPSIG Poster Prize (category: clinical studies) presented at the Second International Congress on Neuropathic Pain, Berlin, Germany, June 2007. 46. de Almeida AN, Fonoff ET, Ballester G, Teixeira MJ, Marino R, Jr. Stereotactic disconnection of hypothalamic hamartoma to control seizure and behavior disturbance: case report and literature review. Neurosurg Rev 2008;31 (3):343-9. 47. Fonoff ET, Lara N, Teixeira MJ. Phrenic nerve stimulation for the treatment of refractory hiccup. Presented at the Meeting of the European Society for Stereotactic and Functional Neurosurgery, Rimini, Italy, Oct 2008. 48. Teixeira MJ, Fonoff ET, Santana PA, Padilha PM, Cescato VAS. Computer-assisted stereotactic fenestration of aqueductal cyst for treatment of hydrocephalus: a technical report. Presented at the, 14th Meeting of the World Society for Stereotactic and Functional Neurosurgery, Rome, Italy, June 2005.

49. Cescato VAS, Teixeira MJ, No´brega, JPS, Almeida GM, Oliveira JO, Jr., Butori S. Tratamento de granuloma blastomico´tico cı´stico talaˆmico por meio de instilaçaõ estereota´xica de quimiotera´picos atrave´s de caˆnula intralesional. Presented at the XVI Brazilian Congress of Neurosurgery, Rio de Janeiro, Brazil, Sept 1986. 50. Hitchcock ER, Teixeira MJ. A comparison of results from center-median and basal thalamotomies for pain. Surg Neurol 1981;15(5):341-51. 51. Teixeira MJ, De Souza EC, Yeng LT, Pereira WC. [Lesion of the Lissauer tract and of the posterior horn of the gray substance of the spinal cord and the electrical stimulation of the central nervous system for the treatment of brachial plexus avulsion pain]. Arq Neuropsiquiatr 1999;57 (1):56-62. (Portuguese). 52. Teixeira MJ, Lepski G, Correia C, Aguiar PH. Interstitial irradiation for CNS lesions. Stereotact Funct Neurosurg 2003;81(1–4):24-9. 53. Rogano L, Teixeira MJ, Lepski G. Chronic pain after spinal cord injury: clinical characteristics. Stereotact Funct Neurosurg 2003;81(1–4):65-9. 54. Rogano LA, Greve JM, Teixeira MJ. Use of intrathecal morphine infusion for spasticity. Arq Neuropsiquiatr 2004;62(2B):403-5. 55. de Siqueira SR, da No´brega JC, de Siqueira JT, Teixeira MJ. Frequency of postoperative complications after balloon compression for idiopathic trigeminal neuralgia: prospective study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102(5):e39-45. 56. Arantes PR, Cardoso EF, Barreiros MA, Teixeira MJ, Gonçalves MR, Barbosa ER, et al. Performing functional magnetic resonance imaging in patients with Parkinson’s disease treated with deep brain stimulation. Mov Disord 2006;21(8):1154-62. 57. Teixeira MJ, de Siqueira SR, Bor-Seng-Shu E. Glossopharyngeal neuralgia: neurosurgical treatment and differential diagnosis. Acta Neurochir (Wien) 2008;150 (5):471-5. 58. de Almeida AN, Teixeira MJ, Feindel WH. From lateral to mesial: the quest for a surgical cure for temporal lobe epilepsy. Epilepsia 2008;49(1):98-107. 59. Pagura JR. Percutaneous radiofrequency spinal rhizotomy. Appl Neurophysiol 1983;46(1–4):138-46. 60. Pagura JR, Rabello JP, Lima WC. Microvascular decompression for trigeminal neuralgia. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. p. 1715-21. 61. Pagura JR, Schnapp M, Passarelli P. Percutaneous radiofrequency glossopharyngeal rhizotomy for cancer pain. Appl Neurophysiol 1983;46(1–4):154-9. 62. Correâ CF, Teixeira MJ. Balloon compression of the Gasserian ganglion for the treatment of trigeminal neuralgia. Stereotact Funct Neurosurg 1998;71(2):83-9. 63. Marino R, Jr., Cukiert A, Pinho E. Epidemiological aspects of epilepsy in Saõ Paulo, Brazil. A prevalence rate study. In:


64.

65.

66.

67.

68.

69.

70.

71. 72.

73.

74. 75.

76.

77.

78.

Wolf P, Dam M, Janz D, Dreifuss FE, editors. Advances in epileptology. New York: Raven Press; 1987. p. 759-64. Cukiert A, Baumel SW, Andrioli MSD, Marino R, Jr. Effects of corpus callosum stimulation on the morphology and frequency of the epileptic bursts in the feline topical penicillin generalized model. Stereotact Funct Neurosurg 1989;52:18-25. Cendes F, Dubeau F, Olivier A, Cukiert A, Andermann E, Quesney LF, et al. Increased neocortical spiking and surgical outcome after selective amygdalo-hippocampectomy. Epilepsy Res 1993;16(3):195-206. Olivier A, Germano I, Cukiert A, Peters T. Frameless stereotaxy for surgery of the epilepsies: preliminary experience. Technical note. J Neurosurg 1994;81:629-33. Cukiert A, Burattini JA, Machado E, Sousa A, Vieira JO, Argentoni M, et al. Results of surgery in patients with refractory extratemporal epilepsy with normal or nonlocalizing magnetic resonance findings investigated with subdural grids. Epilepsia 2001;42:889-94. Cukiert A, Burattini JA, Machado E, Sousa A, Vieira JO, Forster C, et al. Seizure-related outcome after corticoamygdalohippocampectomy in patients with refractory temporal lobe epilepsy and mesial temporal sclerosis evaluated by magnetic resonance alone. Neurosurg Focus 2002;13(4): ecp2 Beraha J, Barcia Salorio JL, Bordes V, Hernandez G. Radiocirurgia estereota´xica./Stereotaxic radiosurgery. Arq Bras Neurocir 1984;3(1):15-22. Beraha J, Feriancic CV, Scaff LA. [Stereotaxic radiosurgery in the treatment of cerebral arteriovenous malformations]. Arq Neuropsiquiatr 1987;45(4):391-6. In memoriam: Dr. Nilton Luiz Latuf (1937–2005). Boletim da SBN (SBN Newsletter). Dezembro; 2005. Siqueira JM. A method for bulbospinal trigeminal nucleotomy in the treatment of facial deafferentation pain. Appl Neurophysiol 1985;48(1–6):277-80. Nashold BS, Jr, Caputti F, Bernard E. Trigeminal DREZ: Caudalis nuclear lesions for relief of facial pain (Abstr). Neurosurgery 1986;19(1):150. Schvarcz JR. Trigeminal DREZ (correspondence to the editor). Neurosurgery 1987;20(2):348. Gorecki JP. Dorsal root entry zone and brainstem ablative procedures. In: Winn HR, editor. Youmans neurological surgery. 5th ed. Philadelphia, PA: Saunders; 2004. p. 3045-58. Tasker RR, Siqueira J, Hawrylyshyn P, Organ LW. What happened to VIM thalamotomy for Parkinson’s disease? Appl Neurophysiol 1983;46(1–4):68-83. Tasker RR, DeCarvalho GTC, Dolan EJ. Intractable pain of spinal cord injury: clinical features and implications for surgery. J Neurosurg 1992;77:373-8. Tasker RR, de Carvalho G, Dostrovsky JO. The history of central pain syndromes, with observations concerning pathophysiology and treatment. In: Casey KL, editor. Pain and central nervous system disease: the central pain syndromes. New York: Raven Press; 1991. p. 31-58.

16

79. Tasker RR, De Carvalho GTC. Central pain of spinal cord injury. In: North RB, Levy RM, editors. Neurosurgical management of pain. New York: Springer; 1997. p. 110-6. 80. Tasker RR, DeCarvalho GC, Li CS, Kestle JRW. Does thalamotomy alter the course of Parkinson’s disease? Adv Neurol 1996;69:563-83. 81. Gusmaõ S, Oliveira M, Tazinaffo U, Honey CR. Percutaneous trigeminal nerve radiofrequency rhizotomy guided by computerized tomography fluoroscopy. Technical note. J Neurosurg 2003;99(4):785-6. 82. Martins LF, Umbach W. Simple determination of the foramen in trigeminus coagulation. Neurochirurgia (Stuttg) 1975;18(5):163-6. 83. Martins LF, Umbach W. [Clinico-anatomical correlation between sighting of the lesion and alleviation of pain following stereotactic surgery]. Neurochirurgia (Stuttg) 1974;17(3):77-83. (German). 84. Martins LF, Umbach W. Size and position of stereotaxic lesions in comparison with clinical pain relief. Confin Neurol 1975;37(1–3):80-5. 85. Damasceno BP, Campos Junior SS, Martins LF, de Melo-Souza SE. [Clinical treatment of resistant epilepsy]. Arq Neuropsiquiatr 1988;46(4):351-8. (Portuguese). 86. Cendes F, Ragazzo PC, da Costa V, Martins LF. [Interhemispheric disconnection syndrome following total callosotomy associated to anterior commissurotomy for the treatment of intractable epilepsy: a case report]. Arq Neuropsiquiatr 1990;48(3):385-8. (Portuguese). 87. Cendes F, Ragazzo PC, da Costa V, Martins LF. Corpus callosotomy in treatment of medically resistant epilepsy: preliminary results in a pediatric population. Epilepsia 1993;34(5):910-7. 88. Zaccariotti VA, Martins LF, da Costa V, Silva NA, da Casas AA, de Melo-Souza SE. [Sneddon syndrome. Report of 3 cases]. Arq Neuropsiquiatr 1995;53(1):82-7. (Portuguese). 89. Vilela Filho O. Estrate´gia para a abordagem de leso˜es intracranianas. Rev Goiana de Medicina 1989;35:59-70. 90. Vilela Filho O, Rocha JCN, Almeida WC, Leite MSB, Silva DJ, Cavalcante JE. Bio´psia de leso˜es intracranianas a maõ livre assistida por tomografia computorizada ou ultra-sonografia. Arq Bras Neurocir 1991;10(2):103-19. 91. Tasker RR, Vilela Filho O. Deep brain stimulation for neuropathic pain. Stereotact Funct Neurosurg 1995;65 (1–4):122-4. 92. Tasker RR, Vilela Filho O. Deep brain stimulation for the control of intractable pain. In: Youmans JR, editor. Neurological surgery. 4th ed. Philadelphia, PA: Saunders; 1996. p. 3512-27. 93. Vilela Filho O. Risk factors for unpleasant paresthesiae induced by parethesiae-producing deep brain stimulation. Arq Neuropsiquiatr 1996;54(1):57-63. 94. Vilela Filho O. Risk factors for paresthesia-producing deep brain stimulation (Abstr.). Resumos de poˆsteres do 3 Simpo´sio Internacional de Dor (Simbidor). Saõ Paulo; p. 19-20.

243

244

16


95. Rezai AR, Francisco AN, Vilela Filho O, Lozano AM, Tasker RR. Deep brain stimulation for intractable neuropathic pain: contemporary management and outcome in 80 patients (Abstr.). J Neurosurg 1998;88:416A. 96. Vilela Filho O, Tasker RR. ‘‘Pathways involved in thalamic ventrobasal stimulation for pain relief: evidence against the hypothesis VB stimulation ! rostroventral medulla excitation ! dorsal horn inhibition’’. Arq Neuropsiquiatr 1994;52(3):386-91. 97. Vilela Filho O. Thalamic ventrobasal stimulation for pain relief: probable mechanisms, pathways and neurotransmitters. Arq Neuropsiquiatr 1994;52(4): 578-84. 98. Vilela Filho O. Pathophysiology of the steady burning, dysesthetic component of neuropathic pain: a hypothesis (Abstr). Stereotact Funct Neurosurg 2001;77:205. 99. Vilela Filho O. Tratamiento Neuroquiru´rgico del Dolor: Estado Actual. Neurotarget 2006;1(2):6-20. 100. Vilela Filho O, Correâ CF. Dor e neuroestimulaçaõ: Estimulaçaõ ele´trica do sistema nervoso para o manejo da dor croˆnica intrata´vel. Biotecnologia 1998;6:56-66. 101. Vilela Filho O, Correâ CF. Tratamento da dor croˆnica refrata´ria por neuroestimulaçaõ. Neurocirurgia Contemporaˆnea Brasileira 1999;2(12): 1–8. 102. Vilela Filho O, Araujo MR, Florencio RS, Silva MAC, Silveira MT. CT-guided percutaneous punctate midline myelotomy for the treatment of intractable visceral pain: technical note. Stereotact Funct Neurosurg 2001;77: 177-82. 103. Vilela Filho O, Pereira CE. Establishment of a reliable landmark to guide percutaneousradiofrequency occipital neurotomy (Abstr). Stereotact Funct Neurosurg 2001;77:204. 104. Vilela Filho O. In vivo evaluation of cordotomy lesions by magnetic resonance imaging: a new approach (Abstr). Stereotact Funct Neurosurg 2001;77:203-4. 105. Vilela Filho O, Carneiro Filho O, Souza HAO, Machado DC, Rodrigues Filho S, Campos JA. SPECT-based tailoring of psychosurgical procedures: is it possible? Stereotact Funct Neurosurg 2001; 76:256-61. 106. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, et al. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45 (5):651-60. 107. Franzini A, Ferroli P, Leone M, Broggi G. Stimulation of the posterior hypothalamus for treatment of chronic intractable cluster headaches: first reported series. Neurosurgery 2003;52(5):1095-9. 108. Vilela Filho O, Souza JT. Sıńdrome de Tourette: uma decorreˆncia da hiperatividade do pa´lido lateral? Rev Psiquiat Clin 1996;22(4):138 (Abstr.). 109. Vilela Filho O, Souza JT. Sıńdrome de Tourette: Anatomofisiologia dos gaˆnglios da base e sistema lı´mbico,

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

circuita´ria e fisiopatologia. In: Santos MGP, editor. Sıńdrome de Gilles de la Tourette: tiques nervosos e transtornos de comportamento associados na infaˆncia e adolesceˆncia. Saõ Paulo: Lemos Editorial; 1998. p. 137-58. Vilela Filho O, Ragazzo PC, Silva DJ, Souza JT, Oliveira PM, Carneiro Filho O, et al. Globus Pallidus externus (GPe) deep brain stimulation (DBS) for the treatment of Tourette Syndrome (TS): evaluation of a new target in a prospective Controlled Study (Abstr.). In Abstract Book of the ASSFN (American Society for Stereotactic and Functional Neurosurgery) 2004 Biennial Meeting – Neuromodulation: Defining the Future. Cleveland: Cleveland Clinic Foundation; 2004. Vilela Filho O, RagazzoPC, Silva DJ, Souza JT, Oliveira PM, Ribeiro TMC. Bilateral Globus Pallidus Externus Deep Brain Stimulation (GPe-DBS) for the Treatment of Tourette Sqndrome: an ongoing prospective controlled study (Abstr.). Stereotact Funct Neurosurg 2007;85: 42-3. Vilela Filho O. DBS for Tourette’s syndrome (Abstr.). In Syllabus – CME Conference of the eighth World Congress of the International Neuromodulation Society & 11th Annual Meeting of the North American Neuromodulation Society Acapulco; (Neuromodulation: Technology at the Neural Interface). 2007. p. 21. Vilela Filho O, Ragazzo PC, Silva DJ, Souza JT, Oliveira PM, Ribeiro TM. Bilateral GPe-DBS for Tourette’s Syndrome (Abstr.). Neurotarget 2008;3(1):66. Vilela Filho O. CT-based stereotactic subcaudate tractotomy (SST): technical note (Abstr). Stereotact Funct Neurosurg 2001;76:287-8. Vilela Filho O, Silva DJ, Souza HAO, Cavalcante JES, Souza JT, Ferraz FP, et al. Stereotactic subthalamic nucleus (STN) lesioning for the treatment of Parkinsons’s disease (PD). Stereotact Funct Neurosurg 2001;77: 79-86. Vilela Filho O, Silva DJ. Unilateral subthalamic nucleus lesioning: a safe and effective treatment for Parkinson’s disease. Arq Neuropsiquiatr 2002;60(4):935-48. Vilela Filho O, Silva DJ. STN lesioning for the treatment of Parkinson’s disease: risks and benefits in a long-term follow-up study (Abstr.). Stereotact Funct Neurosurg 2007;85:24. Vilela Filho O, Silva DJ, Souza JT, Ferraz FAP, Cavalcante JES. STN lesioning-induced dyskinesia in patients with Parkinson’s disease: What is behind it? In Abstract Book of the 2003 Quadrennial Meeting of the American Society for Stereotactic and Functional Neurosurgery. New York: New York University Medical School; 2003. p. 198. Vilela Filho O, Silva DJ, Ferraz FP. Evidence for the involvement of the basal ganglia in the genesis of essential tremor (ET) (Abstr.). Stereotact Funct Neurosurg 2001;77:149-50. Vilela Filho O, Ferraz FAP, Barros BA, Barreto PG, Nobrega MDA, Atayde IB, et al. Effects of Unilateral


121.

122.

123.

124.

125.

126.

127.

128. 129.

130.

131.

132.

Posterior Striatotomy on Harmaline-induced Tremor in Rats (Abstr.). Stereotact Funct Neurosurg 2007;85: 29-30. Vilela Filho O, Souza HAO, Cavalcante JES, Dias PCM, Leite MSB. Bio´psia percutaˆnea assistida por tomografia computadorizada de tumores intramedulares (Abstr.). Programa Final do VII Congresso Nacional da Academia Brasileira de Neurocirurgia. Goiaˆnia: Cegraf; 1997. p. 93. Vilela Filho O, Botter W, Jr, Oliveira SB, Teixeira KS. OVF MRI fiducial: a new and long-lasting material for stereotactic MRI (Abstr.). Stereotact Funct Neurosurg 2001;76:191-2. Vilela Filho O, Carneiro Filho O. Gamma probeassisted brain tumor microsurgical resection: a new technique. Arq Neuropsiquiatr 2002;60(4):1042-37. Vilela Filho O, Carneiro Filho O, Arratia JIC, Ragazzo PC, Arruda JB, Costa V, et al. Gamma probe assisted epilepsy surgery: preliminary results of a new technique (Abstr.). Stereotact Funct Neurosurg 2007;85:46. Vilela Filho O, Carneiro R, Jr. MRI-based frameless technique to determine target coordinates (Abstr.). Stereotact Funct Neurosurg 2007;85:62. Pereira LC, Palter VN, Lang AE, Hutchison WD, Lozano AM, Dostrovsky JO. Neuronal activity in the globus pallidus of multiple system atrophy patients. Mov Disord 2004;19(12):1485-92. Saint-Cyr JA, Hoque T, Pereira LC, Dostrovsky JO, Hutchison WD, Mikulis DJ, et al. Localization of clinically effective stimulating electrodes in the human subthalamic nucleus on magnetic resonance imaging. J Neurosurg 2002;97(5):1152-66. Aguiar LR. Biografia do Prof. Dr. Renato de Muggiati (1932–1996). J Bras Neurocirurg 1996;7(3): 56-7. Muggiati R, Ratton JF, Machado S. Resultado sobre 100 pacientes portadores de sıńdrome de Parkinson tratados cirurgicamente pela talamotomia (Abstr.). In Mattos Pimenta A (Ed). XI Congresso Latino-Americano de Neurocirurgia (Relato´rios Oficiais e Comunicaço˜es Livres). Saõ Paulo; 1965. p. 272. Francisco AN, Lobaõ CA, Sassaki VS, Garbossa MC, Aguiar LR. Punctate midline myelotomy for the treatment of oncologic visceral pain: analysis of three cases]. Arq Neuropsiquiatr 2006;64(2B):446-50. (Portuguese). Meneses MS, Rocha SB, Kowacs PA, Andrade NO, Santos HL, Narata AP, et al. [Surgical treatment of temporal lobe epilepsy: a series of forty-three cases analysis]. Arq Neuropsiquiatr 2005;63(3A):618-24. (Portuguese). Meneses MS, Rocha SF, Blood MR, Trentin A, Jr, Benites Filho PR, Kowacs PA, et al. [Functional magnetic resonance imaging in the determination of dominant language cerebral area]. Arq Neuropsiquiatr 2004;62(1):61-7. (Portuguese).

16

133. Meneses MS, Follador FR, Arruda WO, Santos HL, Yonesawa D, Hunhevicz SC. [Stereotactic implantation of depth electrodes by magnetic resonance image in epilepsy surgery]. Arq Neuropsiquiatr 1999;57 (3A):628-35. (Portuguese). 134. Meneses MS, Arruda WO, Hunhevicz SC, Ramina R, Pedrozo AA, Tsubouchi MH. Comparison of MRIguided and ventriculography-based stereotactic surgery for Parkinson’s disease. Arq Neuropsiquiatr 1997;55 (3B):547-52. 135. Almeida DB, Poletto PH, Milano JB, Leal AG, Ramina R. Is preoperative occupation related to longterm pain in patients operated for lumbar disc herniation? Arq Neuropsiquiatr 2007;65(3B):758-63. 136. Almeida DB, Hunhevicz S, Bordignon K, Barros E, Mehl AA, Burak Mehl AC, et al. A model for foramen ovale puncture training: Technical note. Acta Neurochir (Wien) 2006;148(8):881-3. 137. Salamoni SD, da Costa JC, Palma MS, Konno K, Nihei K, Azambuja NA, et al. The antiepileptic activity of JSTX-3 is mediated by N-methyl-D-aspartate receptors in human hippocampal neurons. Neuroreport 2005;16 (16):1869-73. 138. Paglioli E, Palmini A, Portuguez M, Paglioli E, Azambuja N, da Costa JC, et al. Seizure and memory outcome following temporal lobe surgery: selective compared with nonselective approaches for hippocampal sclerosis. J Neurosurg 2006;104(1):70-8. 139. Isolan GR, Azambuja N, Paglioli Neto E, Paglioli E. [Hippocampal microsurgical anatomy regarding the selective amygdalohippocampectomy in the Niemeyer’s technique perspective and preoperative method to maximize the corticotomy]. Arq Neuropsiquiatr 2007;65 (4A):1062-9. 140. Ferreira NP, Azambuja NA. [Percutaneous cordotomy]. Arq Neuropsiquiatr 1980;38(1):7-17. 141. Linhares MN, Tasker RR. Microelectrode-guided thalamotomy for Parkinson’s disease. Neurosurgery 2000;46 (2):390-5. 142. Dieckmann G, Veras G. High frequency coagulation of dorsal root entry zone in patients with deafferentation pain. Acta Neurochir 1984; Suppl 33:445–50. 143. Brainer-Lima PT, Rao S, Cukiert A, Yacubian EM, Gronich G, Marino Juńior R. Surgical treatment of refractory epilepsy associated with space occupying lesions. Experience and review. Arq Neuropsiquiatr 1996;54(3):384-92. 144. Brainer-Lima PT, Brainer-Lima AM, Azevedo Filho H, Cukiert A. [Partial epilepsy associated to primary brain tumors]. Arq Neuropsiquiatr 2002;60(3-B):797-800. (Portuguese). 145. Brainer-Lima PT, Brainer-Lima AM, Brandt CT, Carneiro GS, Azevedo HC. [Intraoperative mapping of motor areas during brain tumor surgery: electrical stimulation patterns]. Arq Neuropsiquiatr 2005;63 (1):55-60. (Portuguese).

245

246

16


146. Andrade-Souza YM, Zadeh G, Ramani M, Scora D, Tsao MN, Schwartz ML. Testing the radiosurgerybased arteriovenous malformation score and the modified Spetzler-Martin grading system to predict radiosurgical outcome. J Neurosurg 2005;103(4):642-8. 147. Andrade-Souza YM, Schwalb JM, Hamani C, Hoque T, Saint-Cyr J, Lozano AM. Comparison of 2-dimensional magnetic resonance imaging and 3-planar reconstruction methods for targeting the subthalamic nucleus in Parkinson disease. Surg Neurol 2005;63(4):357-62. 148. Andrade-Souza YM, Schwalb JM, Hamani C, Eltahawy H, Hoque T, Saint-Cyr J, et al. Comparison of three methods of targeting the subthalamic nucleus for chronic stimulation in Parkinson’s disease. Neurosurgery 2005;56 (2 Suppl):360-8. 149. Andrade-Souza YM, Zadeh G, Scora D, Tsao MN, Schwartz ML. Radiosurgery for basal ganglia, internal capsule, and thalamus arteriovenous malformation: clinical outcome. Neurosurgery 2005;56(1):56-63. 150. Andrade-Souza YM, Ramani M, Scora D, Tsao MN, TerBrugge K, Schwartz ML. Radiosurgical treatment for rolandic arteriovenous malformations. J Neurosurg 2006;105(5):689-97. 151. Andrade-Souza YM, Ramani M, Scora D, Tsao MN, terBrugge K, Schwartz ML. Embolization before radiosurgery reduces the obliteration rate of arteriovenous malformations. Neurosurgery 2007;60(3):443-51. 152. Steven DA, Andrade-Souza YM, Burneo JG, McLachlan RS, Parrent AG. Insertion of subdural strip electrodes for the investigation of temporal lobe epilepsy. Technical note. J Neurosurg 2007;106(6):1102-6. 153. Andrade-Souza YM, Ramani M, Beachey DJ, Scora D, Tsao MN, Terbrugge K, et al. Liquid embolisation material reduces the delivered radiation dose: a physical experiment. Acta Neurochir (Wien) 2008;150(2):161-4. 154. Holanda MM, Meira UM, Magalha˜es FN, da Silva JA. [Surgical treatment of meralgia paresthetica: case report]. Arq Neuropsiquiatr 2003;61(2A):288-90. (Portuguese). 155. Blond S, Touzet G, Reyns N, Dantas S, Pruvo JP. [Clinical applications of stereotaxic methology]. Ann Fr Anesth Reanim 2002;21(2):162-9. (French). 156. De Salles AA, Katayama Y, Becker DP, Hayes RL. Pain suppression induced by electrical stimulation of the pontine parabrachial region. Experimental study in cats. J Neurosurg 1985;62:397-407. 157. Leichnetz GR, Gonzalo-Ruiz A, De Salles AAF, Hayes RL. The frontal eye field and prefrontal cortex project to the paramedian pontine reticular formation in the cat. Brain Res 1987;416:195-9. 158. Leichnetz GR, Gonzalo-Ruiz A, De Salles AAF, Hayes RL. The origin of brainstem afferents of the paramedian pontine reticular formation in the cat. Brain Res 1987;422:389-97. 159. De Salles AAF, Asfora WT, Abe M, Kjelberg RN. Transposition of target information from the magnetic

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

172.

resonance and CT-scan images to the conventional x-ray stereotactic space. Appl Neurophysiol 1987; 50:23-32. Anzai Y, Lufkin R, De Salles AAF, Hamilton DR, Farahani K, Black KL. Preliminary experience with MR-guided thermal ablation of brain tumors. AJNR Am J Neuroradiol 1995;16:39-48. De Salles AAF, Frighetto L, Behnke E, Sinha S, Bronstein J, Torres R, et al. Functional neurosurgery in the MRI environment. Minim Invasive Neurosurg 2004;47(5):284-9. Lee WY M, De Salles AAF, Frighetto L, Torres R, Behnke E, Bronstein J. Imaging techniques and electrode fixation methods for deep brain stimulation – Intraoperative MRI 0.2T, 1.5 T and fluoroscopy. Minim Invas Neurosurg 2005;48:1-6. De Salles AAF, Solberg T, Medin P, Vassilev V, Cabatan-Awang C, Selch M. Linear accelerator radiosurgery for trigeminal neuralgia. Radiosurgery 1997;2:173-82. De Salles AAF, Melega WP, Lacan G, Steele LH, Solberg TD. Radiosurgery with a 3 mm collimator in the subthalamic nucleus and substantia nigra of the vervet monkey. J Neurosurg 2001;95:990-7. De Salles AAF, Gorgulho A, Selch M, De Marco J, Agazaryan N. Radiosurgery from the brain to the spine: 20 years experience. Acta Neurochir Suppl 2008;101:163-8. Lacan G, De Salles AA, Gorgulho AA, Krahl SE, Frighetto L, Behnke EJ, et al. Modulation of food intake following deep brain stimulation of the ventromedial hypothalamus in the vervet monkey. Laboratory investigation. J Neurosurg 2008;108(2):336-42. Hamani C, Teno´rio F, Mendez-Otero R, Mello LE. Loss of NADPH diaphorase-positive neurons in the hippocampal formation of chronic pilocarpine-epileptic rats. Hippocampus 1999;9(3):303-13. Hamani C, Mello LE. Status epilepticus induced by pilocarpine and picrotoxin. Epilepsy Res 1997;28 (1):73-82. Hamani C, Luer MS, Dujovny M. Microdialysis in the human brain: review of its applications. Neurol Res 1997;19(3):281-8. Andrade DM, Zumsteg D, Hamani C, Hodaie M, Sarkissian S, Lozano AM, et al. Long-term follow-up of patients with thalamic deep brain stimulation for epilepsy. Neurology 2006;66(10):1571-3. Hamani C, Hodaie M, Chiang J, del Campo M, Andrade DM, Sherman D, et al. Deep brain stimulation of the anterior nucleus of the thalamus: effects of electrical stimulation on pilocarpine-induced seizures and status epilepticus. Epilepsy Res 2008;78(2–3):117-23. Hamani C, McAndrews MP, Cohn M, Oh M, Zumsteg D, Shapiro CM, et al. Memory enhancement induced by hypothalamic/fornix deep brain stimulation. Ann Neurol 2008;63(1):119-23.


173. Weinberger M, Hamani C, Hutchinson WD, Moro E, Lozano AM, Dostrovsky JO. Pedunculopontine nucleus microelectrode recordings in movement disorder patients. Exp Brain Res 2008;188(2):165-74. 174. Toda H, Hamani C, Fawcett AP, Hutchison WD, Lozano AM. The regulation of adult rodent hippocampal neurogenesis by deep brain stimulation. J Neurosurg 2008;108(1):132-8. 175. Hamani C, Saint-Cyr JA, Fraser J, Kaplitt M, Lozano AM. The subthalamic nucleus in the context of movement disorders. Brain 2004;127(Pt 1):4-20. 176. Hamani C, Richter E, Schwalb JM, Lozano AM. Bilateral subthalamic nucleus stimulation for Parkinson’s disease: a systematic review of the clinical literature. Neurosurgery 2005;56(6):1313-21. 177. Hamani C, Neimat J, Lozano AM. Deep brain stimulation for the treatment of Parkinson’s disease. J Neural Transm Suppl 2006;(70):393–9. 178. Hung SW, Hamani C, Lozano AM, Poon YY, Piboolnurak P, Miyasaki JM, et al. Long-term outcome of bilateral pallidal deep brain stimulation for primary cervical dystonia. Neurology 2007;68(6):457-9. 179. Hamani C, Schwalb JM, Rezai AR, Dostrovsky JO, Davis KD, Lozano AM. Deep brain stimulation for chronic neuropathic pain: long-term outcome and the incidence of insertional effect. Pain 2006;125 (1–2):188-96. 180. Richter EO, Davis KD, Hamani C, Hutchison WD, Dostrovsky JO, Lozano AM. Cingulotomy for psychiatric disease: microelectrode guidance, a callosal reference system for documenting lesion location, and clinical results. Neurosurgery 2004;54(3):622-28.

16

181. Machado AG, Baker K, Godwin E, Schuster D, Rezai AR. Electrical stimulation of the cerebellar dentate nucleus enhances recovery of motor function following adult ischemic strokes in adult rats. San Diego; SFN, 2007. 182. Machado AG, Azmi H, Rezai AR. Motor cortex stimulation for chronic benign pain. Clin Neurosurg 2005;54:70-7. 183. Machado A, Ogrin M, Rosenow JM, Henderson JM. A 12-month prospective study of gasserian ganglion stimulation for trigeminal neuropathic pain. Stereotact Funct Neurosurg 2007;85(5):216-24. 184. Machado A, Rezai AR, Kopell BH, Gross RE, Sharan AD, Benabid AL. Deep brain stimulation for Parkinson’s disease: surgical technique and perioperative management. Mov Disord 2006;21 Suppl 14: S247-58. 185. Rezai AR, Machado AG, Deogaonkar M, Azmi H, Kubu C, Boulis NM. Surgery for movement disorders. Neurosurgery 2008;62 Suppl 2:SHC809-39. 186. Kopell BH, Machado AG, Rezai AR. Not your father’s lobotomy: psychiatric surgery revisited. Clin Neurosurg 2005;52:315-30. 187. Machado AG, Malone D, Kopell BH, Greenberg B, Rezai AR. Deep brain stimulation for the treatment of refractory obsessive compulsive disorder: assessment of optimal therapeutic cathode location. San Francisco; Gilbenberg Award. AANS Meeting, 2006. 188. Schiff ND, Giacino JT, Kalmar K, Victor JD, Baker K, Gerber M, et al. Behavioural improvements with thalamic stimulation after severe traumatic brain injury. Nature 2007;448(7153):600-3.

247

10 History of Stereotactic and Functional Neurosurgery in Canada A. G. Parrent

Introduction Functional neurosurgery in its broadest sense includes the surgical treatment of pain, movement disorders, epilepsy and psychiatric conditions. Canada has played a rich and pivotal role in the development of these areas. Wilder Penfield, Herbert Jasper, Ted Rasmussen, Claude Bertrand and Ron Tasker – to mention a few – are names that are familiar to most neurosurgeons who are involved in the practice of functional neurosurgery. This chapter will review some highlights in the development of functional neurosurgery in Canada with an emphasis on stereotactic surgery and epilepsy surgery.

Epilepsy Surgery and the Montreal Neurological Institute In 1928 Edward Archibald, professor of surgery at McGill University, brought Wilder Penfield and William Cone to Montreal, to develop neurosurgery at McGill University. Their clinical work started at the Royal Victoria Hospital and the Montreal General Hospital in Quebec [1]. Penfield [2] had wanted to establish an institute for the scientific study and treatment of neurological disorders and in 1932, the Rockefeller Foundation granted 1.2 million dollars to McGill University for this purpose. This money along with contributions from the city, province and other donors allowed construction of the Montreal Neurological Institute (MNI), which opened #


in 1934 as a 50 bed hospital for the investigation and treatment of brain disorders. Penfield’s background included study with Sherrington and Osler at Oxford, Holmes and Greenfield at Queen Square, Cajal and Hortega in Madrid, and Foerster in Breslau. With Foerster he learned the cortical stimulation and excision techniques that he would apply to the treatment of epilepsy. Penfield carried out his first procedure for focal epilepsy at the Royal Victoria Hospital in November 1928, on a young man with posttraumatic epilepsy resulting from a head injury with subdural hematoma and brain contusion after a fall from a horse. At the first operation the motor strip was identified by cortical stimulation and a small adjacent corticectomy was performed. Seizures continued and this man underwent a total of three epilepsy procedures culminating in a wide temporal resection – Penfield’s first temporal lobe resection for epilepsy [3]. Penfield meticulously studied his patients, both clinically and in the operating room and kept detailed records of the results of cortical stimulation. In 1937, Penfield and Boldrey [4] published their results of cortical stimulation mapping in 163 patients marking the first appearance of their motor-sensory homunculus (> Figure 10‐1a, b). Many of our current concepts of cortical localization are based on Penfield’s work [5]. By the early 1950s, Penfield [6] had recognized the unique nature of temporal lobe epilepsy, and introduced the concept of mesial temporal

114

10

History of stereotactic and functional neurosurgery in canada

. Figure 10‐1 (a) Penfield and his surgical team operating on an epilepsy patient who is awake under local anesthesia, in order to identify the epileptic focus by brain mapping. Herbert Jasper, in the glazed gallery, records the electrocorticogram (from Brain vol. 130, figure 10, 2007). (b) The ‘‘homunculus’’ showing areas of cortex related to body movements as evoked by stimulation of the cortex (from Penfield and Boldrey [4])

sclerosis; at that time attributed to herniation of the mesial temporal structures over the edge of the tentorium during difficult childbirth. Penfield and Jasper’s book ‘‘Epilepsy and the functional anatomy of the human brain’’ was published in 1954 and represented a wealth of information on the clinical presentation of epilepsy, observation related to memory, localization of language and surgical aspects of epilepsy treatment [7] (> Figure 10‐2). Penfield was responsible for the recognition of epilepsy as a surgically treatable disease. He utilized electrical stimulation of the cortex in patients undergoing craniotomy under local anesthetic and thereby improved the safety of cortical excisions near eloquent cortex. His technique of subpial dissection resulted in less residual damage and gliosis after corticectomy. This is now a standard technique for epilepsy surgeons. Penfield [1] was Director of the MNI from 1934 to 1960, succeeded by Theodore Rasmussen from 1960 to1972. Rasmussen became Professor

. Figure 10‐2 Penfield and Jasper [7] with manuscript of their book ‘‘Epilepsy and the Functional Anatomy of the Human Brain’’(from Can J Neurol Sci18:540 figure 6)

and Chairman of Neurology and Neurosurgery at the MNI in 1954, coming from the University of Chicago where he was had been Professor of Surgery since 1947 [8].


Rasmussen [9–11] added to the MNI’s contributions to the surgical treatment of epilepsy. In his years there he probably performed more operations for epilepsy than any other surgeon of his time. These cases were meticulously followed and his outcome publications report some of the longest follow-ups of epilepsy surgery patients. Rasmussen [12,13] together with Juhn Wada validated the application of the carotid amytal test for lateralization of language function and extended its value by using it for the assessment of memory in candidates for temporal lobe surgery. This is still an important means of testing memory adequacy and reserve in candidates for temporal lobectomy. Rasmussen’s name became attached to the entity of chronic localized encephalitis associated with intractable focal seizures (Rasmussen’s encephalitis); first described by Rasmussen et al. [14] in 1958. Rasmussen and others [15–17] described the condition of cerebral hemisiderosis

10

as a late complication of anatomical hemispherectomy, and introduced functional hemispherectomy as a strategy for preventing it. The MNI continues in its role, not only in the surgical treatment of epilepsy, but also in its role as educator with the training of numerous Canadian and international neurosurgeons, and neurosurgical fellows in these techniques.

Stereotactic and Functional Neurosurgery at the MNI Gilles Bertrand and John Blundell started the functional stereotactic program at the MNI when they performed the first stereotactic pallidotomy in 1958 and the first thalamotomy in 1959 [18]. Gill Bertrand started at the MNI as a resident in 1951 (> Figure 10‐3). John Blundell was invited to join the MNI group in 1957. He was sent to study with Lars Leksell in Sweden and to bring

. Figure 10‐3 (a) Dr. Gilles Bertrand inserting a microelectrode into a patient with Parkinson’s disease. (from Can J Neurol Sci 14:542 figure 8). (b) Gilles Bertrand (left) with Herbert Jasper (right) in 1988. They collaborated from 1964 to 1966 on pioneer studies of microelectrodes recording from basal ganglia neurons in patients undergoing stereotactic surgery for Parkinson’s disease. (from Can J Neurol Sci 26:228 figure 8)

115

116

10


. Figure 10‐4 The Jasper and Hunter stereotactic instrument. (from Neurosurgery 2004, 54(5):1246 figure 3)

back a Leksell stereotactic instrument. Prior to this, Drs. Herbert Jasper and John Hunter had constructed a stereotactic frame for human use – but this was never used for clinical purposes. This frame was designed to mount on the operating table with the patient in a sitting position. It was held in place by earplugs, orbital bars and a plate under the upper teeth rather than by skull pins (> Figure 10‐4). They settled on the leukotome of Claude Bertrand for lesion production after studying thermocoagulative lesions in animals and finding great variability of lesion size with identical lesion parameters. At the suggestion of Jasper in 1963 they started using microelectrode recording, and the group adopted a system with a straight microelectrode and one with a curved side arm to record from areas around the central exploring electrode (> Figure 10‐5). This group was one of the first to develop a computer program to assist with operative mapping. It allowed operative stereotactic data to be mapped onto digitized versions of various brain atlases that were expanded or shrunk to match the intercommissural distances of the respective patients [19,20]. Oblique trajectories could be followed interactively through the different

. Figure 10‐5 Leukotome (far left) used for lesion making, curved stimulating electrode (second from left) for performing macrostimulation, three-step electrode (center) used for macrostimulation and evoked potentials, straight and curved tungsten microelectrodes (right). (from Neurosurgery 2004, 54(4):1246 figure 4)

parasagittal planes, and operative physiologic data could be registered to stimulation points for later reference. The leukotome was modeled on the computer screen to allow lesions to be accurately planned and mapped [21,22]. As with all neurosurgical centers, the MNI saw a decline in the number of stereotactic procedures performed for Parkinson’s disease after the introduction of L-dopa. Surgery for other


movement disorders patients, mainly essential tremor and dystonia, continued. Abbas Sadikot joined Gilles Bertrand in 1993 and they worked together for a year prior to Bertrand’s retirement from clinical practice. The MNI group continues it work on computerized brain atlases for use in functional neurosurgery. The Leksell stereotactic frame, which was used for their stereotactic procedures, underwent modifications over time. Initial modifications by Bertrand allowed x-rays to be taken through the frame. Later, Andre Olivier made changes to facilitate the insertion of orthogonal electrodes and to make it compatible with angiography and MRI. This became the Olivier–Bertrand–Tipal or OBT frame (OBT frame, Tipal Instruments, Montreal).

10

. Figure 10‐6 Photograph of Claude Bertrand’s pneumotaxic guide with lineated prism to avoid parallax while taking preliminary measurements outside the patient’s head before control x-ray. Also, on the guiding bar, some of the collars could be angulated to allow penetration at an angle or phantom measurements. (from Neurosurgery 2004, 55(3):699 figure 2)

Functional Neurosurgery in Montreal – Hoˆpital Noˆtre-Dame Claude Bertrand was one of the pioneers of stereotactic and functional neurosurgery in Canada. Born in 1917, he obtained his BA (1934), then MD (1940) at the Universite´ de Montreál. Neurosurgical training at the Montreal Neurological Institute was completed in1946. In 1947 Claude Bertrand established the neurosurgical department at Hoˆpital Noˆtre-Dame in Montreal. Very early in his career he was interested in the treatment of involuntary movement disorders and performed anterior choroidal artery ligation on a number of patients, following Irving Cooper’s lead. In 1953, he carried out the first stereotactic lesioning (pallidotomy) for Parkinson’s disease in Canada. Although stereotactic frames designed by Speigel and Wycis, Riechert and Leksell were available; they usually required general anesthetic for application and Bertrand was looking for something better. He designed a stereotactic device called a pneumotaxic guide that could be applied under local anesthetic, and was used for his procedures [23] (> Figure 10‐6). Unsatisfied

with the inconsistency of lesions produced by cauterization or injection of substances into the brain, he used a leukotome to produce lesions. The original Moniz leukotome with a sharp blade was fairly quickly replaced with a leukotome made with blunt fine piano wire after an intracerebral hemorrhage occurred during one of his procedures [24]. The foramen of Munro was initially used as the anterior landmark for localization but Bertrand was convinced by Guiot to change to the anterior commissure. Influenced by Riechert’s reports [25], Bertrand, like many of his contemporaries moved from the pallidum to the thalamus as the target for Parkinsonian tremor. Bertrand was joined by Sonis N. Martinez in 1956, Jules Hardy in 1962 and Pedro MolinaNegro in 1967. Jules Hardy had completed his neurosurgical training in 1960 and took advantage of a McLaughlin traveling fellowship to work with

117

118

10


Gerard Guiot and Denise Albe-Fessard in Paris. There they carried out the some of the first microelectrode recordings in Parkinson’s patients undergoing stereotactic surgery. Following his return to Montreal in 1962, he introduced these techniques to the group and microelectrode recording was used routinely for their functional neurosurgical procedures [24,26] (> Figure 10‐3a). Molina-Negro was the neurophysiologist in the group and involved in the analysis of recordings obtained during stereotactic procedures. He suggested using a proportional measurement system to compensate for variations of individual patient anatomy, and they adopted one in which the AC-PC line was divided into ten equal parts. Using this system, a subthalamic, prelemniscal area target could be delineated with such accuracy that the mere introduction of an electrode was sufficient to arrest tremor [27,28]. One of the major contributions to functional neurosurgery coming from this group was the inception and development of the technique of selective peripheral denervation for the treatment of torticollis [48]. They were unhappy with the results of stereotactic thalamic interventions and wondered whether peripheral denervation could improve the results [18]. They noted that significant though temporary improvement could be produced when active muscles were injected with 1% lidocaine under EMG guidance. Their results of combined thalamotomy and peripheral denervation were reported in 1978 [29]. But they ultimately found that peripheral denervation alone was quite effective for treating torticollis.

The Toronto School The Toronto Neurosurgical program started with Kenneth McKenzie, who broke off from general practice in 1923 to take up a fellowship at the Peter Brent Brigham in Boston. While there he published an article describing sectioning of the anterior cervical spinal nerves along with the

spinal accessory for the treatment of torticollis – a procedure we still call the McKenzie procedure in Canada [30]. He returned to Toronto in 1924 as a surgical resident and went on from there to develop neurosurgery at the Toronto General Hospital. After his retirement in 1952, Dr. Harry Botterell took over the program until 1963 when he left to become Dean of Medicine at Queen’s University in Kingston. The program was then taken over by Dr. Tom Morley. Dr. Ron Tasker (> Figure 10‐7a, b) was a trainee of the Toronto Neurosurgical Program, obtaining his Royal College Fellowship in Neurosurgery in 1959. Tasker spent the following year with Clinton Woolsey at the University of Wisconsin’s Laboratory of Neurophysiology. Woolsey was studying the evolution of the sensorimotor cortex, and carried out cortical mapping in a variety of animals. Tasker developed a great deal of admiration for the ‘simple expressive beauty’ of the figurine charts that were used to represent the results of cortical mapping. Additional time was spent at the National Hospital in Queen Square, London (as was the tradition for Neurosurgical trainees at that time) and in Europe observing Albe-Fessard, Guiot, Riechert, Leksell, Talairach and others. Tasker returned to Toronto September 1961 and started his neurosurgical practice. At that time, functional neurosurgery in Toronto consisted of open cordotomy, the Frazier operation for tic douloureux, open dorsal rhizotomy for pain and the McKenzie operation for torticollis. At that time, stereotactic surgery was used primarily for the treatment of tremor associated with Parkinson’s disease. Pallidotomy was the favored procedure, carried out using ventriculographic guidance and macrostimulation to identify internal capsule. Early on, lesions were made by the inflation of a balloon and instillation of alcohol, or by cryolesioning. Tasker (May 2008, personal communication) rapidly adopted radiofrequency lesioning after its introduction in the early 1960s.


10

. Figure 10‐7 (a) Ron Tasker performing stereotactic surgery with one of the earlier Leksell steterotactic frames (b) Tasker (right) with Hirotaro Narabayashi from Japan

Tasker followed Riechert’s lead in moving the tremor target from pallidum to thalamus sometime in the early 1960s [25]. Tasker espoused the philosophy that the stereotactic operating room offered an opportunity to study the physiology of the brain. He established early collaboration with Raimond Emmers, a physiologist at the College of Physicians and Surgeons of Columbia University and later Leslie Organ of the Department of Physiology at the University of Toronto to study the somatosensory pathways of the thalamus and upper midbrain [31]. A computer program was developed to allow mapping of the stimulation responses using Woolsey-type figurines [32]. In 1982, Tasker et al. [33] published this work – ‘‘The Thalamus and Midbrain of Man’’ – which serves as a remarkable body of data and solid reference material for stereotactic and functional neurosurgeons.

Microelectrode mapping was brought into the operating room sometime in the late 1970s. Dostrovsky and Tasker studied the detailed physiology of the motor and sensory thalamus, mentoring numerous masters and PhD students as well as clinical and research fellows. Between 1964 and 1998, 48 such students and fellows were trained by Tasker, many going on to start up stereotactic programs of their own. Tasker took the surgical treatment of movement disorders in Toronto from pallidotomy to thalamotomy then through the reintroduction pallidotomy in the early 1990s. Tasker suggested that one of the great oversights in functional neurosurgery was the failure to recognize that pallidotomy was effective for bradykinesia in Parkinson’s disease. Deep brain stimulation was introduced in the early 1990s following the work of Benabid in Grenoble.

119

120

10


Tasker (May 2008, personal communication) was interested in the physiology and treatment of chronic pain. When he started practice pain was categorized as benign or malignant, and there were few operations available for the treatment of pain. Cordotomy was one of the favored procedures, was open and performed bilaterally at T1–2. The other procedure was open mesencephalic tractotomy. One of Tasker’s most significant contributions to pain management was the recognition of the difference between nociceptive pain and deafferentation or neuropathic pain and the breakdown of neuropathic pain into its different components: steady pain, shooting pain and allodynia [34,35]. Tasker has been honored with numerous awards in recognition of his contributions to functional neurosurgery including: the Spiegel & Wycis Medal, awarded by the WSSFN in 1993; the Distinguished Service Award by the ASSFN in 1995; Award for Distinguished Contributions to Pain Research and Management in Canada by the Canadian Pain society in 1997. The Tasker Chair in Functional Neurosurgery was created at the University of Toronto in 1999. In the early 1990s Andres Lozano was recruited to the University of Toronto after completing his neurosurgical training at the MNI. Lozano has expanded the functional neurosurgical program in Toronto and has gone on to be a world leader in stereotactic and functional neurosurgery. Lozano now holds the Tasker Chair at the University of Toronto.

Western Canada Edmonton Howard H. Hepburn started the Neurosurgical program in Edmonton, Alberta in 1920, having completed his medical training at McGill University in 1910. His plans for neurosurgical training in Germany were interrupted by the outbreak of

the first world war. In 1919, he returned to North American and obtained training in neurosurgery at the Mayo Clinic in Rochester. Hepburn was joined in 1945 by Guy Morton who, ultimately assumed leadership of the Neurosurgical Division with Dr. Hepburn’s retirement in 1951. Dr. Thomas Speakman joined the division in 1952. Speakman was trained at the MNI and strongly influenced by Penfield. He started performing stereotactic thalamotomy for Parkinson’s disease in the late 1950s [36] and continued until his sudden death in 1969 at the age of 45.

Saskatchewan Krishna (Kris) Kumar started the functional neurosurgical program in Regina, Saskatchewan in 1961. He came to Canada in 1959 after completing his medical training and surgical residency in Indore, India. He obtained Canadian neurosurgical training under W. D. Stevenson in Halifax and Dr. Morton in Edmonton. In Edmonton, Kumar received training in stereotactic surgery under Dr. Speakman. At the start of his career Dr. Kumar (May 2008, personal communication) was involved in the stereotactic treatment of movement disorders, mainly thalamotomy for Parkinson’s disease. After the introduction of L-dopa and the decline in candidates for movement disorder surgery he concentrated his efforts on the treatment of intractable pain. He was involved in the stereotactic implantation of DBS electrodes in periaqueductal grey (PAG) and sensory thalamus from its start in the late 1960s, and was an advocate of the intravenous pain test for assessing surgical candidacy. This was carried out by injecting small repeated amounts of intravenous morphine to assess the degree of pain relief, and the utilization of naloxone to assess reversibility of the response. Patients responding in a reversible manner to i.v. morphine were considered candidates for PAG stimulation, those not responding were


considered better candidates for sensory thalamic stimulation. In general, most patients received simultaneous implantations into both sites followed by percutaneous testing. The most effective electrode was ultimately internalized. The DBS system at the time consisted of a ring electrode activated by a radiofrequencycoupled stimulator (Medtronic). These systems developed slowly; although approved for the treatment of pain in Canada, DBS for pain never obtained FDA approval in the United States. Kumar (May 2008, personal communication) was also involved in spinal stimulation for pain. Starting with the open implantation of intradural and interdural electrodes and through to the epidural systems. The initial percutaneous monopolar electrodes came in straight, sigmashaped and tined varieties. Radiofrequencycoupled systems were eventually supplanted by implanted battery operated systems. Kumar was influential in the development of the fully implantable spinal cord stimulation devices and implantable pumps. Significant contributions to the area of functional neurosurgery include cost-effectiveness studies for spinal cord stimulation and intrathecal drug therapy [37–39], as well as long-term follow-up studies of SCS and DBS for pain [40–42].

British Columbia Frank Turnbull was one of the pioneers of neurosurgery in Canada and the first neurosurgeon in British Columbia. He received his neurosurgical training under K. McKenzie in Toronto, followed by postgraduate work at Queen Square in London, and further work with Foerster in Breslau. He started work at the Vancouver General Hospital in 1933 and devoted special attention to the surgical treatment of intractable pain. Turnbull [43–45] attempted to look at pain syndromes as distinct entities so as to better rationalize management. He published on the pain syndromes associated with

10

pelvic carcinoma and specifically looked at the role of cordotomy in the surgical management of these pain syndromes [45]. The early cordotomies were open procedures, carried out at T1–2. In the mid to late 1960s the switch was made to anterior cervical percutaneous cordotomy. Peter Lehmann joined Frank Turnbull in the late 1940s and started performing thalamotomy for parkinsonian tremor in the late 1950s using the Cooper apparatus and leukotome. Ian Turnbull joined the group in 1966. Ian Turnbull (July 2008, personal communication) obtained his medical training in Vancouver and neurosurgical training in Toronto. While in Toronto he was exposed to Tasker and his work. Residency was followed by a traveling fellowship in Europe gaining exposure to Leksell, Gillingham and others. After his return to Vancouver he rapidly adopted the Todd Wells stereotactic apparatus and used radiofrequency to carry out thalamotomy for tremor disorders. The stereotactic movement disorder program remained active through to his retirement from clinical practice in 1999. In the early 1990s he, like most neurosurgeons switched to pallidotomy for Parkinson’s disease. Ian Turnbull also established and maintained a large pain practice, initially learning the percutaneous cordotomy technique from his father Frank. He favored stimulation of the ventrobasal thalamus for neuropathic pain [46] and reported on combined thalamic or midbrain lesions with cingulumotomy in some patients with nociceptive pain [46].

The Canadian Stereotactic and Functional Neurosurgery Group In 2001 the Canadian Neurosurgical Society formed a Section of Stereotactic and Functional Neurosurgery. The Canadian group is small with only 15 or so neurosurgeons carrying out functional stereotactic procedures in the country. As of 2008, there are active programs in Halifax Nova Scotia,

121

122

10


Quebec City and Montreal in Quebec, Toronto and London Ontario, Winnipeg Manitoba, Regina Saskatchewan, Calgary and Edmonton Alberta and Vancouver British Columbia. The relatively small size of the group has been conducive to collaborative work and has resulted in a multicentre study of DBS for cervical dystonia [47]. Other projects are ongoing.

References 1. Feindel W. Development of surgical therapy for epilepsy at the Montreal Neurological Institute. Can J Neurol Sci 1991;18:549-53. 2. Penfield W. No man alone: a neurosurgeon’s life. Boston: Little, Brown & Co; 1977. 3. Foerster O, Penfield W. The structural basis of traumatic epilepsy and results of radical operation. Brain 1930;53 (2):99-119. 4. Penfield W, Boldrey E. Somatic motor and sensory representation in cerebral cortex of man as studied by electrical stimulation. Brain 1937;60(4):389-443. 5. Penfield W, Erickson TC. Epilepsy and cerebral localization. Springfield, IL: Charles C Thomas; 1941. 6. Penfield W, Earle KM, Baldwin M. Incisural sclerosis and temporal lobe seizures produced by hippocampal herniation at birth. Arch Neurol Psychiatry1953;69:17-42. 7. Penfield W, Jasper H. Epilepsy and the functional anatomy of the human brain. Boston: Little, Brown and Company; 1954. 8. Feindel W. Theodore Brown Rasmussen (1910–2002): epilepsy surgeon, scientist and teacher. J Neurosurg 2003;98:631-7. 9. Rasmussen T. Surgical treatment of complex partial seizures: results, lessons and problems. Epilepsia 1983;24 Suppl 1:S65-76. 10. Feindel W, Rasmussen T. Temporal lobectomy with amygdalectomy and minimal hippocampal resection: review of 100 cases. Can J Neurol Sci 1991;18 Suppl 4:603-5. 11. Rasmussen T, Feindel W. Temporal lobectomy: review of 100 cases with major hippocampectomy. Can J Neurol Sci 1991;18 Suppl 4:601-2. 12. Wada J, Rasmussen T. Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance. Experimental and clinical observations. J Neurosurg 1960;17:266-82. 13. Milner B, Branch C, Rasmussen T. Study of short-term memory after intracarotid injection of sodium amytal. Trans Am Neurol Assoc 1962;87:224-6. 14. Rasmussen T, Olszewski J, Lloyd-Smith D. Focal seizures due to chronic localized encephalitis. Neurology 1958;8: 435-45.

15. Rasmussen T. Postoperative superficial hemosiderosis of the brain, its diagnosis, treatment and prevention. Trans Am Neurol Assoc 1973;98:133-7. 16. Rasmussen T. Hemispherectomy for seizures revisited. Can J Neurol Sci 1983;10:71-8. 17. Falconer MA, Wilson PJ. Complications related to delayed hemorrhage after hemispherectomy. J Neurosurg 1969;30(4):413-26. 18. Bertrand G. Stereotactic surgery at McGill: the early years. Neurosurgery 2004;54:1244-52. 19. Bertrand G, Olivier A, Thompson CJ. Computer display of stereotaxic brain maps and probe tracts. Acta Neurochir Suppl (Wien) 1974;21:235-43. 20. Thompson CJ, Bertrand G. A computer program to aid the neurosurgeon to locate probes used during stereotactic surgery of deep cerebral structures. Comput Programs Biomed 1972;2:265-76. 21. Bertrand G. Computers in functional neurosurgery. In: Rasmussen T, Marino R Jr, editors. Functional neurosurgery. NY: Raven; 1979. 22. Thompson CJ, Bertrand G. A computer program to aid the neurosurgeon to locate probes used during stereotaxic surgery on deep cerebral structures. Comput Programs Biomed 1972;2(4):265-76. 23. Bertrand CM. A pneumotaxic technique for producing localized cerebral lesions, and its use in the treatment of Parkinson’s disease. J Neurosurg 1958;15 (3):251-64. 24. Bertrand CM. Surgery of involuntary movements, particularly stereotactic surgery: reminences. Neurosurgery 2004;55:698-704. 25. Hassler R, Riechert T. lndikationen und lolalisationcmethode der gezielton hirnoperationen. Nervenarzt 1954; 25:441-7. 26. Hardy J. Historical background of stereotactic surgery: reflections on stereotactic surgery and the introduction of microelectrode recording in Montreal. Neurosurgery 2004;54:1508-11. 27. Velasco F, Molina-Negro P, Bertrand C, Hardy J. Further definition of the sub-thalamic target for the arrest of tremor. J Neurosurg 1972;36:184-91. 28. Bertrand C, Hardy J, Molina-Negro P, Martinez SN. Optimum physiological target for the arrest of tremor. In 3rd Symposium on Parkinson’s Disease, Edinburgh, Livingstone, 1969, 251-9. 29. Bertrand C, Molina-Negro P, Martinez SN. Combinedstereotactic and peripheral surgical approach for spasmodic torticollis. Appl Neurophysiol 1978;41: 122-33. 30. McKenzie KG. Intrameningeal division of the spinal accessory and roots of the upper cervical nerves for the treatment of spasmodic torticollis. Surg Gynecol Obstet 1924;39:5-10. 31. Emmers R, Tasker RR. The human somesthetic thalamus. NY: Raven; 1975.


32. Tasker RR, Rowe IH, Hawrlyshyn P, Organ LW. Computer mapping of human subcortical sensory pathways during stereotaxis. J Neurol Neurosurg Psychiatry 1975;38 (4):408. 33. Tasker RR, Organ LW, Hawrylyshyn PA. The thalamus and midbrain of man. A physiological atlas using electrical stimulation. Springfield, IL: Charles C Thomas; 1982. 34. Takser RR. Deafferentation. In: Wall PD, Melzak R, editors. Textbook of pain. London: Churchill Livingstone; 1984. p. 639-55. 35. Tasker RR. Management of nociceptive, deafferentation and central pain by surgical intervention. In: Fields HL editor. Pain syndromes in neurology. 2nd ed. London: Butterworths; 1990. p. 143-200. 36. Speakman TJ. Results of thalamotomy for Parkinson’s disease. Can Med Assoc J 1963;89:652-66. 37. Kumar K, Malik S, Demeria D. Treatment of chronic pain with spinal cord stimulation versus alternative therapies: cost-effectiveness analysis. Neurosurgery 2002;51(1): 106-15. 38. Kumar K, Hunter G, Demeria DD. Treatment of chronic pain by using intrathecal drug therapy compared with conventional pain therapies: a cost-effectiveness analysis. J Neurosurg 2002;97(4):803-10. 39. Hornberger J, Kumar K, Verhulst E, Clark MA, Hernandez J. Rechargeable spinal cord stimulation versus nonrechargeable system for patients with failed back surgery syndrome: a cost-consequences analysis. Clin J Pain 2008;24(3):244-52.

10

40. Kumar K, Toth C, Nath RK, Laing P. Epidural spinal cord stimulation for treatment of chronic pain - some predictors of success. A 15-year experience. Surg Neurol 1998;50(2): 110-20. 41. Kumar K, Hunter G, Demeria D. Spinal cord stimulation in treatment of chronic benign pain: challenges in treatment planning and present status, a 22-year experience. Neurosurgery 2006;58(3):481-96. 42. Kumar K, Toth C, Nath RK. Deep brain stimulation for intractable pain: a 15-year experience. Neurosurgery 1997;40(4):736-46. 43. Turnbull F. The nature of pain in the late stages of cancer. Surg Gynecol Obstet 1960;110:665-8. 44. Turnbull F. Intractable pain. Proc R Soc Med 1954;47(2): 155-6. 45. Turnbull F. A basis for decision about cordotomy in cases of pelvic carcinoma. J Neurosurg 1959;16:595-9. 46. Turnbull IM, Shulman R, Woodhurst WB. Thalamic stimulation for neuropathic pain. J Neurosurg 1980;52(4): 486-93. 47. Kiss ZH, Doig-Beyaert K, Eliasziw M, Tsui J, Haffenden A, Suchowersky O. Functional and Stereotactic Section of the Canadian Neurosurgical Society, Canadian Movement Disorders Group. The Canadian multicentre study of deep brain stimulation for cervical dystonia. Brain 2007;130 Pt 11:2879-86. 48. Bertrand C, Molina-Negro P, Martinez SN. Technical aspects of selective peripheral denervation for spasmodic torticollis. Appl Neurophysiol 1982;45:326-39.

123

15 History of Stereotactic Neurosurgery in Italy A. Franzini . V. A. Sironi . G. Broggi

The history of stereotactic neurosurgery in Italy is strictly linked to the development of the mayor European schools of Paris, Stokholm and Freiburg. In the sixties almost any neurosurgical department in Italy had a stereotactic frame dedicated to thalamotomy for Parkinson disease (the Talairach frame, Reichert frame, Guiot frame, Leksell frame and the Cooper frame were the most popular steretactic devices). From this ‘‘pneumoencephalography era’’ we can remember Franco Migliavacca in Milan, Dalle Ore in Verona, Faust D’Andrea in Naple and Elio Tartarini in Genoa who performed thousands of stereotactic operations for Parkinson disease, mental illness, and pain [4]. After the LDopa discovery and its wide therapeutical application, stereotactic surgery seems to have disappeared from Italy except for a few Institutes which still continued to perform stereotactic operations for tremor, pain, dystonia, and epilepsy. So the real first generation of surgeons mainly devoted to functional neurosurgery in Italy in the ‘‘ventriculography era’’ includes Franco Marossero and Paolo Emilio. Maspes in Milan [12,13,14], Victor Aldo Fasano in Turin [7], Gianfranco Rossi [16] and Beniamino Guidetti [11] in Rome. An original stereotactic frame was also developed in Turin and utilized for the treatment of cerebral palsy in adults and children (> Figure 15-1) utilizing alcoholic lesions and later cryothalamotomy. The main interests of the schools of Milan (F. Marossero) and Rome (G. Rossi) were the functional exploration of the brain by acute and chronic implanted electrodes recording deep EEG activity (SEEG) in epilectic patients candidated to tailored resection of the epileptic focus or to #


deep radiofrequency lesions of nuclei and tracts involved in the origin and diffusion of the epileptic discharge [13]. Also if this methodology was imported from the Paris School of Talairach and Bancaud, they developed many original observations and contributed to the definition of criteria still utilized worldwide for surgical treatment of epilepsy [12,14]. In those years particularly Gianfranco Rossi had an eminent role in the development of functional and stereotactic neurosurgery in Italy; he therefore deserves a more detailed history. Prof. Rossi worked for 4 years in the field of experimental neurophysiology under the leadership of Prof. Giuseppe Moruzzi and his research interest was centered on the anatomic and functional organization of brainstem reticular formation and sleep physiology. Later when he became chairman of the neurosurgical Department at the Catholic University of Rome his neurophisyological background had a strong influence on clinical practice and research projects and original criteria were proposed to improve the interpretation of electrocerebral epileptic signals. The rationales, indications, and relative efficacy of classic surgical resection approaches including callosotomy, multiple subpial transection, and the so-called lesionectomy were studied [17]. In the same years C. A. Pagni in Milan wrote an universally appreciated book and publications about the surgical treatment of central pain [16]. In that period B. Guidetti in Rome performed operations to treat spasticity (Dentatectomy) and pain (Pulvinotomy) [11]. The main interest of the Turin school was the treatment of cerebral palsy and many thalamotomies have been

194

15

History of stereotactic neurosurgery in italy

. Figure 15-1 Stereotactic thalamotomy performed in the ‘‘pneumoencephalographic era’’ with the Italian frame named FasanoSguazzi. Note the contrast medium (lipiodol)in alcoholic solution injected at the target sites on both sides. (Courtesy of Sergio Zeme M.D., University of Turin, Italy)

performed in dystonic children [7]. The second generation of functional neurosurgeons at the beginning of the ‘‘CT era’’ includes De Divitis in Naples who heralded stereotactic surgery for Gilles de la Tourette syndrome [5]; the senior author (G.B.) who published milestone studies about impedance guided biopsy [1], cell kinetics of deep brain tumors, stereotactic treatment of brain abscesses and pioneered in Europe the treatment of cystic components of craniopharingiomas by intracavitary Bleomycin; founded in Milan the first Italian neurosurgical Department dedicated to functional and stereotactic neurosurgery providing fuel for future development of original treatments such as Deep Brain Stimulation for the treatment of chronic refractory cluster headache and disruptive behaviour. Another master of stereotactic neurosurgery was Franco Frank in Bologna who pioneered surgery of mesencephalic structure to treat cancer pain [8]. Prof. Frank was trained in Freiburg by Mundinger and founded in Bologna the school of functional and stereotactic neurosurgery. His

original approach included nearly all fields of interest of neurosurgery with particular regard to management of brain tumors by intracavitary irradiation and management of pain by stimulation of the periacqueductal gray and thalamus [9]. Massimo Scerrati in Rome and Pierligi De Riu in Sassari [6,18] introduced in Italy the stereotactic brachiterapy of brain tumors. Prof. Scerrati developed the ‘‘Scerrati’s arc’’ to transform the Talairach frame in an isocentric frame [19]. Prof. Mario Meglio in Rome introduced the use of spinal cord stimulation to improve blood flow in peripheral vascular diseases [15]. Belonging to this generation is also Claudio Munari who worked in Paris and Grenoble and founded in Italy the first department entirely dedicated to surgery of epilepsy and functional exploration of the brain. Also in the field of radiosurgery the Italian contribution was significant, particularly due to the contribution of Federico Colombo who applied the linear accelerator to the stereotactic frame and widened considerably the field of application of radiosurgery [3]. Dr. Colombo

History of stereotactic neurosurgery in italy

built up a huge casuistic which includes more than 800 arteriovenous malformations treated by radiosurgery and is one of the larger series of the world. Also in the field of neuroimaging, the Italian contribution was highly represented by Cesare Giorgi who built one the first digitalized stereotactic atlases [10] and developed original equipments in the field of robotics and tridimensional neuronavigation. The ‘‘neuromodulation era’’ heralded at the beginning of the eighties by the senior author [2] belong to the present.

References 1. Broggi G, Franzini A. Value of serial stereotactic biopsies and impedance monitoring in the treatment of deep brain tumours. J Neurol Neurosurg Psychiatry 1981;44(5): 397-401. 2. Broggi G, Franzini A, Giorgi C, Servello D, Spreafico R. Preliminary Results of Specific Thalamic Stimulation for Deafferentation Pain. Acta Neurochir (Wien) Suppl 1984;33:497-500. 3. Colombo F, Benedetti A, Pozza F, Zanardo A, Avanzo RC, Chierego G, Marchetti C. Stereotactic radiosurgery utilizing a linear accelerator. Appl Neurophysiol 1985;48 (1–6):133-45. 4. Dalle Ore G, Da Pian R. Intractable pains and destruction of the latero-ventral nucleus of the thalamus. Presentation of a case of ‘‘phantom limb’’ operated upon with complete success. Minerva Med 1960;29;51:2771-3. 5. de Divitiis E, D’Errico A, Cerillo A. Stereotactic surgery in Gilles de la Tourette syndrome. Acta Neurochir (Wien) 1977;(Suppl 24):73. 6. De Riu PL, Rocca A. Interstitial irradiation therapy of supratentorial gliomas by stereotaxic technique. Long term results. Ital J Neurol Sci 1988;9(3):243-8. 7. Fasano VA, Broggi G, Schiffer D, Urciuoli R. Experimental lesions induced by localized cooling of the brain. Morphologic and histochemical study. Neurochirurgie 1965;11(6):519-28.

15

8. Frank F, Fabrizi AP, Gaist G. Stereotactic mesencephalic tractotomy in the treatment of chronic cancer pain. Acta Neurochir (Wien) 1989;99(1–2):38-40. 9. Frank F, Frank G, Gaist G, Galassi E, Sturiale G, Fabrizi A. Deep brain stimulation in the treatment of chronic pain syndromes. Riv Neurobiol 1982;28 (3–4):309-16. 10. Giorgi C, Garibotto G, Cerchiari U, Broggi G, Franzini A, Koslow M. Neuroanatomical digital image processing in CT-guided stereotactic operations. Appl Neurophysiol 1983;46(1–4):236-9. 11. Guidetti B, Fraioli B. Neurosurgical treatment of spasticity and dyskinesias. Acta Neurochir (Wien) 1977; (Suppl 24):27-39. 12. Marossero F, Ettorre G, Infuso L, Pagni CA. Surgical treatment of non-tumoral epilepsy with stereotaxic methods. Minerva Neurochir 1966;10(4):342-3. 13. Marossero F, Ravagnati L, Sironi VA, Miserocchi G, Franzini A, Ettorre G, Cabrini GP. Late results of stereotactic radiofrequency lesions in epilepsy. Acta Neurochir Suppl (Wien) 1980;30:145-9. 14. Marossero F, Ettorre G, Franzini A, Motti DF. Chronic depth electrodes study of one case of bitemporal epilepsy due to glial tumour. Some physiopathological considerations. Acta Neurochir (Wien) 1978;45(1–2): 123-31. 15. Meglio M, Cioni B, Dal Lago A, De Santis M, Pola P, Serricchio M. Pain control and improvement of peripheral blood flow following epidural spinal cord stimulation: case report. J Neurosurg 1981;54(6):821-3. 16. Pagni CA. Central pain and painful anesthesia. In: Progress in Neurological Surgery vol.8:132-157. (Karger, Basel 1976). 17. Rossi GF. Surgical treatment of partial epilepsy: remarks on the relative role of stereo-EEG and other diagnostic examinations. J Neurosurg Sci 1975;19(1–2):89-94. 18. Scerrati M, Roselli R, Iacoangeli M, Montemaggi P, Cellini N, Falcinelli R, Rossi GF. Comments on brachycurie therapy of cerebral tumours. Acta Neurochir Suppl (Wien) 1989;46:94-6. 19. Scerrati M, Fiorentino A, Fiorentino M, Pola P. Stereotaxic device for polar approaches in orthogonal systems. Technical note. J. Neurosurg 1984;61(6):1146-7.

195

6 History of Stereotactic Neurosurgery in the Nordic Countries B. A. Meyerson . B. Linderoth

In the sixth volume of the classical and comprehensive textbook of neurosurgery by Olivecrona and To¨nnis (1957), Lars Leksell contributed a chapter on ‘‘Targeted Brain Operations’’ (Gezielte Hirnoperationen). In the introduction he refers to the development of his stereotactic instrument – he always preferred to describe his stereotactic frame as a surgical instrument – that followed the pioneering work by Spiegel and Wycis in 1947–1948. The first presentation of Leksell’s sterotactic apparatus dates from 1949 and already in the early 1950s he had gained experience in targeted lesional surgery using electrical current and different types of X-ray beams – radiosurgery. There is no doubt that this makes Leksell, together with a few others, one of the founders of stereotactic surgery. It should also be noted that the principal mode of function and general features of his original instrument have survived in the modern version that is still one of the most commonly used stereotactic systems in the world. That is why the advancement of stereotactic neurosurgery in the Nordic countries is so intimately associated with the name of Lars Leksell and his contributions. Lars Leksell (1907–1986) started his neurosurgical training with Herbert Olivecrona in 1935 at the Serafimer Hospital, one of the oldest hospitals in Sweden founded in 1752. The Olivecrona neurosurgical service enjoyed a solid international reputation and attracted a large number of trainees from all over the world. For a short period Leksell served as a volunteer #


medical doctor in Finland when it was attacked by the Soviet Union in November 1939. Later he told that he during that war often speculated on the possibility of extracting bullets from the brain with minimal damage to the surrounding brain tissue using a mechanically guided instrument. In the early 1940s Leksell joined Ragnar Granit, Nobel Laureate 1967, for experimental studies in neurophysiology. In 1945 he presented a PhD dissertation, a monograph on the motor gamma system titled ‘‘The action potential and excitatory effects of the small ventral root fibers to skeletal muscle.’’ This was a major milestone in the understanding of muscle control and has now become part of basic neurophysiology. It should be noted that during these years he, together with Granit and Skoglund, made another major contribution by describing the phenomenon of ephapsis, ‘‘artificial synapses,’’ caused by local pressure on a nerve, as a possible mechanism involved in trigeminal neuralgia. After resuming clinical work, he started work on the development of a stereotactic instrument and in 1947 he visited Wycis in Philadelphia. Leksell described his instrument in a publication in 1949 and this was the first example of a stereotactic system based on the principle of ‘‘center-of-arc’’ in contrast to the Spiegel and Wycis orthogonal, rectangular frame (> Figure 6-1). The use of a movable semi-arc with an electrode carrier implies that the tip of a probe can reach the target regardless of the position of the carrier or the angling of the arc relative to the skull

66

6

History of stereotactic neurosurgery in the nordic countries

. Figure 6-1 The first Leksell stereotactic instrument, based on the ‘‘center-of-arc’’ principle, was described in a publication from 1949

fixation device, i.e., a frame or base plate with bars for bone fixation screws. This construction permits also transphenoidal, straight lateral and suboccipital probe approaches. Leksell was in many respects a perfectionist and for the rest of his life he continued to change and revise the design of virtually every small part of his instrument though the basic semicircular frame was retained. He focused not only on upgrading the function of the instrument but also on its aesthetic appearance. An important feature was that ‘‘the apparatus should be easy to handle and practical in routine clinical work’’ and ‘‘a high degree of exactitude is necessary.’’ An oft-cited quotation is ‘‘Tools used by the surgeon must be adapted to the task and where the human brain is concerned, no tool can be too refined.’’ The first, documented clinical application of Leksell’s stereotactic system was a case of a craniopharyngioma cyst that was punctured and treated with injection of radioactive phosphorus – that patient was probably the first patient in the world to undergo this form of therapy (1948) (> Figure 6-2). Before the advent of modern imaging techniques (CT, MRI), ventriculography was, and in some centers still is, routinely utilized for target

. Figure 6-2 The first practical application of the stereotactic instrument was for puncture of craniopharyngioma cysts

coordinate determination. Already in the late 1940s neuroradiology was a well-developed speciality at the Serafimer Hospital and angiography and pneumoencephalography were routinely practiced. Leksell performed pneumoencephalography, first in the sitting and then in the supine position to visualize the anterior and posterior commissures, respectively. In order to compensate for the divergence of the X-rays, he constructed a diagram of tightly packed concentric circles, approximated to spirals, geometrically related to the divergence and the distance between the X-ray tube and the film, and frame planes, for determining the target coordinates; it has to be admitted, however, that in contrast to Leksell’s other inventions many surgeons found it difficult to understand and use this diagram. Beside the passionate interest in the technical aspects of stereotaxy, Leksell was in the 1950s and 1960s very active in the operation theatre. He performed a large number of pallidotomies, and later also thalamotomies, in Parkinson’s disease and capsulotomies in various forms of


mental disorders (> Figure 6-3). The results of a series of 81 patients subjected to pallidotomy was published in 1960, and 116 patients treated with capsulotomy were reported in 1961. The term and concept of radiosurgery were introduced by Leksell already in 1951 when he reasoned that the ‘‘center-of-arc’’ principle and his first stereotactic instrument were suitable for replacing a probe (needle electrode) by cross-firing intracerebral structures with narrow beams of radiant energy. X-rays were first tried but both gamma rays and ultrasonics were included as alternatives (> Figure 6-4). Initial experiments were performed on cats and then a few patients with pain and chronic psychosis were treated with a 280 kV X-ray tube attached to the arc. Of particular interest is that in 1953 two cases of trigeminal neuralgia were treated and at followup in 1971 they were still free of pain. In 1946 Leksell was appointed head of a neurosurgical unit in Lund in southern Sweden where he became professor in 1958 and remained so until 1960. In those days there were very few neurosurgeons around the world who were active in stereotactic surgery and the international network was very small; it is of interest that the Schaltenbrand and Bailey’s Stereotactic Atlas was partly based on some brain specimens supplied by Leksell. While in Lund, Leksell was apparently able to evade many of his clinical obligations because he was able to initiate a close collaboration with a team of physicists led by Bo¨rje Larsson at the University of Uppsala (north of Stockholm) where a synchrocyclotron was available. They conducted experiments with stereotactic high-energy proton irradiation in goats resulting in a seminal publication in Nature in 1958 (> Figure 6-5). This technique was also applied in a few patients with Parkinson’s disease (pallidotomy), psychiatric disorder (capsulotomy) and pain (mesencephalotomy). Although precisely placed and well-limited lesions could be produced by the focused proton beams, as demonstrated in a few autopsy cases, the synchrocyclotron proved to be too complicated

6

. Figure 6-3 Leksell in the early 1960s with the second generation of his stereotactic frame

. Figure 6-4 Application of the first stereotactic instrument for radiosurgery using cross-firing of 280 kV X-rays. Photo from the early 1950s

for general clinical use. This compelled Leksell to consider other radiation sources and he started designing the 60Co Gamma Unit that was fully integrated with the stereotactic system. The

67

68

6


. Figure 6-5 The first clinical trials with radiosurgery using proton beams from the cyclotron at Uppsala University were performed in the late 1950s

. Figure 6-6 Leksell treating the first case of acoustic neurinoma with the first version of the Gamma Knife in 1968

development of the ‘‘Beam-knife’’ took place after Leksell had been appointed successor to Olivecrona in 1960 and the first unit was inaugurated in 1967 (> Figure 6-6). Later the same year reports of the two first cases, patients with cancer related pain subjected to radiosurgical thalamotomy, were published.

Originally, radiosurgery and the Gamma Unit were developed with the hope that it would offer a bloodless, and less risky, method to be applied principally in functional neurosurgery, for example in thalamotomy for Parkinson’s disease. On the other hand, Leksell had always considered his sterotactic instrument a


surgical tool that should also be utilized in general neurosurgery in order to enhance precision and minimize hazards. This idea had to some extent been realized by the extensive use of stereotactic technique in puncturing cysts and also in performing biopsies in critical regions. However, the Gamma Unit soon proved to be very useful in the treatment of some typical diseases requiring neurosurgery, such as pituitary adenomas, acoustic neurinomas and arteriovenous malformations. This successful application of radiosurgery has indeed revolutionized the management of these conditions but was also met with much skepticism from the neurosurgical community. Albeit the topic of this chapter is the history of stereotaxis, two other examples of Leksell’s exceptionally innovative mind deserve to be mentioned. He was the first to apply ultrasound in neurosurgical diagnosis by the development of echoencephalography as early as 1955. Moreover, his elegantly designed, double action rongeur has become an indispensable tool in the hands of most neurosurgeons.

Leksell’s Disciples Erik-Olof Backlund (1931) started his neurosurgical training in 1960 at the Serafimer Hospital and he soon became a pupil and a close associate of Leksell. At that time neuroradiology could visualize pathological mass processes only indirectly, and therefore stereotactic biopsy was rarely performed, the exceptions being trials to obtain specimens of the cystic wall when performing craniopharyngioma cyst punctures. This became part of Backlund’s doctoral thesis and to facilitate biopsy sampling he constructed a new device consisting of a spiral that could be screwed into the often quite tough craniopharyngioma tissue from which a specimen could then be harvested by using an outer sharp cutting needle as a sleeve punch.

6

This biopsy needle proved to be very useful and is still routinely used. Backlund also contributed other innovative needle instruments for aspiration biopsy, hematoma evacuation, and aqueduct reconstruction. Backlund’s interest in craniopharyngiomas led him to reconsider the choice of radioactive phosphorus for cystic intracavity irradiation and he demonstrated that a colloid 90 Y was more effective. Subsequently, he developed a multimodal treatment program for craniopharyngiomas, and the evacuation of cysts and installation of radioactive substances were in some cases supplemented by radiosurgery of the solid parts (> Figure 6-7). This treatment strategy is still practiced at the Karolinska University Hospital. Another contribution was the application of radiosurgery for capsulotomy in patients with anxiety and OCD; this was first published, with Leksell as co-author, in 1978. As noted by Backlund himself in an autobiographical vignette, the most spectacular phase of his career was the first neurotransplantation of adrenal chromaffin cells in a patient with Parkinson’s disease. This trial was initiated by a group of researchers in basic sciences at the . Figure 6-7 Erik-Olof Backlund (right) and the research engineer Bengt Jernberg (left) preparing a patient for capsulotomy. Note, at that time the head was fixed with a Orthoplast cap. Photo from the mid 1970s

69

70

6


Karolinska Institute and Backlund developed the stereotactic technique for precise depositing of the cellular specimen. Ladislau Steiner (1920) is presumably the most well known of Leksell’s pupils. From 1962 he first trained with Olivecrona who was then still active in spite of retirement. His prime interest was vascular and tumor microsurgery rather than endeavors within the stereotactic and functional field. His first exposure to radiosurgery was when he took part in a major study thalamotomy for cancer-related pain. It then appeared that radiosurgery might be useful for some common neurosurgical diseases also, and together with Leksell he became the first to demonstrate that centrally located arteriovenous malformations, otherwise inaccessible to surgery, could be miraculously obliterated. Steiner, along with Christer Lindquist (1944) who later became one of the leading names in radiosurgery with many important publications, subsequently performed a large number of AVM treatments. Following the first publication many patients from all over the world were referred to him. The AVM experiences triggered a surge of trials on new applications for radiosurgery (acoustic neurinomas, pituitary adenomas, pituitary hormone suppression in Cushing’s disease, pinealomas, meningiomas, and later cerebral metastases). In 1987 Steiner left the Karolinska University Hospital and became professor and director of the Lars Leksell Gamma Knife Center at the University of Virginia. There he created a very active team that has made this center one of the leading ones for radiosurgery. Scientifically also, he has continued to be very active and is co-author of numerous publications covering the entire field of radiosurgery. A few others of Leksell’s disciples who were active in the field of radiosurgery and have made important, more or less independent contributions should be mentioned. Georg Noreń (1943) was the first to demonstrate, in an extensive and long-lasting study, that the outcome of

radiosurgical treatment of acoustic neurinomas is at least as favorable as that with open surgery performed by the most experienced neurosurgeons. Since 1992, Noreń has been director of the Gamma Knife Center at Brown University, Providence. Tiit Ra¨hn (1940) became involved in stereotactic techniques and radiosurgery in the late 1960s and he pioneered the management of hormone secreting pituitary adenomas. His career, until retirement, has been entirely devoted to radiosurgery. The senior author of this paper (B.A.M 1933), Meyerson joined Leksell’s team in 1968, shortly after the inauguration of the first Gamma Unit. He soon became involved in ‘‘gammathalamotomy’’ for cancer-related pain and the preliminary results were published in 1969. Because of the relatively unsatisfactory outcome of this form of surgery a project with intracerebral electrical stimulation (deep brain stimulation, DBS) for pain was initiated in 1974, and with electrodes of our own design stimulation was applied simultaneously in several brain targets. This was the beginning of a life-long career in pain research. At an early stage Meyerson was also introduced to surgical treatment of mental disorders and subsequently he performed the great majority of capsulotomies, as well as a large number of interventions for movement disorders, at the Karolinska University Hospital. The second most prominent stereotactic neurosurgeon in the Nordic countries was Lauri Laitinen (1928–2005) and it is notable that he pursued his own career independent of Leksell. He had his basic neurosurgical training in Helsinki and was confronted with stereotactic surgery when working in London, and visiting Gillingham in Edinburgh. He started operating on Parkinsonian patients in 1961 using first the Cooper instrument and later a RiechertMundinger frame. He was, however, dissatisfied with both these instruments and towards the end of the 1960s he had constructed his own system


that was later acquired by many stereotactic centers (> Figure 6-8). This instrument has the Leksell center-of-arc design combined with some features of the Riechert-Mundinger frame. Throughout his career he was very active in clinical practice and performed about 200 thalamotomies annually. During these operations he made extensive trials with impedance monitoring which he found very helpful for differentiating white from grey matter, for example to identify the border between the internal capsule and pallidal tissue. In 1963 Laitinen published his first paper in the field of stereotactic surgery, on the treatment of torticollis. Later he joined with a young neuropsychologist, Juhani Vilkki, and in a series of papers they documented, for the first time, the deterioration of motor, verbal, and visuospatial functions that may occur after thalamotomy. Laitinen was also very interested in pain, and inspired by Gabriel Mazars in Paris, he performed the first trial of implanting a stimulating electrode in a patient with phantom limb pain already in 1968. In 1980 Laitinen moved to Umea˚, a university town in northern Sweden, where he revived stereotactic and functional neurosurgery to such an extent that Umea˚ became a center of excellence. There he developed a device (the

6

‘‘Stereoadapter’’), a light frame that in a reproducible way could be repeatedly attached to the skull during CT and MRI examinations and then utilized for coordinate determination together with the original frame. In 1984 he, along with Marwan Hariz (who later became his successor) and Tommy Bergenheim, began to test Leksell’s old and almost forgotten version of pallidotomy from the early 1950s. For many of the younger neurosurgeons this endeavor is directly associated with the name of Laitinen. It should be recalled that with the advent of L-dopa in the early 1970s the use of surgery for movement disorders, in particular for Parkinson’s disease, was reduced to just a few cases each year even in the major centers. The first presentation in 1992 of the favorable outcome of Laitinen’s pallidotomy cases resulted in a dramatic change in attitude toward surgery among the neurologists. A contributing, important factor was that this type of intervention had proven to be effective for L-dopa induced hyperkinesias also, which by that time had become a major problem. No doubt, the acceptance of pallidal surgery following Laitinen’s reports paved the way for the ensuing development of stimulation of central brain areas (DBS) as a novel mode of therapy for Parkinson’s disease.

. Figure 6-8 Lauri Laitinen standing by a case displaying his own frame and stereoadapter. Photo from 2003

71

72

6


During the 1960–1970s stereotactic surgery was practised occasionally in a few other neurosurgical units in the Nordic countries but these activities cannot be regarded as pioneering or innovative. One exception was the contributions made by Kjeld Vaernet (1920–2006) in Copenhagen who in 1974 reported on a trial with ‘‘stereotactic stimulation and electrocoagulaton of the lateral hypothamus’’ in five patients with gross obesity. He also performed many stereotactic surgeries for epilepsy and mental disorders. Even though the contributions of Leksell are exceptional and outstanding, the importance of the achievements of his disciples and of Laitinen in making the Nordic countries strong in stereotactic open neurosurgery and radiosurgery should not be underestimated.

Selected Bibliography 1. Backlund EO, Granberg PO, Hamberger B, Knutsson E, Martensson A, Sedvall G, Seiger A, Olson L. Transplantation of adrenal medullary tissue to striatum in parkinsonism: first clinical trials. J Neurosurg 1985;62: 169-73. 2. Backlund EO. Reflections: a historical vignette. Neurosurgery 2004;54:734-41. 3. Backlund EO, Johansson L, Sarby B. Studies on craniopharyngiomas. II. Treatment by stereotaxis and radiosurgery. Acta Chir Scand 1972;138:749-59. 4. Bingley T, Leksell L, Meyerson BA, Rylander G. Stereotaxic anterior capsulotomy in anxiety and obsessivecompulsive states. Third World Congress of Psychosurgery, Cambridge: University of Cambridge Press; 1972. p. 159-64. 5. Boe¨thius J, Lindblom U, Meyerson BA, Wideń L. Effects of multifocal brain stimulation on pain and somatosensory functions. In: Zotterman Y, editor. Sensory functions of the skin in primates with special reference to man. New York: Pergamon Press; 1976. p. 531-48. 6. Hirsch A, Noreń G, Anderson H. Audiologic findings after stereotactic radiosurgery in nine cases of acoustic neurinomas. Acta Otolaryngol 1979;88:155-60. 7. Laitinen LV. A new stereoencephalotome. Zentralbl Neurochir 1971; 32: 67-73. 8. Laitinen LV, Bergenheim AT, Hariz MI: Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61.

9. Laitinen LV. Personal memories of the history of stereotactic neurosurgery. Neurosurgery 2004;55:1420-8. 10. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson B. The high-energy proton beam as a neurosurgical tool. Nature 1958;182(4644):1222-3. 11. Lindquist C, Kihlstro¨m L. Department of Neurosurgery, Karolinska Institute: 60 years. Neurosurgery 1996;39: 1016-21. 12. Leksell L. The action potential and excitatory effects of the small ventral root fibres to skeletal muscle. Acta Physiol Scand 1945;10 Suppl 31:1-79. 13. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta Chir Scand 1949;99:229-33. 14. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-9. 15. Leksell L, Lideń K. A therapeutic trial with radioactive isotopes in cystic brain tumor. In: Radioisotope techniques. Proceedings of the isotope techniques conference, Vol 1. Oxford: H.M. Stationary Office 1953; 1951. p. 76-8. 16. Leksell L. Gezielte Hirnoperationen. In: Olivecrona H and To¨nnis W, editors. Handbuch der Neurochirurgie, Vol 6., Berlin: Springer-Verlag; 1957. p. 178-99. 17. Leksell L. Some principles and technical aspects of stereotaxic surgery. In: Knighton RS and Dumke PR, editors. Pain. Boston: Little, Brown and Company; 1966. p. 493-502. 18. Leksell L. Cerebral radiosurgery. I. Gammathalamotomy in two cases of intractable pain. Acta Chir Scand 1968;134:585-95. 19. Leksell L, Backlund EO, Johansson L. Treatment of craniopharyngiomas. Acta Chir Scand 1967;133:345-50. 20. Leksell L. A note on the treatment of acoustic tumours. Acta Chir Scand 1971;137:763-5. 21. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand 1971;137:311-4. 22. LeksellL, Backlund EO. [Radiosurgical capsulotomy – a closed surgical method for psychiatric surgery] Lakartidningen 1978;75:546-7 (Swedish). 23. Noreń G, Arndt J, Hindmarsh T. Stereotactic radiosurgery in cases of acoustic neurinoma: further experiences. Neurosurgery 1983;13:12-22. 24. Quaade F, Vaernet K, Larsson S. Stereotaxic stimulation and electrocoagulation of the lateral hypothalamus in obese humans. Acta Neurochir (Wien) 1974;30:111-7. 25. Ra¨hn T, Thoreń M, Hall K, Backlund EO. Stereotactic radiosurgery in Cushing’s syndrome: acute radiation effects. Surg Neurol 1980;14:85-92. 26. Steiner L, Leksell L, Greitz T, Forster DMC, Backlund EO. Stereotaxic radiosurgery for cerebral arteriovenous malformations: report of a case. Acta Chir Scand 1972; 138:459-64. 27. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotatic thermolesions in the pallidal region. A clinical evaluation of 81 cases. Acta Psychiatr Scand 1960; 35: 358-77.

History of Stereotactic Surgery

1 History of Stereotactic Surgery P. L. Gildenberg . J. K. Krauss

Preface In order to define when stereotactic surgery started, it is first necessary to define what stereotactic surgery is [1]. ‘‘Stereotaxic’’ surgery began in 1908, when Sir Victor Horsley and Robert Clarke [2] introduced their new apparatus to allow them to insert a probe, blade, or needle under accurate control into a subcortical structure of a monkey or other experimental animal. They specified the use of a Cartesian coordinate system [3] that makes it possible to define a point in space by specifying three coordinates, anterior-posterior (AP), lateral and vertical, and that Cartesian system remains the hallmark of stereotactic systems. Sir Victor Horsley was both a neurosurgeon and neurophysiologist, and is generally recognized as the father of human functional neurosurgery. He had collaborated with Robert Clarke, a mathematician and also a surgeon, and recruited him to help him design their device. The design and mathematics of the instrument was mainly Clarke’s, and the details of how it might be used were Horsley’s [4]. Their initial report of an experiment conducted with this apparatus concerned making electrolytic lesions in the dentate nucleus of the cerebellum of the monkey to study its structure and functions [2] (> Figure 1-1). The 1908 article that introduces stereotaxic surgery is a magnificent illustration of a literary scientific journal article, and the original should be read by anyone with an appreciation of the history of medical literature. The article is divided into four parts. One part describes the first animal stereotaxic apparatus, which has remained a model for most such devices, both for animals and patients, over this past century (> Figure 1-1). #


In order to define a target point, they used Cartesian coordinates, that is, defining a point in space by its relationship to three planes, each at right angles to the other two, and all meeting at a common point as shown in > Figure 1-2. A location can be defined by three coordinates, each of which tells the relationship (in this case in millimeters) to one of the three planes – x mm anterior or posterior to the coronal plane, y mm lateral to the midsagittal plane, and z mm above or below the horizontal plane. For that to have relevance to the structures within the animal brain, the three planes must be registered, or accurately aligned, to the position of the head of the experimental animal. This was accomplished by relating the three planes to the same parts of the device that held the head securely – [6] a horizontal plane (similar to the Frankfort plane used in anthropology [6]) that passed through both ear plugs and a tab that held the left inferior orbital rim (or both if there is no asymmetry) from which to define the vertical coordinate – [7] a mid-sagittal plane that passes through the mid-point between the ear plugs at right angles to the horizontal plane to define the left or right lateral coordinate – and [8] a coronal plane that passes through the ear plugs and is perpendicular to the other two planes to define the AP coordinate. The second part of the Horsley-Clarke paper is a technique for making a stereotactic atlas registered to this same Cartesian system, in which the stereotaxic coordinates of a particular structure could be found. The third part was an excellent treatise about making direct current electrolytic lesions. The fourth and least remembered part was the experiment itself. Clarke suggested to Horsley that the technique might be useful in humans, and even

4

1

History of stereotactic surgery

. Figure 1-1 The original Horsley-Clarke apparatus [2]

. Figure 1-2 The concept of Cartesian planes to measure coordinates aligned with third ventricle

patented the idea of a human stereotactic apparatus, still based on using the boney landmarks to establish the landmarks from which Cartesian coordinates might be determined [9,10]. The

pair separated because of other disagreements, so neither pursued the idea of a human stereotactic any further [11]. In 1918, however, Mussen, an engineer who had been involved in making the Horsley-Clarke apparatus designed a similar device for use with the human skull [12]. There is no evidence that he persuaded any of his neurosurgical colleagues to use it, so he eventually wrapped it in newspaper and stored it in a box in his attic. When his family discovered the box almost 60 years after the Mussen frame had been made, they could determine when it had been stored by the date on the newspaper wrapping. It is perhaps best that it was not used, because it has since become known that the relationship between the boney landmarks on the human skull and the cerebral structures within are so variable that accurate targeting would not have been possible. It took most of the 40 years after the 1908 report that concepts of human physiology became well enough known to have determined the appropriate intracerebral subcortical targets. Let us digress for a moment to confront two other issues involving [6] devices prior to Horsley and Clarke and [7] the spelling of ‘‘stereotaxic’’ versus ‘‘stereotactic.’’ There were reports of a variety of devices to guide a probe into the brain or spinal cord prior to 1908, with consequent claims of a variety of ‘‘first’’ stereotactic devices. We are unaware of any prior system that used a Cartesian coordinate system. Most used overlying landmarks to approximate the location of various structures. Probably the first such system was that presented by Dittmar [13] in 1873, in which he used a guided probe to insert a blade into the medulla oblongata of the rat to perform physiological studies. From the Russian perspective, the beginnings of human stereotactic surgery are ascribed to Zernov [14], who in 1889 described an ‘‘encephalometer’’ that helped localize cortical areas, and Altukhov [15] who used it clinically 2 years later. This was not Cartesian in principle and did


not address the localization of deep subcortical structures. In regard to the spelling of the technology addressed throughout this book, there is a logic to it. Horsley and Clarke called their technique ‘‘stereotaxic,’’ from the Greek ‘‘stereos,’’ threedimensional, and ‘‘taxus,’’ an arrangement, as in taxonomy. In a meeting in 1973 when it was decided to change the name of the ‘‘International Society for Research in Stereoencephalotomy’’ (meaning three-dimensional study of the encephalon or brain), it became necessary to agree on a spelling. Since the advent of human stereotactic surgery, some authors, mainly in Europe, used the spelling ‘‘stereotactic,’’ rather than ‘‘stereotaxic,’’ although the origins of that spelling have been lost. A vote was taken as to whether the newly named society would be spelled ‘‘stereotaxic’’ (‘‘three-dimensional’’ and ‘‘arrangement,’’ both from Greek), or ‘‘stereotactic’’ (a mongrel word meaning ‘‘three-dimensional’’ from Greek and ‘‘tactus’’ from Latin ‘‘to touch’’). It was felt that the object of the surgery was actually to touch the desired structure with a probe or electrode, so by vote the decision was made to spell the society ‘‘Stereotactic.’’ In a prescient moment, the decision was also made to call the societies which followed the International Society for Research in Stereoencephalotomy, the World Society for Stereotactic and Functional Neurosurgery, the American Society for Stereotactic and Functional Neurosurgery, and the European Society for Stereotactic and Functional Neurosurgery, ‘‘stereotactic’’ to denote the manner of localization (how you get to the target) and ‘‘functional,’’ meaning to change the function of the brain (what you do after you get there) [16]. The word ‘‘stereotaxis’’ refers to both animal and human procedures. The convention of referring to human surgery as ‘‘stereotactic’’ and animal surgery as ‘‘stereotaxic’’ has been generally adopted and will be maintained throughout this review. History is not a straight line. Different topics that relate to stereotactic surgery developed at

1

the same time, but often with different chronology. More often, advances in other fields became incorporated into stereotactic technique at various stages of development while other things were happening within stereotaxis itself. Consequently history cannot be presented as a continuum. We will digress as occasion demands, take side trips if they appear to be interesting, and double back on the calendar as we switch from one topic to another. The major topics will have separate titles, and subtopics will be announced in bold type, which signifies a change in gears and often a change in direction. After introducing the technology, we will take movement disorder surgery up to 1968, where we can pause that story, discuss other indications for functional neurosurgery, and then come back to the reawakening of stereotactic movement disorder surgery in 1992.

Birth of Human Stereotactic Surgery The problem of establishing accurate intracerebral landmarks from which specific brain structures could be measured and building a device to insert an electrode accurately to a chosen anatomical structure is credited to Ernest A. Spiegel and Henry T. Wycis (> Figure 1-3). Spiegel was a neurologist and neurophysiologist who fled the Nazis in Vienna and emigrated to Temple Medical School in Philadelphia as Professor of Experimental Neurology [17]. (His wife, Mona Spiegel-Adolph, was simultaneously recruited as Professor of Colloid Chemistry). Wycis began to work in Spiegel’s laboratory when he was a medical student, and their collaboration developed throughout his neurosurgical residency and appointment to the faculty [18]. In 1947, they reported on the first use of a human stereotactic apparatus they had designed (still spelled ‘‘stereotaxic’’ at that time) [5].

5

6

1


. Figure 1-3 Spiegel and Wycis with their Model V apparatus [19]

The time was right for development of human stereotactic surgery (> Figures 1-4 and > 1-5). The motivation was there: Prefrontal lobotomy had been popular and was indicated in many cases, prior to the development of psychotropic medication, but then it became abused by some [20]. One main motivation to develop stereotactic surgery was to provide a controlled lobotomy effect with less risk of unintended neurological deficit. Once stereotaxis was available, however, the first patients were those with movement disorders, and its use in psychosurgery was not reported for several years. The technology had just been developed: One key to this field was the first introduction of X-ray equipment into the operating room with rapid film development. The key to human stereotactic surgery is to identify landmarks within the brain by X-ray or by other imaging and calculating where a target lay in relation to those landmarks. Ventriculogram X-rays could be taken to visualize internal cerebral landmarks about the third ventricle, from which measurement could be made to localize any cerebral

. Figure 1-4 The original Spiegel-Wycis stereotactic apparatus [5]

. Figure 1-5 Spiegel and Wycis in the operating room. Note the Faraday cage to shield electrical noise so recording could be done


structure [5]. This made it possible to use intracerebral landmarks, usually around the third ventricle, from which to measure the three coordinates, each of which was based on one of the three Cartesian planes, hence the AP, lateral, and vertical coordinate (> Figure 1-2). The location of a specific structure required reference to a human stereotactic atlas, which they published soon after stereoencephalotomy was introduced [21] (> Figure 1-6). The science was there: The field of neurophysiology had advanced to the point where targets could be selected on a rational basis. An appreciation of the extra-pyramidal system was emerging, and it was recognized that many movement disorders involved those circuits. Surgical feasibility was demonstrated by Meyers [22] (> Figure 1-7) – series of 38 patients reported in 1942 (> Figure 1-8). There were also attempts to interrupt pathways in the brain for management of pain. Many of these techniques involved surgical interruption of conveniently superficial pathways and are still in use [24–26]. There was a good concept

. Figure 1-6 The orientation of slices in a stereotactic atlas

1

of the primary pain pathway as it ascends through the brainstem, and the concept that the limbic system participated in pain perception was becoming well developed [27,28]. The engineering was there: Spiegel and Wycis [21] called their new technique ‘‘stereoencephalotomy,’’ a three-dimensional system based on brain measurements. The device was essentially a Horsley-Clarke apparatus suspended over the patient’s head, secured to a ring held by a plaster cap made for each patient individually (> Figure 1-3). The ring was aligned with the horizontal plane by ear plugs so that AP and lateral X-rays could be taken accurately and repeatedly. A microdrive supported by the ring was mounted above the patient’s head to hold the electrode and advance it either vertically or obliquely into the target within the brain. They originally selected the anterior commissure and the pineal gland as the two basic landmarks, and produced a stereotactic atlas based on measurements between those structures and the desired target [21]. The use of the anterior and posterior commissure was made a decade later

7

8

1


. Figure 1-7 Russel Meyers, 1904–1999 [23]

. Figure 1-8 Meyers’ open surgical approach to the basal ganglia [23]

by Talairach and his associates [29], which, with some variations, were adopted by most stereotactic neurosurgeons. Originally, X-ray visualization of the landmarks about the ventricle required a pneumoencephalogram [21,30], which made it mandatory that the stereotactic apparatus could be accurately reapplied, which was the protocol when

I (plg) first started working as a medical student with Spiegel and Wycis [18,31] in 1956. Patients had the device attached on Tuesday, pneumoencephalogram performed, and measurement made. The apparatus was removed, because the patient by that time was too sick to have a procedure performed under local anesthesia. The patient was returned to the operating room 2 days later to replace the head ring and for the actual surgery. When positive contrast agents became available, it became possible to do the surgery at a single session [18]. A philosophy that was begun with the first Spiegel-Wycis procedure and maintained through the years by most stereotacticians is that each time an electrode is inserted into the brain it produces a unique opportunity to study human neurophysiology, so each procedure included both physiological confirmation of the electrode position and neurophysiologic studies [32]. Not only did this provide a confirmation of the intended target, but led to advances in our understanding of the human brain and the diseases being treated. Originally, the lesions were made with alcohol injection, which theoretically was more likely to affect the neurons in the intended nucleus while sparing the fibres en passage. The spread of the alcohol was unpredictable, however, so soon lesions were made with the same electrolytic direct current that Horsley and Clarke had reported almost half a century before [2]. However, that carried the risk of a sudden sharp stimulation if the current varied. Other techniques were soon developed, such as oil-procaine or oil-procaine-wax injection [33,34], alcohol injection, sometimes with a balloon cannula or coagulating substance [35,36], mechanical damage with a leukotome [37], and later radiofrequency [38] and cryoprobe methods [39]. The original report by Spiegel and Wycis [5] ended by enumerating a list of potential indications for stereotactic surgery. It is almost certain that they had used this new procedure for all those indications prior to their first publication.


The interest was there: The first decade after human stereotactic surgery was introduced was particularly productive. There was a stream of neurosurgeons from throughout the world visiting Spiegel and Wycis’ service at Temple Medical School in Philadelphia, learning this new technique and returning home to enter this new field. It was necessary for each to design and manufacture his own stereotactic frame, since there were none commercially available. A variety of apparatus was rapidly introduced, including several advanced designs by Spiegel and Wycis, so that they did most of their work with the Model V [19]. All told, during the 1950’s, at least 40 other stereotactic apparatus were designed along the three basic types. The Spiegel-Wycis system required translational adjustments, as did the Horsley-Clarke apparatus, in that the position of the electrode was changed by sliding the electrode carrier anterior-posterior and laterally along a base plate, to adjust to the proper coordinates, the vertical adjustment was made by a microdrive system holding the electrode. The predetermined trajectory might require two separate angular adjustments. Other devices rapidly followed [40]. Lars Leksell [41] returned home to Sweden after a trip to Philadelphia and designed the first arc centered device in 1948. The three coordinates indicated the center of a semicircular arc along which an electrode carrier moved, so it always pointed to the isocenter at which the target lay. Since the target always lies in the center of the arc, insertion along any angle would bring the probe to the target. The following year, Talairach [42] in Paris designed a system which involved the insertion of electrodes through a fixed grid system, which in turn invited an elaboration of the cerebral circulation. In Germany, Riechert and Wolff [43] reported in 1951 their translational system, which included the first phantom base to verify the settings mechanically.

1

Bailey and Stein [44] exhibited a burr hole mounted apparatus in the United States that same year. It was essentially a pointing device mounted on a ball and socket joint. The electrode was aligned with the intended trajectory drawn on both the AP and lateral films, resulting in an angular adjustment pointing at the target. Parallax was corrected by repeatedly adjusting to repeated films as the electrode was further advanced, a system of successive approximations. Special note must be made of the development of Narabayashi’s [45,46] system in Japan. He had been cut off from the western literature at the end of World War II, but independently developed an apparatus under difficult circumstances [47]. After finding out that his system was not unique, he conceded that Spiegel and Wycis had preceded his contribution. In the first decade after stereotactic surgery was born, a number of centers were developed throughout the world. Leksell in Sweden [41]. Talairach and associates [48] and Guiot and colleagues [49] in France. Riechert and Mundinger [50] in Germany. Gillingham [51] in Great Britain, Laitinen and Toivakka [52] in Finland, Rossi [53] in Italy, Bertand and colleagues [54] in Canada, Velasco Suarez and Escobedo [55] in Mexico, Obrador [37] in Spain, and Bechtereva and colleagues [56] and Kandel [57] in Russia, among others. Each investigator in turn added his own embellishments and indications, and the field grew rapidly. Within 20 years, stereotactic surgery was practiced throughout the world [58,59]. It is estimated that by 1965 more than 25,000 stereotactic treatments had been done worldwide [60], and 37,000 patients had been treated by 1969 [61]. The need for communication among this handful of pioneers required meeting together, and the stereotactic societies were born. The first meeting of the International Society for Research in Stereoencephalotomy (which became the World Society for Stereotactic and Functional Neurosurgery in 1973) was held in Philadelphia

9

10

1


in 1966, 20 years after the first human stereotactic procedure. The emphasis was on instrumentation, since there were still few commercially available apparatus. Everyone was invited to bring their stereotactic apparatus, or at least the portable portion. Over 40 systems were laid out on the benches in the chemistry laboratory and demonstrated. The second meeting was held in Atlantic City the following year and the emphasis was on the growing number of indications for stereotactic surgery. During the 1960’s, the Todd-Wells stereotactic apparatus became the most popular in the United States, and the Leksell and RiechertMundinger systems in Europe [62,63]. There was little further modification of apparatus during the next decade, when the emphasis shifted to indications and results. The indications were there: There are four fields in which stereotactic functional neurosurgery has found a place, and all four were established during the first decade after stereotactic surgery was introduced: movement disorders, pain, epilepsy and psychiatric abnormalities. The most cited is the use of stereotactic techniques for treatment of movement disorders. Historically stereotactic surgery was developed to offer a more refined way of doing psychosurgery, without the gross inaccuracies and complications of pre-frontal lobotomy. Treatment of pain was addressed very early in the history of this new technique. Although there were early attempts at treating epilepsy with stereotactic surgery, they were less successful and became more important later, after imaging and invasive monitoring were introduced.

Movement Disorders – the Early Years Although the origin of functional neurosurgery was motivated by a need to improve psychosurgery, the first patient had a motor disorder (Huntington’s chorea).

There was growing understanding that interruption of the extrapyramidal system might benefit patients with movement disorders, particularly Parkinson’s disease, but pre-existing surgical techniques had unacceptable morbidity and mortality. The evolution of surgery for Parkinson’s disease deserves mention. Before the introduction of stereotactic techniques, the most common surgery for movement disorders involved resection of the motor or pre-motor cortex, since it was proposed as early as 1932 by Bucy and Buchanan [64] that it was necessary to interrupt the pyramidal system to obtain relief of athetosis. Shortly after that, Bucy [65,66] also claimed relief of tremor after interruption of the pyramidal system, even though the procedure was fraught with severe post-operative deficits. In 1933, Putnam [67] interrupted the proprioceptive input by posterolateral cordotomy in an attempt to lessen tremor and rigidity with less risk. Walker [68], on the other hand, who had reported mesencephalotomy for pain in 1942, attacked the extrapyramidal pathways at that same mesencephalic level in 1949 by incising the peduncle [68,69], which was reported also about that same time by Guiot and Pecker [70] (and will have some historical importance later in this story). It was the often under-appreciated Russel Meyers [22,23] at the University of Iowa who provided the surgical observations that led to the definition of targets that would soon thereafter become targets for stereotactic surgery. Up to the early 1940s it was thought that a lesion involving the extrapyramidal system would result in irreversible coma, a conclusion that was based on Dandy’s observations that frontal lobe stroke involving the basal ganglia was invariably lethal [71,72]; because Dandy has so stated it was not challenged. Meyers [73] defied that convention and reported in 1939 benefit in a single patient with Parkinson’s disease from surgical extirpation of the head of the caudate nucleus. He was encouraged to investigate further, and he devised


intricate non-stereotactic interhemispheric and transventricular approaches to interrupt the ansa lenticularis at the base of the globus pallidus, which he reported in 1942. Since there were no drugs effective against the manifestations of Parkinson’s disease, adventurous surgery was often indicated, but even Meyers cautioned against its general use, as he reported a mortality rate of 15.7%. In contrast, within a decade, after stereotactic surgery was introduced, Spiegel and Wycis [74] reported operative mortality of 2% to attain an even more accurate interruption of the extrapyramidal system. Riechert and Mundinger [75] soon after reported a mortality rate less than 1%, where it remains today. Although the first reported stereotactic case involved pallidotomy for Huntington’s chorea [5], Spiegel and Wycis were initially reluctant to lesion that structure for Parkinson’s disease for fear that the hypokinesia that is seen after experimental pallidotomy might make Parkinson akinesia worse. However, Hassler and Riechert [76] reported in 1954 that they had successfully treated Parkinson’s disease as early as in 1952 by ventrolateral thalamotomy. This encouraged Speigel and Wycis [77] to lesion the ansa lenticularis fibers as they emerged from the pallidum, which they termed pallido-ansotomy. At about the same time, Narabayashi and Okuma [33] made a lesion within the pallidum with procaine-oil injection. The first patient who had stereotactic surgery in 1947 had two lesions made by alcohol injection, one in the globus pallidus and one in the dorsomedial thalamus [5]. The reasons for using two targets were [6] to interrupt the extrapyramidal circuit in the pallidum and [7] to lessen the emotional tone of the patient by interrupting the thalamic projections to the frontal lobe, since it was recognized that tension made the chorea worse. Although the patient had only moderate but temporary improvement, it had been demonstrated that pathways could be interrupted with minimal risk and provide

1

improvement in motor control while alleviating involuntary movements. During the next 15 years, a number of targets were identified for a variety of movement disorders. Patients with intention tremor and Parkinson tremor were often considered together. Intention tremor was treated with lesions in the ventrolateral nucleus [78,79], as was hemiballismus [80]. Substantia nigra lesions were also used for hemiballismus [19,81] and hypertonus [82]. There were reports of improvement with targets in the globus pallidus [74,83]. The most common target for tremor at that time was in the ventrolateral thalamus [78,84,85]. Hyperkinesia was treated with a variety of targets [75,86,87]. Hassler [88] had not yet published his nomenclature of the subdivisions of the thalamus. The most common indication for stereotactic surgery was Parkinson’s disease. This was prior to the age of l-dopa, and pharmacologic management was not adequate. Consequently, patients were referred to the surgeon earlier in the course of the disease than now, although with more severe symptoms. The later stage bradykinesia was seen less, and patients more often presented with tremor as the primary symptom. There was no such thing as dopamine dyskinesia, and one must remember in reading the literature that the patients who were treated up to the end of the 1960s were different from the patients we see today [40]. Also some patients suffered from postencephalitic parkinsonism after the big Spanish flu epidemic in 1918, a condition which is seen only exceptionally nowadays. Irving Cooper [89] first appeared in the functional neurosurgery literature in the early 1950s. During a Walker pedunculotomy [69] for Parkinson’s disease, Cooper inadvertently interrupted the anterior choroidal artery in a now famous ‘‘surgical accident,’’ so he clipped the vessel and stopped the surgery. The patient awoke with relief of symptoms and no deficit, so, as he reported in 1953, Cooper advocated anterior choroidal ligation as a treatment for parkinsonism [89]. However, both

11

12

1


the results and complications were very variable. He correctly concluded that he had often produced a stroke in an area involved with the production of parkinsonian symptoms, and the globus pallidus had been cited by others as a stereotactic target, so in 1955 he began to inject free-hand that structure with alcohol, so-called chemopallidectomy [90]. He did not use stereotactic localization (his injection guide which he introduced later was not Cartesian nor based on cerebral localization) but he reported anecdotally a relatively large series of patients. The trajectory that Cooper had selected for chemopallidectomy was through the temporal lobe, aiming upwards toward the pallidum. One patient who had a good result died of other causes, and the autopsy revealed that the needle had advanced further than intended and the lesion was actually in the ventrolateral thalamus, so Cooper moved his intended target to that area which had been introduced earlier by Hassler and Riechert. In 1958 Cooper reported a series of 700 patients contrasting chemopallidectomy and chemothalamectomy [91,92] and the series had grown to 1,000 patients by 1960 [93]. He still used alcohol injection, which was a problem, since it would spread in an uncontrolled fashion. Consequently, he changed from injection of alcohol with a brain needle to a cannula with a balloon at the tip that was intended to leave a cavity that would retain the injection material, but it did not [36]. This fared no better and its blunt tip caused damage throughout the insertion tract, so he introduced a thick injection material, Etopalin [93], but still encountered the problem of uncontrolled spread. This led him to introduce the Cryoprobe [39] in 1961 – although it had a relatively large diameter and blunt tip, it made a lesion by destroying the tissue by freezing a ball of tissue at the end of the probe, which he continued to use for the remainder of his surgical career. If we might digress, the story of the Cryoprobe provides a little-known but fascinating coincidence. The original Spiegel-Wycis report introducing human stereotactic surgery [5] in

1947 had four coauthors. In addition to Spiegel and Wycis, there was Marks, the machinist from Temple Medical School who actually made the device. The fourth author was A.St.J.Lee, who was a college student who worked as a handyman in Spiegel’s laboratory. He soon after quit his job and returned to school, completed an engineering degree and became a freelance engineering consultant. The second author on the original Cooper report of ‘‘cryostatic congelation’’ is the same Arnold St. J. Lee [39], who was coincidentally hired by Cooper to design and build the Cryoprobe. I (plg) understand that a dispute arose over who owned the patent, which was never fully resolved. By 1954, Hassler and Riechert [76] had defined their thalamic targets more precisely, with the Vop recommended for tremor and the Voa recommended for rigidity. This was made possible by Hassler’s [88] subdivision of the thalamic nuclei, so that the results of lesioning more precisely localized targets allowed better correlation with clinical results. Spiegel and Wycis [74] still preferred the pallidum as the target, and in 1958 they observed that the lesion would be more effective against rigidity if it were placed more posteriorly than the tremor target in the emerging ansa lenticularis fibers. In 1959, Svennilson and coauthors [52] reviewed Leksell’s cases and also advocated a pallidal lesion more ventral and posterior than that used by other authors (in a study later cited by Laitinen [94] that led to the rebirth of VP pallidotomy). Another interesting study was published in 1959 by Levy who was a stipendiary of the Swiss Academy of Science spending several years under the guidance of Riechert in Freiburg [95]. He compared the differences in the pallidal target used by various groups at that time. This study did not only consider the target coordinates but also the angle of the trajectory in relation to the line between the foramen of Monro and the posterior commissure. There was a large variety considering both the target in the pallidum itself


but also in the attempts to include pallidofugal fibers. Hassler and colleagues considered pallidotomy to be not effective for alleviation of hypokinetic symptoms because it was thought that a negative motor symptom could not be improved by a lesion: ‘‘for theoretical reasons alone, one could hardly expect an improvement of akinesia by a stereotactic operation’’ [7,96]. Nevertheless, despite the theoretical difficulties it was occasionally acknowledged that pallidotomy indeed improved akinesia. In the late 1950s, however, most stereotacticians followed the lead of Hassler and Riechert to the thalamus, since the most dramatic effect on the early Parkinson’s disease before the l-dopa era was prompt and complete relief of tremor, the dominant symptom seen at that time. The optimal target for tremor was eventually accepted to be the Vim nucleus [75,97–99], and the pallidum was almost abandoned [100]. Meanwhile, Spiegel and Wycis [101–103] moved their lesion for Parkinson’s disease to Forel’s field, a procedure they called ‘‘campotomy,’’ for campus Foreli. In 1961, microelectrode recording (MER) from the human brain was introduced by AlbeFessard [104]. It was adopted as a surgical tool to the point where many consider it an essential technique [105–107]. A number of authors used MER to confirm localization in the Vim nucleus [54,108,109], but MER was not yet a routine part of stereotactic surgery, in part for the lack of adequate recording in the unfriendly operating room electrical environment. (It later became important in defining the subthalamic nucleus [110,111]). Both thalamotomies and pallidotomies were also popular for the treatment of dystonia in the 1960s. The results were more variable than those reported for parkinsonism [8,112] but overall beneficial results were achieved in more than 50% of patients including generalized and cervical dystonia. There is published experience with more than 300 patients with cervical dystonia [104,113]. Hassler elaborated a complex model

1

to explain the deviation of the head along its different axes including the interstitial nucleus of Cajal, the pallidum and thalamic nuclei [114,115]. According to his pathophysiological concepts, Hassler suggested choosing different targets depending on the pattern of cervical dystonia in the individual patient. In cases with rotatory torticollis, the primary target was the nucleus ventralis oralis internus and its interstitiothalamic pathway, while for treatment of horizontal torticollis pallidofugal pathways in Forel’s field H1 and the ventro-oral thalamic nuclei were approached. Working further along this concept Sano even started to use the interstitial nucleus of Cajal as a target for cervical dystonia [15,116]. The elaborate selection of surgical targets within the basal ganglia circuitry according to the phenomenology of dystonia in individual patients certainly was fascinating, although it was never conclusively shown that such superselective approaches indeed resulted in improved clinical outcome. It was estimated that by 1965 more than 25,000 functional stereotactic procedures for Parkinsonism had been performed worldwide. Things changed drastically in 1968, when l-dopa became generally available [59]. Within a few months, the number of patients with Parkinson’s disease presenting for surgery plummeted. Only a few patients with primary tremor went to stereotactic surgeons during the next few years. The number of neurosurgeons doing stereotactic surgery declined, and the field was mainly practiced by a few neurosurgeons in academic centers who enjoyed the challenge. There were three principal factors that lead to the abandonment of functional stereotactic surgery for dystonia in the late 1970s, about 10 years later than for parkinsonism. First, the general decline of movement disorders surgery at that time, second, the introduction of selective peripheral denervation by Bertrand, and third, the widespread use of botulinum toxin thereafter. It was not until the introduction of CT-directed targeting,

13

14

1


then MRI targeting, and later the reintroduction of deep brain stimulation that stereotactic surgery rebounded, and it is now more active than ever [59]. Eventually, it was recognized that l-dopa is not a permanent answer to Parkinson’s disease. As the disease progresses with long-term l-dopa administration, the symptoms become more refractory and dopa dyskinesias may limit the tolerated dose [117,118]. The response to medication may vary significantly and abruptly, producing repeated on-off episodes during the day. Rigidity and akinesia may progress, and freezing episodes may intervene. By the early 1980s, a search was on for a surgical answer to these problems. In 1985, Backlund and associates [119] reported their first clinical trials of autologous transplantation of adrenal medullary tissue into the head of the caudate nucleus in two patients. They were encouraged enough to conclude that ‘‘the results merit further clinical trials.’’ Two years later, Madrazo and colleagues [120,121] reported two additional patients and cited more favorable results. Later that year they reported that they had operated on 18 patients [120]. This provoked great interest and programs were begun at a number of institutions. Unfortunately, the results were modest and of limited duration, [122–124], the surgery to retrieve one adrenal gland was stressful to this fragile group of patients, and the adrenal was sometimes found to be atrophic [125]. In addition, there were a number of complications resulting from the craniotomy, if open surgery were used [126], but few complications arose when the adrenal tissue was injected using stereotactic methods [127]. By 1991 the procedure was essentially abandoned. The idea of tissue transplantation continued to garner interest, and attention was shifted to the transplantation of fetal nigral tissue. As early as 1984, a symposium was devoted to the potential neurosurgical use of fetal cells [128]. The experimental groundwork had been done in rats in the

late 1980s [129] or MPTP-injected monkeys [130, 131]. In 1989, the first two patients to receive fetal cell transplantation for the treatment of Parkinson’s disease were reported by Lindvall [132], a multi-institutional consortium was established and others patients followed soon afterward [124]. The problems that existed with such a complex program caused interest to wane. There was considerable difficulty obtaining fetal tissue at the proper age and in the proper amounts, identification of the proper tissue in the fetal brain was difficult, coordination between the service obtaining the tissue and the implantation surgery team was expensive and elaborate, and the ultimate goal of the tissue remained in doubt [122,123,133]. The field of functional neurosurgery was otherwise quiet, through the 1970s and 1980s, except for long-established procedures primarily for pain and epilepsy [59]. There was little psychosurgery being done, mostly because of a public campaign against it [134] and the development of new psychotropic drugs, rather than any scientific reason. Let us pause in the story about stereotactic surgery for movement disorders, as we might catch up to this date the story of other fields of progress, to a great extent in technology. Stereotactic and functional neurosurgery has by its nature been dependent on advances in other fields, such as radiology and later imaging, radiotherapy, computer science, and miniaturization and implantation of electronic devices.

Persistent Pain Let us return to the birth of stereotactic surgery in 1947. Many of the earliest patients had intractable pain [135]. An aside – Spiegel always referred to such pain as ‘‘so-called intractable pain.’’ If there were a chance it might be treated, it would not be intractable, by definition, so truly ‘‘intractable’’ pain was not an indication for the treatment


under discussion. I (plg) prefer not to use the term ‘‘chronic pain,’’ since it includes all longlasting pain, whether from cancer or from a noncancer (or unknown) etiology, which are two different clinical conditions managed in very different ways. Instead, I insist on the use of ‘‘persistent pain’’ if both cancer and non-cancer etiologies are discussed together [136]. Functional neurosurgical management of pain had been considered for almost a century prior to the advent of stereotactic surgery [137]. One of the prime indications was facial pain, particularly trigeminal neuralgia. In 1853, Trousseau [138] suggested that the paroxysmal activity of trigeminal neuralgia resembled epilepsy and suggested gasserian ganglionectomy, but because surgery in that area was too adventurous at that time, it was not reported formally until 1890 by Rose [139]. Horsley and coauthors [140] advocated epidural total gasserian ganglionectomy by a transsphenoidal approach in 1891, but a mortality rate of 20–25% was prohibitive, even in his excellent hands. In 1900, Cushing [141] modified the procedure, and in 1920 he reported 298 consecutive cases without mortality [142]. Ramonede [143] had already introduced the suboccipital approach in 1903, and Dandy [144] modified it in 1925. In 1931, Kirschner [145,146] introduced electrocoagulation of the gasserian ganglion. He developed an apparatus to guide the electrode through the foramen ovale, which was not Cartesian and therefore not stereotactic. His use of electrocoagulation in the human was encouraging to those who later used it in the brain. At the time that stereotactic surgery was introduced, the philosophy for pain management was to interrupt the primary pain pathway, and the most inviting target was the spinothalamic fiber bundle. Although the first patient operated with stereotactic surgery had a motor disorder, Huntington’s chorea, the second had persistent pain. Walker [26] had earlier described the surgical section of the pain pathway at the level of the mesencephalon, and Spiegel and

1

Wycis made lesions in a mesencephalotomy target [135,147], although again they made a second lesion in the dorsomedial nucleus of the thalamus, as well. Walker had reported dysesthesia from the medial lemniscus being involved with his mesencephalotomy incision, but stereotactic techniques avoided that complication. If the lateral spinothalamic tract is sectioned by open surgery in the spinal cord, the procedure is anterolateral cordotomy [148–150]. If sectioned in the brain, the spinothalamic fibers that have not already synapsed in the lower brain stem lie just behind the medial lemniscus at the level of the mesencephalon. Unfortunately, when the surgical section included the medial lemniscus, severe dysesthesia might occur [151], and that was more easily avoided with stereotactic than with open surgical techniques [152,153]. The quintothalamic fibers, which descended from the trigeminal area to join the spinothalamic bundle, provided a more compact target, so a stereotactic procedure was done for facial pain quite early [135]. Again, dorsomedial thalamotomy was often recommended to relieve the emotional component of persistent or chronic pain [154]. Thalamic pain was managed with lesions in the ventral posteromedial nucleus and the centre median nucleus [42]. In 1949, Hećaen, Talairach and associates [42,155] reported on a thalamotomy for persistent pain, but they interrupted the diffuse pain projection system in the centrum medianum plus a lesion in the ventrobasal complex, demonstrating successful relief of pain without interrupting the primary pain pathway. Spiegel and Wycis [156] reported similar findings for facial pain in 1953 and further refined the procedure in 1964 by defining several distinct targets within the mediobasal thalamus [157]. The concept of interrupting the reticulospinothalamic tract for pain management led Nashold [158] in 1969 to extend the mesencephalic lesion into that bundle. This in turn led him to explore the mesencephalon with implanted electrodes to define the multisynaptic pathway [159].

15

16

1


The result was to leave the primary pain pathway intact and to confine the lesion to the extralemniscal area, which provided equivalent pain relief with minimal risk of dysesthesia [151,160]. This demonstration led to the limbic system becoming a primary target for pain management. Foltz and White [161] (as reported by Sweet and Gybels [136], who previously had considerable experience with the use of cingulotomy for psychiatric disorders, began in 1962 to use that procedure for pain management. Long-term relief was reported in over half the patients with cancer pain and one-third with ‘‘failed back syndrome.’’ Although it is functional but not ordinarily stereotactic, percutaneous cervical cordotomy must be mentioned. In 1959, Mullan [162] introduced the implantation of a radioactive seed as a method for producing lesions for motor disorders, such as Parkinson’s disease. In 1963, he introduced the use of a radioactive strontium needle inserted percutaneously between the arch of the first and the lamina of the second cervical vertebrae, lain against the anterolateral surface of the spinal cord for a measured duration, in order to make a cordotomy lesion to manage pain without the need for surgical exposure [163]. This approach was used in 1965 by Rosomoff [164] who made a controlled lesion with a radiofrequency electrode, which was much more practical. This more practical procedure was adopted by many neurosurgeons. There was one grave risk, however, that of Ondine’s curse or sleep-induced apnea after bilateral lesions were made [165]. Unilateral percutaneous cervical cordotomy, however, had good results with reasonable risk, particularly in cancer patients but not particularly in patients with other persistent pain. Nevertheless, since many of the cancer patients required bilateral pain management, Lin, Polakoff and I [166] introduced in 1966 a technique to introduce the electrode through a disk to the lower cervical spinal cord, which provided good relief of bilateral pain but with occasional weakness of the hand.

Back to 1937, when Sjo¨qvist [167] presented a technique, later modified by White and Sweet [168] for surgical interruption of the descending trigeminal tract at the level of the medulla, limiting the analgesia to the distribution of the trigeminal nerve. In 1970, Hitchcock [169] made the procedure stereotactic by inserting the electrode through the foramen magnum, an approach that was uniquely suited to his stereotactic apparatus, and reported good results in postherpetic facial neuralgia. He also used that approach to modify percutaneous cervical cordotomy by inserting the electrode from dorsally through the uppermost spinal cord into the anterolateral spinothalamic tract [170], which was similar to the approach Crue and his colleagues [171] reported in 1968. This led to an inadvertent discovery of a new pain pathway [172]. One of Hitchcock’s patients reportedly moved as the electrode was inserted, and the lesion was made in the midpoint of the spinal cord at the cervico-medullary junction, with an excellent clinical result and no complications [173]. He theorized that he had interrupted a pain pathway, probably the spinoreticulothalamic multisynaptic path, and began to use the same target for pain management. He and later Schvarcz [174] termed the procedure extralemniscal myelotomy. In an attempt to avoid making a cervical lesion to treat pain confined to the pelvic area, Gildenberg and Hirshberg [175] made a mechanical lesion under surgical exposure at the thoracolumbar spinal cord level with good relief of particularly visceral cancer pain, a procedure they dubbed limited myelotomy. The spinal cord of one of the patients who had later died from his cancer was studied by Al-Cher, Willis and his group [8], who demonstrated that there was indeed a pathway, previously undescribed, not multisynaptic, transmitting pelvic visceral pain. This brings us to the point in the discussion where pain management intersects with chronic stimulation of the nervous system, the beginning of so-called neuromodulation.


Neuromodulation-Stimulation There is a long history of application of electrical stimulation to the nervous system. It is reported that Scribonius treated gout pain by application of an electric torpedo fish as early as 15 AD [176]. One of the earliest to perform controlled stimulation with close observation of muscle contraction was Benjamin Franklin [177] as early as 1774, several years before Galvani [178] demonstrated electrical contraction of frog muscle in 1780. Perhaps the earliest treatise on potential physiologic use of electricity was written by Mary Shelly [179] in her novel Frankenstein, which was based on scientific speculation previously published by Dr. Erasmus Darwin, Charles Darwin’s grandfather. In 1870, Fritz and Hitzig [180] demonstrated that limb movement occurred on stimulation of the motor cortex of the dog. Electrical stimulation was applied to the brain of an awake patient soon after, in 1874, when Barthalow [181] stimulated the motor cortex that was exposed after debridement for osteomyelitis. It was left to Sir Victor Horsley [182] to first use intraoperative stimulation in 1884, when he stimulated an occipital encephalocele and noted conjugate eye movement. The first electrical stimulator designed specifically to treat pain appeared in the early 1900s when the Electreat was advertised to relieve not only pain, but innumerable physical maladies, as well. It was battery operated and bore an uncanny resemblance to TENS units that appeared 70 years later. Stimulation during stereotactic surgery was used by Spiegel and Wycis from the very first case in order to obtain physiologic localization of the electrode. Since the first lesions were in the pallidum and soon after the thalamus, stimulation was used to assure that the electrode or needle did not lie in the internal capsule by observing for a contralateral involuntary motor response. Although various frequencies were used, there

1

was not originally a distinction made between high frequency and low frequency stimulation, nor were the responses to stimulation recorded with detail. Hassler [183] was the first to suggest that observations made on stimulation in the operating room might have long-term effects. He had been a graduate student of Rudolph Hess in Switzerland during the late 1940s, where chronic stimulation was routinely administered to cats with permanently implanted electrodes. They recognized that stimulation might produce the same effects as making a lesion at the same site, but changing the frequency might give the contrasting effect. Throughout the stereotactic community, little attention was paid to high frequency versus low frequency stimulation. Those terms were not universally defined, so some institutions tested the patients with significantly different parameters. In addition, stimulation caused seizures in some rare instances, so higher frequencies were sometimes avoided. Even so, intraoperative stimulation became the norm prior to lesion production. For instance, Spiegel and Wycis [103] made lesions in Forel’s field for the treatment of Parkinson’s disease, but stimulated critically prior to lesion production. The desired target lay just above the emerging oculomotor fibers, so they stimulated at successively deeper levels until uni-ocular deviation demonstrated that they had impinged on the oculomotor fibers. The electrode was then withdrawn 2 mm and a lesion was made. Tasker [184] performed an extensive stimulation study of the human thalamus, which he published as a physiological atlas in 1982. The technique was very similar to that he used as a graduate student with Clinton Woolsey from 1961 to 1963 to map the cortex and subcortex of animals [185]. Since there was no way to apply stimulation for prolonged periods, the therapeutic potential of stimulation of the brain was rarely pursued.

17

18

1


Perhaps the earliest was Pool [186] who stimulated frontal tracts rather than performing prefrontal lobotomy for psychosurgery in older patients as early as 1948. In the early 1950s, Heath [187] stimulated a variety of subcortical areas and made detailed observations about behavioral changes. In 1954, Olds and Milner [188] observed that rats would aggressively seek stimulation of the septal area, which they concluded provided the rats with intense pleasure. That same year, Heath [189] concluded that, since pleasure is the opposite of pain, septal stimulation might be used to treat pain. There were no implantable stimulators, so he stimulated the septum for intervals of 15 min daily to weekly and was able to alleviate cancer pain in one patient. A decade later, when chronic stimulation was available, Gol [190] implanted stimulators in the septal area and had relief of cancer pain in several patients. In 1972, Bechtereva [191] reported what may have been the first therapeutic use of stimulators in motor disorders. However, implantable stimulators were not available in Russia, so again the stimulation was applied to the pallidum or thalamus only intermittently. In the meantime, significant advances had been made in the understanding of pain perception. The concept that provided the basis for the use of stimulation in persistent pain was the presentation of the Melzack-Wall [192] gate theory in 1965. They proposed that pain perception involves a ‘‘gate’’ that can be opened or closed to allow pain sensation to pass through or to block it. If the activity of the small pain fibers predominates, the gate will open. If the large non-pain proprioception and touch fibers predominate, te gate will close and pain will decrease (‘‘If you rub it, it feels better!’’). Since the large non-pain fibers are isolated in the dorsal columns, stimulation of that part of the spinal cord should close the gate and provide relief of pain. Stimulating the skin with TENS unit electrodes applied to the skin and adjusting the sensation so it is below

pain threshold [193]. It could also be done by stimulating through an electrode placed behind the spinal cord (and in other positions, as it turned out), which provided the impetus to provide an implantable stimulator. The gate theory was tested in 1967 by Wall and Sweet [194] who stimulated their own infraorbital nerves. Sweet recruited Roger Avery, an engineering colleague at MIT, to make an implantable stimulator, which he and Wepsic [195] used to provide peripheral nerve stimulation for pain management. At about the same time in 1967, Norm Shealy [196] stimulated the large nerves where they were uniquely gathered in the dorsal columns of the spinal cord. He theorized that the impulse would travel retrograde down the dorsal columns to inhibit the small nerve input at each level of the spinal cord to close the gate and diminish pain sensation. He recruited Thomas Mortimer, a graduate engineering student, to design an implantable stimulator. The first mode required an external power supply connected to the stimulator by needles inserted through the skin. By coincidence, Mortimer had interviewed for a job at Medtronic 2 years before that. He contacted Norm Hagfors, one of the engineers he had met, to see if their cardiovascular stimulator might be adapted to stimulate the spinal cord. Mortimer designed an implantable electrode that Shealy used with the Medtronic cardiac stimulator, which provided relief of pain for the last several months of a cancer patient’s life. Shealy contacted Medtronic to improve and provide the system for more patients. Medtronic had previously been in the business of manufacturing implantable cardiovascular stimulators. In 1963, they were making the Barostat, which was used to stimulate the carotid sinus for treatment of hypertension. In 1965, they released the Angiostat, which stimulated the carotid sinus for angina. Shealy’s second patient used that stimulator attached to the electrode that Mortimer had designed and had good relief of chronic pain for


4 years. In 1968, Medtronic provided this system for spinal cord stimulation as the Myelostat. The early spinal cord stimulators came in two parts. An internal implantable system had no internal power supply, since implantable batteries were just being developed. It was attached to a circular wound antenna implanted subcutaneously. The external unit had the controls and a battery, and transmitted both the power and the control signal transcutaneously. By 1981, battery technology had advanced to the point where the entire unit could be implanted. Avery had kept pace with Medtronic until that time, but when Roger Avery retired, the company no longer provided implantable spinal cord stimulators. In 1971, I (plg) was working at the Cleveland Clinic just down the street from where Shealy and Mortimer had introduced spinal cord stimulation. I was able to obtain stimulators from Medtronic modified to provide frequencies of 800–1,200 Hz. They were implanted to stimulate at the C2 level for the treatment of spasmodic torticollis. Half the patients had significant relief, which was the first use of implanted stimulators for motor disorders [197]. In 1976, both Cook [198] and Dooley [199] recognized improvement in spasticity in multiple sclerosis patients who had had spinal cord stimulators implanted for pain of muscle spasm. In addition, Dooley [200] recognized improvement in peripheral blood flow in patients undergoing spinal cord stimulation for pain management. It was in 1973 that Hosobuchi [201] implanted a stimulator attached to an electrode in the somatosensory thalamus for treatment of denervation pain with anesthesia dolorosa, and the field of deep brain stimulation (DBS) was born. There were few attempts during the 1970s to introduce DBS also for the treatment of movement disorders. In 1977, Mundinger reported on the benefits of unilateral thalamic DBS in seven patients with cervical dystonia [202]. Intermittent stimulation for 30–40 min with frequencies up to 150 Hz resulted in improvement of dystonia

1

for up to 7 h. The longest follow-up record was 9 months. The results were never published in the English literature and the study was completely forgotten for decades. Let us go back to 1969. Reynolds [203] demonstrated in the rat that periventricular stimulation produced profound analgesia, sufficient to perform surgery with no apparent pain perception. In 1977, Richardson and Akil [204] provided chronic periventricular stimulation to patients and produced marked reduction in pain. The following year, they demonstrated that application of such stimulation was associated with release of endorphin into the ventricular fluid [7]. Also in 1977, spinal cord stimulation and DBS were being used so extensively that the Food and Drug Administration held a symposium on safety and efficacy [205]. It was felt that the use of stimulators for pain had been documented, but not for other uses, which at that time included movement disorders, epilepsy, cerebral palsy and bladder control. At about that same time, the FDA was given the charge to regulate devices as well as their historical charge to regulate drugs. They felt that relief of pain by deep brain stimulation had not been sufficiently documented and gave the manufacturers several years to provide the data to document that such devices should be continued, especially for pain management. Of the three manufacturers, only Avery provided data, but just then Roger Avery retired so that product was discontinued, and the use of DBS for pain management was de-approved. In 1991, Tsubokawa [206] reported on the stimulation of the motor cortex, but not the sensory cortex, for the management of central pain. In 1995, Migita [207] reported on the use of extracranial magnetic stimulation of the motor cortex for pain management. As a general rule, ablative procedures are more usually indicated for cancer pain, but stimulation procedures for non-cancer persistent pain [208,209].

19

20

1


In 1979, the senior author commented that the engineers could give us any stimulation parameters we wanted, but it was up to the surgeons and scientists to let them know what we need [210]. That has not changed.

The Return of Movement Disorder Surgery When l-dopa became widely used for treatment of Parkinson’s disease in 1968, stereotactic surgery was rarely used for that indication. Although there were other indications for stereotaxis, they were not enough to maintain a critical mass of centers, so stereotactic surgery went into a marked decline [59]. It was a seminal report by Laitinen [94,211] in 1992 that re-awakened interest once again in the use of pallidotomy for Parkinson’s disease, and with that awakened the field of stereotactic and functional neurosurgery. It had become obvious that many of the patients who took l-dopa for years were still disabled and needed a more aggressive therapy. Even more, it was recognized that chronic l-dopa treatment resulted in disabling dyskinesias and unpredictable motor fluctuations. Laitinen recalled the report of Leksell’s pallidotomy that had been published in 1960 by Svennilson and associates [52] who had reported more improvement in bradykinesia and rigidity with a lesion that Leksell had made in the ventral posterior area of the pallidum than with the more central lesion. He attributed this observation to the inclusion of the pallidofugal fibers in the lesion, similar to Spiegel and Wycis [212] pallido-ansotomy. After a cautious start in 1987, Laitinen and his colleagues [94,211] reported in 1992 a significant improvement in rigidity, bradykinesia and dopa-induced dyskinesia in patients who had been taking l-dopa for years for their Parkinson’s disease. There was an enthusiastic return to stereotactic pallidotomy for Parkinson’s disease [213]. The setting was more advanced than had

existed in the original years of surgical management of Parkinson’s disease. The evaluation of results was more sophisticated [214], localization of the target had presumably become more accurate with the introduction of imaging techniques [215], and microelectrode confirmation of the target [216] may have helped to localize the lesion. In addition, it was found that pallidotomy managed not only the symptoms of Parkinson’s disease, but the side-effects of l-dopa, such as dyskinesia. The first ‘‘modern’’ pallidotomy in the United States was performed in December, 1991, in the Hospital for Joint Diseases, New York, by Michael Dogali [217] in the presence of Laitinen and Tasker. Dogali initially considered the possibility of placing lesions in the subthalamic nucleus after reading reports on the experimental studies of the Atlanta group and discussed this issue with Ransohoff. As a neurosurgeon who had actively participated in the first wave of functional stereotactic surgery, however, Ransohoff strongly opposed using the subthalamic nucleus because of the fear of producing hemiballism and instead suggested using the pallidum. Many pallidotomy series were reported over the next 5 years [213,218]. Many of the neurologists who referred such patients became involved in the intraoperative microelectrode evaluation of targets [216], further encouraging referral of patients. The resurgence in stereotactic activity also led to a search for new targets. Improvements in imaging during the prior decade made target visualization more feasible. Improvements in physiological localization allowed the surgeons to be more secure in identifying targets. Prior to the lull, lesions had been placed by a few adventurous neurosurgeons in the zona incerta [219] and the subthalamic nucleus [220], which became a target for DBS. In 1980, during the lull, Brice and McLellan [221] used thalamic stimulators in two patients for management of tremor of multiple sclerosis. Also during the lull, the following year, Benabid and his colleagues [222] reported on management of tremor by deep brain stimulation


(DBS) in the Vim nucleus. In 1994, when Siegfried and Lippitz [223] reported on chronic electrical stimulation of the VL-VPL complex and of the pallidum, they indicated that this represented their experience since 1982. Benabid carried the banner for the use of DBS in motor disorders and popularized the field by his results and the study of the effects of DBS on various targets. Based on laboratory studies that stimulation of the subthalamic nucleus improves the symptoms of Parkinson’s disease, they have concentrated on this target, which has become the primary target for such stimulation [224]. Since the effects of moderately high frequency stimulation are equivalent, the targets for stimulation are for the most part the same than the targets where lesions are made. In those areas where DBS is not generally available, the success of subthalamic stimulation has led to renewed interest in lesioning targets [225]. Other targets have presented themselves, in part because the risk of permanent complications are less with stimulation than with lesions, so surgeons can be more adventurous. Velasco and his group [226] stimulate the prelemniscal radiations with success, a target just lateral to the old campotomy target [227]. DBS has become increasingly popular in Europe during the past decade and in the US since it was approved in 2002 [72,228]. The renaissance of functional stereotactic surgery for dystonia occurred with a delay of about a decade after that of Parkinson’s disease. With that consideration, we have to remember that also the decline of dystonia surgery happened a decade later than that for Parkinson’s disease in the 1970s. Encouraged by the improvement of dyskinesias in Parkinson’s disease patients after pallidotomy, the GPi was reintroduced as a target for radiofrequency lesioning in the early and mid 1990s [112], While it was shown that pallidotomy yielded beneficial results, it became also clear very soon, however, that bilateral surgery which is necessary in dystonia

1

was burdened with a higher frequency of side effects and that the improvement of dystonia lessened after years. Therefore, pallidal DBS was introduced in the mid 1990s for cervical dystonia [229] and generalized dystonia [230,231].

Epilepsy Surgery Let us go back in time again. Epilepsy also invited a variety of approaches [232–234]. Lesions were often made in the targets that had been defined from movement disorders, on the theory that both represented uncontrolled activation that might be propagated through known pathways. Forel’s field was one of those areas, with significant decrease in seizures being reported as early as 1963 [235]. Thalamotomy was used for myoclonic epilepsy [236]. Fornicotomy was used for epilepsy by several investigators [232,237]. Starting in the 1960s, electrodes were inserted for prolonged recording over several days to identify epileptic foci [238]. The hallmark of surgery for epilepsy remained temporal lobectomy in selected patients [239,240], or the more selective amygdalotomy [241], which was made even more precise with image guidance [242], which is discussed below. With image guidance, it was possible to insert a multi-contact depth electrode from occipitally just lateral to the hippocampus through the temporal lobe to its tip, as well as subdural electrode strips over the anterolateral surface and just beneath the temporal lobe. Prolonged recording identified which contact lay at the origin of the seizures. That contact could then be approached also with image guidance to obtain the optimal temporal lobe resection [243].

Psychiatric Surgery Again we go back in time. In the late 1930s and 1940s, pre-frontal lobotomy was introduced as being beneficial for psychiatric disorders [20].

21

22

1


The indications, however, were not clear and it is very likely that many of those patients had schizophrenia, which does not respond to psychosurgery. There were no psychotropic medications other than barbiturates, and often the only alternative to surgery was for permanent commitment to a mental institution. Pre-frontal lobotomy produced poorly controlled damage to the frontal lobes, and it became over-used and abused [244]. During the 1950s, pre-frontal lobotomy became less desirable than it had been, in part because new medications became available and stereotaxis provided a better controlled alternative. The first stereotactic patient had dorsomedian thalamotomy [5]. Dorsomedian lesions were combined with lesions in the anterior nuclei for anxiety and aggressive disorders [245]. Lesions of the internal medullary lamina were again combined with hypothalamic lesions for aggressive behavior [246–248]. Surgery for psychiatric symptoms almost came to a halt in the late 1970s and through the 1980s for what appeared to a great extent for political rather than scientific reasons. As occurred during that period, protests were used to interrupt scientific meetings where psychiatric surgery was being discussed, and all surgery for psychiatric symptoms was equated in the media with the non-anatomical and abused psychosurgery of the pre-stereotactic era [134,249]. Although psychiatric surgery has resumed on a more limited basis, the availability of psychotropics and more specific psychiatric diagnosis has put it on a more scientific basis. The main indications are now recognized as obsessive-compulsive disorder [250,251] and intractable depression [252,253]. As with motor disorders, there appears to be a significant role for deep brain stimulation [254,255].

Image Guided Surgery After the first generation of stereotactic frames, there was little further modification until the late

1970s, with the introduction of image guided surgery [59]. The idea of using X-ray visualization to identify the location of radio-opaque masses is older than you think. Perhaps the birth of image guided surgery can be traced to the work of Elizabeth Fleishmann in San Francisco in 1900 [256]. When she was 28 years old, Roentgen’s discovery of X-rays captured the public’s attention, and do-it-yourself articles appeared in the newspapers. She quit her job as a bookkeeper and somehow managed to set up a complete X-ray system in an office in Sutter Street the following year. She became perhaps the best radiologist on the west coast. Soldiers returning from the Spanish-American War in 1900 were referred to her to identify bullets and shrapnel that they still carried. She developed the skill to triangulate on bullets lodged in the skull or body and to steer surgeons to them so they might be removed. She certainly qualifies as being the mother of image guided surgery. Up until CT scanning was introduced, intracerebral masses were deduced from shifts in blood vessels in angiography or the position of the ventricles in ventriculography. Now for the first time it was possible to see the mass itself, which invited stereotactic intervention as early as 1956 [257]. New techniques had to be developed to relate the target to stereotactic coordinates in order to guide the stereotactic frame with CT images rather than X-ray [258,259]. Each slice of the CT scan presented a measurable two-dimensional picture from which AP and lateral coordinates could be measured. The relative position in space of the target slice provided the third vertical coordinate. By localizing that slice on the head in order to register it to the stereotactic apparatus, the third coordinate could be determined. A biopsy cannula or other probe could be inserted in order to direct the surgeon to the pathology [260]. New stereotactic devices were developed to automate that process. The Leksell apparatus was


modified with a base plate that could be secured to the scanner and then to the head frame in the operating room to transpose the coordinates to the stereotactic apparatus [261]. Another system employed a series of wires of varying lengths incorporated into plastic plates attached to the head frame. The number of wires appearing in cross-section in the target slice indicated its vertical position [262]. One ingenious system used three acrylic screws with lead markers that were secured to the patient’s skull prior to scanning in order to establish a reference plane. The base ring of the stereotactic apparatus was attached to the screws, and coordinates were calculated [263,264]. Several stereotactic devices were developed especially for use with CT (and later MRI) scanning. Some depended on the calibrated movement of the CT scanner table to establish the vertical coordinate [262,265]. A CT scanner that incorporated a stereotactic apparatus was designed [266]. In some busy neurosurgical services, a CT scanner was installed in the operating room [267]. The breakthrough in marrying to stereotactic frames CT or MRI scans, as well as more recent scanners, involved the development in 1980 of a fiducial system that contains all the three-dimensional information for targeting independently on each CT slice. The idea was invented by a medical student, who is the ‘‘B’’ in the BRW apparatus [268]. Three sets of three rods, with each set in an N-shaped configuration, are attached to a frame that is in turn attached to the stereotactic head ring. Since the center rod of each set is diagonal, the height of the slice scan be determined by the position of the center rod relative to the vertical rods on each side of it. Since there are three sets of rods, three points are used to determine the position of the plane bearing the target, and the AP and lateral coordinates can be established on that target slice. This system was first incorporated into the BRW system, a frame consisting of interlocking arcs, and

1

similar systems have been used with other apparatus as well [115,269,270]. The availability of image-based stereotactic surgery opened up new possibilities for management of brain tumors and other masses, and brought many neurosurgeons into the stereotactic arena. The most common stereotactic procedure became biopsy [63]. In addition, aspiration of abscesses [271], hematomas [272,273], and cysts [274] became commonplace. A stereotactic frame could be used to guide the patient to a tumor at craniotomy [275], with a virtual reality program visualizing the tumor beneath the surface prior to making the skin incision. Visualization of structures such as the ventricle and blood vessels can guide the surgeon. The image is upgraded as the resection proceeds, so the surgeon can see when the optimal resection has been done. Cannulas [276] or isotope seeds [277] can provide brachytherapy of various types. Kelly [269,278] is considered by many to be the father of image guided brain tumor resection. His development between 1980 and 1983 of a stereotactic frame that integrated visual guidance technology antedated frameless systems and opened the door to guided resection. The data from a CT scan or MRI was fed into a computer workstation. The computer reconstructed the volume of the target, ordinarily a tumor, into a three-dimensional volumetric object that could be registered to the stereotactic head frame, rather than a target point-in-space, as had been the practice until then. The digital manipulations of those data were even more remarkable when one considers the capability of computers of that time, which required an entire room to manage the data for a single surgery. The registration of the microscope to the patient was done by aligning the view with a cylindrical retractor. The outline of the tumor was seen on a heads-up display superimposed on the microscope view. The image was manipulated so only the crosssection of the part of the tumor being resected was seen on the display. Because of the restricted

23

24

1


access through the cylinder retractor, the resection was usually done with the CO2 laser [279–281]. Several technologies for frameless image guided surgery emerged at about the same time. Since most were developed or at least used commercially early in development, much of the relevant information does not appear in any detail in the scientific literature. The first frameless systems, which became generally available about 1990, consisted of an articulated arm attached to the operating table. Each of the joints contained a potentiometer to measure the angle of each single joint. The information from all the joints could then be used to calculate the position of the tip of the pointer at the end of the arm. It was used first to register the position of the patient’s head to the arm, after which the location of the tip of the pointer in relation to the head and the target could be displayed to the surgeon in real time [282]. The original use of the operating microscope as a frameless guidance instrument was developed by Roberts [283] as early as 1986. It localized its position in space by ultrasound triangulation. However, ultrasound localization had the drawback of being subject to error from air currents or temperature gradients in the operating room. Kelly [284] introduced frameless electromagnetic localization to image guided surgery, first with the Regulus Navigator, then with his Compass system which eventually required only a lap top computer for neuronavigation. The most commonly used localization system at present involves the localization in space of a series of computer identifiable fiducials by a pair of video-like cameras overlooking the surgical field, reported by Heilbrun [285] in 1992 as machine vision. By 1994, his team reported the use of this system for multiple imaging modalities [286,287]. This system was used by Smith and Bucholz [288] in 1994 in a system they called the NeuroStation that grew up to be the Stealth Station,

now the most widely used neuronavigation system in the US (the most widely used system in Europe is the BrainLab system). Fiducials are secured to the patient’s head prior to scanning. A three-dimensional reconstruction of the head, including the fiducials, is computed. In the operating room, the surgeon localizes each of the fiducials sequentially by touching each in turn with the image localized pointer, which registers the location of the patients head to the preoperative volumetric scan. Thereafter, the surgeon can use the tip of the pointer to localize any structure of interest or to localize the tip of the pointer in relation to the target even before the incision is made. More recently, the development of computer-generated skin surface rendering has become an alternative to the use of fiducials. Other authors have used similar threedimensional volume reconstruction as early as 1994 to guide craniotomy and tumor resection by a system that visualizes the entire volume of the target, volume-in-space rather than a pointin-space [275,289–291]. The system that has been used by Gildenberg [275,292,293] localizes a video camera stereotactically, so that the video image of the surgical field is registered to the reconstructed volume of the tumor. The superimposition of both images provides virtual reality to surgical resection.

Further Steps The history of stereotactic and functional neurosurgery has been documented in the ‘‘stereotactic journal’’ as it occurred. When Spiegel first reported the birth of human stereotactic surgery in 1947, he was editor of the journal Confinia neurologica, where a preponderance of early articles was published. When Spiegel retired in 1968, Gildenberg became editor and changed the name to Applied Neurophysiology, since that described almost all of the functional procedures that constituted the field. Within a few years,


image guidance, computers, and stereotactic radiosurgery became part of the field, and the name was changed to Stereotactic and Functional Neurosurgery. Half of the issues between 1968 and 2001 were devoted to proceedings of the World and American Societies for Stereotactic and Functional Neurosurgery. The philosophy was based on the feeling that if anyone wanted to keep up with the field, he or she would know what was happening in the field by reading just this one journal, a repository of stereotactic history as it happened [294]. The journal is stronger than ever under the editorial leadership of David Roberts since 2001. Stereotactic and functional neurosurgery nowadays is a major column supporting the building of neurosurgery. Its concepts have penetrated essentially all subspecialties of neurosurgery and they reach far beyond to other disciplines. Therefore, the question arises how the core of contemporary stereotactic and functional neurosurgery would be defined. In an attempt to establish a Training Chart in Movement Disorders Surgery Added Competence by the ESSFN the first request by the public authorities was to give a precise and pragmatic definition of the field nowadays. After a long discussion the following definition came up as a consensus based on the input of Drs. Blond, Broggi, Gildenberg, Hariz, Krauss, Lazorthes and Lozano: "

Stereotactic and functional stereotactic surgery is a branch of neurosurgery that utilizes dedicated structural and functional neuroimaging to identify and target discrete areas of the nervous system and to perform specific interventions (for example neuroablation, neurostimulation, neuromodulation, neurotransplantation, and others) using dedicated instruments and machinery in order to relieve a variety of symptoms of neurological and other disorders and to improve function of both the structurally normal and abnormal nervous system. The practice of stereotactic and functional neurosurgery mainly extends

1

into the fields of movement disorders, pain, epilepsy, psychoaffective disorders, neoplastic diseases of the nervous system and the restoration of function in degenerative disorders.

References 1. Gildenberg PL. General concepts of stereotactic surgery. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1988. p. 3-12. 2. Horsley V, Clarke RH. The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45-124. 3. Descartes R. Discours de la me´thod pour bien conduire sa raison et chercher la ve´rite´ dans les sciences. Leyden, Holland: Maire; 1637. 4. Pereria EAC, Green AL, Nandi D, Aziz TZ. Stereotactic surgery in the United Kingdom: the hundred years from Horsley to Hariz. Neurosurgery 2008 (in press). 5. Spiegel EA, Wycis HT, Marks M, Lee AS. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 6. Frankfort plane. Dorlandaˆ s Illustrated Medical Dictionary. Philadelphia, PA: W.B.Saunders; 1994, p 1300. 7. Akil H, Richardson DE, Hughes J, Barchas JD. Enkephalin-like material elevated in ventricular cerebrospinal fluid of pain patients after analgetic focal stimulation. Science 1978;201:463-5. 8. Al Chaer ED, Lawand NB, Westlund KN, Willis WD. Pelvic visceral input into the nucleus gracilis is largely mediated by the postsynaptic dorsal column pathway. J Neurophysiol 1996;76:2675-90. 9. Clarke RH. Part I. Investigation of the central nervous system, methods and instruments. Baltimore: Johns Hopkins; 1920. 10. Levy R. A short history of stereotactic neurosurgery. Park Ridge, IL: American Association of Neurological Surgeons; 1992. 11. Schaltenbrand G. Personal observations on the development of stereotaxy. Confin Neurol 1975;37:410-6. 12. Olivier A, Bertrand G, Picard C. Discovery of the first human stereotactic instrument. Appl Neurophysiol 1983;46:84-91. 13. Dittmar C. Ueber die Lage des sogenannten Gefaesszentrums in der Medulla oblongata. Ber saechs Ges Wiss Leipzig (Math Phys) 1873;25:449-69. 14. Zernov DN. Encephalometer. Device for estimation of parts of brain in human (Russian). Proc Soc Physicomed, Moscow Univ 1889;2:70-80. 15. Altukhov NV. Encephalometric investigations of the brain relative to the sex, age and skull indexes. Moscow; 1891.

25

26

1


16. Gildenberg PL. ‘‘Stereotaxic’’ versus ‘‘stereotactic’’. Neurosurgery 1993;32:965-6. 17. Gildenberg PL. Ernest A. Spiegel 1895–1985. Appl Neurophysiol 1985;48:I-VI. 18. Gildenberg PL. Spiegel and Wycis – the early years. Stereotact Funct Neurosurg 2001;77:11-16. 19. Spiegel EA, Wycis HT. Sterencephalotomy, Part II. Philadelphia, PA: Grune & Stratton; 1962. 20. Freeman W, Watts JW. Psychosurgery: intelligence, emotional and social behavior following prefrontal lobotomy for mental disorders. Springfield, IL: Charles C Thomas; 1942. 21. Spiegel EA, Wycis HT. Stereoencephalotomy, Part I. New York: Grune & Stratton; 1952. 22. Meyers R. The modification of alternating tremors, rigidity and festination by surgery of the basal ganglia. Res Publ Ass Res Nerv Ment Dis 1942;21:602-65. 23. Meyers R. Historical background and personal experiences in the surgical relief of hyperkinesia and hypertonus. In: Fields W editor. Pathogenesis and treatment of parkinsonism. Springfield, IL: Charles C Thomas; 1958. p. 229-70. 24. Dogliotti M. First surgical sections, in man, of the lemniscal lateralis (pain-temperature path) at the brain stem, for the treatment of diffuse rebellious pain. Anesth Analg 1938;17:143-5. 25. Schwartz HG, O’Leary JL. Section of the spinothalamic tract in the medulla with observations on the pathways for pain. Surgery 1941;9:183-93. 26. Walker AE. Relief of pain by mesencephalic tractotomy. Arch Neurol Psychiat 1942;48:865-83. 27. Keele KD. Anatomies of pain. Baltimore, MD: Williams & Wilkins; 1949. 28. Livingston WK. Pain mechanisms. New York: Macmillan; 1943. 29. Talairach J, David M, Tournoux P, Corredor H, Kvasina T. Atlas d’anatomie stereotaxique. Paris: Masson; 1957. 30. Spiegel EA, Wycis HT. Pallido-thalamotomy in chorea. Arch Neurol Psychiatry (Chicago) 1950;64:495-6. 31. Gildenberg PL. The birth of stereotactic surgery: a personal retrospective. Neurosurgery 2004;54:199-208. 32. Spiegel EA. Guided brain operations. Basel: Karger; 1982. 33. Narabayashi H, Okuma T. Procaine oil blocking of the globus pallidus for the treatment of rigidity and tremor of parkinsonism. Proc Jpn Acad 1953;29:310-8. 34. Narabayashi H, Shimazu H, Fujita Y, Shikiba S, Nagao T, Nagahata M. Procaine-oil-wax pallidotomy for double athetosis and spastic states in infantile cerebral palsy. Neurology 1960;10:61-9. 35. Cooper IS. Chemopallidectomy and chemothalamectomy for parkinsonism and dystonia. Proc R Soc Med 1959;52:47-60. 36. Gildenberg PL. Studies in stereoencephalotomy. X. Variability of subcortical lesions produced by a heating electrode and Cooper’s balloon-cannula. Confin Neurol 1960;20:53-65.

37. Obrador S, Dierssen G. Cirurgia de la region palidal. Revta Clin Esp Ano XVII 1956;61:229-37. 38. Cosman ER, Nashold BS, Jr, Bedenbaugh P. Stereotactic radiofrequency lesion making. Appl Neurophysiol 1983;46:160-6. 39. Cooper IS, St Lee AJ. Cryostatic congelation. J Nerv Ment Dis 1961;133:259-63. 40. Gildenberg PL. Functional neurosurgery. In: Schmidek HH, Sweet WH, editors. Operative neurosurgical techniques, vol. II. New York: Grune & Stratton; 1982. p. 993-1043. 41. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta chir scand 1949;99:229-33. 42. Hećaen H, Talairach J, David M, Dell MD. Coagulations limiteés du thalamus dans les algies du syndrome thalamique. Rev Neurol (Paris) 1949;81:917-31. ¨ ber ein neues Zielgeraet zur 43. Riechert T, Wolff M. U intrakraniellen elektrischen Abteilung und Ausschaltung. Arch Psychiatr Z Neurol 1951;186:225-30. 44. Bailey P, Stein SN. A stereotaxic apparatus for use on the human brain. Atlantic City: AMA Scientific Exhibit; 1951. 45. Narabayashi H. Stereotaxic instrument for operation on the human basal ganglia. Psychiatr Neurol Jpn 1952;54:669-71. 46. Uchimura Y, Narabayashi H. Stereoencephalotom. Psychiatr Neurol Jpn 1951;52:265. 47. Ohye C, Fodstad H. Forty years with Professor Narabayashi. Neurosurgery 2004;55:222-6. 48. Talairach J, Hećaen M, David M, Monnier M, Ajuriaguerra J. Recherches sur la coagulation therapeutique des structures sous-corticales chez l’homme. Rev Neurol 1949;81:4-24. 49. Guiot G, Brion S, Fardeau M, Bettaier A, Molina P. Dyskinesie volitionelle d’attitude suppremee par la coagulation thalamo-capsulaire. Rev Neurol 1960; 102:220-9. 50. Riechert T, Wolff M. Die Entwicklung u. klinische Bedeutung der gezielten Hirnoperationen. Med Klin 1951; 46:609-11. 51. Gillingham FJ. Small localized lesions of the internal capsule in the treatment of the dyskinesias. Conf Neurol 1962;22:385-92. 52. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotactic thermolesions in the pallidal region. A clinical evaluation of 81 cases. Acta Psychiatr Neurol Scand 1960;35:358-77. 53. Rossi GF, Colicchio G, Gentilomo A, Scerrati M. Discussion on the causes of failure of surgical treatment of partial epilepsies. Appl Neurophysiol 1978;41:29-37. 54. Bertrand G, Jasper H. Microelectrode recording of unit activity in the human thalamus. Confin Neurol 1965;26:205-8. 55. Velasco Suarez MM, Escobedo FR. Stereotaxic intracerebral instillation of dopa. Confin Neurol 1970; 32:149-57.


56. Bechtereva NP, Bondartchuk AN, Smirnov VM, Meliutcheva LA, Shandurina AN. Method of electrostimulation of the deep brain structures in treatment of some chronic diseases. Confin Neurol 1975;37:136-40. 57. Kandel EI. Experience with the cryosurgical method in production of lesions of the extrapyramidal system. Confin Neurol 1965;26:306-9. 58. Bullard DE, Nashold BSJ. Evolution of principles of stereotactic neurosurgery. Neurosurg Clin N Am 1995;6:27-41. 59. Gildenberg PL. Whatever happened to stereotactic surgery? Neurosurgery 1987;20:983-7. 60. Spiegel EA. Methodological problems in stereoencephalotomy. Confin Neurol 1965;26:125-32. 61. Spiegel EA. History of human stereotaxy (stereoencephalotomy). In: Schaltenbrand G, Walker AE, editors. Stereotaxy of the human brain. Anatomical, physiological and clinical applications. Stuttgart/New York: Georg Thieme Verlag; 1982. p. 3-10. 62. Gildenberg PL. Survey of stereotactic and functional neurosurgery in the United States and Canada. Appl Neurophysiol 1975;38:31-7. 63. Gildenberg PL, Franklin P. Survey of CT-guided stereotactic surgery. Appl Neurophysiol 1985;48:477-80. 64. Bucy PC, Buchanan DN. Athetosis. Brain 1932;55:479-92. 65. Bucy PC. Cortical extirpation in the treatment of involuntary movements. In: Putnam TJ, editor. The diseases of the basal ganglia. New York: Hafner Publishing Co; 1966. p. 551-95. 66. Bucy PC, Case TJ. Tremor. Physiologic mechanism and abolition by surgical means. Arch Neurol Psychiat 1939;41:721-46. 67. Putnam TJ. Treatment of athetosis and dystonia by section of the extrapyramidal motor tracts. Arch Neurol Psychiatry 1933;29:504-21. 68. Walker AE. Cerebral pedunculotomy for the relief of involuntary movements. I. Hemiballismus. Acta Psychiatr Neurol Scand 1949;24:712-29. 69. Walker AE. Cerebral pedunculotomy for the relief of involuntary movements. Parkinsonian tremor. J Nerv Ment Dis 1952;116:766-75. 70. Guiot G, Pecker J. Tractotomie mesencephalique anterieure pour tremblement parkinsonien. Rev Neurol Suppl 1949;1:387. 71. Dandy WE. Changes in our conceptions of localization of certain functions in the brain. Am J Physiol 1930;93:643-7. 72. Gildenberg PL. Evolution of basal ganglia surgery for movement disorders. Stereotact Funct Neurosurg 2006;84:131-5. 73. Meyers R. Dandy’s striatal theory of the ‘‘center of consciousness’’; surgical evidence and logical analysis indicating its improbability. Arch Neurol Psychiat 1951;65:659-71. 74. Spiegel EA, Wycis HT, Baird HW, III. Long range effects of electropallido-ansotomy in extrapyramidal and convulsive disorders. Neurology 1958;8:734-40.

1

75. Mundinger F, Riechert T. Die stereotaktischen Hirnoperationen zur Behandlung extrapyramidaler Bewegungssto¨rungen (Parkinsonismus und Hyperkinesen) und ihre Resultate. Postoperative und Langzeitergebnisse der stereotaktischen Hirnoperationen bei extrapyramidalmotorischen Bewedunasstro¨rungen. Teil B. Fortschr Neurol Psychiatr 1963;31:69-120. 76. Hassler R, Riechert T. Indikationen und Lokalisationsmethode der gezielten Hirnoperationen. Nervenarzt 1954;25:441-7. 77. Spiegel EA, Wycis HT. Ansotomy in paralysis agitans; demonstration of results by motion pictures and electromyograms. Trans Am Neurol Assoc 1953;3:178-80. 78. Krayenbuhl H, Yasargil MG. Relief of intention tremor due to multiple sclerosis by stereotaxic thalamotomy. Conf Neurol 1962;22:368-74. 79. Obrador S, Dierssen G. Observations on the treatment of intentional and postural tremor by subcortical stereotaxic lesions. Confin Neurol 1965;26:250-3. 80. Tsubokawa T, Moriyasu N. Lateral pallidotomy for relief of ballistic movement – its basic evidences and clinical application. Confin Neurol 1975;37:10-15. 81. Andy OJ. Diencephalic coagulation in the treatment of hemiballism. Conf Neurol 1962;22:346-50. 82. Meyers R, Fry WJ, Fry FJ, Dreyer LL, Schultz DF, Noyes RF. Early experiences with ultrasonic irradiation of the pallidofugal and nigral complexes in hyperkinetic and hypertonic disorders. J Neurosurg 1959;16:32-54. 83. Obrador S. A simplified neurosurgical technique for approaching and damaging the region of the globus pallidus in Parkinson’s disease. J Neurol Neurosurg Psychiatr 1957;20:47-9. 84. Van Buren JM, Li CL, Shapiro DY, Henderson WG, Sadowsky DA. A qualitative and quantitative evaluation of parkinsonians three to six years following thalamotomy. Confin Neurol 1973;35:202-35. 85. Yoshida M, Okada K, Watanabe M, Kuramoto S. Analysis of tremulous movements after thalamotomy correlated to intrathalamic therapeutic lesions. Appl Neurophysiol 1976;39:311-4. 86. Krayenbuhl H, Siegfried J. Dentatotomies or thalamotomies in the treatment of hyperkinesia. Confin Neurol 1972;34:29-33. 87. Riechert T. The stereotactic technique and its application in extrapyramidal hyperkinesia. Confin Neurol 1972;34:325-30. 88. Hassler R. Architectonic organization of the thalamic nuclei. In: Schaltenbrand G, Walker AE, editors. Stereotaxy of the human brain. Stuttgart/New York: Georg Thieme Verlag; 1982. p. 140 -80. 89. Cooper IS. Ligation of the anterior choroidal artery for involuntary movements of parkinsonism. Psychiat Quart 1953;27:317-9. 90. Cooper IS. Chemopallidectomy. Science 1955;121:217. 91. Cooper IS. Clinical results and follow-up studies in a personal series of 300 operations for parkinsonism. J Am Geriat Soc 1956;4:1171-81.

27

28

1


92. Cooper IS, Bravo GJ. Implications of a five year study of 700 basal ganglia operations. Neurology 1958;8:701-17. 93. Cooper IS. Results of 1000 consecutive basal ganglia operations for parkinsonism. Ann Intern Med 1960;52:483-99. 94. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. 95. Levy A. Die Pallidotomie beim Parkinson syndrom. Eine vergleichende anatomoradiologische Studie. Arch Psychiatr Nervenkr 1959;199:487-507. 96. Hassler R, Mundinger F, Riechert T. The future of therapy in parkinsonism. In: Hassler R, Mundinger F, Riechert T, editors. Stereotaxis in Parkinson syndrome. New York: Springer; 1979. p. 236-48. 97. Kelly PJ, Ahlskog JE, Goerss SJ, Daube JR, Duffy JR, Kall BA. Computer-assisted stereotactic ventralis lateralis thalamotomy with microelectrode recording control in patients with Parkinson’s disease. Mayo Clin Proc 1987;62:655-64. 98. Riechert T. Long term follow-up of results of stereotaxic treatment of extrapyramidal disorders. Conf Neurol 1962;22:356-63. 99. Tasker RR, Siqueira J, Hawrylshyn P, Organ LW. What happened to Vim thalamotomy for Parkinson’s disease? Appl Neurophysiol 1983;46:68-83. 100. Levy A. Stereotaxic brain operations in Parkinson’s syndrome and related motor disturbances. Comparison of lesions in the pallidum and thalamus with those in the internal capsule. Confin Neurol 1967;291-70. 101. Spiegel EA. Significance of Forel’s field for neurosurgery and neurophysiology. Acta Neurochir (Wien) 1965;13: 292-304. 102. Spiegel EA, Wycis HT, Szekely EG, Baird HW, III, Adams J, Flanagan M. Campotomy. Trans Am Neurol Assoc 1962;87:240-2. 103. Spiegel EA, Wycis HT, Szekely EG, Soloff L, Adams J, Gildenberg P, Zanes C. Stimulation of Forel’s field during stereotaxic operations in the human brain. Electroencephalogr Clin Neurophysiol 1964;16:537-48. 104. Albe-Fessard D, Arfel G, Guiot G, Hardy J, Hertzog E, Aleonard P. Identification et de´limitation prećise de certaines structures souscorticales de l’homme par l’electrophysiologie. CR Acad Sci (Paris) 1961;243:2412-4. 105. Narabayashi H, Gildenberg PL, Franklin PO (eds.). Proceedings of the meeting on the use of microphysiological recordings during stereotactic neurosurgery. Stereotac Funct Neurosurg 1989;52:77-261. 106. Narabayashi H, Ohye C. Importance of microstereoencephalotomy for tremor alleviation. Appl Neurophysiol 1980;43:222-7. 107. Ohye C, Narabayashi H. Physiological study of presumed ventralis intermedius neurons in the human thalamus. J Neurosurg 1979;50:290-7. 108. Kelly PJ. Microelectrode recording for the somatotopic placement of stereotactic thalamic lesions in the

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

treatment of parkinsonian and cerebellar intention tremor. Appl Neurophysiol 1980;43:262-6. Tasker RR, Lenz F, Yamashiro K, Gorecki J, Hirayama T, Dostrovsky JO. Microelectrode techniques in localization of stereotactic targets. Neurol Res 1987;9:105-12. Falkenberg JH, McNames J, Favre J, Burchiel KJ. Automatic analysis and visualization of microelectrode recording trajectories to the subthalamic nucleus: preliminary results. Stereotact Funct Neurosurg 2006;84:35-44. Starr PA, Christine CW, Theodosopoulos PV, Lindsey N, Byrd D, Mosley A, Marks WJ, Jr. Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imagingverified lead locations. J Neurosurg 2002;97:370-87. Ondo WG, Krauss JK. Surgical therapies for dystonia. In: Brin MF, Comella CL, Jankovic J, editors. Dystonia: etiology, clinical features and treatment. Philadelphia, PA: Lippincott Williams & Wilkins; 2004. p. 125-47. Loher TJ, Pohle T, Krauss JK. Functional stereotactic surgery for treatment of cervical dystonia: review of the experience from the lesional era. Stereotact Funct Neurosurg 2004;82:1-13. Hassler R, Dieckmann G. Stereotactic treatment of different kinds of spasmodic torticollis. Confin Neurol 1970;32:135-43. Alker G, Kelly PJ. An overview of CT based stereotactic systems for the localization of intracranial lesions. Comput Radiol 1984;8:193-6. Sano K, Sekino H, Tsukamoto Y, Yoshimasu N, Ishijima B. Stimulation and destruction of the region of the interstitial nucleus in cases of torticollis and see-saw nystagmus. Confin Neurol 1972;34:331-8. Crossman AR. A hypothesis on the pathophysiological mechanisms that underlie levodopa- or dopamine agonist-induced dyskinesia in Parkinson’s disease: implications for future strategies in treatment. Mov Disord 1990;5:100-8. Narabayashi H, Yokochi F, Nakajima Y. Levodopa-induced dyskinesia and thalamotomy. J Neurol Neurosurg Psychiatr 1984;47:831-9. Backlund EO, Granberg PO, Hamberger B, Knutsson E, Martensson A, Sedvall G, Seiger A˚. Transplantation of adrenal medullary tissue to striatum in parkinsonism. First clinical trials. J Neurosurg 1985;62:169-73. Drucker-Colin R, Madrazo I, Diaz V. Open microsurgical autograft of adrenal medulla to caudate nucleus of patients with Parkinson’s disease. Schmitt Neurological Sciences Symposium, Rochester, NY, June 30–July 3, 1987. Madrazo I, Drucker-Colin R, Diaz V. Open microsurgical autograft of adrenal medulla to the right caudate nucleus in two patients with intractable Parkinson’s disease. N Engl J Med 1987;316:831-4. Bakay RA, Herring CJ. Central nervous system grafting in the treatment of parkinsonism. Stereotact Funct Neurosurg 1989;53:1-20.


123. Goetz CG, De Long MR, Penn RD, Bakay RA. Neurosurgical horizons in Parkinson’s disease. Neurology 1993;43:1-7. 124. Hitchcock ER, Clough CG, Hughes RC, Kenny BG. Transplantation in Parkinson’s disease: stereotactic implantation of adrenal medulla and foetal mesencephalon. Acta Neurochir Suppl (Wien) 1989;46:48-50. 125. Stoddard SL, Ahlskog JE, Kelly PJ, Tyce GM, van Heerden JA, Zinsmeister AR, Carmichael SW. Decreased adrenal medullary catecholamines in adrenal transplanted parkinsonian patients compared to nephrectomy patients. Exp Neurol 1989;104:218-22. 126. Jankovic J, Grossman R, Goodman C, Pirozzolo F, Schneider L, Zhu Z, Scardino P, Garber AJ, Jhingran SG, Martin S. Clinical, biochemical, and neuropathologic findings following transplantation of adrenal medulla to the caudate nucleus for treatment of Parkinson’s disease. Neurology 1989;39:1227-34. 127. Gildenberg PL, Pettigrew LC, Merrell R, Butler I, Conklin R, Katz J, DeFrance J. Transplantation of adrenal medullary tissue to caudate nucleus using stereotactic techniques. Stereotact Funct Neurosurg 1990;54:268-71. 128. Mark VH, Gildenberg PL, Franklin PO (eds.). Proceedings of the colloquium on the use of embryonic cell transplantation for correction of CNS disorders. Appl Neurophysiol 1984;47:1-76. 129. Freed WJ, Cannon-Spoor HE, Krauthamer E. Intrastriatal adrenal medulla grafts in rats. Long-term survival and behavioral effects J Neurosurg 1986;65(5):664-70. 130. Bakay RA, Fiandaca MS, Barrow DL, Schiff A, Collins DC. Preliminary report on the use of fetal tissue transplantation to correct MPTP-induced Parkinson-like syndrome in primates. Appl Neurophysiol 1985;48:358-61. 131. Redmond DE, Jr, Roth RH, Spencer DD, Naftolin F, Leranth C, Robbins RJ, Marek KL, Elsworth JD, Taylor JR, Sass KJ. Neural transplantation for neurodegenerative diseases: past, present, and future. Ann NY Acad Sci 1993;695:258-66. 132. Lindvall O, Rehncrona S, Brundin P, Gustavii B, Astedt B, Widner H, Lindholm T, Bjorklund A, Tan M. Human fetal dopamine neurons grafted into the striatum in two patients with severe Parkinson’s disease. A detailed account of methodology and a 6-month followup. Arch Neurol 1989;46:615-31. 133. Bakay RA. Neurotransplantation: a clinical update. Acta Neurochir Suppl (Wien) 1993;58:8-16. 134. Breggin PR. The return of lobotomy and psychosurgery. Congressional Record 1972;118:24-6. 135. Spiegel EA, Wycis HT. Mesencephalotomy in intractable facial pain. Trans Am Neurol Assoc 1952;56:186-9. 136. Gybels JM, Sweet WH. Neurosurgical treatment of persistent pain. Pain and headache, vol. 11. Basel: Karger; 1989. 137. Gildenberg PL. History of pain management. In: Greenblatt SH, Dagi TF, editors. A history of neurosurgery. Park Ridge, IL: American Association of Neurological Surgeons; 1997. p. 465-88.

1

138. Trousseau A. De la ne´vralgie épileptiforme. Arch Gen Med 1853;1:33-44. 139. Rose W. Removal of the gasserian ganglion for severe neuralgia. Trans Med Soc Lond 1890;14:35-40. 140. Horsley V, Taylor J, Colman WS. Remarks on the various surgical procedures devised for the relief or cure of trigeminal neuralgia (tic douloureaux). Brit Med J 1891;2:1139-43. 141. Cushing H. A method for total extirpation of the Gasserian ganglion for trigeminal neuralgia. JAMA 1900;34:1035-41. 142. Cushing H. The major trigeminal neuralgias and their surgical treatment based on experiences with 332 Gasserian operations. I. Varieties of facial neuralgia. Am J Med Sci 1920;160:157-84. 143. Ramonede L. Exerese de trijumeau. Presse Med 1903;11:789-91. 144. Dandy WE. Section of the sensory root of the trigeminal nerve at the pons. Bull Johns Hopkins Hosp 1925;36:105-6. 145. Kirschner M. Zur Electrochirurgie. Arch Klin Chir 1931;161:761-8. 146. Kirschner M. Die Punktionstechnik u. Elektrokoagulation des Ganglion Gasseri. Arch Klin Chir 1933;176:581-620. 147. Murtagh F, Jr, Wycis HT, Spiegel EA. Relief of thalamic pain by mesencephalotomy. AMA Arch Neurol Psychiatry 1951;65:255-7. 148. Frazier CH. Section of the anterolateral columns of the spinal cord for the relief of pain. Arch Neurol Psychiatry 1920;4:137-47. 149. French LA. Cordotomy in the high cervical region for intractable pain. J Lancet 1953;73:283-7. 150. Spiller WG, Martin E. The treatment of persistent pain of organic origin in the lower part of the body by division of the anterolateral column of the spinal cord. J Am Med Assoc 1912;58:1489-90. 151. Gildenberg PL. Mesencephalotomy. In: Burchiel KJ, editor. Surgical management of pain. New York: Thieme; 2002. p. 786-94. 152. Iskandar BJ, Nashold BSJ. History of functional neurosurgery. Neurosurg Clin N Am 1995;6:1-25. 153. White JC, Sweet WH. Pain, its mechanism and neurosurgical control. Springfield, IL: Charles C Thomas; 1955. 154. Spiegel EA, Wycis HT, Szekely EG, Gildenberg PL. Medial and basal thalamotomy in so-called intractable pain. In: Knighton RS, Dumke PR, editors. Pain. Boston, MA: Little Brown; 1966. p. 503-17. 155. Hećaen H, Talairach T, David M, Dell MB. Me´moires originaux. Coagulations limiteés du thalamus dans les algies du syndrome thalamique. Rev Neurol 1949;81:917-31. 156. Spiegel EA, Wycis HT. Mesencephalotomy in treatment of intractable facial pain. AMA Arch Neurol Psychiatry 1953;69:1-13.

29

30

1


157. Spiegel EA, Wycis HT, Szekely EG, Gildenberg P, Zanes C. Combined dorsomedial, intralaminar and basal thalamotomy for relief of so-called intractable pain. J Int Coll Surg 1964;42:160-8. 158. Nashold BSJ, Wilson WP, Slaughter DG. Stereotaxic midbrain lesions for central dysesthesia and phantom pain. Preliminary report. J Neurosurg 1969; 30:116-26. 159. Nashold BS, Jr., Wilson WP, Slaughter DG. Sensations evoked by stimulation in the midbrain of man. J Neurosurg 1969;30:14-24. 160. Nashold BS, Jr. Extensive cephalic and oral pain relieved by midbrain tractotomy. Conf Neurol 1972;34:382-8. 161. Foltz EL, White LE, Jr. Pain ‘‘relief ’’ by frontal cingulumotomy. J Neurosurg 1962;19:89-100. 162. Mullan S, Moseley R, Harper PV. The creation of deep cerebral lesions by small beta ray sources implanted under guidance of fluoroscopic image intensifiers (as used in the treatment of Parkinson’s disease). Am J Roentgen 1959;82:613. 163. Mullan S, Harper PV, Hekmatpanah J, Torres H, Dobben G. Percutaneous interruption of spinal pain tracts by means of a strontium-90 needle. J Neurosurg 1963;20:931-9. 164. Rosomoff HL, Carrol F, Brown J, Sheptak P. Percutaneous radiofrequency cervical cordotomy: technique. J Neurosurg 1965;23:639-44. 165. Tenicela R, Rosomoff HL, Feist J, Safar P. Pulmonary function following percutaneous cervical cordotomy. Anesthesiology 1968;29:7-16. 166. Lin PM, Gildenberg PL, Polakoff PP. An anterior approach to percutaneous lower cervical cordotomy. J Neurosurg 1966;25:553-60. 167. Sjo¨qvist O. Eine neue Operationsmethode bei Trigeminusneuralgie: durchschneidung des Tractus spinalis trigemini. Zentralbl f Neurochir 1937;2:274-81. 168. White JC, Sweet WH. Pain and the neurosurgeon. A forty year experience. Springfield, IL: Charles C Thomas; 1969. 169. Hitchcock E. Stereotactic trigeminal tractotomy. Farmatsiia 1970;19:131-5. 170. Hitchcock E. Stereotactic cervical myelotomy. J Neurol Neurosurg Psychiatry 1970;33:224-30. 171. Crue BL, Todd EM, Carregal EJA. Posterior approach for high cervical percutaneous radiofrequency cordotomy. Confinia Neurol 1968;30:41-52. 172. Gildenberg PL. Evolution of spinal cord surgery for pain. Clin Neurosurg 2006;53:11-17. 173. Hitchcock E. Stereotactic myelotomy. Proc R Soc Med 1974;67:771-2. 174. Schvarcz JR. Stereotactic extralemniscal myelotomy. J Neurol Neurosurg Psychiatry 1976;39:53-7. 175. Gildenberg PL, Hirshberg RM. Limited myelotomy for the treatment of intractable cancer pain. J Neurol Neurosurg Psychiatry 1984;47:94-6.

176. Stillings D. The first use of electricity for pain treatment. Medtronic Archive on Electro-Stimulation 1971. 177. Isaacson W. Benjamin Franklin. An American life. New York: Simon & Schuster; 2003. 178. Pruel MC. A history of neuroscience from Galen to Gall. In: Greenblatt SH, Dagi TF, Epstein MH, editors. A history of neurosurgery. Park Ridge, IL: American Association of Neurological Surgeons; 1997. p. 99-130. 179. Shelly M. Frankenstein. Oxford: Oxford Press; 1818. ¨ ber die elektrische Erregbarkeit 180. Fritsch G, Hitzig E. U des Grosshirns. Arch Anat Physiol wiss Med 1870;37:300-32. 181. Morgan JP. The first reported case of electrical stimulation of the human brain. J Hist Med Allied Sci 1982;37:51-64. 182. Vilensky JA, Gilman S. Horsley was the first to use electrical stimulation of the human cerebral cortex intraoperatively. Surg Neurol 2002;58:425-6. 183. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger JA. Physiological observations in stereotaxic operations in extrapyramidal motor disturbances. Brain 1960;83:337-50. 184. Tasker RR, Organ LW, Hawrylshyn PA. The thalamus and midbrain of man. A physiological atlas using electrical stimulation. Springfield, IL: Charles C Thomas; 1982. 185. Woolsey CN. Cortical localization as defined by evoked potential and electrical stimulation studies. In: Schaltenbrand G, Woolsey CN, editors. Cerebral localization and organization. Madison, WI: University of Wisconsin Press; 1964. p. 17-26. 186. Pool JL. Psychosurgery in older people. J Am Geriatr Soc 1954;2:456-65. 187. Heath RG. Exploring the mind-brain relationship. Baton Rouge, LA: Moran Printing, Inc.; 1996. 188. Olds J, Milner P. Positive reinforcement produced by electrical stimulation of the septal area and other regions of the rat brain. J Comp Physiol Psychol 1954;47:419-27. 189. Heath RG. Psychiatry. Annu Rev Med 1954;5:223-36. 190. Gol A. Relief of pain by electrical stimulation of the septal area. J Neurol Sci 1967;5:115-20. 191. Bechtereva NP, Bondarchuk AN, Smirnov VM. Therapeutic electrostimulations of the deep brain structures. Vopr Neirokhir 1972;1:7-12. 192. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971-9. 193. Laitinen L. Placement of electrodes in transcutaneous stimulation for chronic pain. Neurochirurgie 1976; 22:517-26. 194. Wall PD, Sweet WH. Temporary abolition of pain in man. Science 1967;155:108-9. 195. Sweet WH, Wepsic JG. Treatment of chronic pain by stimulation of fibers of primary afferent neuron. Trans Am Neurol Assoc 1968;93:103-7.


196. Shealy CN, Mortimer JT, Reswick JB. Electrical inhibition of pain by stimulation of the dorsal columns. Preliminary clinical report. Anesth Analg (Cleve) 1967;46:489-91. 197. Gildenberg PL. Treatment of spasmodic torticollis with dorsal column stimulation. Acta Neurochir (Wien) 1977;24:Suppl65-6. 198. Cook AW. Electrical stimulation in multiple sclerosis. Hosp Pract 1976;11:51-8. 199. Dooley DM. Spinal cord stimulation. AORN J 1976;23:1209-12. 200. Dooley DM, Kasprak M. Modification of blood flow to the extremities by electrical stimulation of the nervous system. South Med J 1976;69:1309-11. 201. Hosobuchi Y, Adams JE, Rutkins B. Chronic thalamic stimulation for the control of facial anesthesia dolorosas. Arch Neurol 1973;29:158-61. 202. Mundinger F. New stereotactic treatment of spasmodic torticollis with a brain stimulation system (author’s transl). Med Klin 1977;72:1982-6. 203. Reynolds DV. Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 1969;164:444-5. 204. Richardson DE, Akil H. Pain reduction by electrical brain stimulation in man. Part 1: Acute administration in periaqueductal and periventricular sites. J Neurosurg 1977;47:178-83. 205. Gildenberg PL. Symposium on the safety and clinical efficacy of implanted neuroaugmentive devices. Appl Neurophysiol 1977;40:69-240. 206. Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation for the treatment of central pain. Acta Neurochir Suppl (Wien) 1991;52:137-9. 207. Migita K, Uozumi T, Arita K, Monden S. Transcranial magnetic coil stimulation of motor cortex in patients with central pain. Neurosurgery 1995;36:1037-9. 208. Gildenberg PL. History of neuroaugmentative procedures for chronic pain. Neurosurg Clin N Am 2003;14:327-37. 209. Gildenberg PL. History of electrical neuromodulation for chronic pain. Pain Med 2006;7Suppl 1:S7-13. 210. Gildenberg PL. The use of pacemakers (electrical stimulation) in functional neurological disorders. In: Rasmussen T, Marino R, editors. Functional neurosurgery. New York: Raven Press; 1979. p. 59-74. 211. Laitinen LV, Bergenheim AT, Hariz MI. Ventroposterolateral pallidotomy can abolish all parkinsonian symptoms. Stereotact Funct Neurosurg 1992;58:14-21. 212. Spiegel EA, Wycis HT. Ansotomy in paralysis agitans. AMA Arch Neurol Psychiatry 1953;69:652-3. 213. Goetz CG, Diederich NJ. There is a renaissance of interest in pallidotomy for Parkinson’s disease. Nat Med 1996;2:510-4. 214. Iacono RP, Shima F, Lonser RR, Kuniyoshi S, Maeda G, Yamada S. The results, indications, and physiology of

215.

216.

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

1

posteroventral pallidotomy for patients with Parkinson’s disease. Neurosurgery 1995;36:1118-27. Favre J, Taha JM, Nguyen TT, Gildenberg PL, Burchiel KJ. Pallidotomy: a survey of current practice in North America. Neurosurgery 1996;39:883-92. Lozano A, Hutchison W, Kiss Z, Tasker R, Davis K, Dostrovsky J. Methods for microelectrode-guided posteroventral pallidotomy. J Neurosurg 1996;84: 194-202. Krauss JK, Grossman RG. Historical review of pallidal surgery for treatment of parkinsonism and other movement disorders. In: Krauss JK, Grossman RG, Jankovic J, editors. Pallidal surgery for the treatment of Parkinson’s disease and movement disorders. Philadelphia, PA: Lippincott-Raven; 1998. p. 1-23. Vitek JL, Bakay RA. The role of pallidotomy in Parkinson’s disease and dystonia. Curr Opin Neurol 1997;10:332-9. Nandi D, Chir M, Liu X, Bain P, Parkin S, Joint C, Winter J, Stein J, Scott R, Gregory R, Aziz T. Electrophysiological confirmation of the zona incerta as a target for surgical treatment of disabling involuntary arm movements in multiple sclerosis: use of local field potentials. J Clin Neurosci 2002;9:64-8. Meier MJ, Story J, French LA, Chou SN. Quantitative assessment of behavioral changes following subthalamotomy in the treatment of Parkinson’s disease. Confin Neurol 1966;27:154-61. Brice J, McLellan L. Suppression of intention tremor by contingent deep-brain stimulation. Lancet 1980; 1:1221-2. Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM, Hommel M, Perret JE, de Rougemont J. Longterm suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991;337:403-6. Siegfried J, Lippitz B. Chronic electrical stimulation of the VL-VPL complex and of the pallidum in the treatment of movement disorders: personal experience since 1982. Stereotact Funct Neurosurg 1994; 62:71-5. Benabid AL, Pollak P, Gross C, Hoffmann D, Benazzouz A, Gao DM, Laurent A, Gentil M, Perret J. Acute and long-term effects of subthalamic nucleus stimulation in Parkinson’s disease. Stereotact Funct Neurosurg 1994;62:76-84. Alvarez L, Macias R, Guridi J, Lopez G, Alvarez E, Maragoto C, Teijeiro J, Torres A, Pavon N, Rodriguez-Oroz MC, Ochoa L, Hetherington H, Juncos J, Delong MR, Obeso JA. Dorsal subthalamotomy for Parkinson’s disease. Mov Disord 2001;16:72-8. Jimenez F, Velasco F, Velasco M, Brito F, Morel C, Marquez I, Perez ML. Subthalamic prelemniscal radiation stimulation for the treatment of Parkinson’s disease: electrophysiological characterization of the area. Arch Med Res 2000;31:270-81.

31

32

1


227. Spiegel EA, Wycis HT, Szekely EG, Adams DJ, Flanagan M, Baird HW, III. Campotomy in various extrapyramidal disorders. J Neurosurg 1963;20:871-84. 228. Gildenberg PL. Evolution of neuromodulation. Stereotact Funct Neurosurg 2005;83:71-9. 229. Krauss JK, Pohle T, Weber S, Ozdoba C, Burgunder JM. Bilateral stimulation of globus pallidus internus for treatment of cervical dystonia. Lancet 1999; 354:837-8. 230. Coubes P, Roubertie A, Vayssiere N, Hemm S, Echenne B. Treatment of DYT1-generalised dystonia by stimulation of the internal globus pallidus. Lancet 2000;355:2220-1. 231. Kumar R, Dagher A, Hutchison WD, Lang AE, Lozano AM. Globus pallidus deep brain stimulation for generalized dystonia: clinical and PET investigation. Neurology 1999;53:871-4. ¨ ber einen Fall von doppelseitiger 232. Hassler R, Riechert T. U Fornicotomie bei sogenannter temporaler Epilepsie. Acta Neurochir (Wien) 1957;5:330-40. 233. Pagni CA. Stereo-electroencephalographic observations on the psychomotor seizure. Confin Neurol 1966; 27:137-43. 234. Umbach W, Riechert T. The stereotactic coagulation of the fornix as treatment of temporal lobe epilepsy. Proc First Int Congress Neurol Sci 1961;3:565-78. 235. Jinnai D, Nishimoto A. Stereotaxic destruction of Forel H for treatment of epilepsy. Neurochirurgia(Stuttgart) 1963;6:164-76. 236. Laitinen L. Thalmotomy in progressive myoclonus epilepsy. Acta Neurol Scand 1967;43:31-47. 237. Umbach W. Long-term results of fornicotomy for temporal epilepsy. Confin Neurol 1966;27:121-3. 238. Munari C, Hoffmann D, Francione S, Kahane P, Tassi L, Lo Russo G, Benabid AL. Stereo-electroencephalography methodology: advantages and limits. Acta Neurol Scand Suppl 1968;152:56-67. 239. Mundinger F, Becker P, Groebner E, Bachschmid G. Late results of stereotactic surgery of epilepsy predominantly temporal lobetype. Acta Neurochir (Wien) 1976;23Suppl:177-82. 240. Nashold BS, Jr, Flanigin H, Jr, Wilson WP, Stewart B. Stereotactic evaluation of bitemporal epilepsy with electrodes and lesions. Confin Neurol 1973; 35:94-100. 241. Vaernet K. Stereotaxic amygdalotomy in temporal lobe epilepsy. Confin Neurol 1972;34:176-80. 242. Olivier A, Gloor P, Andermann F, Quesney LF. The place of stereotactic depth electrode recording in epilepsy. Appl Neurophysiol 1985;48:395-9. 243. Gildenberg PL. Unpublished data. 1990. 244. Fins JJ. From psychosurgery to neuromodulation and palliation: history’s lessons for the ethical conduct and regulation of neuropsychiatric research. Neurosurg Clin N Am 2003;14:303-6. 245. Spiegel EA, Wycis HT. The central mechanism of emotions. Am J Psychiatry 1951;109:426-31.

246. Orchinik C, Koch R, Wycis HT, Freed H, Spiegel EA. The effect of thalamic lesions upon the emotional reactivity (Rorschach and behavior studies). Res Publ Assoc Res Nerv Ment Dis 1949;29:172-207. 247. Sano K. Sedative neurosurgery with special reference to posteromedial hypothalamotomy. Neurol Med Chir (Tokyo) 1962;4:112. 248. Spiegel EA, Wycis HT, Freed H, Orchinik C. The central mechanism of the emotions; (experiences with circumscribed thalamic lesions). Am J Psychiatry 1951;108:426-32. 249. Gildenberg PL (ed.). Functional neurosurgery. Neurosurg Clin N Am 1995;6:1-181. 250. Fodstad H, Strandman E, Karlsson B, West KA. Treatment of chronic obsessive compulsive states with stereotactic anterior capsulotomy or cingulotomy. Acta Neurochir (Wien) 1982;62:1-23. 251. Martuza RL, Chiocca EA, Jenike MA, Giriunas IE, Ballantine HT. Stereotactic radiofrequency thermal cingulotomy for obsessive compulsive disorder. J Neuropsychiatry Clin Neurosci 1990;2:331-6. 252. Andy OJ, Jurko F. Thalamic stimulation effects on reactive depression. Appl Neurophysiol 1987;50:324-9. 253. Diering SL, Bell WO. Functional neurosurgery for psychiatric disorders: a historical perspective. Stereotact Funct Neurosurg 1991;57:175-94. 254. Benabid AL, Wallace B, Mitrofanis J, Xia R, Piallat B, Chabardes S, Berger F. A putative generalized model of the effects and mechanism of action of high frequency electrical stimulation of the central nervous system. Acta Neurol Belg 2005;105:149-57. 255. Kopell BH, Greenberg B, Rezai AR. Deep brain stimulation for psychiatric disorders. J Clin Neurophysiol 2004;21:51-67. 256. Lienhard JH. Inventing modern. Oxford Press; Oxford: 257. Murtagh F, Wycis HT, Robbins R, Spiegel-Adolph M, Spiegel EA. Visualization and treatment of cystic brain tumors by stereoencephalotomy. Acta Radiol 1956;46:407-14. 258. Gildenberg PL, Kaufman HH. Direct calculation of stereotactic coordinates from CT scans. Appl Neurophysiol 1982;45:347-51. 259. Gildenberg PL, Kaufman HH, Murthy KS. Calculation of stereotactic coordinates from the computed tomographic scan. Neurosurgery 1982;10:580-6. 260. Gildenberg PL. Computerized tomography and stereotactic surgery. In: Spiegel EA, editor. Guided brain operations. Basel: Karger; 1982. p. 24-34. 261. Leksell L, Jernberg B. Stereotaxis and tomography: a technical note with six figures. Acta Neurochir 1980;52:1-7. 262. Barcia-Salorio JL, Broseta J, Hernandez G, Roldan P, Bordes V. A new approach for directed CT localization in stereotaxis. Appl Neurophysiol 1982;45:383-6. 263. Colombo F, Angrilli F, Zanardo A, Pinna V, Alexandre A, Benedetti A. A universal method to employ


264.

265. 266.

267. 268.

269.

270.

271.

272.

273.

274.

275. 276. 277.

278.

279.

280.

CT scanner spatial information in stereotactic surgery. Appl Neurophysiol 1982;45:352-65. Colombo F, Angrilli F, Zanardo A, Pinna V, Benedetti A. A new method for utilizing CT data in stereotactic surgery: measurement and transformation technique. Acta Neurochir (Wien) 1981;57: 195-203. Patil AA. Compute tomography-oriented stereotactic system. Neurosurgery 1982;10:370-4. Koslow M, Abele MG. A fully interfaced computerized tomographic-stereotactic surgical system. Appl Neurophysiol 1980;43:174-5. Lunsford LD. A dedicated CT system for the stereotactic operating room. Appl Neurophysiol 1982;45:374-8. Brown RA, Roberts TS, Osborn AG. Stereotaxic frame and computer software for CT-directed neurosurgical localization. Invest Radiol 1980;15:308-12. Kelly PJ, Kall B, Goerss S, Alker GJ, Jr. Precision resection of intra-axial CNS lesions by CT-based stereotactic craniotomy and computer monitored CO2 laser. Acta Neurochir (Wien) 1983;68:1-9. Leksell L, Leksell D, Schwebel J. Stereotaxis and nuclear magnetic resonance. J Neurol Neurosurg Psychiatry 1985;48:14-18. Broggi G, Franzini A, Peluchetti D, Servello D. Treatment of deep brain abscesses by stereotactic implantation of an intracavitary device for evacuation and local application of antibiotics. Acta Neurochir (Wien) 1985;76:94-8. Backlund EO, von Holst H. Controlled subtotal evacuation of intracerebral haematomas by stereotactic technique. Surg Neurol 1978;9:99-101. Tanikawa T, Amano K, Kawamura H, Kawabatake H, Notani M, Iseki H, Shiwaku T, Nagao T, Iwata Y, Taira T. CT-guided stereotactic surgery for evacuation of hypertensive intracerebral hematoma. Appl Neurophysiol 1985;48:431-9. Zamorano L, Chavantes C, Dujovny M, Malik G, Ausman J. Stereotactic endoscopic interventions in cystic and intraventricular brain lesions. Acta Neurochir Suppl (Wien) 1992;54:69-76. Gildenberg PL, Labuz J. Use of a volumetric target for image-guided surgery. Neurosurgery 2006;59:651-9. Bernstein M, Gutin PH. Interstitial irradiation of brain tumors: a review. Neurosurgery 1981;9:741-50. Gildenberg PL, Woo S, Butler B, Grant W, Berner B. Fractionated brachytherapy: catheter insertion and dosimetry. Stereotact Funct Neurosurg 1994;63:246-9. Kelly PJ, Alker GJ, Jr. A method for stereotactic laser microsurgery in the treatment of deep seated CNS neoplasms. Appl Neurophysiol 1980;43:210-5. Kall BA, Kelly PJ, Goerss SJ. Interactive stereotactic surgical system for the removal of intracranial tumors utilizing the CO2 laser and CT-derived database. IEEE Trans Biomed Eng 1985;32:112-6. Kelly PJ. Tumor stereotaxis. Philadelphia, PA: Saunders; 1991.

1

281. Kelly PJ, Kall BA, Goerss S. Transposition of volumetric information derived from computed tomography scanning into stereotactic space. Surg Neurol 1984; 21:465-71. 282. Kosugi Y, Watanabe E, Goto J, Watanabe T, Yoshimoto S, Takakura K, Ikebe J. An articulated neurosurgical navigation system using MRI and CT images. IEEE Trans Biomed Eng 1988;35:147-52. 283. Roberts DW, Strohbehn JW, Hatch JF, Murray W, Kettenberger H. A frameless stereotaxic integration of computerized tomographic imaging and the operating microscope. J Neurosurg 1986;65:545-9. 284. Rousu JS, Kohls PE, Kall B, Kelly PJ. Computer-assisted image-guided surgery using the regulus navigator. Stud Health Technol Inform 1998;50:103-9. 285. Heilbrun MP, McDonald P, Wiker C, Koehler S, Peters W. Stereotactic localization and guidance using a machine vision technique. Stereotact Funct Neurosurg 1992;58:94-8. 286. Day R, Heilbrun MP, Koehler S, McDonald P, Peters W, Siemionow V. Three-point transformation for integration of multiple coordinate systems: applications to tumor, functional, and fractionated radiosurgery stereotactic planning. Stereotact Funct Neurosurg 1994;63:76-9. 287. Heilbrun MP, Koehler S, MacDonald P, Siemionow V, Peters W. Preliminary experience using an optimized three-point transformation algorithm for spatial registration of coordinate systems: a method of noninvasive localization using frame-based stereotactic guidance systems. J Neurosurg 1994;81:676-82. 288. Smith KR, Frank KJ, Bucholz RD. The NeuroStation – a highly accurate, minimally invasive solution to frameless stereotactic neurosurgery. Comput Med Imag Graphics 1994;18:247-56. 289. Gildenberg PL. The Exoscope system for three-dimensional craniotomy. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: Mc-Graw-Hill; 1997. p. 485-90. 290. Gildenberg PL, Ledoux R, Cosman E, Labuz J. The Exoscope – a frame-based video/graphics system for intraoperative guidance of surgical resection. Stereotact Funct Neurosurg 1994;63:23-5. 291. Zamorano L, Dujovny M, Chavantes C, Malik G, Ausman J. Image-guided stereotactic centered craniotomy and laser resection of solid intracranial lesions. Stereotact Funct Neurosurg 1990;54/55:398-403. 292. Gildenberg PL, Labuz J. Stereotactic craniotomy with the exoscope. Stereotact Funct Neurosurg 1997;68:64-71. 293. Pillay PK, Gildenberg PL, Labuz J. Computer workstation applications: stereoplan, virtual reality, and the Gildenberg exoscope. In: Pell MF, Thomas DGT, editors. Handbook of stereotaxy using the CRW apparatus. Baltimore, MD: Williams & Wilkins; 1994. p. 219-33. 294. Gildenberg PL. Confinia neurologica/applied neurophysiology. 50 years. Appl Neurophysiol 1988;51:1-5.

33

11 History of Stereotactic Surgery in China F.-C. Lee . B. Sun . J. Zhang . K. Zhang . F.-G. Meng

Prehistoric and Feudal China China is unique among the world civilizations in that it has a well-established indigenous medical treatment philosophy dating back thousands of years. This was epitomized by the well known Huang Ti Nei Jing (Yellow Emperor’s Inner Classic), a seminal medical text of ancient China where a legendary physician Yu Fu was alleged to possess the skill for surgical exposure of the brain. The Neijing, Shennong Ben Cao Jing (Divine Husbandman’s Classic of the Materia Medica), and a few other ancient texts laid the foundation of the so-called Traditional Chinese Medicine (TCM), an alternative treatment being actively practiced even today in mainland China and among overseas Chinese alongside modern Western medical and surgical therapies. A good depiction appeared in the popular History of the Three Kingdoms where ‘‘cranial surgery’’ was proposed by Hua Tuo (Hua Lun), a famous physician, on one of the three reigning kings Cao Cao who suffered persistent headaches, presumably due to battle trauma–related intracranial hematoma (or perhaps even intracranial tumor) during the turbulent warring period of Eastern Han and Three Kingdoms (222–280 A.D.). Hua Tuo (> Figure 11‐1) had also reportedly performed surgeries with anesthesia, 1,600 years ahead of similar endeavors in Western Civilization, using wine and a herbal concoction of cannabis boil powder. Incidentally, Hua Tuo came from the ancient Qiao City of Pei State in the modern day Anhui Province, where China’s first

#


stereotactic neurosurgery institute and training center was established in 1983 and the first National Stereotactic and Functional Neurosurgical Conference was held on June 8–14, 1987. The oldest prehistoric evidence of human trephination in European Civilization was probably the Ensisheim Stone Age skull unearthed from a burial site in France which was dated to 5100 B.C. [1]. Archaeological evidence also surfaced in China in 1995 with the excavation in Shandong Province of an adult male skull, aged 35–45 years, bearing a 31 25 mm round parietal calvarial defect with smooth border. The perimeter showed evidence of scrapping and bone regeneration. This belonged to an ancient inhabitant of the Dawen Kou Culture (third to fifth millennium B.C.) and was 14C-dated to about 5000 B.C. (> Figure 11‐2). The opening appeared to have been made with tools resembling a trephine. This lends credence to the conclusion that trephinations were probably performed in Neolithic China. Similar finds were discovered at other archaeological sites in the Qinghai, Heilongjiang, and Henan Provinces of China, variously radiocarbon-dated to between 2000 and 4000 B.C. [2]. The motivation for these skull openings remained varied and speculative, just like in other European archaeological discoveries of the Paleolithic and Neolithic period. Whether for curative or ritual purposes to heal and alter behavior, these could have been early men’s attempt to relate structure to functions and behavior, perhaps a primitive version of modern day functional neurosurgery.

126

11

History of stereotactic surgery in china

Birth of Neurosurgery in Modern China An understanding of the development of stereotactic and functional neurosurgery in China will . Figure 11‐1 Hua Tuo (Hua Lun), born 208 B.C. in Qiao (present Anhui Province, China)

be incomplete without due reference to the history of general neurosurgery, the foundation on which the former has evolved. Modern neurosurgery did not develop much until the last century of Chinese history. Like the rest of the Chinese civilization, it was shrouded in mystery behind the ‘‘bamboo curtain’’ as a result of the cultural and diplomatic isolation of this monolithic, inward-looking nation. The country was plagued by the ineffective, feudalistic Imperial Manchurian rule, subsequently overthrown in 1912 in a popular revolution led by Dr Sun Yat-sen, the father and founder of modern China. Due credit should be given to Dr Song-Tao Guan (ST Kuan) and Dr Cha-Li Zhang (Charles Chang) from Beijing and Shenyang, respectively, for their pioneering works in neurosurgery in China [3]. Guan completed his residency program at the Peking Union Medical College Hospital (PUMC) in 1926. PUMC was jointly founded in 1906 by the American Board of Commissioner for Foreign Missions, the Presbyterian Church in the USA, the Methodist Episcopal Church, and the London Missionary Society, among others. From 1926 to1930 Guan received neurosurgery training at the University of

. Figure 11‐2 Archaeological find at Dawen Kou 5000 B.C. (picture courtesy of Cheng-Yuan Wu)


Pennsylvania under Dr Charles H. Frazier, who was among the first group of American neurosurgeons who established neurosurgery as a new discipline from general surgery. In 1938 Yi-Cheng Zhao, another graduate from the same medical college, was sent by the Rockefeller Foundation (established 1913), through its subsidiary, the China Medical Board, for formal neurosurgical training under Dr Wilder Penfield at the Montreal Neurological Institute at McGill University. Upon his return to China in 1940, Zhao practiced neurosurgery at the PUMC Hospital with Guan. In 1952, three years after the Chinese civil war ended with the establishment of the People’s Republic of China, Zhao founded the first Brain Department comprising neurology and neurosurgery in the port city of Tianjin, one of the very few cities with early Western influence in the ethnophobic China of the nineteenth century. It was reluctantly opened by the Qing Dynasty of China as a treaty port accessible to France, Britain, and others in 1860. Formal concessions with foreign settlements were ceded to Western powers and Japan in 1903 after the Imperial Qing Dynasty’s defeat in the war during the Boxer Rebellion of 1899. These three neurosurgeons, along with Tong-He Zhang of Xian City, performed the first neurosurgical procedures in modern China. Zhang performed prefrontal lobotomy for a psychiatric disorder [4]. From 1932 to 1949, only 16 neurosurgical articles were published in the indigenous Chinese Medical Journal (CMJ), which was initially established in 1887 in Shanghai as China Medical Missionary Journal for publications mainly by Western medical missionaries serving in China. The name was changed to China Medical Journal in 1907, thereafter adopted its current name of CMJ in 1932, when it merged with the English language section of National Medical Journal of China. Only about 60 brain tumors were reported in the Chinese literature. With his training under Dr Frazier who pioneered subtemporal retrogasserian neurotomy in

11

1910 as a permanent cure for tic douloureux, Guan reported his experience with Frazier’s operation for trigeminal neuralgia in 1932 in the CMJ [5]. Charles Chang published on similar procedure and alcohol block three years after Guan. He also drew attention to the occurrence of anesthesia dolorosa and abducens nerve palsy and reported on neurofibroma of the Gasserian ganglion [6]. However, further progress in the development of neurosurgery was repeatedly halted when the country was plunged into two devastating civil wars between the ruling Nationalist Army (Kuomintang) and the Communist People’s Liberation Army (PLA) from 1927 to 1936 and 1946 to 1949, with an intervening period of brutal invasion of China by the Japanese Imperial Army that lasted from 1937 to 1945 when World War II finally ended (> Figure 11‐3). Before the formation of the communist People’s Republic of China (PRC) in 1949, international assistance in the development of neurosurgery as in many other fields was predominantly rendered by the United States and European nations mainly through the China Medical Board. After 1949, however, international collaborations were confined to fellow Communist Bloc countries, in particular the USSR. A 6 month neurosurgery training course was conducted in Beijing by Dr A.E. Arutiunov from the Kiev Neurosurgical Institute of the former Soviet Union in October 1954. Participants of this program included Chung-Cheng Wang (Zhong-Cheng Wang), Ya-Du Zhao, and Da-Jie Jiang, while others were sent to the prestigious Moscow Burdenko Neurosurgical Institute. Among them, Tong-Jin Tu set up a neurosurgery service for the People’s Liberation Army in the Fourth Military Medical University in Xian City (ancient capital of Qin dynasty 221–206 B.C.) upon his return to China in 1956. The T J Tu Award for excellence in neurosurgery was established in honor of his invaluable contributions to military neurosurgery. DrYi-Cheng Zhao (> Figure 11‐4) founded the first Brain Department in China, comprising

127

128

11


. Figure 11‐3 Some early Chinese textbooks, journals, and monographs on stereotactic and functional neurosurgery

. Figure 11‐4 Dr Yi-Cheng Zhao (1908–1974), Father of Neurosurgery of China

neurology and neurosurgery services at the Tianjin Municipal General Hospital in 1952 aided by his students Chung-Cheng Wang and Qing-Cheng Xu. The Neurosurgery Department

at the Beijing Tong-Ren Hospital was established two years later. It was later shifted to the XuanWu Hospital, the predecessor of the Beijing Neurosurgical Institute, the largest neurological center in China. A one-year intensive formal postgraduate neurosurgery training program was started in Tianjin in 1953. The graduates from these programs were to form the core group and foundation of modern neurosurgery in China. Dr Zhao established the Beijing Neurosurgical Institute in March 1960, holding the position of director until his untimely demise in1974, whereupon his student Chung-Cheng Wang took over. The Beijing Neurosurgical Institute, which is now housed within the Beijing Tiantan Hospital, is currently the national clinical, research, and training center with over 300 neurosurgical beds divided into craniocerebral injury, cerebrovascular disease, spinal injury, skull base surgery, intracranial tumors, pediatric neurosurgery, stereotactic and functional neurosurgery, and neurointensive care units in addition to ten research departments. Its basic neuroscience research departments cover neuroanatomy, neurotransmitter, neuropathology, neuropharmacology, neuroepidemiology,


cytobiology, neurophysiology, immunology, neurochemistry, and electron microscopy. The faculty is led by prominent neurosurgeons like Chung-Cheng Wang, Ya-Do Zhao, Shi-Qi Luo, and Ji-Zong Zhao. More than 1000 neurosurgeons have benefited from its training programs till date. Further south in Shanghai, yet another city subjected to early Western influence after the Opium War with Britain in 1840–1843, neurosurgery service was established in 1953 in the Shanghai Red Cross Society Hospital, the predecessor of the Shanghai Huashan Hospital, by Yu-Quan Shi and Zhen-Qin Zhu. This prestigious institution has been credited with the first Chinese-made stereotactic frame and the first indigenously designed operating microscope and pioneered hemispherectomy for infantile hemiplegia in 1959. Postgraduate neurosurgery training programs have been offered since 1958. Shi retired from the Chair of the Department of Neurosurgery in 1989. The Shanghai Huashan Hospital was incorporated into the Shanghai Huashan Institute of Neurological Surgery (SHIN) in March 2000, with a total annual operative statistics exceeding 4,000. After the inception of the PRC in 1949, there were periodic sociopolitical upheavals, in particular the notorious Chinese Cultural Revolution from 1966 to 1976, causing widespread social, political, and economic chaos throughout the country. Scholastic aptitude was considered bourgeois and decadent and could constitute the basis for severe persecution. Not surprisingly, there was a striking absence of neurosurgical publications during this ‘‘Dark Age’’ of modern Chinese history. Until the beginning of the long anticipated economic reform in 1978 by leader Deng XiaoPing, the country endured diplomatic and cultural isolation, compounded further by economic embargo from Western nations due to ideological difference, a legacy of the Cold War. Language barrier and paucity in resources meant that innovative Chinese neurosurgeons had to develop

11

skills de novo and design diagnostic and surgical equipment, often under conditions harsh by Western standards. Their works have been documented in monographs and native Chinese journals such as the Chinese Medical Journal, Chinese Journal of Neurosurgery, Chinese Clinical Neurosurgery, and subspecialty journals, e.g., Chinese Journal of Stereotactic and Functional Neurosurgery, and Chinese Journal of Minimally Invasive Neurosurgery. It is only perhaps in the last decade or so that more and more Chinese neurosurgeons armed with better command of foreign languages have made their works known to the rest of the world through international journals such as Journal of Neurosurgery, Neurosurgery, British Journal of Neurosurgery, Stereotactic and Functional Neurosurgery, etc. (> Figure 11‐5). . Figure 11‐5 The first issue of the Chinese Journal of Stereotactic and Functional Neurosurgery (published in Anhui Province, China, 1986)

129

130

11


Stereotactic and Functioning Neurosurgery Historical Developments in China In the 1940s, psychosurgery was reportedly being carried out by Dr T.H. Zhang. Stereotactic neurosurgery has advanced with strides in North America and Europe since 1946 when SpiegelWycis performed the first human stereotactic pallidotomy [7,8] and Prof. Lars Leksell invented of the Leksell stereotactic frame using polar coordinates in 1949 [9]. Nevertheless, the first surgical treatment for Parkinson’s disease in China was a freehand transorbital pallidotomy using Novocain and Iodipamide by Chung-Cheng Wang in 1959 [10]. Chung-Cheng Wang and fellow contemporaries Jian-Pin Xu, Mao-Shan Wang, and Da-Jie Jiang were the forerunners of stereotactic and functional neurosurgery in China.

In 1963, Dr Jian-Ping Xu, currently in the Guangdong Hospital of Traditional Chinese Medicine, completed a two-year training at the Moscow Neurosurgical Institute under the tutelage of Dr Edvard I. Kandel, who was the first Soviet neurosurgeon trained in stereotactic surgery. In the same year, Xu embarked on stereotactic surgery on patients with Parkinson’s disease using his self-designed Cartesian coordinates–based stereotactic frame in Anhui [11]. Sheer necessity and lack of recourse to expensive Western products motivated many Chinese surgeons, notably MaoShan Wang and Da-Jie Jiang of the Shanghai Medical University, in the early 1960s to design stereotactic devices to be used in, e.g., pallidotomy and thalamotomy for extrapyramidal disease [12–14]. Da-Jie Jiang’s frame (> Figure 11‐6 left) was an attempt to incorporate features of the two main groups of stereotactic devices available in the West then. Devices in the first category were

. Figure 11‐6 Stereotactic head frame designed by Da-Jie Jiang, Shanghai, 1964 (left) and with Li Pan 1989 (right) (with permission)


structurally and mathematically complex, bulky, and difficult to operate. Specialized X-rays were needed. However, they offered a high degree of accuracy. These were exemplified by the SpiegelWycis, Leksell, Riechert, and Talairach equipment [7,9,15,16]. The second was represented by the Cooper and Austin devices [17,18]. These were structurally simple and easy to use but were compromised by a lower degree of accuracy even with repeated intraoperative adjustments with X-rays. Complications were also more prevalent. Limitations in target precision, control of lesion size, and higher degree of complications resulted in the waning of enthusiasm for these early stereotactic procedures (> Figure 11‐7). During the first decade of Western stereotactic neurosurgery, surgeons designed and custom-made their own stereotactic apparatus, as commercially produced frames were not available. The Spiegel-Wycis stereotactic frame was in fact built by Davis, an English machinist, in the workshop of the Temple University Medical Center in Philadelphia. There was also a whole array of Chinese-designed stereotactic equipment, the most prominent among these being

11

the XZ-I to XZ-V brain stereotactic apparatus by Jian-Pin Xu of the Anhui Provincial Hospital (1964–1977). This was not unlike the Guiot– Gillingham frame, small and simple in design, being anchored to the calvarium, and not requiring sophisticated supportive equipment [11,19,20]. However, only procedures within the small confines of one cerebral hemisphere were possible and a lower degree of precision was afforded. The FY85II, designed by the Xian Fourth Military Hospital in 1985, was quite similar in principle to the Todd-Wells and Leksell frames. The patient’s head was secured within the base ring. It was based on the arc principle and had a wider range of maneuverability, permitting bilateral hemispheric procedures with a higher degree of accuracy of within 2 mm. Other stereotactic equipment invented included the DZY-A from Nanjing and a frame by Wu SL from the Guangdong Minimally Invasive Neurosurgery Medical Center, Guangzhou, during the same period (> Figure 11‐8). In 1989, Da-Jie Jiang and Li Pan designed a stereotactic instrument incorporating computerized tomography(CT)andmagneticresonanceimaging(MRI) guidance (> Figure 11‐6 right). The HB-set

. Figure 11‐7 Dr Jian-Ping Xu with his stereotactic device in 1986 (left) and operating with his computer-assisted CT-guided equipment (right)

131

132

11


. Figure 11‐8 Stereotactic device used by Jiang (left) and Wu SL in 1983 (right)

stereotactic equipment, the fruit of a collaboration between the Beijing Science and Engineering University and the Beijing Naval General Hospital, was launched in 1995 [21]. This later phase of intense activity in research and development of instruments was made possible by the modern economic and sociopolitical reform of the late Chinese leader Deng Xiao-Ping in 1978. Rapid acquisition of wealth, prosperity, and affluence in China after 1978 resulted in enhanced purchasing power, allowing practically unhindered access to Western technology. Accelerated growth in a very technology-dependent field such as neurosurgery inevitably ensued. Many hospitals have since acquired CTand high-resolution MRI instruments. Positron emission tomography (PET), single photon emission computerized tomography (SPECT) and magnetoencephalography (MEG) machines were also installed in some of the leading centers. Computer-assisted frame-based and frameless neuronavigational surgical procedures were performed. Some hospitals employed homemade systems. In the West, computer-assisted stereotactic neurosurgery was available as early as

1969. Digital CT- and MRI-image-guided stereotactic surgery was introduced in China in 1973 and 1980, respectively. Following Schaltenbrand-Wahren’s publication of the Atlas for Stereotaxy of Human Brain in 1977 [22] and Ronald Tasker’s release of the Physiological Atlas of the Thalamus and Midbrain Using Electrical Stimulation [23], Jia-Qing Yao et al. produced the Stereotactic Anatomy of Brain Gray Matter (in Chinese) for use by Chinese neurosurgeons in 1983. In 1992, the same author published the Applied Stereotactic Anatomy of the Human Brain with specific reference to the Chinese population [24,25].

Training Programs, Professional Organizations, and National Conferences The earliest generation of Chinese neurosurgeons received their stereotactic neurosurgery training abroad. Jiang-Ping Xu was trained in Moscow under the distinguished Russian


pioneer stereotactic surgeon Dr Edvard I. Kandel from 1960 to1963. Cheng-Yuan Wu was a visiting scholar in stereotactic surgery for one year at the University of Utah, USA, with Dr M. Peter Heilbrun. Others like LZ Cheng, XM Fu, and ZP Ling were subsequently trained at Karolinska Hospital, Sweden, and Henri Mondor Hospital, Paris. Many of the leading stereotactic and functional neurosurgeons currently holding senior positions in major Chinese neurosurgical centers have received training in well-known international centers including UCLA, John Hopkins Hospital, Karolinska Hospital, and Hannover Hospital, among many others. After the first Chinese Stereotactic Neurosurgery Institute was established in Anhui Province in 1983 (> Figure 11‐9) by Jian-Ping Xu and Ye-Han Wang, indigenous stereotactic neurosurgery training program became available to native Chinese neurosurgeons with Dr Xu as the program director. A Computer-assisted Stereotactic Training Course was organized the next year. The Annual National Workshop on Stereotactic and Functional Neurosurgery was hosted by the

11

Anhui Provincial Stereotaxic Neurosurgical Institute in 1987. By 1986, the total number of Chinese-made stereotactic devices in use had reached 300, and total number of stereotactic surgeries exceeded 1,000. It is estimated that as many as 80% of the current practicing stereotactic and functional neurosurgeons have benefited from the Anhui workshops. At various intervals, other leading medical centers, e.g., Beijing Neurosurgical Institute, Beijing General Hospital of Armed Police Force, Tianjin General Hospital, Shanghai Huashan Hospital, and hospitals in Guangzhou and Harbin, offer training in stereotactic neurosurgical techniques. The Anhui Provincial Stereotaxic Neurosurgical Institute, in particular, has trained more than 400 neurosurgeons in stereotactic and functional neurosurgery to date. The Neurosurgery Specialty Society of the Chinese Medical Association (CMA) was established in March 1986 with Chung-Cheng Wang as the Chairman. Wang, born 1925, played a very pivotal role in modern neurosurgery in the PRC, having mentored more than one-third of the

. Figure 11‐9 The first Stereotactic Neurosurgery Institute in Anhui Province, China

133

134

11


current generation of Chinese neurosurgeons. Apart from a very illustrious neurosurgical career, being the pioneer in brain stem microneurosurgery and the first to perform cerebral angiography in China during the impoverished postliberation and Korean War era of 1950s, he served in the Medical Corps in the Korean War against the US-led Allied Forces in 1952 and has been elected as representative to the People’s Congress of China (the Chinese equivalent of the Parliament). The Medal of Honor was bestowed upon him during the XII World Congress of Neurosurgery held in September 15–20, 2001 in Sydney, Australia. The Chinese Society of Stereotactic and Functional Neurosurgery was formed in July 1987 under the umbrella of the CMA. The first National Stereotactic and Functional Neurosurgery Conference in June 6–8, 1987 at Hefei, Anhui Province, was organized by the Neurosurgery Specialty Society of the CMA and attracted 155 participants with 85 papers presented (> Figure 11‐10). It has since been held seven times: May 16–20, 1990, in Chendu; May 9–13, 1993, in Dalian; May 12–16, 1997, in Beijing; August 1–4, 2001, in Harbin; June 6–10, 2004, in Yinchuan; June 26–30, 2006, in Huangshan.

In November 16–19, 2007, the First National Congress of Functional Neurosurgery was held in Xiamen. This was organized by the Stereotactic and Functional Neurosurgery Specialty Committee, Neurosurgery Association of the Chinese Medical Doctors Association (CMDA), a parallel organization to CMA. It was attended by more than 400 participants. The first National Congress for Brain Tissue and Neural Cell Transplantation was held in Kunming in January 1990. The Chinese Association of Epilepsy Surgery was established in 1990 and organized its first National Epilepsy Surgery Work-shop in the following year. In December 1998, the Guangdong Minimally Invasive Neurosurgery Medical Center organized the first National Minimally Invasive Neurosurgery Conference in Guangzhou. The biannual Shanghai International Symposium on Functional Neurosurgery is hosted by the Center for Functional Neurosurgery of the Shanghai Medical University. The Shanghai Medical University, founded in 1927 as the National Fourth Zhongshan University Medical College, was incorporated into the Shanghai Huashan Institute of Neurological Surgery (SHIN) in March 2000. Invited, distinguished international speakers to these national conferences and other meetings

. Figure 11‐10 The historic First National Stereotactic and Functional Neurosurgery Conference. Hefei City, Anhui Province


have included Bjorn Meyerson, Bengt Linderoth (Sweden); Andres Lozano (Canada); Philip L. Gildenberg, Weiser, Philip Starr, Antonio.A.F.De Salles (USA); and J.P. Nguyen, Pr Pierre Cesaro (France). In recent years, Chinese neurosurgeons have increasingly made their presence felt in regional and international conferences such as the Congress of the Asian Society for Stereotactic, Functional and Computer-assisted Neurosurgery (ASSFCN), Congress of European Society for Stereotactic and Functional Neurosurgery (ESSFN), World Congress of the World Society for Stereotactic and Functional Neurosurgery (WSSFN), WFN World congress on Parkinson’s Disease and Related Disorders, Congress of Neurological Surgeons (CNS), and the American Association of Neurological Surgeons (AANS) meetings. Frequent scholastic exchange with overseas counterparts has further enhanced international collaboration.

Modern Chapter of Chinese Stereotactic and Functional Neurosurgery Within the half a century since stereotactic and functional neurosurgery began in China, remarkable progress has been made. However, there is relatively less emphasis on research, be it laboratory or clinical, no doubt due to the immense workload and the difficulty of gathering follow-up information from patients spread over such a vast countryside. Leading neurosurgery centers can now offer stereotactic and functional neurosurgical treatments with cutting edge technology comparable to the West, by neurosurgeons with impeccable credentials. The Beijing Tiantan Hospital with its Beijing Neurosurgical Institute is one of the world’s top three neurosurgical institutes and Asia’s largest center for neurosurgical treatment, training, and research. It was designated by the World Health

11

Organization (WHO) in 1982, together with the Shanghai Huashan Hospital, as the WHO Collaborating Centers for Research and Training in Neurosciences in China. The Shanghai Huashan Hospital was incorporated into the Shanghai Huashan Institute of Neurological Surgery (SHIN) with the Gamma Knife Hospital and two other affiliated hospitals in March 2000. The large number of practicing stereotactic and functional neurosurgeons became self-evident when close to 500 registered participants attended the First National Congress of Functional Neurosurgery in November 2007 in the southern city of Xiamen which overlooks the Taiwan Strait. Sophisticated modern neurosurgical treatments however remain largely the exclusive and privileged domain of the elite urban rich, as health insurance has yet to gain a foothold in this part of the world. Huge disparity still exists in wealth, education, and social welfare between the affluent cities of the eastern coastal provinces, and the relatively more deprived rural areas in the interior. The paradox exists that while urban China can boast of world-class hospitals and treatment, the rural poor do not have easy access to sometimes even basic medical facilities. It is commendable that the problem is being aggressively addressed by the Chinese central government through its emphasis on preferential developmental programs for the western regions. Providing adequate healthcare to one-fifth of the world’s population spread over such a vast country is a daunting task for any government indeed. Nevertheless, the huge population with seemingly unstoppable economic progress offers a golden opportunity and a rich substrate for a quantum leap of development of this technologydriven division of medical science. Acquisition of abundant clinical data permits derivation of useful observations and conclusions on treatment modalitiesfrom huge pools of patients. J.N. Zhang published 580 case reports on stereotactic intracranial procedures for Parkinson’s disease in 2000 [26], and S.B.Yuan reported on 1,431 cases

135

136

11


of intracranial lesions treated with the Chinesedesigned rotating gamma system (OUR-XGD) in the Chinese Journal of Stereotactic and Functional Neurosurgery in 2001 [27], while 3,094 patients were treated with the Leksell Gamma Knife by Liang JC in Guangzhou up to 1999 [28]. The extreme degree of subspecialization within many of the large metropolitan or university units, the envy of many foreign neurosurgical centers, permits rapid development of subspecialty neurosurgery in China. Unencumbered by legislative regulatory constraint and working within a much less litigious environment compared to the West, Chinese neurosurgeons are quick to embrace new modalities of treatments, admittedly some still controversial, ahead of their European or American counterparts. The Beijing-based General Hospital of Armed Police Force, for example, established its Department of Neural Stem Cells Transplantation in March 2004 and has since treated close to 200 patients with strokes, traumatic brain injury, spinal cord injury, motor neuron disease, and cerebral palsy [29]. The Xishan Hospital in Beijing has an ongoing program on the implantation of olfactory ensheathing cells (OEC) from the olfactory bulb of aborted fetuses on patients with diagnoses ranging from amyotrophic lateral sclerosis and spinal cord injury to Parkinson’s disease. It has reportedly treated over 600 cases since 2001 [30,31], while in Europe, for example, the laboratory research by Geoffrey Raisman at the Institute of Neurology, Queen Square, London, on adult autotransplanted OEC on rats is currently still awaiting human clinical trials [32]. These nascent therapies have not yielded the same consistent and reproducible results of, for example, deep brain stimulation (DBS) in Parkinson’s disease. Psychiatric neurosurgery has largely fallen into disrepute under public and political scrutiny and clouded by controversy in the West for the past two decades. This was due partly to the risk of litigation, ironically sometimes from patients

rendered well enough by psychosurgery to initiate legal proceedings, and refusal of reimbursement by insurance companies. The 1986, the U.S. Office of Health Technology Assessment report which stated that ‘‘psychosurgery should be considered experimental as it has never been studied in a scientific manner’’ further discouraged its more widespread acceptance. However, we have witnessed its unabated development in China in the same period. For example, psychosurgery for the alleviation of drug dependence gained popularity in China from 2000 to 2004.

Frame-Based Surgery, Neuronavigation, and Robotics This has seen perhaps the most impressive growth since digital imaging, and the advent of stereotactic instrumentation has brought stereotactic surgery into the realms of many Chinese neurosurgeons, who like their Western counterpart have hitherto confined their works to general neurosurgery [33,34]. Whereas previously stereotactic imaging had to rely on angiography and ventriculography, identification of intracranial landmarks can now be made by noninvasive CT or MRI. Diagnostic tissue biopsies, evacuations of hematomas and abscesses, intracavitatory instillations of radioisotopes for tumors [35,36], e.g., combined 32P and Methotrexate chemotherapy for deep-seated gliomas, interstitial 192Ir brachytherapy for glioblastomas [37,38], and transplantations of neural tissues with minimally invasive techniques have been widely carried out in China. CT-guided neuroendoscopies have also been performed. In 1998, Tian ZM from the Beijing Navy General Hospital of People’s Liberation Army reported on 1,300 cases of CT-guided operations under local anesthesia using both the Leksell stereotactic frame and the indigenous HB-III instrument [39]. Intratumoral instillation of radionuclide and chemotherapeutic agent (BCNU) accounted


for 780 and 24 cases, respectively. Lesioning of functional targets (101 cases), brain tissue biopsy (171 cases) and evacuation of intracerebral hematomas (155 cases) constituted the remainder. The Chinese-made NDY stereotactic frame has also been employed by some neurosurgeons. Frameless stereotaxy or neuronavigation permits neurosurgeons to perform stereotacticassisted volumetric excision of intracranial space occupying lesions accurately, greatly minimizing procedure-related risks [40–42]. It is now being used with intraoperative MRI at the Beijing Tiantan Hospital and Shanghai Huashan Hospital. Neuronavigation with functional MRI (fMRI) has been used in surgery near eloquent areas [43]. Neuronavigation-assisted neuroendoscopies have also been carried out with a high degree of success [44]. It is has become a standard inventory in neurosurgery operating suites in major neurosurgical units in Beijing, Shanghai, Guangzhou, Tianjin, Xian, and other cities. Neuronavigational systems available in China include the SurgiScope (Elekta), StealthStation (Medtronics), VectorVision (BrainLab), and the China-made ASA-610V and ASA-630V (Shenzhen Anke High Tech Co.) [45]. Imageguided, robot-assisted stereotactic system (CRAS HB1) developed by the Beijing Aeronautical and Astronautical Institute was used by Tian ZM of the Beijing Naval General Hospital in 1997 on 32 patients comprising 23 cases of intratumoral brachytherapy, 3 cases of tissue biopsy, and 2 cases each of evacuation of abscess and hematoma, achieving an instrument efficacy of 86% [46].

Surgery for Movement Disorders The first pallidotomy for Parkinson’s disease was done freehand in1959 by Zhong-Cheng Wang, who is the president of the Chinese Neurosurgical Society and the Beijing Neurological Institute. Wang designed the transorbital approach using trocar puncture of orbital roof, contralateral to

11

the side of tremor or rigidity, 4 cm from midsagittal plane aiming 10 superiorly and medially. An FG 20–22 needle was then advanced to the frontal horn, replacing 30–40 ml of cerebrospinal fluid (CSF) with filtered air. The same needle after being withdrawn, was then reinserted a few degrees vertically inclined from Reid’s base line (inferior orbital to superior external auditory meatus) and 10 deviated towards contralateral side to a depth of 9 cm. The pneumoventriculogram thus created confirmed placement of needle tip 2 cm posterior and 1 cm inferior to foramen of Munro, 2 cm lateral to mid-sagittal plane. After 0.5 ml of 1% Novocain confirmed favorable response, 0.5 ml of 40% Iodipamide was introduced. This simplified procedure was intended to bring pallidotomy into the hands of many doctors who may have to practice under relatively Spartan circumstances [10] (> Figure 11‐11). Two years later, Mao-Shan Wang’s stereotactic device was put to fruition and he carried out stereotactic alcohol ablation of globus pallidus, quite akin to the original chemopallidectomy that I.S. Cooper published in Science in 1955. (Cooper subsequently switched the target to the thalamus and reportedly performed over 3,000 cases of chemo-thalamectomies.) [47,48]. MS Wang published his stereotactic procedure ‘‘Surgical Treatment for Parkinsonian Syndrome’’ in the Chinese Journal of Surgery in 1961 [12]. Similar endeavors using self-designed stereotactic apparatus were undertaken by the pioneer surgeon Da-Jie Jiang, who published his twoyear preliminary reports on 20 cases of extrapyramidal system disease in the Chinese Journal of Neurology and Psychiatry in 1964. Jiang’s targets included globus pallidus, ventro-lateral nucleus, centro-median nucleus, nucleus reticularis, and internal capsule depending on the predominance of tremors, rigidity, and upper or lower limb involvement [14]. Localization was based on the Spiegel–Wycis atlas and coordinates [49]. Intraoperative confirmation of electrode placement was

137

138

11


. Figure 11‐11 Dr Wang’s patient feeding with previously tremulous left hand. Handwriting in dots before (above) and after procedure of another patient, 1959 (with permission from CC Wang)

carried out in three stages: using 0.5 ml 1% Novocain instilled into target through electrode needle, electrical stimulation, and depth electrode recording. Parkinson’s disease constituted the largest category of conditions treated, and the early procedures were predominantly ablative in nature. A large series of papers was published by Guo-Dong Gao from the Xian Fourth Military University (1,478 cases) and by Zhang JN (580 cases) [26,50]. Following Backlund’s first clinical trials in 1982 and Madrazo’s 1987 report on open microsurgical adrenal medullary tissue transplantation to the striatum for parkinsonism demonstrating therapeutic benefit [51,52], Wa-Cheng Zhang et al. of the Beijing Xuanwu Hospital embarked on transplantation of adrenal medulla to the head of caudate nucleus for six patients with Parkinsonian tremors using a XZ-III stereotactic device in 1987, using postoperative CT scan to confirm placement of transplant tissue carriage in the head of caudate nucleus. CSF dopamine and normetanephrine levels were elevated to 1–2 times of presurgical assay, and clinical symptomatic improvements of tremors were documented [53]. In 1994, Wu CYet al. published their results of combined fetal substantia nigra

tissue transplantation and stereotactic thalamotomy in the British Journal of Neurosurgery [54]. Gamma knife thalamotomy (Vim, VO) using a 4-mm collimator delivering 130 Gy for the treatment of 58 cases Parkinson’s disease and other movement disorders was reported by Zhang Jie et al., claiming 80% good response over a follow-up period of up to nine years [55]. The first human microelectrode recording (MER) during functional brain surgery was reported by Albe-Fessard in 1961. MER was introduced in China in 1998 by Yong-Jie Li of the Beijing Institute of Functional Neurosurgery, Xuanwu Hospital. More than 2,000 cases of pallidotomies with CT or MR guidance have been carried out annually in China [50,56,57]. A significant proportion of those employed intraoperative MER for physiologic confirmation of electrode placement. Li published a review of 1,135 cases of surgical treatment of movement disorders in 2001 [58]. In 1972, Bechtereva N.P. of the former Soviet Union carried out therapeutic stimulation of the deep brain structures using intermittent external stimulation [59]. Siegfried and Benebid started DBS surgery for Parkinson’s disease in 1985 after the former unexpectedly noticed


amelioration of Parkinsonian tremors from the implanted thalamic deep brain stimulator meant for chronic pain [60,61]. The U.S. FDA has approved DBS as a treatment for essential tremors, Parkinson’s disease, and dystonia in1997, 2002, and 2003, respectively. Led by Chung-Cheng Wang, the Beijing Tiantan Hospital performed the first DBS surgery for Parkinson’s disease in China in September 1998. Visits by Dr Philip A. Starr, University of California San Francisco (UCSF), and Dr Andres Lozano, Toronto Western Hospital, to China in 1999 helped DBS surgery, which gained further popularity. By 2003, more than 300 personal cases of DBS of subthalamic nucleus of Luys were performed by leading functional neurosurgeons like Jian-Guo Zhang, Yong-Jie Li (Beijing), and Bomin Sun (Shanghai). Other functional neurosurgeons like Kang-Yong Liu, Xiao-Wu Hu (Shanghai), Zhi-Pei Ling (Anhui), Shi-Zhong Zhang (Guangzhou), and Guo-Dong Gao (Xian) have also accumulated considerable

11

personal experience in DBS surgery. Their works have been variously published in Chinese Journal of Neurosurgery in 2002 [50,58,62,63]. There are more than 30 hospitals in China where DBS for Parkinson’s disease are regularly performed (> Figures 11‐12 and > 11‐13). The total number of cases done since 1998 has exceeded 1,000. In the Beijing Tiantan Hospital, Xuanwu Hospital, and Shanghai Ruijin Hospital, DBS treatment has also been extended to dystonia, Tourette syndrome, Hallervorden–Spatz disease, Meige’s syndrome, and chorea using subthalamic nucleus as target [26,64]. In 2002, Bo-min Sun used subthalamic nucleus of Luys for primary and tardive dystonia and achieved 90% 3 months to 3 years postoperative improvement based on Unified Parkinson Disease Rating Scale (UPDRS) and the Burke–Fahn–Marsden Scale [65]. Microelectrode recording were generally used in DBS surgery for intraoperative neurophysiologic targeting like in other overseas centers [66,67].

. Figure 11‐12 Deep brain stimulation surgery at the Shanghai Medical University (Fudan University) Huashan Hospital

139

140

11


. Figure 11‐13 DBS electrodes in subthalamic nucleus. (Shanghai Huashan Hospital)

Stereotactic Radiosurgery Stereotactic radiosurgery was first introduced into China in 1993 when the Leksell gamma knife was installed in Wanjie Hospital in the northeastern port city of Qingtao. This was soon followed by the Beijing Tiantan and Shanghai Huashan Hospitals. There are more than 50 stereotactic radiosurgery centers in China. The Chinese-designed rotating gamma system OURXGD was launched in 1994 and relied on only 30 cobalt-60 sources instead of 201 in the Leksell system. The SGS-I stereospecific gamma ray whole body treatment system (> Figure 11‐14) has since been launched. It is estimated that more than 8,000 cases of stereotactic radiosurgery and radiotherapy are treated annually, with Beijing Tiantan Hospital alone accounting for more than 10,000 cases till December 2007. Currently, there are 17 Leksell gamma knife radiosurgery installations (compared to 35 in the USA), almost an equal number of LINAC X-Knife units, over 20 rotating gamma system (OUR-XGD) units, 1 Novalis (Brain Lab) ststem, and 4 CyberKnife (Accuray Inc.) systems in mainland China. The OUR-XGD rotating gamma

system and the SGS-I have been exported to countries such as Egypt, India, and Hungary. In 1999, Liang JC and Wu HX of the Guangzhou General Hospital of PLA reported on the gamma knife treatment of 3,094 cases of intracranial lesions including brain tumors (with or without cytoreductive surgery), arteriovenous malformations, and even functional targets [28,68]. From July 1995 to May 1998, they had treated 280 cases of arterio-venous malformations, of which 42 were Spetzler–Martin grade I, 68 grade II, 95 grade III, and 7 grade IV and grade V 4, with a high obliteration rate for grade I and II. Another large clinical series of 1,431 cases was published by Yuan SB of the Sichuan Gamma Knife Center, Chengdu Army Hospital, using the OUR-XGD rotating gamma system over a four-year period from January 1997 to January 2001 [27]. The results of the LINAC X-Knife treatment on 510 cases were documented by Wang LG in 2000 [69]. Gamma knife radiosurgery have been utilized for thalamotomy in Parkinson’s disease and other functional disorders [55]. It is worth noting that Lars Leksell actually conceived stereotactic radiosurgery in 1951 for the treatment of the functional disorder


11

. Figure 11‐14 SGS-I: Chinese-made whole-body rotating gamma ray treatment system

of obsessive compulsive neurosis apart from tic douloreaux [9]. X-Knife radiosurgery for intractable epilepsy had been carried out by Qi ST on focal functional areas after preoperative electroencephalography (EEG), MRI, and PET evaluations [70].

Epilepsy Surgery Treatment of epilepsy in China can be traced to ancient times. Apart from Neijing (Yellow Emperor’s Inner Classic) 2000 years ago, there have been many medical classics dating back to Tang Dynasty (Qian Jin Yao Fang – The Invaluable Medical Script 652 A.D.), Sung Dynasty (Ji Sheng Fang Medical Script 1253 A.D.), Yuan Dynasty (Danxi Miscellaneous Comprehensive Medical Text 1481 A.D.) and Ming Dynasty (Treatment Yardstick 1600 A.D.), where traditional (TCM) treatments of epilepsy were described. Victor Horsley introduced surgery for medically intractable epilepsy in 1886. It is estimated that about one-quarter of epilepsy patients

become medically intractable, rendering them candidates for surgical intervention. Epilepsy surgery was accepted as an alternative treatment by the U.S. National Institute of Health Consensus Development Conference on Surgery for Epilepsy, March 19–21, 1990. In China, with a population of 1.3 billion, the total number of surgeries performed for medically refractory epilepsy had quadrupled from 600 before the year 2000 to 2500 in the year 2005 with a quarter from Beijing alone [71]. Review of the medical literature and hospital records indicate that the earliest documented works on epilepsy in China were in the 1950s, when Guo-Sheng Duan reported on lesionectomy for post-traumatic epilepsy with the aid of electrocorticography [72]. Krynauw RA first reported the removal of one cerebral hemisphere as the treatment for infantile hemiplegia [73]. Yu-Quan Shi of Shanghai First Medical College performed hemispherectomy under general anesthesia on four cases of severe infantile hemiplegia aged 3 years 5 months to 9 years in 1956. Pneumoencephalography demonstrated

141

142

11


ipsilateral ventricular and cisternal enlargement, ventricular shift to contralateral side, as well as porencephaly. A normal contralateral ventricle was a prerequisite. EEG was used in cooperative patients. Hemispherectomy was done en bloc as opposed to Krynauw’s cruciate excision of the hemisphere in four blocks, sparing the caudate nucleus and thalamus (> Figure 11‐15). Shi observed improved seizure control, alleviation in aggressive behavior, and slight improvement in developmental IQ, but persistence of hemiplegia. Spasticity and hypertonia, though reduced, did not translate into improvement in dexterity of hand movements [74]. Ya-Du Zhao of Beijing Xuanwu Hospital also reported his experience on 12 cases of hemispherectomy and anterior temporal lobectomy from 1959 to 1963, employing intraoperative electrocortical recording and bipolar direct electrical cortical stimulation [75] (> Figure 11‐16). Shi and Zhao were considered the pioneers who laid the foundation of modern epilepsy surgery in China. (Ya-Du Zhao was the younger son of Yi-Cheng Zhao, the ‘‘founder of neurosurgery in China’’). Subsequent works were done by

Chen-Ji Liu of the Nanjing Military General Hospital and Zhi-Xun Wu of the Kunming Medical College in 1963. Jian-Ping Xu and Ye-Han Wang of the Anhui Provincial Hospital popularized stereotactic ablative epileptic surgery using homedesigned stereotactic equipment in the 1970s. In 1978, surgical treatment for epilepsy was finally made available in the western interior province of Sichuan by Li-Da Gao, Ge Wu, and ChangGui Zhou. After a quiescence of 10 years during the unfortunate Chinese Cultural Revolution (1966–1976), resurgence of epileptic surgery was made possible by the rapid socioeconomic development commencing 1978. Many procedures were undertaken, including temporal lobectomy, corpus callosotomy, neuroaugmentation (vagal nerve and cerebellar stimulation), and multiple vertical subpial transaction (MST) (Frank Morell, Rush-Presbyterian-St Luke’s Medical Center) [71,76–78]. Techniques in stereotactic corpus callosotomy and hippocampal resection were published by Jian-Ping Xu and Ye-Han Wang of the Anhui Provincial Stereotaxic Neurosurgical Institute in 1984 [79], while in 1994 Xiao AP et al. from

. Figure 11‐15 Pneumoencephalography showing left cerebral hemispheric atrophy, compensatory ipsilateral ventriculomegaly (left picture), and the medial surface of the excised left cerebral hemisphere (right picture). (DrYu-Quan Shi, 1956, Shanghai, with permission)


11

. Figure 11‐16 Electrocorticography – unipolar lead tracing. Site A showed positive spike waves, electrical stimulation reproduced a subdued form of the original seizure. (Dr YD Zhao, Beijing, 1959, with permission)

the Nanjing Medical University reported 51 cases of stereotactic ablation of Forel H. Field (Mukawa J 1966) [80,81] and unilateral or bilateral amygdalotomy for 51 patients, aged 7–54 years, over a seven-year period since 1987 [82]. Bilateral cingulotomy and anterior capsulotomy were supplemented on 16 patients who had mental disorder. Procedures were done under local anesthesia except for four children. Intraoperative target localization was aided by pulsed electrical stimulation and acoustic impedance. Through similar works by C.C. Jiang and DJ Jiang of the Shanghai Huashan Hospital (1989) [83], Tang YL, and Zhang ZX, 82 cases (1990) [84] have been published. Gui-Sing Wang in 2004 used depth electrode for hippocampal EEG monitoring and stereotactic guidance in temporal lobe epilepsy surgery [85]. Neuronavigation and electrocorticography have been used in the resection of tumors causing secondary epilepsy [86,87]. Notable contributions were made by Qi-Fu Tan from the Hospital of Nanjing Military Area Command, who proposed guidelines on patient selection, preoperative assessment, operative results, and evaluation protocols, and also improved on the techniques of temporal lobectomy.

He performed corpus callosotomy in 1983 and Rasmussen’s functional hemispherectomy (Theodore Rasmussen, Montreal Neurological Institute) [88]. His results were published in Stereotactic and Functional Neurosurgery [89]. The spectrum of operated cases currently ranges from focal or partial epilepsy due to neoplasm to focal cortical dysplasia, hippocampal hemiatrophy, epileptogenic foci in temporal or frontal lobes, and multifocal epileptic zone located at or near eloquent cortex. Standard presurgical workup includes comprehensive neurosurgical, neurological, and neuropsychiatric assessment, electrophysiological studies including video EEG monitoring, digital imaging with MRI, and fMRI and MR spectroscopy where indicated [90,91]. Alterations in cerebral metabolism in the ictal and interictal period are documented with PET or SPECT [92]. MEG and implanted electrodes have also been employed [85,93]. Zhang G.J. of Beijing Xuan Wu Hospital [94] reported his experience on the application of long-term intracranial EEG monitoring using rectangular grids, linear strips, and stereotactically implanted flexible depth electrodes in epilepsy surgery in the Chinese Journal of Neurosurgery in

143

144

11


2005 (> Figure 11‐17). Luan G.M. from the Beijing San-Bo Brain Science Institute pioneered low-power bipolar electrocoagulation on layer III of cerebral cortex for epileptogenic foci on or near eloquent cortex for patients in whom MST was otherwise indicated, and reported good results [95]. This has been increasingly adopted by other epilepsy surgeons. However, its longterm efficacy and delayed sequelae have yet to be fully elucidated. He and his colleague Yunlin Li carried out various combinations of procedures from lobectomy, hemispherectomy, corpus callosotomy, and vagus nerve stimulation to low-power bipolar electrocoagulation on the eloquent cortex on 77 patients with multiple epileptogenic foci. Intraoperative electrocorticography was employed. Fifty-three patients achieved class I and II seizure control, based on Dr Jerome Engel Jr’s classification on surgical outcomes [96]. Stereotactic radiosurgery in epilepsy surgery was reported by Qi-Fu Tan in 2001 [97]. S.T. Qi used the X-Knife (BrainLab) with EEG, MRI, and PET for intractable seizures from focal functional areas [70]. Treatment on 38 patients diagnosed with medial temporal lobe epilepsy

from January 1998 to December 2004 using the Chinese-made rotating gamma system (OURXGD) was reported by Yuan SB, with 73.68% overall efficacy (Engel I–III) [98]. The Beijing Tiantan Hospital took the lead by setting up the first independent epilepsy surgery unit in China, with a current annual surgical census exceeding 200. The number of hospitals with an epilepsy surgery unit in Beijing City alone has tripled from three in 1990 to nine in 2005. There are between 150 and 200 practicing epilepsy surgeons in China responsible for over 2,500 cases per year. Epilepsy surgery is now available in Nanjing, Guangzhou, Harbin, Chongqing, Chendu, Kunming Anhui, Wuhan, Shandong, Xinjiang, and Shijiazhuang, apart from Beijing and Shanghai. The National Epilepsy Surgery Society was formed in 1990, with the first National Epilepsy Surgery Workshop held in the next year in Qufu (birth place of Confucius). It has since been held four times on annual or biannual basis. The Chinese Newsletter of Epilepsy Surgery was published in 1992 with Qi-Fu Tan as the editor. The China Association Against Epilepsy (CAAE) was formed in June 2005. It has become a member

. Figure 11‐17 Electrocorticography in epilepsy surgery in twenty-first century China


of the International League against Epilepsy (ILAE). Epilepsy surgery has featured prominently during six of the National Stereotactic and Functional Neurosurgery Conferences since 1987. Works by Chinese epilepsy surgeons are published in the Journal of Epilepsy Surgery, Journal of Asian Epilepsy in addition to the Chinese Journal of Stereotactic and Functional Neurosurgery. The second edition of the Textbook of Epilepsy Surgery edited by Qi-Fu Tan was released in 2006 [99].

Psychosurgery Although psychosurgery was performed in China by Tong-He Zhang of Xi’an in the 1940s, barely five years after the Portuguese Nobel Prize winner neurologist Egas Moniz conceived and performed the first operation for psychiatric disorder with neurosurgeon Almeida Lima on November 12, 1935 [100], this was followed by a lull of almost four decades. This was undoubtedly influenced by the events in neighboring fellow communist Soviet Union. Bekhterev and the father of Russian Neurosurgery Puusepp performed frontal leucotomy for maniac depressive psychosis and psychic equivalents of epileptics in 1906–1910, while classical leucotomies of Moniz and Lima for schizophrenia and severe pain were done from 1930s to late 1940s. Psychosurgery was banned in Soviet Union in1950 for ideological reasons [101]. Interests in psychiatric neurosurgery were rekindled in 1980s in a handful of Chinese hospitals in tandem with developments in the West [102]. Wu SL reported on 23 cases of schizophrenia treated with anterior cingulotomy in 1988 [103]. The first National Psychosurgery Conference was held in Nanjing in 1988, during which 542 cases of neurosurgical treatment for psychiatric disorders were presented. Procedures performed were anterior cingulotomy and amygdalectomies for various intractable mental illnesses including chronic schizophrenia and epileptosis. During the conference, the National

11

Clinical Guidelines for Psychosurgery were formulated, spelling out criteria for patient selection, informed consent, procedure types, and perioperative psychological assessments, among other issues. Since schizophrenia constituted the main bulk of operative cases, surgery expectedly did not result in amelioration of symptoms, just as Ballantine concluded in his retrospective report in 1989 on 198 patients who underwent anterior cingulotomy that schizophrenia and personality disorder responded least well to surgical treatment. A similar conclusion was made by Moniz [104,105]. Surgical option was not warmly received by psychiatrists who were never strong advocates of surgical intervention for their patients. The number of psychosurgeries dwindled progressively, and by the later half of 1990s many hospital had discontinued psychosurgery altogether. The revival of psychosurgery in China in the twenty-first century coincided with the widespread application of CT- and MRI-guided stereotactic surgery in the field of neurosurgery, allowing components of the limbic system, e.g., cingulate gyrus, amygdale, etc., to be accurately and safely approached [106]. Stereotactic limbic surgery, a more appropriate term than psychosurgery, was applied to the more responsive affective, anxiety, and obsessive compulsive disorders. A better understanding of the pathophysiological substrate of psychiatric diseases resulted in more widespread acceptance of psychosurgery. The French neurosurgeon Jean Talairach and colleagues first performed anterior capsulotomy for psychiatric disorders in 1949. Radiofrequency (RF) thermocoagulation stereotactic capsulotomy was subsequently developed by Lars Leksell. In 1999, Bomin Sun of Shanghai introduced MRIguided bilateral capsulotomy (> Figure 11‐18) for refractory obsessive-compulsive disorder, which he reported in the Chinese Journal of Neuropsychiatric Disorder in 2003. Capsulotomy-induced localized orbitofrontal subcortical metabolic changes in obsessive compulsive disorder were documented with PET (> Figure 11‐19) [107].

145

146

11


. Figure 11‐18 Postoperative MRI. Capsulotomy in obsessive compulsive disorder. (Shanghai Huashan Hospital 1999)

A few cases of DBS for obsessive compulsive disorder have been carried out by Sun, targeting the anterior capsulum. Unilateral (right) lesion combined with DBS of the dominant (left) side appeared to yield better surgical outcomes. Clinical analyses of 138 cases of multitarget stereotactic thermocoagulation of amygdala, anterior cingulum, and anterior internal capsule for intractable psychosis have been reported by Xiao-Feng Wang and Ke-Ming Jiang in the Chinese Journal of Stereotactic and Functional Neurosurgery in 2003 [106]. Since 2000, psychosurgery for opiate addiction was carried out in some Chinese hospitals. For the few years till 2004, when it was finally banned for general clinical applications by the Chinese Ministry of Health because of public outcry from its socioethical controversy, more than 500 cases have been performed. From July 2000 to November 2004, 272 cases of stereotactic

ablation of nucleus accumbens for opiate addiction were performed at the Tangdu Hospital, Xian Fourth Military Medical University [108]. (Note: The Russians had chosen cingulate gyrus as the target.) Guo-dong Gao published the clinical study on the success and relapse in the initial 28 cases, claiming 65% good to excellent results over a mean follow-up period of 15 months [109]. Other workers have also targeted the medial septal diagonal band complex [110]. One case of successful bilateral DBS on nucleus accumbens was included in Xu Ji-Wen’s 27 case reports of neurosurgical treatment for drug addiction in 2005 [111].

Functional Neurosurgery for Pain Surgical treatment for pain had progressed from interruption of pain pathways to stimulation of


11

. Figure 11‐19 From left: Preoperative and post-PET scan of a Chinese patient with obsessive compulsive disorder

pain-inhibiting pathways. Initial reports in the 1960s by Spiegel-Wycis, Ballantine, and, more recently, Wilkinson consisted of thalamotomy and anterior cingulotomy for chronic, intractable pain [8,112,113]. DBS of periaqueductal-periventricular gray and somatosensory thalamus for the relief of cancer pain and deafferentiation pain was reported by Richardson and Akil in 1977 [114]. Meta-analysis of 13 seriescomprising 1,114patients by Young RF et al. indicated a modest average long-term success rate of 60% [115]. It is currently performed in the United States as an ‘‘off label’’ (physician discretion) procedure. Neuroaugmentation for pain with the exception of Tsubokawa’s motor cortex stimulation (MCS) [116] is rarely performed in China [117]. Instead, the procedures were predominantly ablative in nature. Jiang-Ping Xu and Chung-Cheng Wang pioneered the percutaneous RF thermocoagulation for trigeminal neuralgia in China. A large series

of 1,860 cases was reported by Cheng-Yuan Wu et al. from the Qilu Hospital of Shandong University in 2004 [118]. A stereotactic basal frame for the 3D CT localization of foramen ovale and trigeminal stereoguide was designed. Excellent or good response was noted after selective percutaneous RF thermocoagulation in 96.3% cases, with a two-year recurrence of 24.8%. X-ray guidance, 3D CT, and neuronavigation enhanced the safety and accuracy of the procedure, overcoming the shortcomings of the freehand procedure. Microvascular decompressions for trigeminal neuralgia and hemifacial spasm are regularly carried out. Gamma knife radiosurgery for trigeminal neuralgia was reported by Mian-Shun Pan et al. of the Hefei Gamma Knife Hospital, Anhui Province, on 67 patients from 1997 to 2002, achieving 87.8% relief after an average three weeks [119]. Stereotactic thermocoagulation for central pain has been done since the early 1990s. Yang

147

148

11


FM from Harbin reported on this procedure in 1992 [120].Ventriculogram and, in more recent cases, CT- or MR-guided RF lesion of centromedian nucleus and amygdala for chronic cancer pain on 46 patients have been reported by Jing GJ in 2002 [121,122]. Hu YS and Li YJ concluded in 2005 that a combination of bilateral anterior cingulotomy with mesencephalotomy resulted in more gratifying postoperative results based on the Visual Analog Scale and the McGill Pain Questionnaire assessment of patients [123]. Long-term response is still pending. Gamma knife stereotactic radiosurgery for the treatment of cancer pain has also been reported in a series of 322 patients by Jin Hu and Li-Jie Yi [124]. Frequent psychiatric overlay of chronic pain patients and the exceptional tolerance for pain unique to a race that has endured centuries of tremendous hardship can make objective verification of pain alleviation difficult in China.

Clinical and Laboratory Research Laboratory research on low-power bipolar coagulation of epileptogenic foci has resulted in its clinical application for intractable epilepsy from eloquent cortex [125]. This was first pioneered by Luan GM of the Beijing Brain Science Institute as an alternative to multiple subpial transaction in 2002 [95]. This has proven effective and is being adopted by many epilepsy surgeons especially in China. Neural transplantation research on the survival, migration, and differentiation, and the effects of Nerve Growth Factor (NGF) or Glial cell line derived neurotrophic factor (GDNF) on transplanted embryonal or bone marrow–derived stem cells and olfactory ensheathing cells in laboratory models of cerebral ischemia, spinal cord injury, and substantia nigra are being carried out [31,126,127], hopefully to be translated to clinical application. Effects of limbic leucotomy on quinpirole-induced obsessive compulsive behavior of rats drew parallel conclusions to human clinical results [128].

Further improvement on the robotic system for stereotactic neurosurgery previously launched in 1998 and 2002 is being carried out by the Robotic Institute of Beijing University of Aeronautics and Astronautics [46], while design and upgradation of software for targeting in stereotactic surgery continues [129]. The effect of controlled-release polymers of BCNU-PLA in experimental glioma in vivo and intratumoral radioimmunotherapy by radio-iodinated monoclonal antibodies form part of the cerebral tumor stereotactic brachytherapy research and experimental treatment by Li AB at the Beijing 304 Military Hospital [130].

Literature and Journals As the numerous clinical and scientific research papers published in the local journals each year are in the Chinese language mainly for local consumption, they are not accessible to foreigners outside China. However, with a better command of foreign languages, Chinese medical professionals have begun to publish in international journals over the past decade. The Chinese Journal of Stereotactic and Functional Neurosurgery was first published by the Anhui Provincial Stereotaxic Neurosurgery Institute in 1986, initially irregularly. It is now released bimonthly (> Figure 11‐5). The Chinese Journal of Minimally Invasive Neurosurgery is a monthly publication of the Guangzhou General Hospital of Guangzhou Military Region. The Journal of Chinese Neurosurgery, the official journal of the Chinese Neurosurgical Society, is published quarterly since 1985. Other related journals include the Journal of Epilepsy Surgery, Chinese Journal of Contemporary Neurology and Neurosurgery, Chinese Journal of Clinical Neurosurgery, Chinese Journal of Neurosurgical Disease Research, and Chinese Journal of Neuromedicine. Stereotactic and functional neurosurgery books and monographs (in Chinese) published include Stereotactic Anatomy of Brain Gray


Matter by Yao JQ and Chen YM in 1980, Functional and Stereotactic Neurosurgery by Chen BH [131], Applied Stereotactic Anatomy of the Human Brain by Yao JQ and Dai HR in 1992, Brain Transplantation by WU CY in 1993, Stereotactic Radiosurgery by Chen BH [132], Textbook of Epilepsy Surgery edited by Tan QF in 1995 (2nd edition 2006), Minimally Invasive Neurosurgery by Ma LT in 1999, The Temporal Lobe Epilepsy Surgery by Lin Li in 2003, Modern Stereotactic Neurosurgery by Tian ZM in 1997, and the Chinese translation of Epilepsy Surgery edited by H.O. Luders and Y.G.Gomair (Lippincott, Williams and Wilkins, Baltimore, 2000).

Conclusion Napoleon Bonaparte of France (1804) once said: ‘‘Let China sleep. For when China wakes, it will shake the world.’’ In the last half century, we have witnessed the spectacular rise of China from obscurity to world prominence. As China integrates herself into the world community, Chinese medical professionals have redefined their roles in the international medical fraternity [133–144].

9. 10. 11. 12. 13.

14.

15.

16.

17. 18. 19. 20.

21.

References 1. Alt Kw, Jeunesse C, Buitrago-Tellez Ch, et al. Evidence for stone age cranial surgery. Nature 1997;387:360. 2. Han KX, Chen XC. The archaeological evidence of trephination in early China. Archaeology 1999;7:63-8. 3. Jia ZF. A brief history of the development of neurosurgery in China. Chin Med J (English) 1987;100:503-8. 4. Zhao YD, Zhao YL. Neurosurgery in the People’s Republic of China: a century’s review. Neurosurgery 2002;51:468-77. 5. Kwan (Guan) ST. Trigeminal neuralgia: diagnosis and treatment. Chin Med J (English) 1932;46:260-76. 6. Chang C. Neurofibroma of the gasserian ganglion. Chin Med J 1935;49:412. 7. Spiegel EA, Wycis HT, Marks M, Lee AS. Stereotaxic apparatus for operations on the human. Science 1947;106:349-50. 8. Spiegel EA, Wycis HT, Szekely EG, Gildenberg PL. Combined dorsomedial, intralaminar and basal thalamotomy

22. 23.

24.

25. 26.

27.

28.

11

for relief of so called intractable pain. J Int Coll Surg 1964;42:160-8. Leksell L. The stereotactic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316. Wang CC, Huang KQ. A new operation for parkinsonian syndrome. Chin J Neurol Psychiatr 1960;6(2):80-1. Xu JP. Stereotactic application in neurosurgery. Acta Anhui Medical College 1963;6:228-31. Wang MS. Surgical treatment for parkinsonian syndrome. Chin J Surg 1961;9(7):524-5. Jiang DJ. Investigation of stereotactic brain surgery: design and application of a stereotactic instrument. Chin J Neurol Psychiatr 1964;8:238-43. Jiang DJ. Stereotactic operation for extrapyramidal system disease: a preliminary report of 20 cases. Chin J Neurol Psychiatr 1964;8:370-3. Riechert T, Mundinger E. Beschreibung und Anwendung eines Ziegerates fur sterotaktische Hirnoperationen. Acta Neurochirurg Suppl 1954;3:309. Talairach J, Hecaen H, David M. Lobotomie prefrontal limitee par electrocoagulation des fibres thalamo-frontales a leur emergence du bras anterieur de la capsule interne. In: Congress Neurologique International. Paris: Masson; 1949. p. 1412. Parera C, Cooper IS. A modification of chemopallidectomy guide. J Neurosurg 1960;17:547. Austin A, Lee A. A plastic ball-and-socket type of stereotaxic director. J Neurosurg 1958;15:264. Gillingham FJ. Advances in stereotactic and functional neurosurgery. New York: Springer; 1980. Xu JP. The beginning and development of stereotactic and functional neurosurgery in China in the last 25 years. Chin J Stereotac Funct Neurosurg 1990;3:3-5. Tian ZM. Modern stereotactic neurosurgery (Chinese). Beijing Chinese Science and Technology Press; 1997. Schaltenbrand G, Wahren W. Atlas for stereotaxy of the human brain. Stuttgart: Thieme; 1977. Tasker RR, Hawrylshyn P, Rowe IH, Organ LW. Computerized graphic display of results of subcortical stimulation during stereotactic surgery. Acta Neurochir (Wien) 1969;21:173-180. Yao JQ, Dai HR. Applied stereotactic anatomy of the human brain (Chinese). Shanghai Science and Technology Press; 1992. Yao JQ, Chen YM. Stereotactic anatomy of brain gray matter (Chinese). Beijing Science Press; 1983. Zhang JN, Wu SL, Zhang X, et al. Brain stereotactic procedures for Parkinson’s disease: 580 cases report. Mod Rehabil 2000;4:1212-13. Yuan SB, Lei J, Tang Y, Liao SC, Shun SQ, Deng ZP, Zhang XH, Pang XX. Preliminary reporting of rotating gamma system in 1431 cases of intracranial lesions. Chin J Stereotactic Funct Neurosurg 2001;14:209-12. Liang JC, WU HX, Jun ZH, Li L, Fu XP. Preliminary report of 3094 cases of intracranial diseases treated by

149

150

11 29. 30.

31.

32.

33. 34. 35.

36.

37.

38.

39.

40.

41.

42.

43.

44.


gamma knife. Chin J Minimally Invasive Neurosurg 1999;4:9-12. General Hospital of Armed Police Force Beijing. http:// www.neuralstemcell.com.cn. Huang HY, Liu K. Neural regeneration and functional recovery of spinal cord injury after olfactory ensheathing cell transplantation. J Naval General Hospital (Beijing) 2001;4:65-7. Huang HY, Wang HM, Li BC, et al. Preliminary report of clinical trial for olfactory ensheathing cell transplantation in chronic spinal cord injury. Chin J Stereotac Funct Neurosurg 2004;6:348-50. Raisman J. Repair of spinal cord injury by auto transplantation of adult olfactory mucosal ensheating cells: translation from rat to clinical trials. Queen Square London: Spinal Repair Unit, Institute of Neurology; 2007. Hardy TL, Koch J. Computer assisted stereotactic surgery. Stereotac Funct Neurosurg 1982;45:396-8. Liu ZH. Latest development in stereotactic neurosurgical techniques. Chin J Neurosurg Dis Res 2003;2(1):1-3. Zha WG, Li AM, Fu XP. Stereotactic brachytherapy for cystic glioma in deep or eloquent brain areas. Chin J Minim Invasive Neurosurg 2006;11(10):474-7. Zhu FQ, LI AM, Xu F, Zhou YX, et al. CT guided stereotactic intracavitatory irradiation for intracranial cystic craniopharyngioma. Chin J Neurosurg 1998;14:142-4. Lin YC, Chen WJ, Zhu Ge QX. Combined CT- guided stereotactic interstitial radiotherapy of 32P and methotrexate chemotherapy for treatment of deep brain gliomas. Chin J Neurosurg 1996;12:216-8. Liu ZH, YU X, Guo Y, Cui YH, et al. The experimental and clinical preliminary studies of interstitial brachytherapy using iridium-192 on Glioblastomas. Chin J Neurosurg 1998;14:134-8. Tian ZM, Liu ZH, Li SY. Analysis of 1300 cases with CTguided stereotactic operations. Stereotac Funct Neurosurg 1998;70:147-8. Zhang X, Zhang J, Fei Z, et al. Neuronavigation-guided microsurgery for resection of brain tumors. Natl Med J China 2002;82(4):219-21. Zhao JZ, Cao Y, Lu Z, et al. Clinical evaluation of frameless stereotaxy in minimally invasive neurosurgery. Natl Med J China 2001;81:1042-5. Zhao YL, Wang CC, and Zhao JZ. Clinical use of navigation system in neurosurgical operations with 55 cases reports. Chin J Neurosurg 1998;14:274-6. Li GQ, Niu CS, Wang YH, et al. Functional MRI assisted neuronavigation guided surgery of intracranial tumors located in or near the motor cortex. Chin J Minim Invasive Neurosurg 2003;8(12):532-5. Jiang XF, Wang YH, et al. Neuronavigation assisted endoscopic procedures in neurosurgery. Chin J Stereotac Funct Neurosurg 2004;17(3):160-2.

45. Li GQ, Niu CS, et al. Application of ASA-610V neuronavigation system in minimally invasive neurosurgery. Chin J Contemp Neurol Neurosurg 2002;2(1):17-20. 46. Tian ZM, Wang TM, LI W, Chen MD. Location and mapping in robotic system for stereotactic neurosurgery. Chin J Biomed Engineering 2001;20(5):385-93. 47. Cooper IS. Chemopallidectomy. Science 1955;121:217. 48. Cooper IS, Bravo G, Riklan M, Davidson N, Gorek E. Chemopallidectomy and chemothalamectomy for parkinsonism. Geriatrics 1958;13:127-47. 49. Spiegel EA, Wycis HT: Stereoencephalotomy: thalamotomy and related procedures, vol. I–II. New York: Grune & Strattun; 1952-62. 50. Gao GD, Zhang H, Wang XQ, Liang QC. Indications of stereotactic and functional neurosurgery for Parkinson’s disease. Chin J Neurosurg 2002;18:12-4. 51. Backlund EO, Granberg PO, et al. Transplantation of adrenal medullary tissue to striatum in parkinsonism. First clinical trials. J Neurosurg 1985;62:169-73. 52. Madrazo I, Cuevas C, et al. Open microsurgical autograft of adrenal medulla to the right caudate nucleus in two patients with Parkinson’s disease. N Engl J Med 1987;316:831-4. 53. Zhang WC, Cao JK, et al. Neural transplantation with adrenal medullary tissue for tremors. Chin J Surg 1987;25 (11):650-2. 54. Wu CY, Zhou MD, Bao XF, et al. The combined method of transplantation of foetal substantia nigra and stereotactic thalamotomy for Parkinson’s disease. Br J Neurosurg 1994;8:709-16. 55. Jie Z, Ohye C, Shibazaki T. Gamma knife thalamotomy for the treatment of Parkinson’s disease and other movement disorders. Chin J Stereotac Funct Neurosurg 2005;18(6):354-7. 56. Wu SL, et al. Stereotactic neurosurgery for extrapyramidal disease. Chin J Nerv Ment Dis 1983;16(2):73. 57. Wu SL. Stereotactic surgery for tremors. Report of 507 cases. Chin J Stereotac Funct Neurosurg 1995;8(2):7-9. 58. Li YJ. Surgical treatment for movement disorders: a review of 1135 cases. Chin J Neurosurg 2001;6:350-3. 59. Bechtereva NP, Bondarchuk AN, Smirnov VM. Therapeutic electrostimulation of the deep brain structures. Vopr Neirokhir 1972;1:7-12. 60. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of ventroposterolateral pallidum: a new therapeutic approach for alleviating all parkinsonian symptoms. Neurosurgery 1994;35:1126-9. 61. Benebid AL, Pollak P, Louveau A, Henry S, de Rougemont J. Combined (thalamotomy and stimulation) stereotactic surgery of the VIM thalamic nucleus for bilateral Parkinson’s disease. Appl Neurophysiol 1987;50:344-6. 62. Zhang JG, Wang CC, Zhang XY. Deep brain stimulation in idiopathic Parkinson’s disease. Chin J Neurosurg 2002;18(1):4-7.


63. Sun BM, Liu KL, Lang LQ. Bilateral Deep brain stimulation of subthalamic nucleus in advanced Parkinson’s disease. Chin J Neurosurg 2002;18(1):8-11. 64. Zhang XH, Li YJ. Immediate and short term outcome after pallidotomy for intractable Tourette’s syndrome. Chin J Stereotactic Funct Neurosurg 2004;2:88-90. 65. Sun BM, Chen S, Zhan S. Subthalamic nucleus stimulation for primary dystonia and tardive dystonia. Acta Neurochir Suppl 2007;97(2):207-14. 66. Wang LX, Zhou XP, Hu XW, et al. MRI combining with micro electrode recording technology guided stereotactic surgery of Parkinson’s disease (270 cases). Chin J Stereotac Funct Neurosurg 2000;13(3):137-9. 67. Lee MK, Lee FC, Chee CP. The sunmed deep brain stimulation program for Parkinson’s disease: review after four years. XVII WFN World Congress on Parkinson’s disease and related disorders, Amsterdam. (Suppl); 2007. 68. Xu BT, Liang JC, Wang WM. Long term clinical efficacy of gamma knife radiosurgery for pituitary adenomas: a report of 487 cases. Chin J Minim Invasive Neurosurg 2006;11(6):244-6. 69. Wang LG, Guo Y, Zhang X. Linac X-knife stereotactic radio neurosurgery: 510 cases report. J Fourth Mil Med Univ 2000;21:1121-3. 70. Qi ST, Qiu BH, Yang KJ, Liu CY, Wang KW. Stereotactic radiosurgery treatment on focal area for intractable epilepsy. Chin J Stereotac Funct Neurosurg 2002;15:213-6. 71. Luan GM, Li YL. Epilepsy surgery in China: the history and current development. Neurology Asia 2007;12 Suppl 2:1-3. 72. Tan QF. Current status of epilepsy surgery in China. Chin J Stereotac Funct Neurosurg 2005;18(1):60-2. 73. Krynauw RA. Infantile hemiplegia treated by removing one cerebral hemisphere. J Neurol Neurosurg Psychiat 1950;13:243. 74. Shi YQ. Hemispherectomy for the treatment of infantile hemiplegia. Chin J Nerv Ment Dis 1959;1:48-51. 75. Zhao YD, Yang ZL, Tan YL. Surgical treatment of epilepsy. Chin J Neurol Psychiat 1965;9(4):325-9. 76. Kuang YQ, GU JW, et al. Combination of lesionectomy with multiple subpial transaction (MST) for intractable Epilepsy: 52 cases (Translation). Chin J Stereotac Funct Neurosurg 2004;02. 77. Morrell F, Whisler WW, Bleck TP. Multiple subpial transaction: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70:231-9. 78. Yuan SB, Zhang J, Chen MZ, Zeng FJ, Chen LG. Multiple subpial transaction for intractable functional cortex epilepsy. Stereotac Funct Neurosurg 1998;70:92. 79. Xu JP, Wu HX. Amygdalotomy in temporal epilepsy: 10 case report. Chin J Nerv Ment Dis 1984;10:26-8. 80. Jinnai D, Mukawa J. Forel-H-tomy for the treatment of epilepsy. Confin Neurol 1970;32(2):307-15. 81. Mukawa J, Iwata Y, Kobayashi K. Forel-H field lesion effect on subcorticogenic and limbicogenic

82.

83.

84.

85.

86.

87.

88. 89.

90.

91.

92.

93.

94.

95.

96.

97. 98.

11

seizures- experimental studies. Neurologia Medicochirogica 1966;8:282-3. Xiao AP, Chang Y, Chen YY, et al. Stereotactic therapy for epilepsy. Chin J Stereotac Funct Neurosurg 1994;7: 21-22. Jiang CC, Jiang DJ, Yin SJ. Preliminary report of stereotactic callosotomy and/or amygdale nucleus for intractable epilepsy. Chin J Neurosurg 1989;5:260-2. Tang YL, Zhang ZX. The therapy of intractable epilepsy with stereotactic methods: a report of 82 cases. Chin J Neurosurg 1990;6:305-7. Wang GS, Xu JW, et al. The use of depth electrode in hippocampal EEG monitoring for localizing the epileptogenic zone in temporal lobe surgery. Chin J Stereotac Funct Neurosurg 2004;17(5):271-3. Shi YW, Mu J, Shi JJ, Gu F. Role of intraoperative electrocorticography in epilepsy surgery (Translation). Chin J Stereotac Funct Neurosurg 2007;20(3):165-6. Li TD, Bai HM, Jiang XX. Resection of brain tumors with epilepsy by electrocorticogram. Chin J Stereotac Funct Neurosurg 2004;17(2):85-7. Villemure JG, Rasmussen T. Functional hemispherectomy: methodology. J Epilepsy 1990;3 Suppl:177-82. Tan QF, et al. Long term results of functional hemispherectomy for intractable seizure. Chin J Stereotac Funct Neurosurg 2000;75:90-5. Jin LR, Wu LW, Gao W, et al. Ictal video-EEG monitoring in presurgical evaluation for intractable temporal epilepsy in 15 cases. Chin J Neurol 2002;35(3):132-4. Liang SL, LI AM, et al. Usefulness of 128 channels preoperative prolonged video-EEG in locating epileptogenic zone. Chin J Stereotac Funct Neurosurg 2003;16 (4):206-9. Wang HM, Zhang GP, LI L, et al. Clinical application of interictal SPECT in epilepsy surgery. Chin J Stereotac Funct Neurosurg 2004;17(5):268-70. Zhu D, Li L, Hua G, Guo J. Neuronavigation with magnetoencephalography for epilepsy surgery. Chin J Stereotac Funct Neurosurg 2005;18(3):155-7. Zhang GJ, LI YJ, et al. Application of long term intracranial EEG monitoring in epilepsy surgery. Chin J Neurosurg 2005;8:8-11. Li YL, Luan GM. A new method to treat functional intractable epilepsy: Introduction of bipolar electrocoagulation technique. Chin J Stereotac Funct Neurosurg 2002;3:265-8. Engel J, Jr, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with respect to epileptic seizures. In: Engel J, Jr, Surgical treatment of the epilepsies. New York: Raven Press; 1993. p. 609-621. Tan QF. The stereotactic radiosurgery in epilepsy. Modern Rehabilitation 2001;1:105. Yuan SB, Wu W, Liang X, et al. The rotating gamma knife for medial temporal lobe epilepsy (report of 38 cases). Chin J Stereotactic Funct Neurosurg 2007;20(1):10-14.

151

152

11


99. Tan QF. Epilepsy surgery (Chinese). Nanjing University Press; 1995. 100. Moniz E. Prefrontal leucotomy in the treatment of mental disorders. Am J Psychiatry 1937;93:1379-85. 101. Lichterman BL. History of psychosurgery in Russia. Acta Neurochirugica 1993;125:1-4. 102. Binder DK, Iskandar BJ. Modern neurosurgery for psychiatric disorders. Neurosurgery 2000;47:9-23. 103. Wu SL. Bilateral stereotactic anterior cingulotomy for schizophrenia: a report of 23 cases. Chin J Neurosurg 1988;4:83-6. 104. Ballantine HT, Jr. Historical overview of psychosurgery and its problems. Acta Neurochir Suppl (Wien) 1988;44:125-12. 105. Feldman RP, Goodrich JT. Psychosurgery: a history overview [legacies]. Neurosurgery 2001;48(3):647-59. 106. Wang XF, et al. The clinical analysis of intractable psychosis treated with stereotactic techniques: report of 138 cases. Chin J Stereotac Funct Neurosurg 2003;16(4): 199-202. 107. Sun BM, Guan Y, Lang L, et al. Capsulotomy induces localized orbitofrontal and subcortical metabolic changes in obsessive compulsive disorder (abstract). Am Assoc Neurol Surg 2001;4 Toronto. 108. Wang XL, He SM, Li J, et al. Comparative study of reasons and influential factors on relapse after abstinence. Chin J Min Invas Neurosurg 2006;8:11. 109. Gao GD, Wang XL, He SM, et al. Clinical study for alleviating opiate drug dependence by a method of ablating the nucleus accumbens with stereotactic surgery. Stereotac Funct Neurosurg 2003;81:96-104. 110. Wang KW, Qi ST, Yang KJ. Medial septal diagonal band complex: A new target point of stereotaxic surgery for psychic dependence to opiods (report of 34 cases). Chin J Minim Invasive Neurosurg 2006;11(3):112-5. 111. Xu JW, Wang GS, Zhou HY. Neurosurgical treatment on alleviating heroin psychological dependence. Chin J Neurosurg 2005;21(10):19-22. 112. Ballantine HT, Jr, Cassidy WL, Flanagan NB, et al. Stereotaxic anterior cingulotomy for neuropsychiatric illness and intractable pain. J Neurosurg 1967; 26:488-95. 113. Wilkinson HA, Davidson KM, Davidson RI. Bilateral anterior cingulotomy for chronic non cancer pain. Neurosurgery 1999;45:1129-36. 114. Richardson DE, Akil H. Pain reduction by electrical brain stimulation in man.1. Acute administration in periaqueductal and periventricular sites. J Neurosurg 1977;47:178-83. 115. Young RF, Rinaldi PC. Brain stimulation in pain. In Levy RM, North RB, editors. The neurosurgery of chronic pain. New York: Springer 2003. 116. Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation for the

117.

118.

119.

120.

121.

122.

123. 124.

125.

126.

127.

128.

129.

130.

131. 132.

treatment of central pain. Acta Neurochir Suppl (Wien) 1991;52:137-9. Zhao CS, Luo H, Sun BM. Chronic electrical motor cortex stimulation for intractable central pain. Chin J Stereotac Funct Neurosurg 2004;17(4):345-7. Wu CY, Meng FG, Wang HW, et al. Selective percutaneous radiofrequency thermocoagulation for treatment of trigeminal neuralgia: 1860 cases report. Chin J Neurosurg 2004;20:55-8. Pan MS, Ling ZP, Wang P, Zhen HM. The clinical study on gamma knife radio surgery for trigeminal neuralgia. Chin J Stereotac Funct Neurosurg 2004;17 (6):358-61. Yang FM, Yang SC, Lian YZ, Shu J, Zhao B. Treatment of central pain with stereotactic radiofrequency thermocoagulation. Acta Harbin Med Univ 1992;26:188-9. Jing GJ, Yang KY, et al. CT-guided brain stereotactic therapy of chronic cancer pain: report of 35 cases. Mod Rehab 2001;5:97-98. Jing GJ, Yang KY. Brain stereotaxic therapy for chronic cancer pain by thalamotomy (Chinese). J Sun Yat Sen Univ (Med Sci) 2002;23(2):13-7. Hu YS, LI YJ. Stereotactic neurosurgery for cancer pain. Chin J Pain Medicine 2005;11(4):197-200. Hu J, Yi LJ. The treatment of cancer pain with stereotactic radiotherapy. Chin J Pain Medicine 2002;8 (4): 195-7. Yang ZH, Luan GM, Zhang Y. The modal building and research of temporal lobe epilepsy. Chin J Neurosurg 2005;8:8-11. Wang Y, Duan DY, Xu QY. Survival, migration and differentiation of neural stem cells after transplantation into the brain. Chin J Biotechnology 2004;11: 13-17. Yu LH, Kang DZ, Lin JH. Orient migration of intravenous administration of bone marrow stromal cells after spinal cord injury in adult rats. Chin J Stereotac Funct Neurosueg 2006;19(1):17-20. Guo CL, Zhao AJ, Song J, Liu SJ. The effect of stereotactic limbic leucotomy lesion on obsessive compulsive disorder in rats. Chin J Stereotac Funct Neurosurg 2006;19(1):9-12. Hu ZY, Zhao XZ, et al. Design and application of soft ware for targeting thalamus and basal ganglia in stereotactic surgery (Chinese). Acta Academiae Medicinae Militaris Tertiae 2003;25(11):1019-20. Xu XN, Wang JH, Lu G, et al. Controlled release polymers of BCNU-PLA against experimental glioma in vivo. Chin J Stereotac Funct Neurosurg 2004;17(5): 280-3. Chen BH. Functional and stereotactic neurosurgery (Chinese). Inner Mongolia People’s Press; 1988. Chen BH. Stereotactic Radiosurgery (Chinese). Beijing Press; 1994.


133. Anderson WS, Lenz FA. Surgery insight: DBS for movement’s disorders. Nature Clinical/Practice Neurol 2006;2:310-20. 134. Chou SN. Perspectives in international neurosurgery: a glimpse of neurosurgery in China. Neurosurgery 1978;3:120-2. 135. Gildenberg PL. Stereotactic surgery- the past and the future. Stereotac Funct Neurosurg 1998;70: 57-70. 136. Li P, Wu BJ, Nan Z, et al. Treatment of functional disorders with gamma knife radiosurgery. Stereotac Funct Neurosurg 1998;70:102-3. 137. Ma LT. Minimally Invasive Neurosurgery (Chinese). Wuhan: Huazhong Science Publisher; 1999. 138. Tan QF, Sun KH, Sun KJ. Results of surgical treatment in 76 patients with temporal lobe epilepsy. Stereotac Funct Neurosurg 1998;70:91.

11

139. Wang YH, Dong YJ, and XU JP. Stereotactic neurosurgery: past and present. Chin J Stereotac Funct Neurosurg 1986;1(1):46-52. 140. Wu CY, Wang YH, Meng FG, Liu YG. Development of stereotactic neurosurgery in China. Neurosurgery 2005;56:851-60. 141. Yi SY, Wu CY. Brain tissue transplantation (Chinese). People’s Hygiene Press; Beijing: 142. Zhang K, Zhang JG, Ma Y, et al. Subthalamic deep brain stimulation in the treatment of secondary dystonia. Chin Med J 2006;119(1):2069-74. 143. Zhang YQ, Li YJ, et al. Surgical treatment for adult onset dystonia: analysis of 16 cases. Chin J Min Invas Neurosurg 2006;11(2):121-5. 144. Tian ZM, Lu WS, Wang TM, et al. Application of a robotic telemanipulation system in stereotactic surgery. Stereotact Funct Neurosurg 2008;86:54-61.

153

9 History of Stereotactic Surgery in France A. L. Benabid . S. Chabardes . E. Seigneuret

Introduction The history of stereotaxy is part of a larger perspective of a methodological approach that is therapeutic and focuses on the search for precision. The search for precision implies the recognition and marking of targets (taxonomic version of the etymology of the word Stereotaxy, from taxis: order) as well as to the tactic act, which is the achievement, at least in human stereotaxy, of this approach (tactic version of the etymology of the word stereotaxy). The common denominator of these two definitions is the space (from the Greek Stereos: solid, volume), which by itself would define the methodology, based on the spatial coordinates of a point that will be called the target. This etymological duality corresponds in fact to a historical perspewctive, as for the first time, in 1917, the need to precisely recognize and localize the various spatial structures of the brain led Horsley and Clark [1] to design an instrumental method aimed at quantifying cerebral space, in order to attribute precise coordinates to the different structures that the anatomo physiologists were studying at that time. This taxonomy approach led to the development of an instrument and a method, and then to the elaboration of the concept of the surgical act based on exact location. This corresponded in fact to the tactic version of the etymological definition of stereotaxy, even if it has been, historically, only secondary. Having designed an apparatus and developed a method, one tended to build on these two concrete elements a philosophy, or at least a state of mind. This constitutes the most interesting, maybe the #


most noble, part of the history of stereotaxy. This history has been the reflection of the complex interaction, depending upon circumstances, between the technical means of the moment and the therapeutic needs, as well as on pharmacological alternatives. This process is not specific to stereotaxy. It characterizes every approach of Homo Faber, which tries to solve his current problems, on the bases of the know-how which is available. He often stumbles on technological bottlenecks, which impede the development of methods, and even make it transitorily disappear, until the advance of knowledge in other domains provides the key, which will open the lock. A new momentum of development is therefore observed until the time when a new obstacle stops the process again, or when the need disappears, often because advancements of knowledge in other domains have brought more satisfactory solutions. Stereotaxy is at the crossroads of industrial technology, of surgical technology and therapeutic needs, themselves strongly enclosed in the domain of neurosciences, which, as we know, is undergoing rapid evolution. It is therefore not surprising to see that its history has been chaotic, and it can be foreseen that it will become even more complicated.

Innovation Comes from Necessity: History of the Development of the Stereotactic Frame In technological evolution, as in any kind of method, development is possible only if the need exists. The evolution of the need depends

98

9

History of stereotactic surgery in france

on the evolution of sciences and related disciplines. The need to discover the origin of an idea and identify its conceptual father depends on the value of this idea. There is a rich literature devoted to determining the paternity of stereotactic apparatuses, quite often in a chauvinist perspective. This need witnesses the importance of stereotaxy in the world of neurosciences, as it is easy to recognize its impact on our thinking and practice, in experimental neurophysiology as well as in clinics. This chapter could be included in this international competition, thanks to Professor Grellier in Nice who retrieved a document and was kind enough to give it to us. It proves that the stereotactic frame, and therefore the stereotactic method and practice, was invented in France at the end of the nineteenth century. In a paper published on 27 November 1897, in the review L’ Illustration [2], there is a very detailed description of it, in a scientific style which is not often used in public magazines (but it is probable that the culture of the public at that time was more scholarly than nowadays). The procedure was performed in a laboratory of the faculty of medicine in Paris, where the inventor had set up a system that could be anchored by pins to the skull of the patient presenting with projectiles (bullets or shrapnel pieces). This system featured a topological radiology modality oriented in space, consisting of two Crooke’s tubes and two photographic film supports mounted orthogonally. The radiographic antero-posterior and lateral projections of the head of the patients that were obtained made it possible to localize, as on an analytical geometry diagram, the presence of opaque foreign bodies, generally projectiles, even if they were not in teleradiological conditions. The figure, > Figure 9‐1 presented in this article of L’ Illustration is of remarkable precision and establishes without any doubt the feasibility of this approach, and of its precision, under the appearance of a frame, which could be very well compared to the frame of Horsley and Clark, or even to the Leksell frame, which is wellknown nowadays. The article reports that two interventions were successfully

performed, and reported in the Academy of Medicine by Marey (the physiologist). One cannot say whether this invention inaugurated either stereotaxy or interventional neuroradiology, but it is witnesses that the need to navigate within the brain was manifested a long time ago, whenever it was necessary to perform an intervention inside it. It seems that the encephalometer of Zernov (1889) [3], used in surgery by Altukhov 2 years later (1891) [4], aimed at the same goal of reaching intracranial structures, but without introducing in the approach a precise methodology of localization. It is more classical to quote at the origin of stereotaxy the publication in 1908 of Horsley, neurophysiologist and neurosurgeon, and Clark, mathematician, probably because their publication has given, in a scientific review known and read by their peers, the description of the apparatus, a stereotactic atlas of the monkey brain, and the method of electrolytic lesion by constant current. Ten years later, Mussen, the engineer who had worked on this frame, manufactured a version that he had designed, and then stored in his attic, wrapped in a journal, the headline of which helped to date this invention [5,6]. It was only in 1947 that Spiegel and Wycis [7] reported the use of an apparatus that allowed intraoperative X-rays, and coined the term stereoencephalotomy. This new approach allowed for the first time the mortality after operations to drop to about 2%. Ever since, the evolution of the stereotactic frames has not stopped, following trends and fashions, which accelerated its developments, and its adaptations to new indications. Two main categories of frames were developed, each favoring a modality of approach to the target. The Goniometric frames, putting the target at the pre-established center of an arch, are headed by the Leksell frame, followed by those of Richiert and Mundinger, and Todd and Wells, which finally produced the frames of Brent Robert Wells (BRW), and Cosman Robert Wells (CRW). The Cartesian frames on the contrary have developed orthogonal approaches from the lateral and the anteroposterior sides of the head, such as


9

. Figure 9‐1 Upper right drawing: corresponds to the upper left photograph. The system, is made of:A frame fixed on the head by a cast. Three rods are firmly pressed on the forehead and on the cheek bones–Two photographic films will be successively placed on one side–On the other side two Crookes tubes (X-ray) which can be oriented–X-ray pictures are taken with one of the Crookes tubes. This is repeated with the other tube and another film–The system is removed from the head, the extremities of the three rods are tattooed on the skin, and the spatial position of the bullet is obtained at the intersection of two threads used to join the focus of the tubes to the images on the films, with respect to the three skin fiducials. Lower right drawing: corresponds to the lower left photographThe extremities A, B, C and D of four sliding rods which are adjusted such as to touch the skin fiducials (or face markers) for A, B, and C, and to the target/bullet for D. The length of the D rod represents the depth of the target/bullet and will guide the surgeon during the intracerebral procedure–This “Operating Compass” can be transported to the operating theatre, in the hospital or at the patient’s house. When it is applied on the face markers, the rod D points the direction and shows the distance to the target

99

100

9


the Talairach frame, as well as those of OlivierBertrand, and of Hitchcock. Designs to adapt the goniometric approaches to the Talairach system have been made for individual users [8].

Introduction in France of the State of the Art of the Stereotactic Method After the reports of Remy and Contremoulins in 1897, the next mention of stereotaxy in France is found in the early 1950s about movement disorders (treated by ansotomy, pallidotomy and thalamotomy), [9–13], psychosurgery by lobotomy [14–16] and pain [17–24], using the conventional frames, methods and targets available in the western world. The French neurosurgeons practiced functional neurosurgery as a part of their general activity, showing special interest in these methods.. Some were especially attracted by the field and developed their own methods, such as ansotomy and campotomy for movement disorders [25–26], their frames [27–29] and their anatomical charts and atlases [30–34]. Among them would emerge, in academic institutions and practices, some who would provide such important contributions that they would make a profound mark on French stereotaxy.

Visionary Inputs of Two Giants Against this background of random practice and approaches, two main schools emerged and set the rules which have not only changed French stereotaxy but have had a real worldwide impact. They were inspired and led by two giants: Gerard Guiot at Hospital Foch, in the Parisian suburb of Suresnes, and Jean Talairach at Hospital SanteAnne, in Paris, who have educated and trained most of the French functional neurosurgeons and influenced many of the foreign masters in this field.

The Recognition of the Benefits of Multidisciplinarity: G. Guiot and D. Albe-Fessard Besides his tremendous input in neurosurgery at many levels, including pioneering work in pituitary surgery, Gerard Guiot (1912–1988) extended to all aspects of his practice an obsessional search for elements to improve his practice and results. In his functional neurosurgical practice, from pain to movement disorders, he needed to use every means to improve the precision of localization. He introduced the electrophysiological method into stereotactic procedures with the help of neurophysiologists such as Denise AlbeFessard and G Arfel. The use of intraoperative micro-electrophysiological recording has incomparable value, and was strongly stressed by the French stereotactic school in the third quarter of the twentieth century [11–13,27,35–38]. Its use was validated when it was clear that there was a correlation between efficacy, lack of complications, and the precision of the placement of the lesion, which was then used as the therapeutic tool. However, the importance of intraoperative neurophysiology has further decreased, for several reasons. The interest in stereotaxy has decreased because of the advent of levodopa, the expertise required an equipment not being available for neurosurgeons. Electrophysiology also appeared to be obsolete, or useless, because the target became more visible with the new imaging methods. This wrongly suggested that it was sufficient to aim at the structure to reach its function. This is true for the Pallidum and for the Subthalamic Nucleus, but it is not the case for the subdivisions of the Thalamus and for new targets such as the Pedunculopontine Nucleus (PPN), which are not visible and where the functional site has to be discovered through electrophysiology. Despite the advances in modern imaging, there are distortions in MR images, and displacements of cerebral structures during surgery, particularly during the progression of the


electrodes. The functional target which is related to the efficacy of chronic stimulation cannot be reached without additional investigations. The recent development and the new interest in functional neurosurgery, particularly for movement disorders, the need for precision and safety, and the possibility to mimic the therapeutic effects of chronic stimulation during surgery, have given back intraoperative electrophysiology its double role. As an exploratory method, it provides the characteristic signature of the neuronal set through the morphology of the action potentials, their discharge pattern, and the activity evoked by external stimuli (proprioceptive, auditory, visual). As a therapeutic method, it reproduces accurately the effects which would be observed over a long term and allows the functional recognition of the target. It has become part of the classical panel of methodology in neurosciences. The discovery during those explorations of the differential effects of stimulation at high and low frequency has led to the development of methods of chronic deep brain stimulation at high frequency [39–47]. This development would not have been possible, or at least been long delayed, if implantable stimulators did not already exist, having been developed for the treatment of pain [23,48,49]. There was a long and disappointing inefficacy during which it had slowed down. The development of new applications to the treatment of movement disorders has re-motivated the industry to redevelop flexible electrodes and programmable stimulators.

The Reconsideration of an Unsatisfactory Situation: Jean Talairach Jean Talairach (1911–2007) could be described as a psychiatrist believing in psychosurgery and feeling that surgery was too serious to be left to neurosurgeons. He redesigned a stereotactic frame and raised it to the level of a method, if not a philosophy;

9

he built his own stereotactic atlas. When psychosurgery went out of favor, he had the spirit to refocus his stereotactic doctrine, his skills and his tools, to new applications, thereby opening new avenues such as a totally new vision of the surgical treatment of epilepsy. When movement disorders also went out of the field of stereotaxy, after levodopa was introduced as a treatment, he shifted gears and used his methodology to create a new field.

Stereotactic Biopsies The fate of a method depends on the need for it. When an outside phenomenon, such as a pharmacological newcomer, occurs, which brings, at least for a given period, a satisfactory solution, the need for the method disappears and sends it to oblivion. At the same time, the understanding that some pathologies, such as brain tumors, are not adequately taken care of by classical surgery and could eventually respond correctly to general treatments, such as chemotherapy or radiotherapy, induces the application of this forgotten, or nearlyabandoned, method to a totally new field: this is what Jean Talairach and his associates did when they introduced stereotactic biopsy, at a time where movement disorders needed less surgery as levodopa had come and provided a very attractive treatment. The school led by Jean Talairach at Sainte Anne must take credit for the development of stereotactic biopsies, for multiple reasons [50–54]. The growing awareness in the middle of the twentieth century that brain tumor resection was not always efficient or useful and that it was sometimes sufficient and less invasive to treat brain tumors by radiotherapy or chemotherapy had stressed the necessity to obtain the pathological diagnosis before applying those nonsurgical treatments. In addition to providing the right diagnosis, and sometimes to errors being corrected, multiple staged stereotactic biopsies allowed determining the extent of the lesion, particularly at the time when

101

102

9


computerized tomography did not exist. Currently, using modern imaging, this need for the determination of the extent has diminished, allowing a decrease in the number of samples to the minimal necessity to obtain the histological signature, for instance, taking advantage of the recent demonstration that oligodendroglioma are chemosensitive when they bear the genetic mutation in 1p19q.

Brachytherapy [55–61] The use of ionizing radiations, in all their forms, has always played an important role in the therapeutic function of stereotaxy. The Parisian school of Jean Talairach had very early taken advantage of the radionecrotic properties of isotopes to create lesions with a reasonably foreseeable size: using Yttrium 90Y, mainly beta emitter, it was possible, because of its reduced propagation, to create a lesion of about 5 mm in diameter, in the place where the seed had been deposited under radiological control. This was used to replace electrolytic lesions for functional neurosurgery. This had also allowed Jean Talairach to develop methods of pituitary freination by Gold 198Au in the treatment of diabetic retinopathy by pituitary stereotaxy, based on the empirical observation, following the occurrence of a Sheehan syndrome, of the amelioration of this severe complication of diabetes. These indications of major hormonal surgery were also responsible for the concept, design and realization of the pituitary stereotactic frame. These methods, which led to very satisfactory results, later disappeared, following the advent of various forms of hormonal suppressive chemotherapy, similarly efficient and easier to manipulate. In the 1970s appeared the need to have a more focal treatment of brain tumors. Jean Talairach, along with Gabor Szikla, using Iridium 192Ir, introduced stereotactic Curitherapy, (named in

France from the multiple Nobel Price winner Marie Curie), in parallel with the German school, particularly in Freibourg with Riechert and Mundinger. In the literature in English, particulary in the United States, the term brachytherapy has been coined to better express the short distance effects of this therapeutic approach. This whole field of interstitial irradiation was undertaken along with experimental approaches and careful postoperative control and follow-up, as well as extensive histological studies of the postmortem specimens, to understand the mechanisms and improve the methodology.

Radiosurgery with Osvaldo Betti [62] Similarly, the Swedish school, led by Lars Leksell, combined stereotaxy and external radiotherapy to deliver high ionizing radiation with the precise doze in a well-defined spot. They developed the Gamma knife, made of several cobalt 60 sources, with convergent collimated beams on a determined point in space within the stereotactic frame. Initially designed to be applied in neurosurgery for movement disorders, this method has reached its full development in the treatment of arterio-venous malformations, improving the surgical outcome and reducing morbidity. More recently, this radiosurgery (which creates a lesion by radionecrosis) has been applied to tumors of small diameters and particularly to multiple brain metastases. In order to make this radiosurgery more accessible without being obliged to purchase the expensive Gamma knife system, the Talairach school, through its Argentinian pupil, Osvaldo Betti, has designed another instrument for radiosurgery using a linear accelerator, which he named LINAC. The stereotactic principles are satisfied by a physical connection between the linear accelerator and a chair rocking in two dimensions and to which the stereotactic frame is


attached. The system has been installed at Hospital Tenon in Paris and became the first radiosurgical set-up used in France.

Harvesting the Crop When the Seeds Have Been Spread: The Development of Epilepsy Surgery Through Deep EEG Recording, Extended to Brain Tumor Surgery Jean Talairach, Jean Bancaud, Gabor Szikla, and Claudio Munari created a huge momentum in Sainte Anne, a school like which there are few in the world, and which could be compared to what happened at McGill around Penfield and Jasper. Most of the French stereotactic and functional neurosurgeons came here to be trained, and returned to Lille, Marseille, Grenoble, Rennes, Toulouse and Bordeaux, and later to other academic centers. Those who were to become famous in other parts of the world also visited Talairach’s school, learnt from it and also shared their knowledge, expertise and ideas. They came from and returned to Japan, Spain, Argentina, USA and Canada. Talairach had been trained in the Neurosurgery Department of Sainte Anne Hospital, under the supervision of his mentor, M. David. Since 1946, he had been developing stereotactic approaches to functional neurosurgery for chronic pain, movement disorders, pituitary diseases, and unoperable tumors. His interest was especially focused on the stereotactic definition of deep brain structures and he built up a coordinate system based on the anterior commissure-posterior commissure axis, which allowed him to “normalize” anatomical data from several brains. This led to the publication, in 1957, of his first stereotactic atlas on deep brain nuclei. His second stereotactic atlas, devoted to telencephalic structures, would be published 10 years later while he was already working with J. Bancaud. J. Bancaud, a neurologist

9

and electroencephalographer, was a student of H. Fishgold, a radiologist who was working at La Salpe´trie`re Hospital in Paris. At the beginning of the 1950s, Fishgold’s group joined David’s group in Sainte-Anne Hospital, and this undoubtedly represented a turning point for both Talairach and Bancaud’s careers. They were then gaining experience at the same institution, where epilepsy surgery had began, and they benefited from the high scientific experience of P. Buser and M.D. Dell in neurophysiology, and of H. Hecaen and J. de Ajurriaguera in neuropsychology. Bancaud saw the potential of Talairach’s method to localize the sources of EEG discharges in three dimensional space, and he began working with him to develop applications of stereotactic functional exploration for presurgical evaluation of medically intractable epilepsies. An operating room dedicated to stereotactic neurosurgery was opened at Sainte-Anne Hospital in 1959, and the project to record epileptic seizures by means of stereotactically implanted intracerebral electrodes became a reality. The term StereoElectroEncephaloGraphy (SEEG) [63–68] would be coined in 1962 to describe this new method, which, for many reasons, was revolutionary in the field of epilepsy surgery: SEEG recordings allowed the study of spatiotemporal dynamics of seizure discharges with respect to their clinical features, with a high degree of anatomical precision. The technique of SEEG offered the possibility to dissociate the presurgical investigation phase from the surgical therapeutic act: SEEG recordings and electrical stimulations were carried out in the operating room under “acute” conditions; after removal of the intracerebral electrodes, resection surgery was planned, and performed as a second step. SEEG was directly derived from a new conceptual approach for studying partial epileptic seizures, which was based on the assumption that the chronological occurrence of ictal clinical signs reflects the spatio-temporal organization of the epileptic

103

104

9


discharge within the brain. The same year, Talairach’s team was joined by G. Szikla (1928–1983), an emigrant from Hungary where he had worked with Prof Zoltan, and who devoted his career to anatomy and stereotaxy, and played a major role in the development of stereotactic angiography. More particularly, carefully looking at the close relationships of cortical vessels to the gyri and sulci, Szikla showed that a meticulous analysis of vascular trajectories allowed the extraction of gyral form and dimension, and provided constant landmarks for the interpretation of anatomical variablity. Thus, stereotactic angiography was not only a routine procedure for safe implantation of intracerebral electrodes, but also a valuable tool for deducing the actual location of electrode contacts, particularly before the availability of modern imaging methods. The selection of the structures to be explored was based on a very careful analysis of all the data – notably clinical – collected during the non-invasive presurgical investigations. Such an analysis led to one or several hypotheses concerning the site(s) of seizure onset and the pathways of preferential ictal spread. Electrodes were then placed accordingly, in a way that enabled interpolation of intracerebral EEG activity within the interelectrode space. Proceeding this way, Bancaud and Talairach could precisely study how symptoms accumulated as the epileptic discharge propagated into different cortical structures and, in turn, could confirm or inform their initial hypotheses. The high spatial and temporal resolution, coupled with a high power of localization that the SEEG procedures offered, rapidly provided a large amount of data that had immediate repercussions on the practice of epilepsy surgery. This was mainly oriented towards tailored brain resection, but other avenues were not neglected. Deep brain stimulation for epilepsy in the thalamus (unpublished data) was initiated, following Mondragon and Lamarche’s work [69] on monkeys at Sainte Anne in Talairach’s INSERM unit U97, and this was the beginning of the application of STN stimulation to intractable nonoperable epilepsies

[70,71], in the wake of deep brain stimulation for movement disorders.

The Evolution of the Stereotactic Treatment of Movement Disorders The relation between stereotaxy and the treatment of movement disorders is the most exemplary as their mutual developments have been constantly linked during the last 50 years, in North America as well as in Europe, and particularly in France. In 1950, pallidotomies and ansotomies were performed in France by Fenelon [25,26], Guiot [27] and Gros [28] and by Mazars [72–80] through coagulation of the deep brain target through procedures guided on intraoperative visual landmarks. However, this procedure lacked precision and reliability, and the stereotactic approach was the logical next step. This step was taken by Guiot in France in 1952 [11], and quickly followed by Hassler and Riechert in Germany [81], Leksell in Sweden [82], Gillingham in Scotland [83], and Cooper [84,85] in the US as well as in Japan [86–88]. The lack of medical treatments for Parkinson’s disease, and even more for essential tremor, did no’t provide an alternative to the surgical treatment, which at that time was essentially ablative (microcoagulation, thermolesion, alcohol injection, or isotope insertion). Despite recurrences and complications, and because of its overall efficiency, the treatment of movement disorder remained the main condition for stereotaxy, until the advent of the levodopa era in the late 1960s. Since then, in France as well as in most of the Western countries, the surgical treatment of movement disorders, and therefore stereotaxy, underwent a dramatic decline, a few thalamotomies and pallidotomies still being performed in a limited number of expert centers. When, on the contrary, it appeared that some forms of Parkinson’s disease did not respond to the drugs or were plagued by motor


fluctuations and dyskinesias, there was a new need for surgery. However, complications were unacceptable at that time and the method was required to evolve towards new directions, where efficacy would result in negligible morbidity. This led eventually to new developments such as brain grafting or, more recently, high frequency deep brain stimulation. Stereotaxy, like other methods, has experienced this type of fluctuation in its history and has on each occasion progressed and acquired new skills and possibilities, far beyond what it was initially designed for modern stereotactic surgery. The observation that after the honeymoon, levodopa had side effects such as levodopa-induced dyskinesias, and dopamine agonists induced hallucinatory and psychotic side effects, re-created the need for surgery and for new approaches. The saga of the brain grafts [89–92] did not involve French stereotaxy until the late 1990s when Marc Peschanski’s team [93–110], under a consensual but unspoken nationwide agreement, was considered as the only French team to engage in this difficult and highly skilled adventure. Since that period, he has contributed significantly to the advancement of the field of neural grafts, necessarily associated with the gene therapy domain [93,95,107,110]. He has applied this approach to movement disorders such as Parkinson’s disease, but mainly to Huntington’s disease [93– 96,100,101,103,106], which currently has no treatment. This is a clear example that some very specific fields of research and their applications should be totally assigned to a group, avoiding a sterile spread of financial and technical support. It was the reintroduction of Leksell’s pallidotomy by his two pupils Laitinen and Hariz in 1992 [111], and the introduction in 1987 of high frequency stimulation (HFS) [39–41,44,112– 114], that reinstalled in France, stereotaxy as an armamentarium of movement disorders. Its flexibility, adaptability, reversibility, and low morbidity, which help in performing HFS bilaterally,

9

have justified the quick and large development of this method in France, where the functional neurosurgery community and the neurologists, forming multidisciplinary teams, convinced the government to provide full reimbursement to this method. This allowed the development of several academic centers (up to 17 in 2007), and the extension to other movement disorders such as essential tremor and dystonia, to other targets than the Vim thalamus, such as the Pallidum and the subthalamic nucleus.

Stereotactic Surgery for Pain and the Development of Deep Brain and Cortical Stimulations Cancer pain was the most frequent indication of stereotaxy, particularly in the 1950s [115]. In the 1960s, Gabriel Mazars, also at Sainte Anne hospital in Paris, elaborated on the neurophysiological bases which were different from the gate control theory, a model of sensory perception and pain generation, on the basis of which he aimed at the thalamus and implanted electrodes [17–24], which were sometimes connected to an implantable stimulator, which was made on his ideas and then commercialized by the ELA company. At the same time in France several teams, sometimes in cooperation with other European countries [116], performed thalamic stimulation or periacqueductal gray (PAG) stimulation. Following the reports of the gate control theory by Wall and Melzack (1960), of the positive effects of thalamic stimulation for deafferentation pain and of the PAG for nociceptive pain, pain surgery also took advantage of the fundamental work of the INSERM team of Jean Marie Besson and Gise`le Guilbaud, which was close to Sainte Anne hospital. Quite often the practitioners of stereotactic surgery were very active in the field of functional neurosurgery as a whole and did not dissociate their activity from other methods such as spinal cord stimulation, for instance.

105

106

9


This practice of deep brain stimulation for pain occurred in parallel to the large practice of spinal cord stimulation, which is not truly stereotactic, but part of functional neurosurgery. One must also mention Sedan in Marseille, Lazorthes in Toulouse, and Blond in Lille, who were very active in spinal cord stimulation as well as in intrathecal morphine, in strong correlation with the neurophysiologists. This created a real pain culture, which had a strong influence on the French stereotactic community, by calling for a more scientific approach, a critical evaluation of the results, and an ethical look at the overall field. In this stream of activity to try to treat pain, the work of Tsubokawa and his first reports on the efficiency of motor cortex stimulation for various types of pain, such as postherpetic facial pain or post-stroke (mainly thalamic) pain [117], attracted the attention of the teams led by Yves Keravel in Creteil Mondor Hospital [118], Marc Sindou in Lyon [119] and Yves Lazorthes in Toulouse [120].

The Introduction of Robotics and the Emergence of Neuronavigation in Stereotactic Surgery The quest for precision, reliability and safety never ends and takes advantage of all the available means. The logical extention of stereotaxy, increasingly based on numerical data provided by modern imaging, is to use these digital data to visualize targets on the MRI and CTscan modalities, to provide the possibility to draw trajectories and to plan procedures: this is what neuronavigation has progressively achieved in parallel to the increasing power of computers. The existence in Grenoble of industrial developments in the field of robotized multipurpose tools has naturally led to the concept, design and construction of a robotized stereotactic arm (Neuromate1 Schaerer-Mayfield, see chapter 37 in this

Handbook) [121] following the approaches of Kelly (a former fellow of J Talairach) [122] and of Ronald Young [123], aimed at replacing the grids, arch and goniometers of the stereotactic frame (ref Robots). In the same line of thought, a robotized microscope holder (Surgiscope1, ISIS) has been designed and built. These systems have been installed in several centers and are being currently routinely used and improved upon.

The Development of New Applications of Deep Brain Stimulation at High Frequency and the Rebirth of Psychosurgery in France New Applications of HFS Since the success of HFS in the treatment of Parkinson’s diseases, it has been extended to other targets (Vim, Pallidum, STN, CM-Pf) and also to other movement disorders (essential tremor, dystonia, tardive dystonia), based on the example of STN, which was derived from basic science. Other teams, in Italy, Belgium, and the Netherlands have successfully extended it to cluster headaches, obsessive compulsive disorders (OCD), and Tics de Gilles de la Tourette. In France, the fortuitous observation of the improvement by STN HFS of OCD symptoms in parkinsonian patients [124], led to the initiation of a multicenter clinical trial which has been successful in establishing the limbic portion of STN as a specific target for OCD treatments. This has been possible after the approval of the National Ethics Committee, an appeal to which was filed in December 2000 and which gave conditional approval in June 2002, for a carefully controlled clinical multicentric trial, similar to what happened in movement disorders, due to the growing interest in new indications. Psychosurgery had been almost stopped since the early 1970s in France, as a consequence of


international pressure in the face of the abuses and complications, mainly lobotomies, even when stereotactic leucotomies were considered. There was in France very reduced psychosurgical stereotaxic activity in parallel to the development of the biological psychiatry and psychotherapy. There is now a new trend to return to psychosurgery for Tics and for depression, but the French neurosurgical community is very careful about the risks of jeopardizing one more time the possibility to help mental disorders if mistakes are made again.

Coma and Minimally Conscious States François Cohadon (1993) in Bordeaux, took a very careful and cautious approach to the problem of persisting vegetative post-traumatic states [125,126], at the same time as Tsubokawa [127,128] in Japan. In 25 cases of post traumatic vegetative state persisting 3 months after the initial injury, deep brain stimulation at low frequency has been used to activate the cortex with the hope of producing some degree of functional recovery. Electrodes were stereotactically implanted in the centrum medianum-parafascicularis complex. Bipolar stimulation was provided daily from 8 a.m. to 8 p.m. In 12 cases no changes occurred in the clinical features and overall behavior. DBS was given up after 2 months. All these patients with a follow-up of 1–10 years, remained in a permanent vegetative state, four of them eventually died. In 13 cases, following 1–3 weeks of DBS, a definite improvement was obtained with the recovery of some degree of consciousness and interpersonal relationship. Although some degree of long-term spontaneous recovery was documented, it was concluded that there was no significant effect. The reason why Cohadon, as well as Tubokawa, did not observe the results reported by Schiff et al. might be due to the fact that their patients were truly in vegetative states and not minimally conscious.

9

Conclusion French stereotaxy has reached a very internationally competitive level. Although one might consider that the first stereotactic frame was invented and used in 1897, stereotaxy seemed to start in the 1950s. Through two giant pioneers, it introduced multidisciplinarity as a philosophy. The group of Sante Anne created a whole monument of stereotactic discipline, which was wisely applied to several fields and created several approaches and strategies, developed atlases and frames, and inspired many young surgeons and students. French stereotaxy has remained very innovative since then. This was possible because stereotactic functional neurosurgery was and remained almost exclusively academic, was multidisciplinary with the primary goal of excellence, strong ethical concern and patient’s care, free from financial goals and economic considerations, as the French social security system did not demand economic criteria (length of stay, cost-benefit).

References 1. Horsley V, Clarke RH. The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45–124. 2. Remy M, Contremoulins M. Le chercheur de projectiles, L’Illustration1897;55423 3. Zernov DN. Encephalometer: device for estimation of parts of brain in human. Proc Soc Physicomed Moscow Univ 1889;2:70-80. 4. Altukhov NV. Encephalometric investigations of the brain relative to the age, sex and skull indexes. Moscow; 1891. 5. Picard C, Olivier A, Bertrand G. The first human stereotaxic apparatus: The contribution of Aubrey Mussen to the field of stereotaxis. J Neurosurg 1983;59:673-6. 6. Olivier A, Bertrand G, Picard C. Discovery of the first human stereotactic instrument. Appl Neurophysiol 1983;46:84-91. 7. Spiegel EA, Wycis HT, Marks M, et al. Stereotactic apparatus for operations on the human brain. Science 1947;106:349-50.

107

108

9


8. Sedan R, Duparet R. Stereómetre adaptable au cadre ste´reótaxique de J Talairach. Neurochirurgie 1968;14: 577-82. 9. Fenelon F, Thiebaut F. Re´sultats du traitement neurochirurgical d’une rigidite´ parkinsonienne par intervention striopallidale unilate´rale. Rev Neurol 1950;83:280. 10. Fenelon F. Essai de traitement neurochirurgical du syndrome parkinsonien par intervention directe sur les voies extrapyramidales imme´diatement sous-striopallidales (anses lenticulaires). Rev Neurol 1950;83:580-5. 11. Guiot G Brion S.“[Neurosurgery of choreoathetosic and Parkinsonian syndromes].” Sem Hop 1952;28 (49):2095-9. 12. Guiot G, Hardy J, Albe-Fessard D. De´limitation prećise des structures sous-corticales et identification de noyaux thalamiques chez l’homme par l’e´lectrophysiologie ste´reótaxique. Neurochirurgia 1962;5:1-18 13. Guiot G, Derome P, Trigo JC. Le tremblement d’attitude: indication la meilleure de la chirurgie ste´reotaxique. Presse Med 1967;75:2513-8 14. Petit-Dutaillis, D., MessimyR, et al. “[Effect of prefrontal leucotomy on the behaviour of a maniac].” Rev Neurol (Paris) 1950;82(1):57-60. 15. Gros C,Vlahovitch B.“[Hemispherectomies; neurophysiological considerations].” Rev Neurol (Paris) 1951;85 (6):482-4. 16. Gros C, Vlahovitch B, et al. “[Effects of prefrontal interventions (leukotomy, topectomy) on psychic and kinetic manifestations of Parkinson syndrome].” Sem Hop 1955;31(32):1831-2. 17. Mazars G, Roge´ R, Mazars Y. Re´sultats de la stimulation du faisceau spino-thalamique et leur incidence sur la physiopathologie de la douleur. Revue Neurol 1960;20:136-8. 18. Mazars G, Merienne L, Cioloca G. Stimulations thalamiques intermittentes antalgiques. Revue Neurol 1973;128(4):273-9 19. Mazars G, Merienne L, et al. “[Use of thalamic stimulators in the treatment of various types of pain].” Ann Med Interne (Paris) 1975;126(12):869-71. 20. Mazars G. Intermittent stimulation of nucleus ventralis posterolateralis for intractable pain. Surg Neurol 1975;4:93-5. 21. Mazars G, Merienne L, et al. “[Present state of pain surgery].” Neurochirurgie 1976;22 Suppl 1:5-164. 22. Mazars G, Merienne L, et al. “[Effects of so-called periaqueductal gray substance stimulation].” Neurochirurgie 1979;25(2):96-100. 23. Mazars G, Merienne L, Cioloca G. Control of dyskinesias due to sensory deafferentation by means of thalamic stimulation. Acta Neurochirurgica Suppl (Wien) 1980;30:239-43. 24. Merienne L, Mazars G. Modification du sche´ma corporel sous stimulation thalamique. A propos des stimulations thalamiques pour membre fantoˆmes douloureux. Neurochirurgie 1981;27:121-3.

25. Fenelon F. “[Neurosurgery of parkinsonian syndrome by direct intervention on the extrapyramidal tracts immediately below the lenticular nucleus. Communication followed by film showing patient before and after intervention].” Rev Neurol (Paris) 1950;83(5):437-40. 26. Fenelon F. Bilan de quatre anneés de pratique d’une intervention personnelle pour maladie de Parkinson. Rev Neurol 1953;89:580-5. 27. Guiot G, Brion S. Traitement des mouvements anormaux pour la coagulation pallidale. Technique et re´sultats. Rev Neurol 1953;89:578-80. 28. Gros C, Vlahovitch B, et al. “[the “Target” Zone in the Surgery of Parkinson’s Disease (Study of a Series of 407 Stereotaxic Operations)].” Neurochirurgie 1964;10:413-26. 29. Talairach J, Tournoux P.“[Apparatus for hypophysial stereotaxis by nasal approach].” Neurochirurgie 1955;1(1):127-31. 30. Szikla G, Talairach J.“Coordinates of the rolandic sulcus and topography of cortical and subcortical motor responses to low frequency stimulation in a proportional stereotaxic system.” Confin Neurol 1965;26(3):471-5. 31. Szikla G, Bouvicr G, Hori T, Petrov V. Angiography of the human brain cortex. New York: Springer; 1977. 32. Talairach J, David M, Tournoux P, et al. Atlas d’anatomie stereotaxique des noyaux gris centraux. Paris: Masson; 1957. 33. Talairach J, Szikla G, Tournoux P, et al. Atlas d’anatomie stereotaxique du telencephale. Paris: Masson; 1967. 34. Talairach J, Tournoux P. Coplanar stereotaxic atlas of the human brain: 3-dimensional proportional system: an approach to cerebral imaging. Stuttgart: Thieme Medical Publishers; 1988. 35. Albe-Fessard D, Arfel G, Guiot G, et al. De´rivation d’activite´s spontaneés et e´voqueés dans les structures ce´re´brales profondes de l’homme. Rev Neurol 1962;2:89-105. 36. Albe-Fessard D, Arfel G, Guiot G. Activite´s électriques caracte´ristiques de quelques structures ce´re´brales chez l’homme. Ann Chir 1963;17:1185-214. 37. Talairach J, Hecaen H, David M, et al. Recherches sur la coagulation the´rapeutique des structures sous-corticales chez l’homme. Rev Neurol 1949;81:4-24. 38. Talairach J, Tournoux P, Szikla G, et al. Traitement de la maladie de Parkinson: remarques me´thodologiques et technique chirurgicale. Gaz Med 1961;2:76-91. 39. Benabid A, Pollak P, Louveau A, et al. Combined (thalamotomy and stimulation) stereotactic surgery of the VIM thalamic nucleus for bilateral Parkinson’s disease. Appl Neurophysiol 1987;50:344-6. 40. Benabid A, Pollak P, Gervason C, et al. Long term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991;337:403-6. 41. Benabid A, Pollak P, Gao D, et al. Long term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. J Neurosurg 1996;84:203-14. 42. Blond S, Caparros-Lefebvre D, Parker F. Control of tremor and involuntary movement disorders by chronic


43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54. 55.

56.

57. 58.

stereotactic stimulation of the ventral intermediate thalamic nucleus. J Neurosurg 1992;77:62-8. Caparros-Lefebvre D, Blond S, Vermersch P, et al. Chronic thalamic stimulation improves tremor and levodopa induceddyskinesias in Parkinson’s disease. J Neurol Neurosurg Psych 1993;56:268-73. Limousin P, Pollak P, Benazzouz A, et al. Effect on Parkinsonian signs and symptoms of bilateral subthalamic nucleus stimulation.Lancet 1995;345(8942):91-5. Siegfried J. Effets de la stimulation du noyau sensitif du thalamus sur les dyskine´sies et la spasticite´. Rev Neurol (Paris) 1986;142:380-3. Siegfried J, Lippitz B. Chronic electrical stimulation of the globus pallidus in Parkinson’s disease. Acta Neurochir 1993;124:14-18. Tasker RR. Effets sensitifs et moteurs de la stimulation thalamique chez l’homme. Applications cliniques. Rev Neurol (Paris) 1986;142:316-26. Adams J, Hosobuchi Y, Fields H. Stimulation of internal capsule for relief of chronic pain. J Neurosurg 1974;41:740-4. Hosobuchi Y, Adams J, Rutkins B. Chronic thalamic stimulation for the control of facial anesthesia dolorosa. Arch Neurol 1973;29:158-61. Talairach J, deAJ, et al. “[Therapeutic sub-cortical coagulations; topographic study of the ventricular system in relation to the central gray nuclei].” Presse Med 1950;58 (38):697-701. Talairach J, De Aljuriaguerra J, et al. “[A stereotaxic study of the deep encephalic structures in man; technic; physiopathologic and therapeutic significance.].” Presse Med 1952;60(28):605-9. Talairach J. “[Destruction of the anterior ventral thalamic nucleus in the treatment of mental diseases].” Rev Neurol (Paris) 1952;87(4):352-7. Talairach J, Tournoux P.“[Stereotaxic localization of central gray nuclei].” Neurochirurgia (Stuttg) 1958;1(1): 88-93. Talairach J, Tournoux P, et al. “[Parietal surgery of pain].” Acta Neurochir (Wien) 1960;8:153-250. Talairach J, Ruggiero G, et al. “A new method of treatment of inoperable brain tumours by stereotaxic implantation of radioactive gold; a preliminary report.” Br J Radiol 1955;28(326):62-74. Talairach J, Szikla G, et al. “[Stereotaxic destruction of the nontumoral hypophysis by radioactive isotopes. (Hypophysectomy for hormone-dependent cancer. Interstitial radiotherapy for malignant exphthalmos and Cushing’s syndrome of high origin)].” Presse Med 1962;70:1399-402. Talairach J, Szikla G, et al. “[Pituitary sterotaxic surgery].” Confin Neurol 1962;22:204-16. Talairach J, Bonis A, et al. “[Technic and results of interstitial radiotherapy of the pituitary in various endocrine syndromes: chromophobe adenoma, acromegaly and Cushing’s disease].” Acta Isot (Padova) 1964;4(4):355-85.

9

59. Talairach J, Szikla G.“[Amygdalo-hippocampal partial destruction by yttrium-90 in the treatment of certain epilepsies of rhinencephalic manifestation].” Neurochirurgie 1965;11(3):233-40. 60. Talairach J Szikla G.“[Intrasellar application of radioactive isotopes from a functional standpoint (mammary carcinoma, malignant, exophthalmos, Cushing’s syndrome, diabetes, malignant arterial hypertension)].” Nucl Med (Stuttg) 1965;:2:Suppl 371+. 61. Talairach J, Szikla G, et al. “Therapeutic utilization of radioactive isotopes in pituitary surgery.” Int J Neurol 1965;5(1):78-93. 62. Betti O, Derechinsky V.“[Multiple-beam stereotaxic irradiation].” Neurochirurgie 1983;29(4):295-8. 63. Bancaud J, Talairach J.“[Epilepsy of the supplementary motor area: a particularly difficult diagnosis in children].” Rev Neuropsychiatr Infant 1965;13(6):483-99. 64. Bancaud J, Talairach J.“[Methodology of stereo EEG exploration and surgical intervention in epilepsy].” Rev Otoneuroophtalmol 1973;45(4):315-28. 65. Talairach J, Bancaud J, et al. “[New approach to the neurosurgery of epilepsy. Stereotaxic methodology and therapeutic results. 1. Introduction and history].” Neurochirurgie; 19741:20 Suppl 1-240. 66. Talairach J, Szikla G.“Application of stereotactic concepts to the surgery of epilepsy.” Acta Neurochir Suppl (Wien) 1980;30:35-54. 67. Talairach J, Tournoux P, et al. “Stereotaxic exploration in frontal epilepsy.” Adv Neurol 1992;57:651-88. 68. Talairach J, Bancaud J, et al. “Surgical therapy for frontal epilepsies.” Adv Neurol 1992;57:707-32. 69. Mondragon S, Lamarche M.“Suppression of motor seizures after specific thalamotomy in chronic epileptic monkeys.” Epilepsy Res 1990;5(2):137-45. 70. Benabid AL, Minotti L, Koudsie A, de Saint Martin A, Hirsch E. Antiepileptic effect of high-frequency stimulation of the subthalamic nucleus (corpus luysi) in a case of medically intractable epilepsy caused by focal dysplasia: a 30-month follow-up: technical case report. Neurosurgery 2002;50(6):1385-91; discussion 1391-92. 71. Chabardes S, Kahane P, Minotti L, Koudsie A, Hirsch E, Benabid AL. Deep brain stimulation in epilepsy with particular reference to the subthalamic nucleus. Epileptic Disord 2002;4 Suppl 3:83-93. 72. MazarsMazars G. “[Surgical treatment of Parkinson’s disease].” Gaz Med Fr 1959;66(12):1063-4. 73. Mazars G, Droguet P, et al. “[Volitional dyskinesia of posture. Disappearance of the dyskinesia after thalamic coagulation].” Rev Prat 1960;103:73-4. 74. Mazars G, Pansini A, et al. “[Homolateral responses caused by stimulation of the radiating crown and of the internal capsule].” Rev Prat 1960;103:134-6. 75. Mazars G, Pansini A, et al. “[Contribution to the knowledge of the relations between the lesions produced by surgery of dyskinesias and the principal motor tract].” Rev Prat 1961;104:433-6.

109

110

9


76. Mazars G, Chodkiewicz JP, et al. “[The cephalogyric contingent of the cortico-capsular fibers].” Rev Neurol (Paris) 1965;112(6):553-7. 77. Mazars G, Chodkiewicz JP, et al. “[Nosology and therapeutic indications in cervico-cephalic dyskinesias].” Rev Neurol (Paris) 1967;116(5):441-3. 78. Mazars G, Merienne L, et al. “[Surgery of cephalic orientation dyskinesia].” Neurochirurgie 1968;14(6):745-52. 79. Mazars G, Merienne L, et al. “[Dyskinesias of the upper limb associated with dyskinesias of cephalic orientation. Study of 45 cases and presentation of a film].” Rev Neurol (Paris) 1970;122(4):275-7. 80. Merienne L, Mazars G. Traitement de certaines dyskine´sies par stimulation thalamique intermittente. Neurochirurgie 1982;28:201-6. 81. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger J. Physiological observations in stereotaxic operations inextrapyramidal motor disturbances. Brain 1960;83:337-51. 82. Svennilson E, Torvik A, Lowe R, et al. Treatment of parkinsonism by stereotactic thermolesions in the pallidal region. Acta Psychiatr Neurol Scand 1960;35:359-77. 83. Kelly PJ, Gillingham FJ. The long-term results of stereotacticsurgery and L-dopa therapy in patients with Parkinson’s disease. A 10-year follow-up study. J Neurosurg 1980;53:332-7. 84. Cooper IS, Bravo GI. Chemopallidectomy, and chemothalamectomy. J Neurosurg 1958;15:244-56. 85. Cooper IS. Surgical treatment of parkinsonism. Ann Rev Med 1965;16:309-30. 86. Narabayashi H, Okuma T. Procaine oil blocking of the globuspallidus for treatment of rigidity and tremor of parkinsonism:preliminary report. Proc Jpn Acad 1953;29:134. 87. Narabayashi H, Yokoshi F, Nakajiina Y. Levodopa induced dyskinesia and thalamotomy. J Neurol Neurosurg Psychiatr 1984;47:831-9. 88. Ohye C, Maeda T, Narabayashi H. Physiologically defined VIM nucleus: its special reference to control of tremor. AppliedNeurophysiol 1976;39:285-95. 89. Backlund E, Granberg P, Hamberger B, et al. Transplantation of adrenal medullary tissue to striatum in parkinsonism. First clinical trials. J Neurosurg 1985;62: 169-73. 90. Bjo¨rklund A, Stenevi U. Reconstruction of the nigrostriatal dopamine pathway by intracerebral nigral transplants. Brain Res 1979;177:555-60. 91. Bjo¨rklund A, Stenevi U, Dunnett S. Functional reactivation the deafferented neostriatum by nigral transplant. Nature 1981;289:497-9. 92. Bjorklund A, Dunnett SB, et al. “Neural transplantation for the treatment of Parkinson’s disease.” Lancet Neurol 2003;2(7):437-45. 93. Peschanski’s team Bachoud-Levi AC, Hantraye P, et al. “Prospectives for cell and gene therapy in Huntington’s disease.” Prog Brain Res 1998;117:511-24.

94. Bachoud-Levi AC, Remy P, et al. “Motor and cognitive improvements in patients with Huntington’s disease after neural transplantation.” Lancet 2000;356(9246):1975-9. 95. Bachoud-Levi AC, DeglonN, et al. “Neuroprotective gene therapy for Huntington’s disease using a polymer encapsulated BHK cell line engineered to secrete human CNTF.” Hum Gene Ther 2000;11(12):1723-9. 96. Bachoud-Levi AC, Gaura V, et al. “Effect of fetal neural transplants in patients with Huntington’s disease 6 years after surgery: a long-term follow-up study.” Lancet Neurol 2006;5(4):303-9. 97. Peschanski M, Rudin M, et al. “Magnetic resonance imaging of intracerebral neural grafts.” Prog Brain Res 1988;78:619-24. 98. Peschanski M, Defer G, et al. “Bilateral motor improvement and alteration of L-dopa effect in two patients with Parkinson’s disease following intrastriatal transplantation of foetal ventral mesencephalon.” Brain 1994;117 (Pt 3): 487-99. 99. Peschanski M. “The breaking of embargoes.” Lancet 2001;357(9260):963. 100. Peschanski M, Dunnett SB.“Cell therapy for Huntington’s disease, the next step forward.” Lancet Neurol 2002;1(2):81. 101. Peschanski M, Bachoud-Levi AC, et al. “Integrating fetal neural transplants into a therapeutic strategy: the example of Huntington’s disease.” Brain 2004;127 (Pt 6):1219-28. 102. Peschanski M. “[Stem cells, time for scale-up].” Med Sci (Paris) 2008;24(4):335-7. 103. Lefaucheur JP, Menard-Lefaucheur I, et al. “Electrophysiological deterioration over time in patients with Huntington’s disease.” Mov Disord 2006;21(9):1350-4. 104. Levivier M, Dethy S, et al. “Intracerebral transplantation of fetal ventral mesencephalon for patients with advanced Parkinson’s disease. Methodology and 6-month to 1-year follow-up in 3 patients.” Stereotact Funct Neurosurg 1997;69(1–4 Pt 2):99-111. 105. Mitjavila-Garcia MT, Simonin C, et al. “Embryonic stem cells: meeting the needs for cell therapy.” Adv Drug Deliv Rev 2005;57(13):1935-43. 106. Palfi S, Ferrante RJ, et al. “Chronic 3-nitropropionic acid treatment in baboons replicates the cognitive and motor deficits of Huntington’s disease.” J Neurosci 1996;16 (9):3019-25. 107. Peltekian E, Parrish E, et al. “Adenovirus-mediated gene transfer to the brain: methodological assessment.” J Neurosci Methods 1997;71(1):77-84. 108. Cesaro P, Peschanski M, et al. “Treatment of Parkinson’s disease by cell transplantation.” Funct Neurol 2001;16 (1):21-7. 109. Cochen V, Ribeiro MJ, et al. “Transplantation in Parkinson’s disease: PET changes correlate with the amount of grafted tissue.” Mov Disord 2003;18(8):928-32. 110. Bloch J, Bachoud-Levi AC, et al. “Neuroprotective gene therapy for Huntington’s disease, using


111.

112.

113.

114.

115.

116.

117.

118.

polymer-encapsulated cells engineered to secrete human ciliary neurotrophic factor: results of a phase I study.” Hum Gene Ther 2004;15(10):968-75. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM, Hommel M, et al. Long-term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991;337(8738): 403-6. Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann D, et al. Electrical stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med;1998;339(16):1105-11. Krack P, Batir A, Van Blercom N, Chabardes S, Fraix V, Ardouin C, et al. Five years follow-up of bilateral stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 2003;349:1925-34. Gros C, Vlahovitch B, et al. “[Personal experience with surgery of pain in cancer].” J Radiol Electrol Arch Electr Medicale 1955;36(7–8):618-20. Blond S, Siegfried J.“Thalamic stimulation for the treatment of tremor and other movement disorders.” Acta Neurochir Suppl (Wien) 1991;52:109-11. Tsubokawa T, Katayama Y, et al. “Chronic motor cortex stimulation for the treatment of central pain.” Acta Neurochir Suppl (Wien) 1991;52:137-9. Nguyen JP, Lefaucher JP, Le Guerinel C, Eizenbaum JF, Nakano N, Carpentier A, et al. “Motor cortex stimulation in the treatment of central and neuropathic pain.” Arch Med Res 2000;31(3):263-5.

9

119. Garcia-Larrea L, Peyron R. Motor cortex stimulation for neuropathic pain: From phenomenology to mechanisms.” Neuroimage 2007;37 Suppl 1:S71-9. 120. Lazorthes Y, Sol JC, Fowo S, Roux FE, Verdie JC. “Motor cortex stimulation for neuropathic pain.” Acta Neurochir Suppl 2007;97(Pt 2):37-44. 121. Benabid AL, Hoffmann D, Lavallee S, Cinquin P, Demongeot J, Lebas JF, et al. Is there any future for robots in neurosurgery? In: SymonL, editors. Advances and Technical Standards in Neurosurgery, vol. 18. Springer: Wien New York; p. 3-45 122. Kelly PJ, Alker GJ. A method for stereotactic laser microsurgery in the treatment of deep seated CNS neoplasms. Appl Neurophysiol 1980;43:210-15. 123. Young RJ. Application of robotics to stereotactic neurosurgery. Neurol Res 1887;9:123-8. 124. Mallet L, Mesnage V, et al. “Compulsions, Parkinson’s disease, and stimulation.” Lancet 2002;360(9342):1302-4. 125. Cohadon F,Richer E. “[Post-traumatic vegetative states].” Neurochirurgie 1993;39(5):269-80. 126. Cohadon F, Richer E. “[Deep cerebral stimulation in patients with post-traumatic vegetative state. 25 cases].” Neurochirurgie 1993;39(5):281-92. 127. Tsubokawa T, Yamamoto T, et al. “Deep-brain stimulation in a persistent vegetative state: follow-up results and criteria for selection of candidates.” Brain Inj 1990;4 (4):315-27. 128. Yamamoto T, Katayama Y, et al. “Deep brain stimulation therapy for a persistent vegetative state.” Acta Neurochir Suppl 2002;79:79-82.

111

4 History of Stereotactic Surgery in Germany J. K. Krauss

Stereotactic and functional neurosurgery has a long and rich history in Germany [1,2]. At the time functional and stereotactic surgery was introduced in clinical practice, medical progress had slowed down considerably during the post-World War II depression era. In this article, a brief overview on the pre-war development of neurosurgery in Germany with a special emphasis on issues relevant to functional neurosurgery, followed by a more thorough review on the birth of functional neurosurgery in Freiburg and its further achievements are discussed. The account is based on original historical documents, previous historical reviews and a personal interview, and the reader is referred to those documents for a more complete bibliography on this topic [1,3–6].

Development of Neurosurgery in Germany Ernst von Bergmann was the first German surgeon who devoted his career to surgery of the nervous system [7]. He became professor of surgery in 1878, in Wu¨rzburg, and he moved to the Charite´ in Berlin, in 1882. He published several books focussing on neurosurgical topics. When he died in 1907, several of his pupils continued to carry on his interests. The two major driving forces for the further development ultimately were Fedor Krause and Otfrid Fo¨rster [8,9]. Fedor Krause developed his career first in Hamburg, and later in Berlin. In February, 1892, he explored the trigeminal nerve via an extradural approach for treatment of trigeminal #


pain. He was active in many other aspects of neurosurgery and was highly esteemed when he died in 1937. Otfrid Fo¨rster, who was 18 years younger than Krause, was the first to pioneer the concept that neurosurgery should be an independent discipline in Germany. Fo¨rster was trained as a neurologist and a neurophysiologist, but he was rather an autodidact in surgery. He became well known during the 1920s and had contacts with Bailey, Bucy, Penfield, and Cushing. Fo¨rster was invited to Moscow after Lenin suffered from an apoplectic insult. When Lenin died, it was he who signed the protocol of the autopsy reporting hemorrhage from an aneurysm of the medial cerebral artery. Fo¨rster furthered many innovations in neurosurgery at his institute in Breslau. He was the first to pioneer dorsal rhizotomy for treatment of spasticity, and cordotomy for treatment of pain syndromes. In 1937, when the sky was clouded already by the reign of the Nazi regime and the international reputation of Germany became more and more overshadowed, he still managed to organize a German-British meeting of neurosurgeons in Berlin and became a honorary member of the British neurosurgical society [9]. He died in 1940 at the age of 67. It was Wilhelm To¨nnis who became the first chair of an ‘‘extraordinariate’’ in 1937, in Berlin [10]. Although he is mostly known for his work on trauma surgery, he established fruitful contacts with other neurodisciplines including new concepts such as neuropathology. Spatz and Hallervorden were the first to describe more extensively the disorder which carried their names for decades and was renamed recently in neurodegeneration with brain iron accumulation [11].

54

4

History of stereotactic surgery in germany

The development of neurosurgery in Germany came to a sudden halt when World War II began [10]. When the war was over, neurosurgery soon became an advanced discipline, during the reorganization of healthcare, not devoted mainly to trauma anymore.

Freiburg: the Early Epicentre of German Functional and Stereotactic Neurosurgery Traugott Riechert, born in 1905, was a pupil of To¨nnis who had worked on brain injury during the Second World War. He was the second in Germany to become professor of neurosurgery on June 27, 1946, in Freiburg. At that time, psychiatry and neurology in Freiburg were united under the leadership of the psychiatrist Behringer. Richard Jung, who later became the head of the department of neurology, and Rolf Hassler, who established a laboratory for neuropathology, were members of the faculty. Hassler had come to Freiburg from Berlin after the war was over. He was a pupil of the internationally renowned scientists, Cecile and Oscar Vogt, who had to leave Berlin when the nazis had taken over, to be replaced by Spatz. The Vogts were actually the first to forge the term ‘‘extrapyramidal system.’’ When they left Berlin to move to the Black Forest, they brought with them a large collection of brains (among them was Lenin’s brain) and brain slices which was one of the reason that Rolf Hassler was also attracted to come to Freiburg. When Riechert accepted the call to Freiburg in 1946, he had no rooms available in his department [6]. Therefore, he was housed initially in an attic of the psychiatry department. In the late 1947, the department of neurosurgery assumed its operative activities after Watts had visited Freiburg to demonstrate the technique of Freeman and Watts to perform transorbital leucotomy. This procedure was performed on a few patients in Freiburg and interestingly treated intractable pain also, but it

never became popular. Riechert made rounds in the psychiatry ward, and every Saturday there was a big conference attended by the neurologists and psychiatrists. The first stereotactic frame in Freiburg was developed during this period by Riechert together with the physicist, Wolff [12]. It was quite inaccurate, but it was the prototype for the second model which became to be known as the Riechert-Mundiger frame [13]. When Richard Jung became the interim chair of the department of psychiatry after the death of Behringer, he asked the residents whether one of them would like to join the neurosurgeons [9]. Fritz Mundinger volunteered, and he started his career in neurosurgery in 1951. At that time Hassler had become the head of the neuropathology laboratory. Mundinger first concentrated on the improvement of the accuracy of the stereotactic frame and every afternoon, if possible, he met with the technician F. L. Fischer of the Fischer Company to develop the device further into the shape it later became known and still exists today [13]. Functional stereotactic neurosurgery in Freiburg started with pallidotomies [14]. While it became recognized that rigidity responded quite well to pallidotomy, it was also noted that tremor was ameliorated to a much lesser extent. Also, with the early lesioning systems, hemorrhages were not infrequent and they were thought to be partially related to the vascular supply of the pallidum. By that time Hassler was already known as ‘‘the expert of the thalamus’’ worldwide. He had managed to establish his own nomenclature dividing the thalamus into 34 subnuclei based on their cytoarchitectonic structures and their cortical and subcortical connections [3]. His classification was already published in his habilitation thesis in 1948. It was he who suggested to target the thalamus based on his neuroanatomic concepts, and also to stay out of trouble with the branches of the anterior choroidal artery during pallidotomy. He argued that it would be useful not only to target the pallidothalamic system, but also the dentatothalamic system and their projections to


the cortex. He thought that the ventralis oralis anterior nucleus would be a good target for rigidity, and hence he suggested the ventralis oralis posterior nucleus as an ideal target for tremor. The basic problem, however, was to identify and find the nuclei. The time had come to introduce thalamotomy for treatment of movement disorders. The first thalamotomy ever was actually performed on November 14, 1952, in Freiburg [15]. The patient was a 38-year-old man with postencephalitic parkinsonism with prominent tremor. The target coordinates were determined by comparing the pneumencephalography X-rays, which were made several days prior to the operation with 5 mm brain slices that Hassler had prepared. The algorithm for target calculation taking into account cranio-cerebral correlations and X-ray distortion was quite complicated. The operation was performed by Mundinger in the presence of Riechert, Hassler, and von Baumgarten. The immediate result was that the tremor stopped and the response of both, the patient and the physicians, was overwhelming. The results were published two years later in 1954, by Hassler and Riechert, in German, in ‘‘Der Nervenarzt’’ [15]. Over the next few years, pallidotomy was replaced step-by-step by thalamotomy in Freiburg. Subsequently, Freiburg became the first major center for stereotactic and functional neurosurgery in Germany. In 1960, thousands of functional stereotactic procedures had been performed in the department [16,17]. Over the next few years, the department became extremely busy and up to five thalamotomies for Parkinson’s disease were performed each day. Also, the indications were expanded and other movement disorders such as dystonia, myoclonus, hemiballism, and tics were successfully treated by thalamotomies [18–22]. In the early 1960s, the zona incerta was introduced as a target for ablative surgery [23]. With the development of temperature control and radiofrequency lesioning, lesions could now be made much smaller and side effects were significantly

4

reduced [24,25]. The effects of both low frequency and high frequency stimulation with electric current via the radiofrequency probe were studied systematically as early as in the late 1950s [26,27]. Another early development in Freiburg was the use of radionuclides to treat brain tumors [6]. In 1951, the first radionuclide used was phosphor 32. Later the most frequently used radionuclides were iodine 125 and iridium 192. During the 1950s, stereotactic implantation of radionuclides was used in the treatment of movement disorders in 20 patients. In 1963, Mundinger developed the ‘‘Gammamed’’ for brachyradiotherapy [4]. This machine was constructed for high-dose radiation and it was loaded with iridium 192. Radiation therapy was delivered by stereotactic techniques.

Spread Over Germany While the department of neurosurgery in Freiburg grew and became larger, many visitors and several guests from Germany and other countries (> Figure 4-1) were trained there in functional stereotactic neurosurgery. Many of them spread the technique to other cities, were they worked in the frame of general neurosurgery or in specialized subunits. Some of them were: Dieckmann in Homburg, Umbach in Berlin, Spuler in Wu¨rzburg, Mu¨ller in Hamburg, Thomalske in Frankfurt, Nittner in Cologne, and Yasargil in Zurich [28]. Independent departments for stereotactic and functional neurosurgery were built in Freiburg, Homburg, Frankfurt, and Wu¨rzburg. Many of these neurosurgeons modified the stereotactic techniques they had learned in Freiburg and expanded the indications. Struppler, who became active over decades in Munich, actually was trained as a neurologist. When Germany was divided into two countries, there was little cross-border exchange due to the restrictions of the cold war. Nevertheless, Goldhahn managed to establish functional stereotactic neurosurgery in Leipzig,

55

56

4


. Figure 4-1 Mundlinger (second from left) and Riechert (third from left) at the entrance of the building which housed the neurosurgical department in Freiburg in the 1960s together with guests from Russia including Ugrumov (third from right)

East Germany (formerly communist ‘‘Deutsche Demokratische Republik’’) [29]. The first major atlas for stereotactic surgery of the human brain ‘‘Stereotaktische Operationen/ Stereotaxis’’ was edited by Georges Schaltenbrand from Wu¨rzburg and by Percival Bailey from Chicago [30]. It was published in three volumes by Georg Thieme Verlag, Stuttgart, in 1959. This atlas had a major impact on the further development of functional stereotactic surgery. Strong input into the atlas was supplied by Hassler, Riechert, and Mundinger, and its development was also supported by Leksell. Twenty years later its second edition ‘‘The Atlas for Stereotaxy of the Human Brain’’ was published, this time edited by Georges Schaltenbrand and Waldemar Wahren

[31]. While the first edition ‘‘The Schaltenbrand/ Bailey Atlas’’ was bilingual – German and English, the second edition ‘‘The Schaltenbrand/Wahren Atlas’’ was only in English. Tragically, after writing the foreword to the accompanying textbook edited by Schaltenbrand and Walker [32], Schaltenbrand suffered a stroke and died before the second edition of his monumental work went to press. While there was a strong focus on functional stereotactic neurosurgery for treatment of movement disorders in Germany, other classical indications like pain, psychoaffective disorders, and epilepsy were developed further. As indicated above, classical leucotomy did not have a wide distribution in Germany. Also, after the introduction of stereotactic techniques to target subcortical


4

structures for psychoaffective disorders, psychosurgery was rather limited and such procedures were confined to six centers [33]. As in other countries, psychosurgery came to a complete halt in the 1970s, when it came to the focus of public attention and provoked overwhelmingly unwanted responses in the media.

Charite in ‘‘East’’ Berlin [36]. The revival of movement disorders surgery in Germany was closely related to the introduction of deep brain stimulation [37]. The base for treatment of movement disorders is now widely distributed all over Germany with major centers in Freiburg, Cologne, Berlin, Magdeburg, Kiel and Hannover.

Silent Years and Revival

References

Birkmayer and Hornykiewicz had already published their observations on levodopa in parkinsonism in 1961, in a German journal [34]. After the general introduction of levodopa in clinical practice in the late 1960s, the number of functional stereotactic procedures for treatment of Parkinson’s disease sharply dropped, and over the years many centers completely stopped their activities. Nevertheless, some institutions still remained active. Even when thalamotomy had become completely out of public focus, about 100 thalamotomies were performed each year for treatment of tremor or Parkinson’s disease. When I joined Mundinger in the department of functional and stereotactic neurosurgery as a resident, in Freiburg, in 1988, I had the unique opportunity to learn this procedure. The academic interest in functional and stereotactic neurosurgery was always strong in Germany. During the 1960s and the 1970s networks with colleagues from other European countries became more and more apparent. As detailed in chapter A2 the founding meeting of the European Society of Stereotactic and Functional Neurosurgery was in Freiburg in 1970. While pallidotomy became a major procedure for treatment of movement disorders in the United States and in several European countries after Laitinen had raised it from oblivion [35], it became never really popular in Germany. Also, partially for legal reasons neurotransplantation never had a more widespread application in the 1980s and 1990s with the exception of some patients having been operated on by Vogel in the

1. Krauss JK, Grossmann RG. Historical review of pallidal surgery for treatment of parkinsonism and other movement disorders. In: Krauss JK, Grossmann RG, Jankovic J, editors. Pallidal surgery for the treatment of Parkinson’s disease and movement disorders. Philadelphia: LippincottRaven; 1998. p. 1-23. 2. Sturm V. Stereotaxie. In: Arnold H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 254-7. 3. Hassler R, Mundinger F, Riechert T. Stereotaxis in Parkinson syndrome. Berlin, Heidelberg, New York: Springer; 1979. 4. Mundinger F. Stereotaktische Operationen am Gehirn. Stuttgart: Hippokrates; 1975. 5. Riechert T. Stereotactic brain operations. Bern, Stuttgart, Vienna: Huber; 1980. 6. Krauss JK. Interview with Professor Mundinger. AANS Archives Video Interview series: Leaders in Neuroscience; 1998. 7. Bushe KA, Collmann H. Neurochirurgie von den Anfa¨ngen bis zum spa¨ten 19. Jahrhundert. In: Arnold H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 1-23. 8. Collmann H, Halves E, Arnold H. Neurochirurgie in Deutschland von 1880 bis 1932. In: Arnold H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 25-77. 9. Frowein RA, Dietz H, Rosenow DE, Vitzthum HE. Neurochirurgie in Deutschland von 1932 bis 1945. In: Arnold, H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 79-95. 10. Frowein RA, Dietz H, Franz K. Neurochirurgie in Deutschland von 1945 bis 1970, In: Arnold H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 97-119. 11. Thomas M, Hayflick SJ, Jankovic J. Clinical heterogeneity of neurodegeneration with brain iron accumulation (Hallervorden-Spatz syndrome) and pantothenate kinase-associated neurodegeneration. Mov Disord 2004;19:36-42.

57

58

4


12. Riechert T, Wolff M. Die technische Durchfu¨hrung von gezielten Hirnoperationen. Arch Psychiat Zeitschr Neurol 1953;190:297-316. 13. Riechert T, Mundinger F. Beschreibung und Anwendung eines Zielgera¨tes fu¨r stereotaktische Hirnoperationen (II. Modell). Acta Neurochir 1956;3:308-37. 14. Mundinger F, Riechert T. Ergebnisse der stereotaktischen Hirnoperationen bei extrapyramidalen Bewegungssto¨rungen auf Grund postoperativer und Langzeituntersuchungen. Dtsch Ztschr Nervenheilk 1961;182:542-76. 15. Hassler R, Riechert T. Indikationen und Lokalisationsmethode der gezielten Hirnoperationen. Nervenarzt 1954;25:441-7. 16. Mundinger F, Riechert T. Die stereotaktischen Hirnoperationen zur Behandlung extrapyramidaler Bewegungssto¨rungen (Parkinsonismus und Hyperkinesen) und ihre Resultate. Fortschr Neurol Psych 1963;31:1-66, 69-120. 17. Mundinger F, Riechert T. Indikationen und Langzeitergebnisse von 1400 uni- und bilateralen stereotaktischen Eingriffen beim Parkinsonsyndrom. Wien Zschr Nervenhk 1966;23:147-77. 18. Riechert T. The stereotactic technique and its application in extrapyramidal hyperkinesia. Confin Neurol 1972;34:325-30. 19. Mundinger F, Riechert T, Disselhoff J. Long-term results of stereotactic treatment of spasmodic torticollis. Confin Neurol 1972;34:41-6. 20. Mundinger F, Riechert T, Disselhoff J. Long term results of stereotaxic operations on extrapyramidal hyperkinesia (excluding parkinsonism). Confin Neurol 1970;32:71-8. 21. Krauss JK, Mundinger F. Functional stereotactic surgery for hemiballism. J Neurosurg 1996;85:278-86. 22. Loher TJ, Pohle T, Krauss JK. Functional stereotactic neurosurgery for treatment of cervical dystonia: review of the experience from the lesional era. Stereotact Funct Neurosurg 2004;82:1-13. 23. Mundinger F. 30 Jahre stereotaktische Hirnoperationen beim Parkinsonismus (Ergebnisse im Vergleich pallidosubthalamischer Ausschaltungen und Indikationen). In: Ga¨nshirt H, Berlit P, Haack G, editors. Pathophysiologie, Klinik und Therapie des Parkinsonismus. Basel: Editiones Roche; 1983. p. 331-57.

24. Gross I. Klinische Vergleichsuntersuchungen zwischen der konventionellen und der frequenzreinen, temperaturkontrollierten Hochfrequenzkoagulation zur stereotaktischen subkortikalen Ausschaltung. Inaugural Dissertation. Albert-Ludwigs-Universita¨t, Freiburg, 1964. 25. Hassler R, Mundinger F, Riechert T. Correlations between clinical and autoptic findings in stereotactic operations of parkinsonism. Confin Neurol 1965;26:282-90. 26. Hassler R, Riechert T. Wirkungen der Reizungen und Koagulationen in den Stammganglien bei stereotaktischen Hirnoperationen. Nervenarzt 1961;32:97-109. 27. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger JA. Physiological observations in stereotaxic operations in extrapyramidal motor disturbances. Brain 1960;83:337-54. 28. Krayenbu¨hl H, Wyss OAM, Yasargil MG. Bilateral thalamotomy and pallidotomy as treatment for bilateral parkinsonism. J Neurosurg 1961;18:429-44. 29. Goldhahn G, Goldhahn WE. Experience with stereotaxic brain surgery for spasmodic torticollis. Zentralbl Neurochir 1977;38:87-96. 30. Schaltenbrand G, Bailey P. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Thieme; 1959. 31. Schaltenbrand G, Wahren W. Atlas for stereotaxy of the human brain. Stuttgart: Thieme; 1977. 32. Schaltenbrand G, Walker EA. Stereotaxy of the human brain. Stuttgart, New York: Thieme; 1982. 33. Mu¨ller D. Psychiatrische Chirurgie (sog. Psychochirurgie). In: Arnold, H, et al., editors. Neurochirurgie in Deutschland: Geschichte und Gegenwart. Berlin, Wien: Blackwell; 2001. p. 258-66. 34. Birkmayer W, Hornykiewicz O. Der L-Dioxyphenylalanin (L-Dopa)-Effekt bei der Parkinson-Akinese. Wien Klin Wschr 1961;73:787-8. 35. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. 36. Vogel S. Mo¨glichkeiten und Grenzen der Neurotransplantation. In: Poewe W, Madeja DU, editors. Aktuelle Aspekte der Therapie des Parkinson-Syndroms. Karlsruhe: Braun; 1991. p. 73-81. 37. Krauss JK, Volkmann J: Tiefe Hirnstimulation. Darmstadt: Steinkopff; 2004.

8 History of Stereotactic Surgery in Great Britain E. A. C. Pereira . A. L. Green . D. Nandi . T. Z. Aziz

The Horsley–Clarke Apparatus Stereotaxis was the name given to a procedure and apparatus invented in London by the Englishmen, Horsley and Clarke. Before then, the German Dittmar had used a guided probe to transect rodent medulla in 1873 [1], and Zernov in Russia had described an encephalometer enabling brain surface localization in 1889 [2]. Neither technique enabled targeting with respect to a fixed threedimensional Cartesian coordinate system. Sir Victor Alexander Haden Horsley (1857–1916; > Figure 8-1a) was the first neurophysiologist who was also a neurosurgeon, pioneering a Great British tradition of such hybrid scholars particularly prevalent in the stereotactic comamunity. His uniquely deft approach to neurosurgery – derived from experiments upon over 100 primates – made a reputation such that ‘‘the Staff intended to have Horsley and nobody else’’ when tenure became available at the National Hospital for the Paralyzed and Epileptic in Queen Square in 1886 [3,4]. Robert Henry Clarke (1850–1926; > Figure 8-1b) studied medicine at St. George’s Hospital in London before surgical training in Glasgow. He worked with Horsley in London in the late 1880s [5]. Throughout the following decades, they became intellectually consumed by the experimental challenges of cerebral localization of motor function following the seminal work of Hughlings Jackson and others. Clarke traveled to Egypt in the early 1890s to convalesce after developing aspiration pneumonia from aspirin inhalation. While there, gazing up at the stars, he conceived an apparatus through #


which probing intracranial instruments could be inserted that could be clamped to an animal’s head fixing it to a Cartesian co-ordinate system by skull pins placed laterally, bars attached to plugs inserted into the external auditory meatus and further bars resting upon the nose and orbital margins. The idea was presented to Horsley on return to England [6]. A decade later in 1905 James Swift, a machinist at Palmer & Company in London, was commissioned to construct the first machine from brass, ‘‘Clarke’s stereoscopic instrument employed for excitation and electrolysis,’’ comprising frame, carrier and needle holder and costing £300 (> Figure 8-2) [7]. Results were published in 1906 from experimental use of the first instrument for targeting electrolytic lesions in the deep cerebellar nuclei of non-human primates [8]. In 1908 they described the apparatus and its use in greater depth, coining the term ‘‘stereotaxic’’ from the Greek ‘‘stereos’’ meaning ‘‘solid’’ and ‘‘taxis’’ meaning ‘‘arrangement,’’ commenting that ‘‘by this means every cubic millimeter of the brain could be studied and recorded’’ [9]. Although Clarke suggested that the apparatus might be useful in humans, neither he nor Horsley pursued the idea and shortly afterwards they ceased collaboration. Yet Clarke patented the apparatus including its proposed use in humans in 1914 and devoted much time to its improvement. By 1920, a rectilinear modification enabling needle inclination at different angles in an equatorial frame enabling 360 of movement were described (> Figure 8-2) [10]. Three others used the original apparatus in London for experimental work; first the visiting

78

8

History of stereotactic surgery in great britain

. Figure 8-1 (a) Sir Victor Alexander Haden Horsley; (b) Robert Henry Clarke (courtesy of the Wellcome Library, London)

American surgeon Ernest Sachs who studied the optic thalamus [11], then the neurologist S.A. Kinnier Wilson who studied the basal ganglia of 25 monkeys using the ‘‘Clarke-Horsley machine’’ [12]. Clarke’s original instrument was last utilized by Barrington, a London urologist who used it to study the effect of brain lesions upon micturition in cats [13]. Barrington died suddenly in 1956. Among the contents of his laboratory, ‘‘in true British fashion, was a biscuit tin’’ that contained parts of the original apparatus. After several inquiries, a technician in the Royal Veterinary College where Barrington had once worked produced a mahogany box containing the original model and it was returned to University College London in 1970. It now resides in the Science Museum in London, having been promoted from closed storage to prominent display by Tipu Aziz in 2000 (> Figure 8-3). Two further apparatus designs were made for Clarke by machinists Goodwin and Velacott also of Palmer & Company in London and exported to the United States for animal research soon after the First World War, the latter to Johns Hopkins

after agreement that the Baltimore institution would publish Clarke’s stereotaxic atlas [14]. Another of Horsley’s students, Aubrey Mussen also contemplated translation of Horsley-Clarke stereotaxis to humans. Mussen purchased one of the original Horsley-Clarke apparatus secondhand for £100 while working at the National Hospital in London from 1905 to 1906 and returned to McGill University with it, subsequently publishing results from studies of the hypoglossal nuclei that Horsley traversed with his deep cerebellar lesions [15]. Mussen designed further stereotactic instruments with Clarke, including a ‘‘cyclotome,’’ a probe used to make disk-shaped incisions along its axis and a ‘‘spherotome’’ used to cut spherical volumes bearing much resemblance to Antońio Egas Moniz’ leucotome [10]. Mussen designed and had commissioned a modification of the Horsley-Clarke apparatus for use in humans in 1914 on his return to London. It was completed around 1918, again in brass, and most likely again by Palmer & Company (> Figure 8-4). In the frame, electrode holders slid along horizontal graduated bars or vertical corner


8

. Figure 8-2 Clarke and Horsley’s 1905 primate stereotaxic apparatus, showing also Clarke’s 1920 equatorial modification (top left), (courtesy of the Wellcome Library, London)

posts enabling orthogonal approaches to intracranial structures in anteroposterior and lateral directions. The apparatus required a human brain atlas and Mussen envisaged its use to thermocoagulate tumors using ‘‘Galvanic current. . .through a 5 mm trephine in the skull and puncturing the dura without exposing the brain at all’’ [17]. In the two decades that followed, Mussen neither completed the human atlas nor convinced neurosurgical colleagues to take up use of his frame [18]. He wrapped the unused British made apparatus in newspaper dating from the 1940s and placed it in his loft [19].

After Horsley’s original experiments, Sachs, Wilson and Barrington all had loan of the original Horsley-Clarke frame. Sachs and Mussen utilized Clarke’s second and third frames respectively for animal experiments in North America throughout the 1920s [20,21], as did others during the following decade. However, at least two key challenges remained in translating experimental animal stereotaxis into a clinical tool. Firstly there was great variability between human skull landmarks and cerebral structures and secondly humans could not be sacrificed as animals were to enable confirmatory histology of accurate targeting - and thus experimental validity. Three decades on from

79

80

8


. Figure 8-3 The senior author with one of the original HorsleyClarke frames in 2000

major advances were firstly to create a frame tailored to the individual skull by means of a plaster skull cap and secondly to align their so-called ‘‘stereoencephalatome’’ not just to skull but to brain landmarks like the calcified pineal gland and foramen of Monro by means of intraoperative pneumoencephalography.

Post-War Innovation

. Figure 8-4 Mussen’s circa 1918 human stereotactic instrument (after Picard et al. [16])

Horsley and Clarke’s work, Spiegel and Wycis devised an apparatus for stereotactic neurosurgery in humans, publishing their achievement in 1947 [22]. The North Americans established ‘‘stereotactic’’ as the preferred term, fusing Greek with the Latin word ‘‘tactis,’’ the pluperfect passive form of the verb ‘‘tangere’’ meaning ‘‘to touch.’’ Their

With two World Wars, British stereotactic surgery remained fallow for the half century after Horsley’s discovery, the discipline only reaching the clinic once word had spread of Spiegel and Wycis’ invention. At first, primary applications were for treating psychiatric disorders and later clinical usage diffused to movement disorders in the 1950s and chronic pain soon after. Ahead of the rest, an English crusader and two Scottish pioneers emerged, each a clinical polymath but with an academic focus. In London, Geoffrey Knight developed stereotactic subcaudate tractotomy for psychiatric disorders, treating hundreds of patients while his eminent contemporary Sir Wylie McKissock continued freehand approaches. In Edinburgh, John Gillingham established stereotactic surgery for multiple clinical indications, designing his own stereotactic frame. In turn, his talented associate Ted Hitchcock was inspired first at Edinburgh and then in Birmingham to pioneer stereotactic approaches to the high cervical spinal cord and brainstem. Francis John Gillingham (b.1916; > Figure 8-5) trained in St. Bartholomew’s Hospital in London before entering the neurosurgical faculty at Edinburgh in 1950. Gillingham spent 12 years as first assistant to Norman McOmish Dott, one of the great triumvirate alongside Sir Hugh Cairns in Oxford and Sir Geoffrey Jefferson in Manchester, the Cushing apostles who definitively established neurosurgery as a specialty in Great Britain [24–26]. Gillingham was a brilliant and pioneering aneurysm surgeon


8

. Figure 8-5 Francis John Gillingham (left) preparing for a stereotactic thalamotomy in 1968 (after Housepian 2004 [23])

like his mentor Dott [27,28]. A passionate educator, he also introduced the concept of subspecialty fellowships to British neurosurgical training [29], but stereotactic surgery received his greatest contributions. The Parisian neurosurgeon Gerard Guiot. introduced Gillingham to stereotactic surgery. They had become friends after Guiot visited Edinburgh to learn aneurysmal surgery from Dott and Gillingham. Guiot’s 1953 telegram to Gillingham read ‘‘I have something interesting to show you – come over.’’ Four days were spent performing freehand pallidotomies to treat parkinsonism under local anesthesia using a subfrontal approach to the anterior perforated substance interrupting the ansa lenticularis as described by Fenelon and Thiebaut following the seminal discoveries of Cooper [30–32]. Gillingham returned to Edinburgh to treat two patients in 1955 and 1957, reporting improvements in tremor, rigidity and quality of life. Wishing to reduce risks from the demanding subfrontal approach, he adapted Guiot’s stereotactic method [33]. In 1960 he published results from stereotactic ‘‘thermal electrocoagulation lesions of the globus pallidus, internal

capsule and thalamus either separately or in combination’’ in a further 58 patients operated upon from 1957 to 1959 [34]. In addition to globus pallidus and internal capsule, he began targeting the ventrolateral thalamus for refractory tremor based on work by Hassler [35]. ‘‘Of these (60) patients 53, or 88%, had tremor and/or rigidity abolished or significantly reduced without complications’’ [34]. From his early clinical results, Gillingham drew several conclusions. On targeting he wrote that ‘‘The best type of lesion. . .would seem to be the double one, made at the same time in the ventro-lateral nucleus of the thalamus and in the globus pallidus 16mm from the mid-line, both lesions bordering on the internal capsule. . .. Bilateral lesions in the treatment of bilateral Parkinsonism, provided they are small and strategically placed, would seem to be eminently practicable. . . usually with an interval of 3–6 months between the two operations.’’ On his modification of Guiot’s stereotactic apparatus he felt ‘‘that the merits of this method lie in the relatively short operative procedure and in its accuracy and simplicity. Its principles are based

81

82

8


on the fact that the globus pallidus and thalamus bear a reasonably constant anatomical relationship to the anterior and posterior commissures, the intercommissural line, and the mid-sagittal plane of the head. . .. The method used has evolved progressively, and is unique, in allowing the creation of lesions in the globus pallidus, internal capsule, or thalamus with one electrode track at different depths’’ [34]. In their stereotactic apparatus design, Guiot and Gillingham favored operative principles to prioritize patient comfort, not restricting their movements by clamping their head, and to reduce laborious calculations. Guiot planned a parasagittal approach using intraoperative encephalography to delineate the midline and intercommissural point. Gillingham favored an occipitoparietal entry to avoid striate arteries and horizontal patient positioning to reduce putative brain shift. Thus the Guiot-Gillingham stereotactic apparatus was devised (> Figure 8-6). Radioopaque midline markers were used for the procedure and a 1 mm steel ball placed in each 5 mm lesion for subsequent charting. Over the . Figure 8-6 The Guiot-Gillingham stereotactic apparatus using a posterior rather than a coronal approach (after Gillingham et al. [34])

post-operative weeks, the ball was seen to fall through the necrotic lesion on skull radiographs, elegantly providing an estimate of lesion size. The frame’s conception preceded Hassler’s targeting of the thalamus for tremor and hence Gillingham attributed to serendipity that his posterior approach enabled multiple targets to be lesioned in a single pass [36]. Despite impressive clinical outcomes, Gillingham noted some inaccuracy to his lesions in the context of Brierley and Beck’s demonstration that relationships between basal ganglia structures and commissural landmarks were highly variable [37,38]. David Whitteridge, his neurophysiologist colleague at Edinburgh, had demonstrated to him in 1961 how microelectrode recording could distinguish between grey and white matter and thus delineate the lateral geniculate nuclei in the cat [39]. He immediately saw its utility for distinguishing functionally between deep brain structures and, with his colleague Michael Gaze, developed the technique for humans as did Guiot [40]. Fundamental physiological insights were gained in the quest to improve lesion accuracy and clinical efficacy, including spontaneous rhythmical discharge in the thalamus found to be synchronous with tremor [41]. Using microelectrode recording, target localization could be done accurately with a margin of error less than 1mm. Gillingham evolved the Guiot-Gillingham apparatus throughout the 1960s and 1970s, He added a phantom to allow an oblique track to more medial brain targets for epilepsy and psychiatric disorders, then an inferior extension to the posterior limb of the frame for targeting the cerebellum, brain stem and cervical spine in chronic pain and dystonias [42]. In 1977 he added a motor to automatically drive an electrode in at a slow and measured rate for microelectrode recording. Stereotactic surgery for deep hematomas and tumor biopsies was also performed [43]. Ten year follow-up in the post levodopa era of a second 60 patient parkinsonian cohort of Gillingham’s operated upon between


1965 and 1967 showed decline in efficacy for bradykinesia, but consistent relief of tremor and rigidity [44]. Gillingham remained engaged in academic neurosurgery well into his ninth decade [45]. As Gillingham was Dott’s prote´ge´, so Ted Hitchcock was Gillingham’s. Edward Robert Hitchcock (1929–1993) studied medicine at Birmingham then neurosurgery at Oxford before joining the Edinburgh staff at the then recently opened Western General Hospital in 1965. While there, he received unique exposure to Gillingham’s stereotactic surgery which attracted international renown [23]. Hitchcock’s developed an interest in chronic pain and in particular the concept of percutaneous high spinal stereotactic commissural myelotomy. This procedure aimed to divide the decussating spinothalamic tracts through a targeted lesion and reduce the risks of respiratory paralysis conferred by open cordotomy. It was aimed at chronic cancer pain. It required access below the plane of a versatile frame, thus he invented his own target-centered arc system secured on a hollow square aluminum base ring secured to the skull by three-point fixation (> Figure 8-7). Vertical and horizontal bars determined probe length and laterality. The system was first used both for surgery and for microelectrode recording in the spinal cord by

. Figure 8-7 Hitchock’s stereotactic apparatus for brainstem and cervical spine surgery (after Hitchcock [46])

8

percutaneous approach using portable radiographs [46–48]. Hitchcock reported initial results of good or complete pain relief in 13 out of 19 patients at follow-up ranging from 1 week to 4 years [49]. A stereotactic pontine approach to spinothalamic tractotomy and to the trigeminal nucleus for anesthesia dolorosa was also applied [50–54], as were approaches to the thalamus and dentate nucleus to treat dystonia and in particular the spasticity of cerebral palsy [55]. The rationale behind the stereotactic pontine spinothalamic approach was to provide good analgesia with minimal risks to respiration, micturition and upward gaze [54]. Hitchcock wrote of his apparatus in the early 1970s that ‘‘the design and construction make this one of the most accurate, adaptable and simplest of modern stereotactic instruments’’[56]. Hitchcock became Professor of Neurosurgery at Birmingham in 1978, succeeding Brodie Hughes (1913–1989) who was also an established stereotactic surgeon [57,58]. Hitchcock put his stereotactic frame to many further clinical uses including biopsy of supratentorial, infratentorial and high spinal tumors and intraventricular masses [59–61], foreign body removal [62], realtimeclippingofotherwiseinoperablearteriovenous malformations [56], image-guided craniotomies [63], and in the 1980s in the planning and treatmentstagesofradiosurgery[64].Thesemanyvaried clinical indications in brain and spine earned him the nickname ‘‘Columbus of the brain’’ in the local clinical neuroscience community. At the MidlandHospitalforNeurologyandNeurosurgery, he used his stereotactic expertise to established a programme at first for adrenal medullary and then in the late 1980s for fetal mesencephalic transplantation in Parkinson’s disease, performing the procedure on 55 patients [65–68]. Neurosurgery for psychiatric disorders in Great Britain echoed its popularization in the United States following Freeman and Watts’ simplification of Moniz’ procedure in the 1940s [69,70]. Its foremost British proponent was the

83

84

8


London neurosurgeon Sir Wylie McKissock (1906–1994), founder of the neurosurgical department at Atkinson Morley’s Hospital in Wimbledon [71]. McKissock favored a freehand approach to the frontal lobe from above [72]. He described the rostral leucotomy in 1951 as a rejoinder to Freeman and Watt’s transorbital ‘‘icepick’’ leucotomy which he considered to contravene ‘‘established aseptic surgical principles’’ [73,74]. McKissock’s immense South England practice and his reputation for extraordinary surgical speed inculcated a peripatetic service visiting other hospitals in his car with his surgical instrument set in the boot, drawing parallels with Freeman [75,76]. It is suggested that McKissock alone may have performed one quarter of the 10,365 procedures performed in the United Kingdom from 1942 to 1954 [77]. Geoffrey Cureton Knight (1906–1994) of Hammersmith and Brook Hospitals in London and Woolwich saw more readily than McKissock the merits of stereotactic over freehand approaches in reducing the morbidity and mortality of neurosurgery for psychiatric disorders [74,78]. After his freehand experience [79], Knight created the procedure of stereotactic subcaudate tractotomy in 1961 using a modified stereotactic device that his London colleague the Scottish neurosurgeon Ian Reay McCaul (1916–1989) reported in 1959 [58,80]. His first few hundred orbital undercuttings led him to conclude that lesions extending posteriorly under the caudate were most efficacious and that the last 2 cm was key [81,82]. Knight used bony landmarks on lateral radiographs, and later air encephalography to guide him. In addition, he employed brachytherapy as an ablative tool, implanting radioactive Yttrium (Y90) to create flat lesion approximately 20 by 20 by 7 mm (> Figure 8-8) [83–85]. The treatment proved effective and endured four decades amid decline in use of other psychosurgical treatments. The group described treatment of 1,300 patients with ‘‘non-schizophrenic affective disorders,’’ 40–60% going on to live normal or near normal lives with a reduction in

. Figure 8-8 Anteroposterior radiograph of Geoffrey Knight’s stereotactic subcaudate tractotomy showing yttrium seeds in situ for brachytherapy (after Knight [83])

suicide rates from 15 to 1% [86,87]. Long-term outcomes were published by the psychiatrist Bridges and the London neurosurgeon John Bartlett, Knight’s successor from 1972. Following the retirement of Knight, the unit was named the Geoffrey Knight Unit for Affective Disorders to emphasize Knight’s appreciation of the fundamental importance of psychiatric evaluation both in diagnosis and in full consideration of medical treatments prior to offering surgery. In 1996 the unit moved to the Maudsley Hospital and Y90 production ceased. Bartlett adapted a Leksell frame arc compatible with modern neuroimaging using concepts that underlay the McCaul device. Radiofrequency lesioning replaced radioisotope implantation [88,89]. It is a tribute to Knight that McKissock’s colleague at Atkinson Morley’s Hospital, Alan Richardson (1926–1998), adopted a stereotactic approach for his psychiatric procedures [72]. Together with his psychiatrist colleague Desmond Kelly, he combined Knight’s subcaudate tractotomy with a cingulotomy to invent the procedure of limbic leucotomy in the early


1970s [90,91]. It is interesting to note that cingulectomy for psychiatric disorders was first performed in 1948 in Oxford by Sir Hugh Cairns, albeit freehand [92]. Both Knight’s subcaudate tractotomy and Richardson’s limbic leucotomy continue to be performed in carefully selected cases refractory to medical treatment worldwide including in Great Britain and elswehere [93–95]. Stereotactic neurosurgery was embraced by Dott’s unit in Edinburgh and, at Oxford after Cairns, Pennybacker appointed Watkins to establish a service, but other regions were also keen to commence it. In Manchester, Jefferson’s successor Richard Johnson appointed John Dutton to undertake a high volume of ablations for parkinsonism and other stereotactic procedures throughout the 1960s using a Leksell frame. Soon after, John Gleave (1925–2006) established a stereotactic service in Cambridge also using a Leksell frame, treating parkinsonism with cryosurgery and developing a side-cutting stereotactic biopsy cannula [96,97]. Other British neurosurgeons like McCaul made and modified stereotactic frames. Of particular note was the frame of Alfred Michael Bennett (1920–1996) and his use of a sphere inserted into a burr hole to aid targeting [98,99]. Bennett’s apparatus were popular locally, used by Sid Watkins and later David Thomas in London amongst others [100,101]. Most designs were less radical and therefore perhaps less memorable than those of Gillingham and Hitchcock.

Stereotactic Atlases Horsley and Clarke produced the first stereotaxic atlas, a monkey version appearing in their 1908 publication. Later publications were by Clarke at first for the cat in collaboration with the British ophthalmic surgeon E. Erskine Henderson in 1912 and later by Clarke alone for the monkey in 1920 [10,102,103]. Both atlases comprised brain slices 2 mm thick. The latter atlas showed sections of monkey brain at calibrated intervals

8

with a scale giving slice thickness and height from the base of the apparatus. Sections were registered by a Cartesian coordinate system to the skull landmarks of inferior orbital rim and both external auditory canals to which the frame was fixed. Zero axes were the plane between these structures axially, the mid-sagittal plane and the coronal plane between both external auditory meatus orthogonal to both other planes. The human brain atlases produced outside Great Britain, in particular by Spiegel and Wycis, Schaltenbrand and Bailey and Talairach transformed stereotactic functional neurosurgery. Schaltenbrand detailed anatomical nuclei with an emphasis on the thalamus and adjacent deep brain structures now used in deep brain stimulation for movement disorders and Talairach revealed vasculature relevant to epilepsy surgery [104–106]. The British neurosurgeon Sid Watkins also made rigorous contributions. Eric Sidney Watkins (b.1932) was given the task of starting stereotactic neurosurgery at the Radcliffe Infirmary in Oxford in the 1950s by Joe Pennybacker. Dissatisfied with the variability of basal ganglia structures with respect to the frequently uncalcified and thus radiolucent pineal gland using the Spiegel and Wycis atlases, he used the Schaltenbrand and Talairach atlases that appeared shortly after. Initially globus pallidus and ansa lenticularis were targeted for Parkinsonian rigidity, tremor and dystonia then later the lateral thalamus in the 1960s. The desire to create his own atlas arose in the early 1960s from a wish to commence thalamotomy for pain and a concern at the adequacy of available atlases to accurately enable targeting based upon anatomy alone in the absence of subjective or physiological guidance. Encouragement came from London neurosurgical colleagues John Andrew and Valentine Logue who were also keen to begin such therapies. Another atlas that became available was created by Brierley and Beck who sectioned 40 brains in 3–5 mm slices, relating them in a proportional hypothesis for thalamic nuclear determination to

85

86

8


anterior and posterior thalamic limits and the midthalamic point and describing great individual variations [38]. Watkins found the atlas to be limited clinically as the use of simultaneous positive and air ventriculography using air in the ambient cisterns to outline the pulvinar nuclei and thus the thalamic limits was not consistently reproducible. In the 1960s at the National Hospital for Nervous and Mental Diseases together with Watkins then at the Royal London Hospital, John Andrew produced a greatly enlarged atlas with drawings defining in detail deep brain nuclei including the thalamus and its relations [107]. The atlas was based upon 38 formalin fixed brains. It measured the position of the thalamic centromedian nucleus using 1 mm coronal slices with reference planes between the posteroinferior margin of the foramen of Monro and posterior commissure and the midpoint between the ventricular surfaces of the anterior and posterior commissures. Its utility lies in the presentation of statistical data in a graphic form together with stereotactic coordinates superimposed on simple line drawings of the thalamus. In 1978 at the London Hospital, Fari Afshar detailed brain stem and cerebellar nuclei, again under Sid Watkins’ supervision [108]. The impetus for the Afshar atlas came from an interest in attempting to ameliorate spasticity in cerebral palsy by ablation of the cerebellar dentate nucleus. Approximately 30 brains were prepared using positive-contrast ventriculography with skull and brain mounted in a stereotactic frame in order to accurately correlate structures with coordinates. Again, formalin fixation of 1mm slices was performed. Modified Mulligan stain was used and each section magnified, with drawings made using a camera lucida. Reference plains were the fourth ventricular floor and fastigium and the midsagittal plane. As before, variability profiles were quantified and standard deviations presented. Both Watkins atlases are resources used today to help ratify localization and indeed the

Afshar atlas continues to augment heated debates regarding the targeting of the novel functional neurosurgical treatment of deep brain stimulation of the pedunculopontine region for Parkinson’s disease [109,110]. Watkins has commented upon the major difficulties in measurement due to distortion related to fixation and shrinkage, past atlases suffering from approximately 10% shrinkage. To reduce shrinkage to 2–5%, he utilized Corsellis’ technique – a ten day formalin suspension after removing brain and skull en bloc minus the frontal and facial bones [111]. To appease undertakers’ concern at cosmetic consequences, each cadaver’s scalp was replaced over a plaster of Paris prosthesis fixed to a broom handle on a nail inserted into the cervical canal. However, cremation of the augmented cadavers became suboptimal, precipitating a strike among undertakers serving the London Hospital by the time Afshar’s posterior fossa atlas had reached its completion [112].

Computed Tomography By the late 1970s, British stereotactic functional surgery was a decade on from its first successes, having declined with the advent of neuropsychopharmacology. Levodopa was introduced to relieve Parkinson’s disease [113], chlorpromazine and monoamine oxidase inhibitors to ameliorate schizophrenia and depression respectively, and the case series showing good relief after lesional surgery for chronic pain paled against a background of new analgesics and peripheral neuromodulatory therapies. Stereotactic approaches to tumors had been established but most neurosurgeons did not train in stereotactic neurosurgery. Hitchcock continued psychiatric procedures for medically refractory depression, obsessivecompulsive disorders and anxiety alongside Bartlett and Richardson, but the developing subspecialty sought advances in other domains. Enter the British engineer.


Sir Godfrey Newbold Hounsfield (1919– 2004; > Figure 8-9) joined EMI in Hayes, Middlesex in 1951, having been first a mechanic first of radios then later radars in the Royal Air Force and obtained a diploma from Faraday House Electrical Engineering College in London. At EMI he worked first on radars and guided weapons, then the first all-transistor computers. During a weekend ramble in 1967 he conceived what later became the first EMI-scanner and the technique of computed tomography (CT), which he recounted as ‘‘a realization that you could determine what was in a box by taking readings at all angles through it.’’ Recording multiple pictures from a rotating X-ray source, a series of slices could be photographed and a three-dimensional image reconstructed from the slices. After initial successful experiments with a cylindrical phantom containing radio-opaque objects in his Hayes laboratory using X-rays, Hounsfield forged a collaboration . Figure 8-9 Sir Godfrey Hounsfield at the controls of the EMI scanner in Atkinson Morley’s Hospital in London (after Petrik et al. [114])

8

with James Ambrose, radiologist at Atkinson Morley’s Hospital, to translate the device’s utility to humans. McKissock gave the endeavor his blessing [75]. Hounsfield set to work on bullock’s heads obtained from a kosher slaughterhouse in East London to obviate the traumatic intracranial hemorrhage seen after conventional slaughter [114]. Ambrose interpreted the early scans and suggested the use of sodium iothalamate contrast to highlight tumors [115]. His early interpretations and predictions formed the basis of contemporary diagnostic neuroradiology [116,117]. The first patient was scanned in 1971, revealing a cyst [118]. In 1979 Hounsfield was awarded the Nobel Prize for Physiology or Medicine together with Allan Cormack, the Cape Town physicist whose mathematical theories Hounsfield had realized [119]. The advent of CT breathed new life into stereotactic surgery in Britain. Several neurosurgeons began to experiment with CT compatible apparatus, both imported, usually Leksell, and those of Gillingham and Hitchcock [36,120]. Magnetic resonance imaging (MRI) followed shortly after, again with frames being modified as required. After a quiescent decade, the late 1980s augured for a renaissance. Alongside emerging limitations of drug therapies for movement disorders resurrecting clinical indications for functional neurosurgery, great advances came from increasing computer power enabling the fusion of more spatially robust CT information with the greater soft tissue detail of MRI and comparison to computerized brain atlas images. In Britain, the current generation of senior stereotactic neurosurgeons were gaining their clinical training and conducting their first research, some in classical stereotactic methods abroad, some by subspecialty fellowships in Britain thanks to Gillingham’s enduring influence upon neurosurgical programmes, and others by animal experiments true to Horsley’s hybrid scientist-surgeon mould. The stage was set for two decades of rapid advances in British stereotactic neurosurgery.

87

88

8


Radiosurgery Lars Leksell’s brilliance showed not only in his frame design [121], employing the novel arcquadrant principle, but also in his insight that focused radiation could be used as the tool. Many intersecting radiation beams focused towards a target would result in a high cumulative radiation dose, with radiation intensity declining rapidly with distance from the ‘‘isocenter.’’ Thus, a deep brain structure could be lesioned noninvasively by focused radiation. The technology of ‘‘radiosurgery’’ could be applied to acoustic neuromas, arteriovenous malformations and other discrete pathologies [122]. Britain acquired one of the first gamma knives in 1985 at the same time as Argentina just 3 years on from the Falklands War. David Forster achieved incredible feats in leading the campaign to fund the device on the British National Health Service, building the infrastructure over 2 years to host it and ultimately purchasing the first custom-made unit in 1985. The National Centre for Stereotactic Radiosurgery was refurbished in 1991 and its cobalt sources renewed. The unit receives approximately one half to two thirds of all radiosurgery referrals from all over the United Kingdom. Approximately 500 cases a year are performed, a third of cases treated being ateriovenous malformations, with small to medium sized acoustic neuroma treatment having increased from 10% to one third over the period 1994–2001 and approximately 100 meningiomas and other skull base or recurrent tumors treated per year. Other indications treated have included trigeminal neuralgia, pituitary tumors and metastases, although the latter two indications are treated in smaller proportions than outside the United Kingdom, reflecting more conservative referral patterns [123]. For similar reasons, few epilepsy and functional cases have been performed. Several neurosurgical centers use linear accelerators to perform radiosurgery, each performing up to 40 cases per year. A gamma knife was also

acquired by the Cromwell Hospital in London in 1998, run privately by Christer Lindquist. It has treated 1,000 patients since installation and has recently been refurbished.

Functional Surgery European factors driving British stereotactic surgery included Benabid’s application of thalamic deep brain stimulation to Parkinson’s disease in 1987 and Laitinen’s reapplication of Leksell’s pallidotomy in 1992 [124,125]. Functional neurosurgery was resurrected at the Radcliffe Infirmary in Oxford four decades after Watkins’ departure under the headship of Mr. Christopher Adams [126]. We had already established with Alan Crossman in Manchester by the early 1990s at the same time as DeLong’s team across the Atlantic that lesions made to the subthalamic nucleus in primates reversed the motor symptoms of 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine (MPTP) induced parkinsonism [127–129], At Oxford and Charing Cross Hospital in London, we undertook stereotactic surgery of this target and others [130–135]. At the same time we continued non-human primate research into establishing the pedunculopontine nucleus as a potential target for gait freezing and postural instability [136–139]. The former target is now the target of choice for Parkinson’s disease surgery and initial clinical results of the latter show great promise [140–143]. Other translational research at the University of Oxford and Imperial College London included invasive deep brain electrophysiological insights into tremor and dystonia [144–146], use of single photon emission tomography (SPET) [147], magnetoencephalography (MEG) [148], and diffusion tensor imaging (DTI)[149,150] to study deep brain stimulation and research into deep brain stimulation for pain and blood pressure control and brainstem control of exercise [151,152]. We have used deep brain stimulation to treat 70 patients


with dystonia [153], 60 with chronic pain [154], and we perform one fifth of Britain’s movement disorders surgery. In Bristol, Professor Steven Gill has continued the Great British tradition of innovation, creating a stereotactic frame convenient for radiosurgery under Professor David Thomas’ supervision in London [155,156], and performing several clinical firsts including glial-cell derived neurotrophic factor infusion and pedunculopontine nucleus stimulation for Parkinson’s disease [157–160]. With Mr. Nikunj Patel, he continues to drive the field forward. After the retirement last decade of Mr. John Miles in Liverpool whose tremendous pain practice still left time for several innovations [161– 163], Professor David Thomas also recently

8

retired as the Gough-Cooper Professor of Neurosurgery at the National Hospital of Neurology and Neurosurgery at Queen Square. He had devoted three decades to the improvement of stereotactic surgical techniques with and without frames [164–168]. Britain welcomed Professor Marwan Hariz at Queen Square as the first Edmond J. Safra Chair of Functional Neurosurgery, ‘‘solving’’ the functional neurosurgery service there [169], and establishing a biennial international workshop like its host, unsurpassed for its conviviality and candour (> Figure 8-10). Almost all of the 34 hospitals conducting neurosurgery in Great Britain and Northern Ireland have consultants able to offer stereotactic surgery a century on from Horsley’s first experiments. A third of these hospitals have

. Figure 8-10 The faculty of the International Workshop on Functional Neurosurgery for Movement Disorders and Mental Illness & Commemoration of the 150th Anniversary of the Birth of Sir Victor Horsley, London, 2007 (courtesy of Professor Marwan Hariz). Back Row from left to Right: Lazaro Alvarez La Habana, Peter Brown, Gun Marie Hariz, Paul Krack, Pierre Pollak, Bart Nuttin, Roger Melvill, Steven Gill, Patricia Limousin-Dowsey, Niall Quinn, John Rothwell, Veerle Visser-Vandewalle, Roger Lemon, Rees Cosgrove, Andres Lozano, Laura Cif, Ludvic Zrinzo, Marjan Jahanshahi, Stephen Tisch, Hans Speelman, Philippe Coubes, Pat Forsdick. Front Row Left to Right: Takaomi Taira, Jean-Luc Houeto, Alim-Louis Benabid, Tipu Aziz, Boulos-Paul Bejjani Byblos, Alan Crockard, Marwan Hariz, Carmelo Sturiale

89

90

8


subspecialty trained stereotactic surgeons offering functional procedures, the majority of them affiliated to universities and conducting clinical or translational research. Far from being the reserve of the eccentric scientist-surgeons looked upon with suspicion by the rest of the neurosurgical fraternity, stereotactic surgery has become an established clinical subspecialty and academic discipline in its own right. The Society of British Neurological Surgeons formed in 1929 has recently begun devoting specific sections to stereotactic and functional neurosurgery at its meetings. A further society with a focus upon pain that welcomes stereotactic neurosurgeons, the Neuromodulation Society of the United Kingdom and Ireland (NSUKI) was founded in 2001. Such factors prove good indicators of the definitive establishment of the subspecialty in the United Kingdom. Phil Gildenberg has commented that there are four central tenets of the field of stereotactic surgery [170]. The need to be innovative – that a better way to do something may be more apparent to the most junior member of the team rather than the most senior. That stereotactic surgeons work as a community not in isolation. That stereotactic surgery is, to a large extent, a basic science. Although not appealing to the neurosurgeon interested only in a better way to cut, it is exciting to one appreciating the associated basic science. Finally, there is awed appreciation for the insight and courage of the true pioneers in the field. The British school exemplifies such values. It is hoped that its practitioners will continue to uphold tradition as we look to the future with excitement.

Acknowledgments We thank Mr. John Bartlett for helpful comments, Professor David G. T. Thomas, Professor Anthony J. Strong, Professor John D. Pickard, Mr. Robert Macfarlane and Mr. Colin Watts for advice on historical sources and Professor Marwan I. Hariz

for helpful comments and advice on > Figure 8-10. The authors receive financial support for research from the UK Medical Research Council, Normal Collisson Foundation, Charles Wolfson Charitable Trust and Oxford Partnership Comprehensive Biomedical Research Centre.

References 1. Dittmar C. Ueber die Lage des sogenannten Gefaesszentrums in der Medulla oblongata. Bersaechs Ges Wiss Leipzig (Math Phys) 1873;25:449-69. 2. Zernov DN. Encephalometer: device for estimation of parts of brain in human. Proc Soc Physicomed Moscow Univ 1889;2:70-80. 3. Paget S. Sir Victor Horsley: a study of his life and work. New York: Harcourt, Brace and Howe; 1920. 4. Northfield WC. Sir Victor Horsley, his contributions to neurological surgery. Surg Neurol 1973;1:131-4. 5. Davis RA. Victorian physician-scholar and pioneer physiologist (Robert Henry Clarke). Surg Gynecol Obstet 1964;119:1333-40. 6. Sachs E. Victor Horsley. J Neurosurg 1958;15:240-4. 7. Schurr PH, Merrington WR. The Horsley-Clarke stereotaxic apparatus. Br J Surg 1978;65:33-6. 8. Clarke RH, Horsley V. On a method of investigating the deep ganglia and tracts of the central nervous system (cerebellum). Br Med J 1906;2:1799-800. 9. Horsley V, Clarke RH. The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45-124. 10. Clarke RH. Part I: Investigation of the central nervous system, methods and instruments. Baltimore, MD: Johns Hopkins; 1920. 11. Sachs E. On the structure and functional relations of the optic thalamus. Brain 1909;32:95-186. 12. Wilson SAK. An experimental research into the anatomy and physiology of the corpus striatum. Brain 1914;36:427-92. 13. Barrington FJF. The effect of lesions of the hind- and midbrain on micturition in the cat. Q J Exp Physiol 1925;15:81-102. 14. Fodstad H, Hariz M, Ljunggren B. History of Clarke’s stereotactic instrument. Stereotact Funct Neurosurg 1991;57:130-40. 15. Mussen AT. Notes on the movements of the tongue from stimulation of the twelfth nucleus, root and nerve. Brain 1909;32:206. 16. Picard C, Olivier A, Bertrand G. The first human stereotaxic apparatus. The contribution of Aubrey Mussen to the field of stereotaxis. J Neurosurg 1983;59:673-6. 17. Bertrand G. Stereotactic surgery at McGill: the early years. Neurosurgery 2004;54:1244-51; discussion 1251-2.


18. Jensen RL, Stone JL, Hayne R. Use of the Horsley-Clarke stereotactic frame in humans. Stereotact Funct Neurosurg 1995;65:194-7. 19. Gildenberg PL. The history of stereotactic neurosurgery. Neurosurg Clin N Am 1990;1:765-80. 20. Sachs E, Fincher EF. Anatomical and physiological observations on lesions in the cerebellar nuclei in Macacus rhesus (preliminary report). Brain 1926;50:350-6. 21. Mussen AT. Experimental investigations on the cerebellum. Brain 1927;50:313-49. 22. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on human brain. Science 1947;106:349-50. 23. Housepian EM. Stereotactic surgery: the early years. Neurosurgery 2004;55:1210-4. 24. Fraenkel GJ. Hugh Cairns: first Nuffield professor of surgery, University of Oxford. Oxford: Oxford University Press; 1991. 25. Rush C, Shaw JF. With sharp compassion: Norman Dott Freeman surgeon of Edinburgh. Aberdeen: Aberdeen University Press; 1990. 26. Schurr PH, Royal Society of Medicine (Great Britain). So that was life: a biography of Sir Geoffrey Jefferson, Kt CBE FRS MS FRCS (1886–1961): master of the neurosciences and man of letters. London: Royal Society of Medicine Press; 1997. 27. Todd NV, Howie JE, Miller JD. Norman Dott’s contribution to aneurysm surgery. J Neurol Neurosurg Psychiatry 1990;53:455-8. 28. Gillingham J. Proceedings: twenty-five years’ experience with middle cerebral aneurysms. J Neurol Neurosurg Psychiatry 1975;38:405. 29. Gillingham FJ. Surgical training in the EEC – the training of a specialist. Acta Neurochir (Wien) 1982;61:17-24. 30. Cooper IS. Ligation of the anterior choroidal artery for involuntary movements: parkinsonism. Psychiatr Q 1953;27:317-9. 31. Fenelon F, Thiebaut F. Resultats du traitment neurochirugical du syndrome parkinsonien par intervention direct sur les voies etrapyramidales immediatement sous-striopallidales (anse lenticulaire). Rev Neurol (Paris) 1950;83:437-40. 32. Guiot G, Brion S. Traitement des mouvements anormaux par la coagulation pallidale; technique et resultats. Rev Neurol (Paris) 1953;89:578-80. 33. Guiot G. Le traitement des syndromes parkinsoniens par la destruction du pallidum interne. Neurochirurgia (Stuttg) 1958;1:94-8. 34. Gillingham FJ, Watson WS, Donaldson AA, Naughton JA. The surgical treatment of parkinsonism. Br Med J 1960;2:1395-402. 35. Hassler R, Riechert T. Indikationen und lokalisations: methode de gezeilten hirnoperationen. Nervenarzt 1954;25:441-7.

8

36. Gillingham J. The Guiot-Gillingham Apparatus. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw Hill; 1998. p. 139-44. 37. Gillingham J. Introduction to the scientific sessions. Accidents in stereotaxy – side-effects or bonus? Acta Neurochir (Wien) 1974;Suppl 21:5–12. 38. Brierley JB, Beck E. The significance in human stereotactic brain surgery of individual variation in the diencephalon and globus pallidus. J Neurol Neurosurg Psychiatry 1959;22:287-98. 39. Gillingham FJ. Neurosurgery. Br J Surg 1966;53:833-6. 40. Guiot G, Hardy J, Albe-Fessard D. Delimination precise des structures sous-corticales et identification de noyaux thalamiques chez l’homme par l’electrophysiologie stereotaxique. Neurochirurgia (Stuttg) 1962;5:1-18. 41. Gaze RM, Gillingham FJ, Kalyanaraman S, Porter RW, Donaldson AA, Donaldson IM. Microelectrode recordings from the human thalamus. Brain 1964;87:691-706. 42. Gillingham FJ, Campbell D. Surgical interruption of the conduction pathways for the control of intractable epilepsy. Acta Neurochir Suppl (Wien) 1980;30:67-74. 43. Kalyanaraman S, Gillingham FJ. Stereotaxic biopsy. J Neurosurg 1964;21:854-8. 44. Kelly PJ, Gillingham FJ. The long-term results of stereotaxic surgery and L-dopa therapy in patients with Parkinson’s disease. A 10-year follow-up study. J Neurosurg 1980;53:332-7. 45. Gillingham J. Forty-five years of stereotactic surgery for Parkinson’s disease: a review. Stereotact Funct Neurosurg 2000;74:95-8. 46. Hitchcock E, Lewin M. Stereotactic recording from the spinal cord of man. Br Med J 1969;4:44-5. 47. Hitchcock E. Stereotaxic spinal surgery. A preliminary report. J Neurosurg 1969;31:386-92. 48. Hitchcock E. An apparatus for stereotactic spinal surgery. Lancet 1969;1:705-6. 49. Hitchcock E. Stereotactic myelotomy. Proc R Soc Med 1974;67:771-2. 50. Hitchcock E. Stereotactic trigeminal tractotomy. Farmatsiia 1970;19:131-5. 51. Hitchcock E, Kim MC, Sotelo M. Further experience in stereotactic pontine tractotomy. Appl Neurophysiol 1985;48:242-6. 52. Hitchcock E, Teixeira MJ. Pontine stereotactic surgery and facial nociception. Neurol Res 1987;9:113-7. 53. Hitchcock E, Sotelo MG, Kim MC. Analgesic levels and technical method in stereotactic pontine spinothalamic tractotomy. Acta Neurochir (Wien) 1985;77:29-36. 54. Hitchcock ER. Stereotaxic pontine spinothalamic tractotomy. J Neurosurg 1973;39:746-52. 55. Gornall P, Hitchcock E, Kirkland IS. Stereotaxic neurosurgery in the management of cerebral palsy. Dev Med Child Neurol 1975;17:279-86.

91

92

8


56. Shieff C. The Hitchcock apparatus. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw Hill; 1998. p. 101-4. 57. Hughes B. Involuntary movements following stereotactic operations for parkinsonism with special reference to Hemi-Chorea (Ballismus). J Neurol Neurosurg Psychiatry 1965;28:291-303. 58. Lyle I, Taylor S. Lives of the fellows of the Royal College of Surgeons of England: 1983–1990. London: Royal College of Surgeons of England; 1995. 59. Hitchcock ER, Issa AM, Sotelo MG. Stereotactic excision of deeply seated intracranial mass lesions. Br J Neurosurg 1989;3:313-20. 60. Hitchcock E, Kenny BG. Stereotaxy of third ventricular masses. Acta Neurochir Suppl (Wien) 1989;46:82-5. 61. Hitchcock E, Morris CS. Immunocytochemical techniques in stereotactic biopsy. Stereotact Funct Neurosurg 1989;53:21-8. 62. Hitchcock E, Cowie R. Stereotactic removal of intracranial foreign bodies: review and case report. Injury 1983;14:471-5. 63. Hitchcock E. Open stereotactic surgery. Acta Neurochir Suppl (Wien) 1991;52:9-12. 64. Hitchcock E, Kitchen G, Dalton E, Pope B. Stereotactic LINAC radiosurgery. Br J Neurosurg 1989;3:305-12. 65. Hitchcock E. Stereotactic neural transplantation. Stereotact Funct Neurosurg 1994;62:120-33. 66. Hitchcock ER, Kenny BG, Clough CG, Hughes RC, Henderson BT, Detta A. Stereotactic implantation of foetal mesencephalon (STIM): the UK experience. Prog Brain Res 1990;82:723-8. 67. Hitchcock ER, Clough CG, Hughes RC, Kenny BG. Transplantation in Parkinson’s disease: stereotactic implantation of adrenal medulla and foetal mesencephalon. Acta Neurochir Suppl (Wien) 1989;46:48-50. 68. Kudoh C, Detta A, Meyer C, Byrne P, Hitchcock E. Bilateral intracaudate cografting of fetal ventral mesencephalon and striatum in advanced Parkinson’s disease. Transplant Proc 1999;31:3397-402. 69. Freeman W, Watts JW. Psychosurgery in the treatment of mental disorders and intractable pain. 2nd ed. Springfield, IL: C. C. Thomas; 1950. 70. Moniz E. Tentatives Ope´ratoires dans le Traitement de Certaines Psychoses. Paris: Masson; 1936. 71. Richardson AE. Sir Wylie McKissock. Surg Neurol 1988;30:173-4. 72. Kitchen N. Neurosurgery for affective disorders at Atkinson Morley’s Hospital 1948–1994. Acta Neurochir Suppl 1995;64:64-8. 73. McKissock W. Rostral leucotomy. Lancet 1951;2:91-4. 74. McKissock W. Discussion on psychosurgery. Proc R Soc Med 1959;52:206-9. 75. Bell BA. Wylie McKissock – reminiscences of a commanding figure in British neurosurgery. Br J Neurosurg 1996;10:9-18.

76. El-Hai J. The lobotomist: a maverick medical genius and his tragic quest to rid the world of mental illness. Hoboken, NJ: Wiley; 2005. 77. Tooth J, Newton M. Leucotomy in England and Wales 1942–1954: report on public health and medical subjects no. 104. London: Her Majesty’s Stationery Office; 1961. 78. Knight GC. Discussion on psychosurgery. Proc R Soc Med 1959;52:209-10. 79. Knight GC, Tredgold RF. Orbital leucotomy; a review of 52 cases. Lancet 1955;268:981-6. 80. McCaul IR. Stereotaxic surgery and involuntary movement. Proc R Soc Med 1961;54:378-80. 81. Knight G. The orbital cortex as an objective in the surgical treatment of mental illness. The results of 450 cases of open operation and the development of the stereotactic approach. Br J Surg 1964;51:114-24. 82. Knight G. Stereotactic tractotomy in the surgical treatment of mental illness. J Neurol Neurosurg Psychiatry 1965;28:304-10. 83. Knight GC. Bi-frontal stereotactic tractotomy: an atraumatic operation of value in the treatment of intractable psychoneurosis. Br J Psychiatry 1969;115:257-66. 84. Knight G. Neurosurgical aspects of psychosurgery. Proc R Soc Med 1972;65:1099-104. 85. Knight G. Further observations from an experience of 660 cases of stereotactic tractotomy. Postgrad Med J 1973;49:845-54. 86. Bridges PK, Bartlett JR, Hale AS, Poynton AM, Malizia AL, Hodgkiss AD. Psychosurgery: stereotactic subcaudate tractomy. An indispensable treatment. Br J Psychiatry 1994;165:599-611; discussion 612-3. 87. Hodgkiss AD, Malizia AL, Bartlett JR, Bridges PK. Outcome after the psychosurgical operation of stereotactic subcaudate tractotomy, 1979–1991. J Neuropsychiatry Clin Neurosci 1995;7:230-4. 88. Malhi GS, Bartlett JR. A new lesion for the psychosurgical operation of stereotactic subcaudate tractotomy (SST). Br J Neurosurg 1998;12:335-9. 89. Malhi GS, Bartlett JR. Depression: a role for neurosurgery? Br J Neurosurg 2000;14:415-22; discussion 423. 90. Kelly D, Richardson A, Mitchell-Heggs N. Stereotactic limbic leucotomy: neurophysiological aspects and operative technique. Br J Psychiatry 1973;123:133-40. 91. Kelly D, Richardson A, Mitchell-Heggs N, Greenup J, Chen C, Hafner RJ. Stereotactic limbic leucotomy: a preliminary report on forty patients. Br J Psychiatry 1973;123:141-8. 92. Williams LM. Arousal dissociates amygdala and hippocampal fear responses: evidence from simultaneous fMRI and skin conductance recording. Neuroimage 2001;14:1070-9. 93. Christmas D, Matthews K, Eljamel MS. Neurosurgery for mental disorder. Br J Psychiatry 2004;185:173-4; author reply 174. 94. Feldman RP, Alterman RL, Goodrich JT. Contemporary psychosurgery and a look to the future. J Neurosurg 2001;95:944-56.


95. Kopell BH, Machado AG, Rezai AR. Not your father’s lobotomy: psychiatric surgery revisited. Clin Neurosurg 2005;52:315-30. 96. Marks PV, Wild AM, Gleave JR. The ‘Cambridge’ stereotactic biopsy instrument. Br J Neurosurg 1990;4:127-8. 97. Wild AM, Xuereb JH, Marks PV, Gleave JR. Computerized tomographic stereotaxy in the management of 200 consecutive intracranial mass lesions. Analysis of indications, benefits and outcome. Br J Neurosurg 1990;4:407-15. 98. Bennett. A stereotaxic apparatus for use in cerebral surgery. Br J Radiol 1960;33:343–51. 99. Blandy JP, Greig CM, Gillam SJ, Williams DI. Lives of the fellows of the Royal College of Surgeons of England: 1997–2002: vol. 9 with a cumulated index to vol. 1–8. London: Royal College of Surgeons of England; 2005. 100. Eljamel MS, Forster A, Tulley M, Matthews K. Staged functional neurosurgery using image fusion: electronic atlas and microelectrode recording at Dundee. Stereotact Funct Neurosurg 1999;73:140-2. 101. Pell MF, Thomas DG. The initial experience with the Cosman-Roberts-Wells stereotactic system. Br J Neurosurg 1991;5:123-8. 102. Clarke RH. Part II: Atlas of the frontal sections of the cranium and brain of the rhesus monkey. Baltimore, MD: Johns Hopkins; 1920. 103. Clarke RH, Henderson EE. Atlas of photographs of sections of the frozen cranium and brain of the cat (Felis Domestica). J Psychol Neurol 1912;18:391-409. 104. Spiegel EA, Wycis HT, Baird HW. Studies in stereoencephalotomy. I. Topical relationships of subcortical structures to the posterior commissure. Confin Neurol 1952;12:121-33. 105. Schaltenbrand G, Bailey P. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Thienne; 1959. 106. Talairach J, David M, Tournoux P, Corredor H, Krasina T. Atlas d’Anatomie Stereotaxique. Paris: Masson; 1957. 107. Andrew J, Watkins ES. Stereotactic surgery. A stereotaxic atlas of the human thalamus. Baltimore, MD: Williams & Wilkins; 1969. 108. Afshar F, Watkins ES, Yap JC. Stereotaxic atlas of the human brainstem and cerebellar nuclei: a variability study. New York: Raven Press; 1978. 109. Mazzone P, Insola A, Lozano A, Galati S, Scarnati E, Peppe A, Stanzione P, Stefani A. Peripeduncular and pedunculopontine nuclei: a dispute on a clinically relevant target. Neuroreport 2007;18:1407-8. 110. Zrinzo L, Zrinzo LV, Hariz M. The pedunculopontine and peripeduncular nuclei: a tale of two structures. Brain 2007;130:e73; author reply e74. 111. Corsellis JA. Individual variation in the size of the tentorial opening. J Neurol Neurosurg Psychiatry 1958;21:279-83.

8

112. Watkins ES. On making stereotactic atlases. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw Hill; 1998. p. 235-6. 113. Cotzias GC, Van WM, Schiffer LM. Aromatic amino acids and modification of parkinsonism. N Engl J Med 1967;276:374-9. 114. Petrik V, Apok V, Britton JA, Bell BA, Papadopoulos MC. Godfrey Hounsfield and the dawn of computed tomography. Neurosurgery 2006;58:780-7; discussion 780-7. 115. Ambrose J, Gooding MR, Richardson AE. The Lancet – saturday II october 1975. Sodium iothalamate as an aid to diagnosis of intracranial lesions by computerised transverse axial scanning. Lancet 1975;2:669-74. 116. Ambrose J. CT scanning: a backward look. Semin Roentgenol 1977;12:7-11. 117. Ambrose J, Hounsfield G. Computerized transverse axial tomography. Br J Radiol 1973;46:148-9. 118. Hounsfield GN. Historical notes on computerized axial tomography. J Can Assoc Radiol 1976;27:135-42. 119. Hounsfield GN. Computed medical imaging. Science 1980;210:22-8. 120. Hitchcock E. Stereotactic-computerized tomography interface device. Appl Neurophysiol 1987;50:63-7. 121. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta Chir Scand 1949;99:229-33. 122. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-9. 123. Rowe JG, Radatz MW, Walton L, Kemeny AA. Changing utilization of stereotactic radiosurgery in the UK: the Sheffield experience. Br J Neurosurg 2002;16:477-82. 124. Benabid AL, Pollak P, Louveau A, Henry S, de Rougement J. Combined (thalamotomy and stimulation) stereotactic surgery of the VIM thalamic nucleus for bilateral Parkinson disease. Appl Neurophysiol 1987;50:344-6. 125. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. 126. Aziz TZ, Adams CB. Neurosurgery at the radcliffe infirmary, Oxford: a history. Neurosurgery 1995;37:505-10. 127. Bergman H, Wichmann T, DeLong MR. Reversal of experimental parkinsonism by lesions of the subthalamic nucleus. Science 1990;249:1436-8. 128. Aziz TZ, Peggs D, Sambrook MA, Crossman AR. Lesion of the subthalamic nucleus for the alleviation of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)induced parkinsonism in the primate. Mov Disord 1991;6:288-92. 129. Aziz TZ, Peggs D, Agarwal E, Sambrook MA, Crossman AR. Subthalamic nucleotomy alleviates parkinsonism in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-exposed primate. Br J Neurosurg 1992;6:575-82.

93

94

8


130. Parkin S, Nandi D, Giladi N, Joint C, Gregory R, Bain P, Scott R, Aziz TZ. Lesioning the subthalamic nucleus in the treatment of Parkinson’s disease. Stereotact Funct Neurosurg 2002;77:68-72. 131. Parkin SG, Gregory RP, Scott R, Bain P, Silburn P, Hall B, Boyle R, Joint C, Aziz TZ. Unilateral and bilateral pallidotomy for idiopathic Parkinson’s disease: a case series of 115 patients. Mov Disord 2002;17:682-92. 132. Liu X, Rowe J, Nandi D, Hayward G, Parkin S, Stein J, Aziz T. Localisation of the subthalamic nucleus using radionics image Fusion? and Stereoplan? combined with field potential recording: a technical note. Stereotact Funct Neurosurg 2001;76:63-73. 133. Aziz TZ, Nandi D, Parkin S, Liu X, Giladi N, Bain P, Gregory RG, Joint C, Scott RB, Stein JF. Targeting the subthalamic nucleus. Stereotact Funct Neurosurg 2002;77:87-90. 134. Aziz T, Torrens M. CT-guided thalamotomy in the treatment of movement disorders. Br J Neurosurg 1989;3:333-6. 135. Pereira EA, Green AL, Nandi D, Aziz TZ. Deep brain stimulation: indications and evidence. Expert Rev Med Devices 2007;4:591-603. 136. Munro-Davies LE, Winter J, Aziz TZ, Stein JF. The role of the pedunculopontine region in basal-ganglia mechanisms of akinesia. Exp Brain Res 1999;129:511-7. 137. Nandi D, Aziz TZ, Giladi N, Winter J, Stein JF. Reversal of akinesia in experimental parkinsonism by GABA antagonist microinjections in the pedunculopontine nucleus. Brain 2002;125:2418-30. 138. Jenkinson N, Nandi D, Miall RC, Stein JF, Aziz TZ. Pedunculopontine nucleus stimulation improves akinesia in a parkinsonian monkey. NeuroReport 2004;15:2621-4. 139. Nandi D, Jenkinson E, Miall C, Stein JF, Aziz TZ. Pedunculopontine nucleus. J Neurosurg 2004;100:978-9; author reply 979. 140. Kleiner-Fisman G, Herzog J, Fisman DN, Tamma F, Lyons KE, Pahwa R, Lang AE, Deuschl G. Subthalamic nucleus deep brain stimulation: summary and metaanalysis of outcomes. Mov Disord 2006;21 Suppl 14: S290-304. 141. Stefani A, Lozano AM, Peppe A, Stanzione P, Galati S, Tropepi D, Pierantozzi M, Brusa L, Scarnati E, Mazzone P. Bilateral deep brain stimulation of the pedunculopontine and subthalamic nuclei in severe Parkinson’s disease. Brain 2007;130(6):1596-1607. 142. Schupbach WM, Chastan N, Welter ML, Houeto JL, Mesnage V, Bonnet AM, Czernecki V, Maltete D, Hartmann A, Mallet L, Pidoux B, Dormont D, Navarro S, Cornu P, Mallet A, Agid Y. Stimulation of the subthalamic nucleus in Parkinson’s disease: a 5 year follow up. J Neurol Neurosurg Psychiatry 2005;76:1640-4. 143. Rodriguez-Oroz MC, Obeso JA, Lang AE, Houeto JL, Pollak P, Rehncrona S, Kulisevsky J, Albanese A,

144.

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

Volkmann J, Hariz MI, Quinn NP, Speelman JD, Guridi J, Zamarbide I, Gironell A, Molet J, Pascual-Sedano B, Pidoux B, Bonnet AM, Agid Y, Xie J, Benabid AL, Lozano AM, Saint-Cyr J, Romito L, Contarino MF, Scerrati M, Fraix V, Van Blercom N. Bilateral deep brain stimulation in Parkinson’s disease: a multicentre study with 4 years follow-up. Brain 2005; 128:2240-9. Liu X, Miall RC, Aziz TZ, Palace JA, Stein JF. Distal versus proximal arm tremor in multiple sclerosis assessed by visually guided tracking tasks. J Neurol Neurosurg Psychiatry 1999;66:43-7. Liu X, Yianni J, Wang S, Bain PG, Stein JF, Aziz TZ. Different mechanisms may generate sustained hypertonic and rhythmic bursting muscle activity in idiopathic dystonia. Exp Neurol 2006;198:204-13. Wang S, Liu X, Yianni J, Green AL, Joint C, Stein JF, Bain PG, Gregory R, Aziz TZ. Use of surface electromyography to assess and select patients with idiopathic dystonia for bilateral pallidal stimulation. J Neurosurg 2006;105:21-5. Pereira EA, Green AL, Bradley KM, Soper N, Moir L, Stein JF, Aziz TZ. Regional cerebral perfusion differences between periventricular grey, thalamic and dual target deep brain stimulation for chronic neuropathic pain. Stereotact Funct Neurosurg 2007;85:175-83. Kringelbach ML, Jenkinson N, Green AL, Owen SL, Hansen PC, Cornelissen PL, Holliday IE, Stein J, Aziz TZ. Deep brain stimulation for chronic pain investigated with magnetoencephalography. Neuroreport 2007;18:223-8. Muthusamy KA, Aravamuthan BR, Kringelbach ML, Jenkinson N, Voets NL, Johansen-Berg H, Stein JF, Aziz TZ. Connectivity of the human pedunculopontine nucleus region and diffusion tensor imaging in surgical targeting. J Neurosurg 2007;107:814-20. Aravamuthan BR, Muthusamy KA, Stein JF, Aziz TZ, Johansen-Berg H. Topography of cortical and subcortical connections of the human pedunculopontine and subthalamic nuclei. Neuroimage 2007;37:694-705. Green AL, Wang S, Owen SL, Paterson DJ, Stein JF, Aziz TZ. Controlling the heart via the brain: a potential new therapy for orthostatic hypotension. Neurosurgery 2006;58:1176-83; discussion 1176‐83. Green AL, Wang S, Owen SL, Xie K, Bittar RG, Stein JF, Paterson DJ, Aziz TZ. Stimulating the human midbrain to reveal the link between pain and blood pressure. Pain 2006;124:349-59. Yianni J, Bain P, Giladi N, Auca M, Gregory R, Joint C, Nandi D, Stein J, Scott R, Aziz T. Globus pallidus internus deep brain stimulation for dystonic conditions: a prospective audit. Mov Disord 2003;18:436-42. Owen SL, Green AL, Nandi DD, Bittar RG, Wang S, Aziz TZ. Deep brain stimulation for neuropathic pain. Acta Neurochir Suppl 2007;97:111-16. Kitchen ND, Thomas DG. Minimally invasive stereotaxy: clinical use of the Gill-Thomas-Cosman (GTC)


156.

157.

158.

159.

160.

161.

162.

repeat stereotactic localiser. Minim Invasive Neurosurg 1994;37:61-3. Sofat A, Kratimenos G, Thomas DG. Early experience with the Gill Thomas locator for computed tomographydirected stereotactic biopsy of intracranial lesions. Neurosurgery 1992;31:972-4. Gill SS, Patel NK, Hotton GR, O’Sullivan K, McCarter R, Bunnage M, Brooks DJ, Svendsen CN, Heywood P. Direct brain infusion of glial cell line-derived neurotrophic factor in Parkinson disease. Nat Med 2003;9:589-95. Love S, Plaha P, Patel NK, Hotton GR, Brooks DJ, Gill SS. Glial cell line-derived neurotrophic factor induces neuronal sprouting in human brain. Nat Med 2005;11:703-4. Patel NK, Bunnage M, Plaha P, Svendsen CN, Heywood P, Gill SS. Intraputamenal infusion of glial cell line-derived neurotrophic factor in PD: a two-year outcome study. Ann Neurol 2005;57:298-302. Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine nucleus for Parkinson’s disease. Neuroreport 2005;16:1883-7. Page RD, Miles JB. Validation of CT targeting for functional stereotaxis with postoperative magnetic resonance imaging. Br J Neurosurg 1994;8:461-7. Dervin JE, Miles J. A simple system for image directed stereotaxis. Br J Neurosurg 1989;3:569-74.

8

163. Dervin JE, Miles JB. Development of an analogue method to link stereotactic surgery to computed tomography. Neurochirurgia (Stuttg) 1984;27:162-5. 164. Dorward NL, Alberti O, Palmer JD, Kitchen ND, Thomas DG. Accuracy of true frameless stereotaxy: in vivo measurement and laboratory phantom studies. Technical note. J Neurosurg 1999;90:160-8. 165. Thomas DG, Kitchen ND. Stereotactic techniques for brain biopsies. Arch Dis Child 1993;69:621-2. 166. Kitchen ND, Lemieux L, Thomas DG. Accuracy in frame-based and frameless stereotaxy. Stereotact Funct Neurosurg 1993;61:195-206. 167. Thomas DG, Nouby RM. Experience in 300 cases of CTdirected stereotactic surgery for lesion biopsy and aspiration of haematoma. Br J Neurosurg 1989;3:321-5. 168. Thomas DG, Anderson RE, du Boulay GH. CT-guided stereotactic neurosurgery: experience in 24 cases with a new stereotactic system. J Neurol Neurosurg Psychiatry 1984;47:9-16. 169. Powell M, Kitchen N. The development of neurosurgery at the National Hospital for Neurology and Neurosurgery, Queen Square, London, England. Neurosurgery 2007;61:1077-90; discussion 1090. 170. Gildenberg PL. Stereotactic surgery: the early years (comment). Neurosurgery 2004;55:1214.

95

12 History of Stereotactic Surgery in India P. K. Doshi

Ancient Indian Literature The earliest reference to any functional neurosurgery in world could be found in Indian mythology script Shiva Purana (http://is1.mum.edu/vedicreserve/puran.htm). It reflects a transplantation of elephant head on a human being (Ganesha). Ganesha – the elephant-deity riding a mouse – has become one of the most common mnemonics for anything associated with Hinduism. The son of Shiva and Parvati, Ganesha has an elephantine countenance with a curved trunk and big ears, and a huge pot-bellied body of a human being. (> Figure 12-1) He is worshipped as Lord of success and destroyer of evils and obstacles. He is also worshipped as the god of education, knowledge, wisdom, and wealth. The story of the birth of this zoomorphic deity, as depicted in the Shiva Purana, goes like this: Once goddess Parvati, while bathing, created a boy out of the dirt of her body and assigned him the task of guarding the entrance to her bathroom. When Shiva, her husband returned, he was surprised to find a stranger denying him access, and struck off the boy’s head in rage. Parvati broke down in utter grief and to soothe her, Shiva sent out his squad (gana) to fetch the head of any sleeping being who was facing the north. The company found a sleeping elephant and brought back its severed head, which was then attached to the body of the boy. Shiva restored its life and made him the leader (pati) of his troop, hence his name ‘‘Ganapati.’’ Shiva also bestowed a boon that people would worship him and invoke his name before undertaking any venture. #


This narration highlights two important aspects of functional neurosurgery. One is the reproduction of a human being from dermis derived stem cell (‘‘created a boy out of the dirt of her body’’) and second, the ultimate aim/ achievement that could ever happen – ‘‘whole head transplant.’’ In another epic, it is documented that in 1800 BC, Jivaka (physician to the Lord Buddha) removed intracranial mass lesion through trephination.

Development of Stereotactic Surgery Neurosurgery in India is a post World War II development, resulting from the keen desire of the new rulers of independent India, that the country should keep up with all the modern advances in every field of Neurosurgery [1]. It is interesting to note that many of the pioneers of Indian neurosurgery were exposed to stereotactic and functional neurosurgery, during their training abroad, which helped them to develop stereotactic neurosurgical specialty parallel with the international advances. Stereotactic surgery developed in parallel across several centers in India. In this chapter, the development of each subset of functional neurosurgery is described separately along with the contributions of each center, rather than following the chronology of date. Most important references have been included but these are not necessarily all encompassing.

156

12

History of stereotactic surgery in india

. Figure 12-1 Lord Ganesha (Elephant God)

In the 1940s bold pioneers like Chintan Nambiar, a surgeon at Stanley medical college, Madras, used to perform freehand stereotactic lesions by using a template in the temporal region. While mentioning this in an oration, the pioneering stereotactic surgeon B Ramamurthi quotes ‘‘The surgeon was bold and the patients bolder.’’ He performed 74 cases of chemopallidectomy using this free hand technique, out of which 27 had excellent results [2]. The first neurosurgical set up was established at the Christian Medical College (CMC), Vellore, in Tamilnadu, by Jacob Chandy. Chandy had obtained 2 years of training under Wilder Penfield at the Montreal Neurological Institute (MNI). In January 1949, equipped with neurosurgical equipments brought from Canada and an EEG machine, Chandy started the neurosurgical unit at CMC. He started with ten beds spread across medical and surgical wards. In 1950, he

was joined by Baldev Singh, a neurologist. In 1962, CMC acquired the Bertrand stereotactic guide, with the help of which they performed surgeries for Parkinson’s disease and epilepsies. Later in 1987, KV Mathai procured BRW frame. CMC became an established stereotactic unit, where till date around 1,800 stereotactic biopsies, 400 stereotactic craniotomies, 100 functional neurosurgical procedures, and 700 radiosurgical procedures have been performed. Vedantam Rajashekar, the present Head of the Department of Neurosurgery, and past president of the Indian Society of Stereotactic and Functional Neurosurgery, has been conducting stereotactic workshop to train young neurosurgeons [1]. Following on the heels of the department at Vellore was that at the Madras Medical College and Government Hospital; with the joining of Ramamurthi, in October 1950. Ramamurthi started his neurosurgical training with Rowbotham, in Newcastle upon Tyne, England. Subsequently he visited numerous centers across Europe and USA to gain wider experience. He also visited MNI and observed Rasmussen’s and Penfield’s work. Ramamurthi started neurosurgical unit with four beds which were increased to ten after 18 months [1]. Inspired by Irving Cooper’s use of an inflatable balloon to make lesions in pallidum for the treatment of Parkinson’s diseases, V Balasubramaniam and Ramamurthi performed surgeries using Cooper’s balloon (1962) [3]. Under radiological guidance a balloon was introduced and left in place for 48 h. This was followed by alcohol ablation. After performing surgeries on 12 patients of movement disorders, they were discouraged with the results. Import restrictions and bureaucratic hurdles made further imports of these balloons difficult and hence they gave up this method. In 1960 Ramamurthi received an invitation for dinner with the Governor of Hyderabad, General Shrinagesh (> Figure 12-2). It happened that the Governor was suffering from Parkinson’s disease and had undergone unilateral lesion in London by Lawrence Walsh at Atkinson Morley Hospital.


12

. Figure 12-2 Ramamurthi having dinner with, General Shrinagesh (Black Suit), Governor of Andhra Pradesh

He asked Ramamurthi if such facilities are available in India as he had started experiencing symptoms on the opposite side. Ramamurthi explained that though he had the necessary expertise, the equipments were not available. In those days it was very difficult to import any equipment or obtain foreign exchange to travel abroad. General Shrinagesh immediately called up the Prime Minister of India, Pandit Jawaharlal Nehru; it was 11 PM. The Prime Minister immediately agreed to Shrinagesh’s suggestion to bring in Lawrence Walsh and Denis Williams (Neurologist) from England to conduct a workshop and training for movement disorders surgery. They also obtained permission for Walsh to leave his equipment behind after the workshop. Walsh and Williams came and stayed in Madras for 3 weeks during which they performed 40 surgeries and 30 neuroscientists took advantage of their expertise (> Figure 12-3). On completion of the program as decided, they left the Leksell stereotactic apparatus and the lesion generator back in Madras [4]. This was a major impetus and from there on Madras became a leading stereotactic center where more than 1,700 procedures were performed between 1959 and 1975 (> Table 12-1). Presently very little work is being

done at this center after Ramamurthi and his team retired from the Madras Medical College. In 1970, S Kalyanaraman, a young neurosurgeon in Ramamurthi’s team from the Madras Medical College, reported simultaneous use of two stereotactic apparatus on the same patient. He used the Leksell stereotactic equipment in combination with the Sehgal stereotactic equipment to perform simultaneous targeting of intracranial structures. Sehgal’s stereotactic equipment is a compact, burr hole based stereotactic device designed by Arjun Sehgal in India. The Leksell frame was used to align Sehgal’s apparatus to the target on one side and on the other it was used to approach the target. The purpose was to obtain simultaneous recording from the thalamus, thus reducing the operative time. They also observed that the number of X-rays required to localize the targets were reduced [5,6]. RM Varma started a neurosurgical unit in the (then) All India Institute of Medical Health, Bangalore, in 1958. Varma’s efforts led to the formation of the National Institute of Mental Health and Neurosciences (NIMHANS) out of this unit [7]. Varma was trained in Bristol and started performing lesions for Parkinson’s disease using a unique free hand technique, through

157

158

12


. Figure 12-3 Lawrence Walsh, Chief Nurse, Ramamurthi and Dennis Williams

. Table 12-1 Stereotactic Surgery at Madras Medical College (1959– 1975) Thalamotomy Amygdalotomy Hypothalamotomy Cingulumotomy Basofrontal tractotomy Dentatectomy Leucotomy Thalamolaminotomy Capsulotomy Pulvinotomy Mesencephalc reticulotomy Hypophysectomy

858 480 122 143 56 73 5 11 16 4 2 2

foramen ovale. Later on in 1980s, they obtained the Leksell stereotactic equipment and recently a gamma knife unit. HM Dastur, started stereotactic surgery at King Edward Memorial (KEM) Hospital, Bombay (Mumbai) in 1959. The initial surgeries were performed using Oliver’s guide. Narabayashi visited KEM hospital in 1962 and lent the design of his stereotactic frame for fabricating a local frame on similar lines. Later on in 1975, Dastur joined Jaslok Hospital, Bombay (Mumbai), where he continued to perform stereotactic surgery using

a Reichert-Mundinger frame. In another center at Bombay Hospital, SN Bhagwati started stereotactic surgery with Mckinney’s apparatus from 1962 and this was replaced by Leksell’s frame in 1964. Multiple centers started practicing stereotactic surgery in the 1970s. These include Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh (1974), with Mckinney frame, All India Institute of Medical Sciences (AIIMS), New Delhi (1977), with Leksell frame and Shri Chitra Tirunal Institute for Medical Sciences and Technology (SCTIMST) with Leksell frame [2,8]. In 1991 Mr. Bose, an engineer and Apte, a neurosurgeon from Pune, developed an indigenous arc centered stereotactic apparatus. Three revisions have been made to this initial model, which is now compatible for CT and MRI guided procedures. They followed this by manufacturing a radiofrequency lesion generator in 1996. The prices of these equipments are considerably lower than standard stereotactic frames and hence have become popular for performing biopsies and other stereotactic procedures. Presently 40 neurosurgical units have stereotactic equipment and


they perform various levels of stereotactic procedures from biopsy to functional neurosurgery. The Indian Society of Stereotactic and Functional Neurosurgery was formed in 1997. Its first meeting was held in Delhi with Balasubramaniam as president and Rajashekar as Secretary. In 2007 the tenth meeting of the society was held in Kolkata. This is a rapidly expanding society with the current membership of more than 100 members. Stereotactic radiosurgery was first introduced in India at the Apollo Hospitals, Chennai, using Linac based X-knife system. Linac based radiosurgery is offered at multiple centers including Bombay Hospital and Jaslok Hospital, Mumbai. Gamma Kinfe was introduced at Hinduja Hospital, Mumbai, in 1997 [9]. This was soon followed by similar units at AIIMS and VIMHANS in New Delhi, and later on at other centers including Vellore, NIMHANS, R&R (Army Hospital, New Delhi), PGIMER (Chandigarh) etc.

Epilepsy Surgery In the early days of stereotactic and functional neurosurgery the enthusiasm of the neuroscientists could not be contained. Virtually lesions were placed in every part of the brain for varied disorders ranging from epilepsy to psychiatric illness [10]. To further understand the development drivers of these surgeries, we need to look into India’s social and economical background of that period. A recent (2006) survey (http://www. prb.org/Articles/2006/CommunityBasedHealth InsuranceShowsPromiseinIndia.aspx) showed that only 11% of Indian population had some kind of health insurance, this could not have been more than 2–4% in 1960s. Thus the medical treatment had to be funded by the patient themselves or they had to use government or municipal hospitals to treat them. Though these hospitals were supposed to be providing free medical treatment, the medicines (especially the expensive ones)

12

had to be purchased by the patient. Paradoxically the cost of surgeries in these hospitals was far less than prolonged medical treatment. In a report on epilepsy research, Mathai [11] mentions that follow-up studies indicated that where the person was receiving more than two drugs the cost of the therapy would come >2 $/month and only 10–15% of the patients had the resources to afford this. Hence in developing countries financial insufficiency formed an added indication for surgical therapy in the control of seizures. Chandy, who had trained at the MNI, started epilepsy surgery in CMC, Vellore in 1949. He was joined by Baldev Singh, a neurologist, who had interest in epilepsy and was equipped with an EEG machine to diagnose and manage epilepsy. Singh and Chandy [12] in an exhaustive study of delta waves in 800 EEG records at CMC, commented on their characteristics and localizing value. Those were the days when EEG was the only non-invasive investigation for brain disorders (including tumors). During 1949–1990, 141 epilepsy surgeries were performed at CMC [13]. They performed topectomy and lobectomy for suprasylvian epilepsy; for temporal lobe epilepsy (TLE) the surgical procedures done were topectomy, temporal lobectomy with amygdalectomy, temporal lobectomy with amygdalohippocampectomy, and only amygdalectomy. Hemispherectomy was done for cases with multilobar epilepsy. Temporal lobe resections were done based on the scalp EEG, sphenoidal studies, neuropsychological assessment and the intraoperative ECoG and depth electrode studies. Total or near total seizure control was obtained in 53% patients and a satisfactory outcome in 20% patients. They found that mental retardation, pre operative scalp EEG and post excision electrocorticography were predictors of outcome. In Madras, Ramamurthi developed an excellent team. He along with Balasubramaniam, Kalyanaraman and Kanaka as neurosurgical colleagues; Arjundas, Jaganathan and later on Sayeed as neurological colleagues, Vriddhagirinathan the

159

160

12


psychologist and Valmikinathan, neurochemist developed a comprehensive epilepsy surgical program. They reviewed the literature and noted that Falconer had obtained good results by standard temporal lobectomies in three fourth of the patients [14]. Narabayashi and Chitanondh had reported 50% seizure control and improvement in behavioral disorders in patients undergoing amygdalectomy. Ramamurthi felt that massive temporal lobectomy, only to ablate these medial structures, is a mutilating procedure. They treated complex partial seizures with proved medial temporal focus with stereotactic lesions rather than by a full temporal lobectomy [15–19]. They established the details of localization of epileptic focus by careful pre and postoperative observations of their neurological colleagues. In a paper published in 1979 [20], Ramamurthi shares his experience of 56 cases that he operated. The patients included, suffered from purely TLE, TLE with secondary generalization or TLE with focal seizures. EEG studies including sphenoidal and depth electrode recordings were performed prior to surgery. Based on these findings, stereotactic lesions of 600–800 m3, were made in the area of maximum abnormality. He found that 26 of the 56 patients became seizure free and eight cases required reoperation. He commented that this procedure was useful in patients who had bilateral TLE or contralateral previous temporal lobectomy, as it preserved the hippocampus and thus the memory. Mathai, in CMC Vellore, also performed amygdalectomy through a craniotomy for similar reasons. They performed intraoperative corticography and depth studies from amygdala to plan their surgical resections. They concluded that when the seizure discharges are from the cortex rather than amygdala, excision of the amygdala only reduces the intensity of cortical activity. However, when the epileptiform discharges are primarily from the amygdala, amygdalectomy alone might suffice; especially when an electrocorticographic seizure can be produced by stimulating the

amygdala [11]. In Bombay (Mumbai), at KEM hospital Dastur, assisted by neurologist Anil Desai, performed epilepsy surgery including temporal lobectomy and hemispherectomy. He continued his work after joining Jaslok Hospital, where he was assisted by Mrs. PN Wadia, neurophysiologist in performing epilepsy surgery. Corticography and depth recordings during the surgery were also performed. During 1960s depth electrodes and corticography were used routinely for epilepsy surgeries. Arjundas observed that the depth electrode supplements information obtained from scalp EEG in identifying the epileptogenic focus and also reveals foci not evident in scalp EEG records [21]. Kanaka and Balasubramaniam found that depth studies were useful in understanding the propagation of epileptic discharges. They found that depth electrode study along with electrocorticography also provided precise localization of the epileptic focus. They used this information for planning appropriate surgery [22]. Kalyanaraman had obtained PhD degree from Edinburgh, University on ‘‘Anatomical and Physiological studies on the internal capsule and adjacent diencephalic structures during human stereotaxy.’’ Based on his work and the observations of Gillingham [23], Kalyanaraman postulated that in an area in the posterior limb of the internal capsule, medial to the pyramidal tract where lie the corticospinal fibers that form the common pathway for the epileptic seizure. A bilateral internal capsulotomy should control clinical seizures without causing pyramidal signs and without producing electroencephalographic improvement. To test this hypothesis, they performed bilateral internal capsulotomy on seven patients with intractable grand mal epilepsy. They had good outcome (comparable to Engel grade I and II) in three patients, considerable reduction in seizure frequency in one and poor outcome in two. One patient died of pneumonia postoperatively. None of these patients had significant permanent morbidity [24,25].


Mathai postulated that in patients suffering from generalized seizures or focal seizures arising from significant (eloquent) cortical areas, which cannot be excised without producing serious neurological deficit, interruption of propagation pathways may modify the frequency and pattern of the seizures. They performed stereotactic lesioning of ansa and fasciculus lenticularis. Bertrand stereotactic guide was used and the target chosen was 1 cm behind AC and 1.5 cm lateral. A leucotome was used to make a lesion of 0.5–0.8 cm. Scalp and depth EEG recordings were obtained during this stage. They had variable results, ranging from good seizure control to reduction in seizure frequency and severity. They concluded that because of the multiple propagation pathways in epilepsy these procedures may only temporarily modify the clinical seizure pattern. However, in seizures where the frequency is once a day or more such procedures are of value. Destruction of ansa and fasciculus lenticularis bilaterally for generalized cerebral seizures and unilaterally for focal cortical seizures seems to alleviate, although not fully, the intensity and frequency of seizures [26]. Few epilepsy surgeries were performed during 1970–1990. Radhakrishnan, epileptologist, started an epilepsy surgery program at the SCTIMST. This was a complete program with facilities for invasive recording, neuropsychology, video EEG and nuclear medicine. Their initial focus was on temporal lobe epilepsy. From March 1995 through February 2002, they performed 394 epilepsy surgeries, 370 of them were anterior temporal lobectomy with amygdalohippocampectomy for refractory temporal lobe epilepsy. They reported 78% seizure freedom at 2 years follow-up [27]. Epilepsy surgery was started in AIIMS, by VP Singh [28], at Jaslok Hospital by P Doshi and at Hinduja Hospital by CE Deopujari and BK Misra in the late 1990s. All these centers have facilities of state of the art neuroradiology, neurophysiology

12

(including prolonged video EEG), nuclear medicine departments for SPECT and PET scans, neuropsychologist, epileptologist and functional neurosurgeon.

Movement Disorders Surgery Movement disorders surgery has had the most colorful history in any account of functional neurosurgery. This is true for India as well. Enterprising neurosurgeons used varied techniques from free hand technique to frame based systems, unilateral and bilateral lesions, even simultaneous bilateral lesions; alcohol to radiofrequency lesioning and from pallidum to field of Forel, to ameliorate movement disorders. Balasubramaniam and Ramamurthi started performing chemopallidectomy using Cooper’s balloon in 1962 as mentioned earlier [3]. From 1964 they used the Leksell’s apparatus to perform thermal lesions for movement disorders [29]. Kalyanaraman, performed bilateral simultaneous thalamotomies in patients with various movement disorders. They observed that the complication rate was acceptable and no greater than staged procedures in patients with advanced Parkinson’s disease. In patients with bilateral intention tremors this formed a good surgical option as it avoided double hospitalization [30]. Using Sehgal’s stereotactic apparatus he used to perform bilateral simultaneous recordings as described earlier. During surgery an opaque marker was introduced into the site of the lesion. Immediate postoperative X-rays were taken and the exact location of the lesion was charted with the help of the atlas prepared by Schaltendrand and Bailey. This served not only to assess the accuracy of lesion placement but also to correlate the result of surgery with the site of the lesion. They noted minor differences in the anatomical calculations of deep brain structures in different races and groups [31].

161

162

12


Though in 1970s, the number of surgeries for Parkinson’s disease started decreasing, due to the cost constraints of prolonged Levodopa therapy excellent surgical benefits (especially for unilateral disease) Madras group continued to perform surgeries for Parkinson’s disease. Another unique area of interests was the treatment for cerebral palsy. Following the work of Narabayashi [32], Balasubramaniam and his colleagues operated on a large number of cerebral palsy patients. As their experience evolved they chose different targets depending on the predominant symptom complex. For Rigidity they made a lesion in the area below the ventrolateral nucleus (VL); for dyskinesias they used variety of targets including the ventralis intermedius nucleus (VIM), the centromedian nucleus (CM) and the dentate nucleus of the cerebellum. As most of these surgeries were done under general anesthesia verification of the electrode placement by stimulation, as done in Parkinson’s disease was not possible. Stimulation was still done to exclude electrode placement in the corticospinal tract [33]. They later on introduced stereotactic dentatectomy for patients with predominant spasticity. They found that VL and sub VL lesions were effective for rigidity, whereas for patients with a mix of rigidity and spasticity these lesions had to be supplemented by dentatectomy. Patients with sensory induced involuntary movements benefited from centromedian thalamotomy [34]. For severe hyperkinetic disorders, Kanaka found hypothalamotomy to play a distinct role in their management. She observed that it works because the area destroyed forms part of the limbic system. It seemed to be more on the ‘‘effector’’ side. It does not cause any morbidity. In the management hyperkinetic behavior disorders the first target to be destroyed must be the amygdaloid nucleus. If this operation fails, then hypothalamotomy may be done as the next operation [35]. Varma (1964) in Bangalore developed a free hand technique of lesioning the thalamus for

Parkinson’s disease. He modified the technique of Arthur Ecker and Theoder Perl [36] for this surgery. The technique involved use of two needles, outer 19G, to cannulate the foramen ovale; and the inner, 26G to perform chemothalamotomy. The outer needle had a slight curve at the end to help direct the inner needle to the desired position (> Figure 12-4). Varma used external landmarks to do away with ventriculography. He used lead pellets placed on the external canthus of the eye and in the internal auditory meatus to serve as markers. Lateral and AP radiographs were obtained after the introduction of the needle. A topograph was created outlining the above landmarks in relation to the venterolateral thalamus (based on an atlas). This was then overlain on the lateral x-ray and the distance between the needle and the target obtained. Appropriate radiographic and anthropometric corrections were applied to calculate the relationship of the needle to the target and the final position of needle adjusted. Varma notes that in many patients the tremor used to get arrested when the needle reached the target. This was then followed by chemothalamotomy with absolute alcohol [37]. His work was later on reviewed with MRI imaging of the patients operated, by . Figure 12-4 Varma’s Foramen Ovale ‘‘Thalamotomy’’ for PD. Outer and inner needle seen in situ, with cranial landmarks outlined


Uday Muthane [38,39]. Muthane analyzed the site of the lesion by MRI and, in one case, postmortem examination. He noted that the lesion was actually placed 1.5 cm below the thalamus and it coincided with the subthalamic nucleus. Stereotactic surgery was performed initially at CMC, Vellore by the free hand technique and later from 1961 onwards using Bertrand’s frame. Parkinson’s disease and dystonia were the main indications. The initial target used was globus pallidum, which was, later on changed to the thalamus. The free hand technique involved localizing the foramen of Monro using pneumoenchephalogram and pallidal target was calculated based on stereotactic atlas derived coordinates. AP and lateral radiographs were used for localization. The lesions were made using absolute alcohol in incremental methods whilst checking for neurological deficits (Mathai KV, personal communication). In Bombay (Mumbai), Dastur started performing lesions in thalamus, pallidum and field of Forel for dystonia at the KEM hospital in 1959 (Dastur HM, personal communication). Desai, neurologist associated with KEM neurosurgical department, visited Narabayashi in 1962, to assist Dastur in performing movement disorders surgery. The program received further impetus following Narabayashi’s visit. Dastur recounts a very interesting experience. They planned to perform a Cooper’s lesion for a patient with severe ‘‘hyperkinesis.’’ Following surgery, the patient had a remarkable improvement. However, on postoperative x-ray analysis they found that their lesions were not in the intended area but the lesions were more in the ventrolateral thalamus above the CA-CP plane. Gajendra Sinh, neurosurgeon at the Jaslok Hospital, used to call this a KEM lesion. The lesions were made using myodil and wax (> Figure 12-5). In 1974 Sinh invited Laitinen to visit Jaslok Hospital. Laitinen demonstrated his pallidotomy operations, which were then followed up by Sinh. After the 1980s movement disorders surgery

12

. Figure 12-5 Notings of ‘‘KEM thalamotomy’’ by Dastur

almost stopped following remarkable improvement in medical management. In another unit in Bombay (Mumbai), Bhagwati and his colleagues performed thalamotomies for Parkinson’s disease in 110 cases. They were also convinced about the efficacy of bilateral lesions and 30 of these patients had undergone bilateral lesions. They reported an interesting observation of reactivation of successfully abolished tremors on thalamic stimulation whilst performing surgery for the second side in Parkinson’s disease. In five of the thirty patients who underwent staged (minimum interval 6 months) bilateral thalamotomy, tremors recurred on the side of the second thalamotomy, i.e., on the ipsilateral side. In three patients, they performed a repeat lesion for the control of these tremors, but the patients had a rather stormy convalescence and developed drowsiness, confusion,

163

164

12


dysarthria with one of them developing pseudobulbar palsy, in the next two patients simultaneous bilateral lesions were not made. The recurrent tremors persisted in these two patients [40]. Kalyanaraman also had similar experience in patients undergoing bilateral thalamotomy. Bhagwati earlier used diathermy to make lesions but later on switched over to cryo lesioning [41]. Once again, like other places in the world, interest in Parkinson’s disease surgery waned after the introduction of Levodopa. Surgery for movement disorders surgery was revived in 1997. Doshi after his training in Europe started performing pallidotomy. However, within a short period (1998) he switched over to deep brain stimulation surgeries [42]. He and his colleagues observed that STN DBS can produce depression in an occasional patient [43], which was later on accepted as an important side effect in almost 8–18% of STN DBS patients [44,45]. SCTIMST also started performing pallidotomy and deep brain stimulation around the same time. Presently movement disorder surgery is being performed in Mumbai, Hyderabad, Banglore, Trivandrum and Delhi.

Surgery for Chronic Pain Roy offered stereotactic cingulotomy for intractable terminal cancer pain. He used Oliver’s apparatus. The needle was positioned to produce a lesion 2–4 cm posterior to anterior tips of the ventricles and 2 cm in vertical height from above the ventricles. 0.1 ml of carbolic acid with myodil was used in making the lesion. The maximum relief was noted approximately for 2 months after the operation. After this period, the relief of pain was not complete and analgesics were again required. They advocated this surgery for terminally ill cancer patients [46]. Dorsal cordotomy was the preferred procedure in patients suffering from pain of incurable

malignancy in the lower half of the body. However, when the upper half of the body was involved intracranial targets were chosen. Ramamurthi [47] notes that intractable pain due to lesions other than malignancy was not often seen in Indian neurosurgical practice. The incidence of oropharyngeal cancer was very common in South India and advanced cases reported with intractable pain. In such cases section of the trigeminal or the glossopharyngeal nerve was fraught with grave risks especially due to deglutition difficulties with resultant risk of aspiration. Kalyanramana and Ramamurthi studied the neurophysiology of the sensory relay nucleus [48]. From the atlas of Schaltenbrand and Bailey [49] they calculated that the facial area of the sensory relay nucleus of the thalamus was centered on a point 4 mm in front of the center of the posterior commissure, 4 mm above the intercommissural line and 13 mm lateral to the midsagittal plane. Whereas, the termination of the quintothalamic tract into the sensory relay nucleus was calculated to be centered around a point 3 mm in front of the posterior commissure, on the intercommissural line and 10 mm lateral to the midsagittal plane. They used neurophysiological guidance to further refine their target localization. They noted that though sensory responses could be obtained from the internal capsule or other parts of thalamus, the threshold of this response was lowest when the electrode was in the sensory relay nucleus. They also found that microelectrode recordings showed evoked potentials from peripheral stimulation. They initially made large lesions of 8 mm, five times, in the first few patients. However, later on they made only one lesion of 8 mm in the sensory thalamus and an additional lesion in the quintothalamic tract region if pain relief was not adequate. As most of these patients suffered from terminal cancer, they died after a few months and had adequate pain relief till they lived. Two patients who had post herpetic neuralgia


continued to have pain relief at a follow-up of 6 months [50]. They also performed cingulotomy and hypothalamotomy for pain relief [51].

Psychiatric Disorders Surgery There was a great interest and enthusiasm amongst Indian neurosurgeons in the field of psychiatric disorders surgery as early as 1940. Balkrishna Rao in Bangalore, performed prefrontal leucotomies on patients selected by Govindaswami [52]. BK Anand, a neurophysiologist, participated in the study of the role of the hypothalamus and limbic system in the regulation of feeding behavior conducted by John Fulton and JR Brobeck [53,54]. They discovered, what is since known as, the hypothalamic feeding center. On returning to India, he worked at the Lady Hardinge Medical College, New Delhi and later, at the Department of Physiology, AIIMS, New Delhi [55]. He introduced new experimental techniques and approaches for the study of brain and behavior in India. These include the methods of stereotactic placement of electrodes for making local electrolytic lesions, electrical stimulations, recording of depth EEG, evoked potentials and single unit potentials with microelectrodes and the usage of unanesthetized and free moving animals for behavioral experiments. Manchanda et al. [56] observed that electrical stimulation of perifornical regions of hypothalamus in the carnivore cat evoked different varieties of aggressive behavior: flight, defense, attack. These led other researchers and neurosurgeons to explore the vast field of psychiatric disorders surgery. During his training with Rowbotham, Ramamurthi used to visit St. Lukes Hospital at Middlesborough, where Rowbotham used to perform prefrontal leucotomies. He found that it was successful in more than 60% of the patients. He decided to undertake this upon return to India.

12

Ramamurthi found that the psychiatrists in South India were forward looking and readily referred cases for surgery [10]. For severe depression, Ramamurthi performed a stamp size lesion, extending from 0.7 to 3.0 cm from the midline and 0.8 to 2.8 cm in front of the tip of the anterior clinoid process in the subfrontal region using the diathermy [57]. There was an instantaneous improvement, with some patients describing that ‘‘a great load has been lifted off my chest.’’ Most patients benefitted [58]. In Obsessive Compulsive neurosis, lesions were made in the cingulum. The results were good and long lasting. In some patients, where the results were not as good as expected another lesion was made in the subfrontal region or the cingulum [59–61]. Another interesting indication for psychosurgery was drug addiction. Thirty two cases suffering from drug addictions (alcohol, morphine and pethidine) were operated by Balasubramaniam during 1970–1972. Surgery was done under general anesthesia. Stereotactic localization was performed using pneumoencephalography and carotid angiography. These investigations were essential to determine the thickness of the corpus callosum for making precise lesion in the cingulum, clearing the corpus callosum. The target was selected in line with the foramen of Monro, midway between the pericallosal and callosomarginal arteries. In the coronal plane, the center of the target was 7 mm from the midline. Destruction was done in all cases by injection of myodil, oil, and wax mixture prepared according to the formula of Narabayashi [62]. Both sides were done at one sitting. Postoperative x-rays were taken to confirm the accuracy of the lesion. Of the twenty eight cases followed up for more than 6 months, 22 had been addiction free [63]. Recounting his experience Ravi Ramamurthi says that he found that this was most effective for pethidine addicts. He also mentioned that it was only offered to the patients who were inclined to

165

166

12


be weaned off their addiction but had failed. The cingulotomy would take away the affective part of their withdrawal symptoms. Balasubramaniam and his colleagues also performed stereotactic amygdalotomy for aggressive behavior in children and adults. Such behavior ranged from continuous severe violent acts towards others, pyromania, destructive tendencies, to episodic attacks of behavior disorders or severe degrees of restlessness. The aim of the operation was to destroy the amygdaloid nucleus or its connections so as to make the patient more manageable either with or without drugs. During 1964–1967, they performed 50 operations on 44 patients; most of them bilateral. The center of the amygdaloid nucleus was considered to be 4 mm in front of the apex of the temporal horn. Two reference points were taken. The first 18 mm below the CACP plane, 6 mm anterior to midcommisural plane and 22 mm from the midsagittal plane. The second reference point was 4–5 mm anterior to the apex of the temporal horn. The author’s preferred the second reference point. In case there was significant cortical atrophy, the first reference point was used. The lesion was made either by diathermy coagulation or Bertrand loop. Diathermy coagulation was done with an 8 mm. electrode, the area of destruction being approximately 200 m3. For the amygdaloid nucleus nine lesions were made. The total volume would be about 1,800 m3 which was slightly greater than the volume of the amygdaloid nucleus which is about 1,200 m3. The lesions made with Bertrand loop were much smaller measuring 500 m3. In 30 operations diathermy was employed and in 19 the Bertrand loop. In one operation both were used. They found that patients having aggressive behavior associated with epilepsy had a better outcome as compared to those suffering from post encephalitic illness aggression [64]. Another target of interest was the posterior hypothalamus. Stereotactic intervention into the posterior hypothalamus was noted to give

satisfactory results for controlling both aggressive, violent behavioral disorders and intractable pain. From the endocrinological point of view, this procedure activates the hypothalamic-hypophyseal axis only temporarily, without causing any serious dysfunctions [65]. Based on these observations Balasubramaniam and Kanaka performed hypothalamotomy on patients who failed to improve after amygdalotomy [66–68]. During the subsequent years, 522 surgeries were performed for aggressive behavior disorder, 402 were bilateral amygdalotomies, and 120 posteromedian hypothalamotomy [69].

Neural Transplantation Stimulated by the reports of A Bjorklund and GD Das at the first congress of the International Brain research organization held at Laussane, Switzerland in 1982, Gopinath (neuroanatomist), Tandon and Mahapatra (neurosurgeons) with Nayar and Mohan Kumar (neurophysiologist) set up a unit to study neural transplantation, with the help of the Department of Science and Technology. They studied neural transplants in rat and primates. Transplantation of embryonic neocortex was performed into cerebellum, lateral ventricle, third ventricle, striatum, hippocampus, tectum and the anterior chamber of the eye in rat. Behavioral and electrophysiological studies were carried out before and after transplantation. Transplantations were also performed in rhesus monkey’s striatum, neocortex and cerebellum for standardization. Parkinson’s disease model was produced in rhesus monkeys using MPTP. As it was difficult to maintain a bilateral Parkinson’s disease model, a unilateral disease model was created. After stabilizing the signs and symptoms for 4–12 weeks, they were grafted with fetal substantia nigra into the striatum using stereotactic techniques. Four monkeys were transplanted. Gopinath [70] notes that three of the four monkeys improved sufficiently to handle food


while one had to be sacrificed due to complications [71]. Similar transplant program was also initiated at NIMHANS [72] in 1989 and at the post graduate institute of basic medical sciences, Madras [73,74]. Presently, various centers in Bangalore and Delhi are performing studies on Mesenchymal stem cell transplantation for Parkinson’s disease and spinal cord injury. Currently, India is experiencing a renaissance of stereotactic and functional neurosurgery. Imports of equipment has become easier and the present generation of neurosurgeons have the best of both the worlds; extensive clinical experience in India as well as advanced training in specific fields from various centers around the world. There is an increase in understanding and interest in developing India into a global health care provider. Realizing this, a large number of private hospitals have begun to invest in state-of-the-art equipment, thus providing an important platform for the development of this subspecialty.

Acknowledgment I would like to acknowledge the help of NH Wadia and Ravi Ramamurthi for critically reviewing the manuscript. Ravi Ramamurthi, RM Varma, Uday Muthane, KV Mathai, V Rajashekar, and HM Dastur for providing information and personal inputs.

References 1. Sridhar K. Prof. Ramamurthi: the legend and his legacy. Neurol India 2003;52:27-31. 2. Kak V. Neurosurgery. In: Pandya SK, editor. Neurosciences in India retrospect and prospect. New Delhi: The Neurological Society of India; 1989. p. 577-641. 3. Cooper IS. Chemopallidectomy: investigative technique in geriatric parkinsonism. Science 1955;121:217.

12

4. Ramamurthi B. The Governor and the Prime Minister. In: Ramamurthi B, editor. Uphill all the way – an autobiography. Chennai, India: Guardian Press; 2000. p. 206-16. 5. Kalyanraman S, Ramamurthi B. Two machine stereotaxy. Neurol India 1970;18(Suppl 1):53-5. 6. Kalyanaraman S, Ramamurthi B. Simultaneous bilateral stereotaxic lesions in the diencephalon. In: Second international symposium stereoencephalotomy, Copenhagen. Confin Neurol 1965;26:310-14. 7. Nadkarni TD, Goel A, Pandya SK. Neurosurgery in India. Neurol India 2002;48:332-5. 8. Bhatia R, Tandon PN, Banerji AK. Brain abscess – an analysis of 55 cases. Int Surg 1973;58(8):565-8. 9. Kannan V, Deopujari CV, Misra BK, Shetty P, Shroff MM, Pendse AM. Gamma-knife radiosurgery for trigeminal neuralgia. Australasian Radiol 1999;43(3): 339-41. 10. Ramamurthi B. Stereotactic surgery in India: the past, present and the future. Neurol India 2000;48:1-7. 11. Mathai K. Investigations into methods for the rehabilitation of persons disabled by convulsive disorders. Project no. 19-P-5113-F-OI. Division of International activities, Office of research demonstrations and training, social and rehabilitation service. Department of Health, Education and Welfare. Washington D.C: 20201. 12. Singh B, Chandy J. Electroencephalographic study of delta waves. Neurol India 1955;3:5-9. 13. Daniel RT, Chandy MJ. Epilepsy surgery: overview of 40 years of experience. Neurol India 1999;47(2):98-103. 14. Falconer F, Serafetinides FA. A follow up study of surgery in temporal lobe epilepsy. J. Neurol Neurosurg Psychiat 1963;26:300-7. 15. Balasubramaniam V, Kanaka TS. Stereotaxic surgery of the limbic system in epilepsy. Acta Neurochir suppl 1975;23:225-34. 16. Ramamurthi B. Stereotaxic surgery in temporal lobe epilepsy. Neurol India 1970;18:42-5. 17. Balasubramaniam V, Kalyanaraman S, Arjundas G, et al. Stereotaxic ablation of the irritable focus in temporal lobe epilepsy Confin Neurol 1970;32:316-21. 18. Kalyanaraman S, Sayeed ZA, Dharmapal N. Surgical treatment of epilepsy. In: Harris P, Mawdsley C, editor. Epilepsy, Proceedings of the Hans Berger centenary symposium. Edinburgh: Churchill Livingstone; 1974. p. 203-8. 19. Balasubramaniam V, Kanaka TS. Why hemispherectomy? Appl Neurophysiol 1975;38:197-205. 20. Ramamurthi B, Sayeed ZA, Rao DB. Stereotactic surgery of medial temporal structures in temporal lobe epilepsy. Neurol India 1979;27(4):192-5. 21. Arjundas G. Usefulness of depth electrode studies in epilepsies. In: Mani KS, Walker AE, Tandon PN, editors. Proceedings of national seminar on epilepsy. Bangalore:

167

168

12 22.

23.

24. 25.

26.

27.

28.

29. 30. 31.

32.

33.

34.

35.

36. 37.

38.


Indian epilepsy association, Bangalore Chapter; 1976. p. 75-7. Kanaka TS, Balasubramaniam V. Cortical and depth electrode studies. Neurol India 1980;28:150-4; Mathai KV. Epilepsy – some epidemiological experimental and surgical aspects. Neurol India 1986;14:299-314. Gillingham FJ, Watson WS, Donaldson AA, Naughton JAL. The surgical treatment of parkinsonism. Brit Med J 1960;2:1395. Kalyanaraman S, Ramamurthi B. Stereotaxic surgery for generalized epilepsy. Neurol India 1970;18:34-41. Kalyanaraman S, Ramamurthi B. Stereotaxic capsulotomy for epilepsy. Proc Australian Assoc Neurol 1968;5: 311-12. Mathai KV, Taori GM. Stereotaxic destruction of ansa and fasciculus lenticularis in the control of seizures. Neurol India 1972;20(Suppl 2):169-74. Rao MB, Radhakrishnan K. Is epilepsy surgery possible in countries with limited resources? Epilepsia 2000;41: S31-4. Bhatia M, Singh VP, Jain S, et al. Epilepsy surgery in India: AIIMS experience. J Assoc Phys India 1999;47: 492-5. Balasubramaniam V, Ramamurthi B. Stereotactic surgery. Neurol India 1965;13(3):93-6. Kalyanaraman S, Ramamurthi B. Simultaneous bilateral stereotaxic lesions. Neurol India 1966;14(3):151-3. Kalyanaraman S, Ramamurthi B. The size of the third ventricle in British and Indian subjects. Neurol India Suppl 1:60-4. Narabayashi H. Physiological analysis of ventrolateral thalamotomy for rigidity and tremor. Confin Neurol 1965;26:264-8. Kanaka TS, Balasubramaniam V, Ramanujam PB, Ramamurthi B. Stereotaxic surgery in cerebral palsy – a study of 57 cases. Neurol India 1970;18(1):56-9. Balasubramaniam V, Kanaka TS, Ramanujam PB. Sterotaxic surgery for cerebral palsy. J Neurosurg 1974;40:577-82. Balasubramaniam V, Kanaka TS, Ramanujam, PB, Ramamurthi B. Stereotaxic hypothalamotomy. Stereotact Funct Neurosurg 1973;35:138-143. Ecker A, Perl T. Percutaneous injection of the thalamus in parkinsonism. Arch Neurol 1960;8:271-8. Varma RM. Percutaneous chemothalamectomy for parkinsonism. Neurol India 1964;12(2):54-60; Varma RM. Percutaneous technique of chemothalamotomy for parkinsonism. Paper presented at third world congress of neurological surgeons, Brussels, 1965. Muthane UB, Varma RM, Sundarajan P, et al. Percutaneous trans-foramen ovale approach to subthalamic nucleus (Varma’s Technique): a shorter route and economical technique in surgical treatment of Parkinson’s disease. Mov Disord 1998;13:71.

39. Muthane UB, Bhatt MH, Wadia NH. Parkinson’s disease and other akinetic disorders. In: Wadia NH, editor. Neurological practice an Indian perspective. Elsevier: Amsterdam; 2005. p. 353-66. 40. Bhagwati SN, Singhal BS. Reactivation of successfully abolished tremors on thalamic stimulation whilst performing surgery for the second side in Parkinson’s disease. Neurol India 1972;20(Suppl 2):166-8. 41. Bhagwati SN. Stereotaxic surgery for Parkinsons’s disease. Bombay Hosp J 1969;2:39-42. 42. Doshi PK, Chhaya NA, Bhatt MH. Bilateral subthalamic nucleus stimulation for Parkinson’s disease. Neurol India 2003;51:43-8. 43. Doshi PK, Chhaya NA, Bhatt MH. Depression leading to attempted suicide following bilateral subthalamic nucleus stimulation for Parkinson’s disease. Mov Disord 2002;17 (5):1084-5. 44. Tir M, Devos D, Blond S, et al. Exhaustive, one-year follow-up of subthalamic nucleus deep brain stimulation in a large, single-center cohort of parkinsonian patients. Neurosurgery 2007;61(2):297-304. 45. Temel Y, Kessels A, Tan S, et al. Behavioural changes after bilateral subthalamic stimulation in advanced Parkinson disease: a systematic review. Parkinsonism Relat Disord 2006;12(5):265-72. 46. Roy TK. Stereotactic cingulate lesions for intractable pain in malignant conditions – a report of 5 cases. Neurol India 1983;31(1):55-8. 47. Ramamurthi B, Davidson A. Progress in stereotactic surgery in Madras. Proc Inst Neurol, Madras 1975;5:81-97. 48. Kalyanaraman S, Ramamurthi B. Studies on the sensory relay nucleus of the thalamus during stereotaxic surgery. Neurol India 1972;20(Suppl 2):155-7. 49. Schaltenbrand G, Bailey P, editors. Introduction of Stereotaxis with an atlas of the human brain. Stuttgart: George Thieme; 1959, p. 2. 50. Kalyanaraman S, Ramamurthi B. Stereotaxic surgery for intractable pain neurology India September 1969;17 (3):109-115. 51. Kalyanaraman S, Logamuthukrishnan S. Stereotaxic posteromedial hypothalamotomy for intractable pain. Proc Inst Neurol Madras 1974;4:95. 52. Govindaswamy MV, Rao B. Bilateral prefontal leucotomy in Indian patients. Lancet 1944;1:466. 53. Anand BK, Brobeck JR. Hypothalamic control of food intake in rats and cats. Yale J Biol Med 1951;24:123-40. 54. Anand BK, Brobeck JR. Loacalization of a feeding center in the hypothalamus of rat. Proc Soc Exp Biol Med 1951;77:323-4. 55. Desiraju T. Fundamental neurophysiology. In: Pandya SK, editor. Neurosciences in India: Retrospect and Prospect, Neurological society of India, 1989, p. 113-52.


56. Manchanda SK, McBrooks C. Homeo-static controls in the regulation of autonomic nervous system function. In: McBrooks C, Koizumi K, Sato A, editors. Integrative Function of the Autonomic Nervous System. University of Tokyo; 1979. p. 427-430. 57. Kalyanaraman S, Ramamurthi B. Stereotaxic basofrontal tractotomy. Neurol India 1973;21:113-18. 58. Ramamurthi B, Ravi R, Narayanan R. Long term follow-up of functional neurosurgery in psychiatric disorders. experience of 30 cases. The World Conference of Stereotactic and Functional Neurosurgery. Zurich: 1981. 59. Balasubramanium V, Kanaka TS. Stereotaxic surgery of the limbic system in epilepsy. Acta Neurochir 1975;23: 225-34. 60. Balasubramaniam V, Ramanujam PB, Kanaka TS, et al. Stereotaxic surgery for behaviour disorders. In: Hitchcock E, Laitinen L, Vaernet KC, editors. Psychosurgery. Springfield, IL: Charles C. Thomas; 1972. p. 156-63; Ramamurthi B, Ravi R, Narayanan. Functional neurosurgery in psychiatric illness. Indian J Psychiatry 1980;22:261-4. 61. Ramamurthi B, Ravi R, Narayanan. Functional neurosurgery in psychiatric illness. Indian J Psychiatry 1980;22:261-4. 62. Balasubramaniam V. Stereotaxic amygdaloid lesions in behaviour disorders. Ph D. Thesis, Madras University; 1969. 63. Balasubramaniam V, Kanaka TS, Ramanujam PB. Stereotaxic cingulumotomy for drug addiction. Neurol India 1973;21(2):63-6. 64. Balasubramaniam V, Ramamurthi B, Jagannathan K, Kalyanaraman S. Stereotaxic amygdalotomy. Neurol India 1967;15(3):119-22.

12

65. Mayanagi Y, Hori T, Sano K. The posteromedial hypothalamus and pain, behavior, with special reference to endocrinological findings. Appl Neurophysiol 1978;41: 223-31. 66. Arjundas G, Balasubramaniam V, Reddy EC, Ramamurthi B. Hypothalamus and its effects on the viscera. J Assoc Physicians India 1971;19:477-81. 67. Balasubramaniam V, Kanaka TS, Ramanujam PB, Ramamurthi B. Stereotactic hypothalamotomy. Confin Neurologica 1973;35:138-43. 68. Balasubramaniam V, Ramamurthi B, Jagannathan K, Kalyanaraman S. Stereotaxic amygdalotomy. Neurol India 1967;25:119-21. 69. Ramamurthi B. Surgery for aggressive behavior disorders. International Congress Series No 433. Neurological surgery. Proceedings of the Sixth Congress of Neurological Surgery. Sao Paulo; 1977. 70. Gopinath G. Neural transplantation. In: Pandya SK, editor. Neurosciences in India retrospect and prospect. Neurological society of India. 1989. p. 219-30. 71. Mahapatra AK, Gopinath G, Tandon PN. Neural transplantation. Prog Clin Neurosci 1987;1:45-8. 72. Murthy SK, Desiraju T. Quantitative assessment of dendritic branching and spine densities of neurons of hippocampal embryonic tissue transplanted into juvenile neocortex. Dev Brain Res 1989;46:33-46. 73. Muthuswamy R, Sheeladevi A, Namasivayam A, et al. Transplantation of monkey embryonic cortical tissue into the brain of bonet monkey (Macaca radiata). J Anat Soc India 1988;37:27. 74. Muthuswamy R, Sheeladevi A, Namasivayam A, et al. Heterospecific transplantation of human embryonic cortical tissue into the cerebellum of bonnet monkey (Macaca radiata). Neuroscience 1987;22:S765.

169

5 History of Stereotactic Surgery in Japan C. Ohye

Production of the First Stereotactic Instrument and its Application in Human It was post-World War II, when a young bud of stereotactic surgery sprouted in Japan as in other countries. Hirotaro Narabayashi, a postgraduate psychiatrist, inspired by Prof. Ogawa’s suggestion, explored the destroyed and almost burned out Tokyo with a drawing of a stereotactic instrument, looking for a factory that had survived the war. Finally, he found one small factory still working, and with the help of a kind engineer there, he produced the first model of a stereotactic instrument in 1949 [1]. His medical student days during the War were miserable; they were shortened because of heavy air strikes and academic school life almost stopped. In fact, as many Professors and young doctors were summoned for military service, lectures were often cancelled. So Narabayashi spent all his free time in the University Library, where he found an interesting book ‘‘Die Extrapyramidales Erkrankungen’’ by the German pathologist Alfons Jacob [2]. The book proved to be a source of great inspiration and perhaps his idea of using the stereotactic instrument to invade the basal ganglia was born around this time. After graduating from the University of Tokyo in 1945, he joined Professor Uchimura’s Department of Psychiatry in 1946. At that time, there was no separate department of Neurology, so patients with Parkinson’s disease and other involuntary movements were treated in the department of Psychiatryand he noticed that there was no effective treatment for several cases. When Professor #


Teizo Ogawa, Professor of Neuroanatomy talked to him about a stereotactic instrument used in animal experiments in Ranson’s Lab in the United States, young Narabayashi was excited and immediately took steps to see if it could be applied to the human brain. However, at that time, with Tokyo almost completely destroyed by bombardment, he had difficulty finding a factory to make such an instrument, as mentioned earlier. Being a passionate young man he had the first stereotactic instrument made in 1949 [3,4]. In 1947, Spiegel and Wycis, Philadelphia, United States, reported the first case of stereotactic surgery on thalamic dorsomedial nucleus for a patient with a neurological problem and movement disorder [5,6]. At that time in Japan, communication with other countries was very limited, and Narabayashi was not aware of the exciting news. Unaware of the work of Spiegel and Wycis, Narabayashi continued to improve on his stereotactic instrument and simultaneously worked on a map of the brain(not published). In 1951, he operated on a patient with cerebral palsy with athetoid movement [1]. But his most exciting success was on June 4, 1952 [7,8], when the first pallidotomy on Parkinson’s disease was successfully completed. After the injection of a small amount of oil wax (procaine-oil and honey wax) the tremors and rigidity disappeared almost immediately and completely. Seeing this miraculous result, Professor Uchimura, chairman of the Department, recognizing the scientific worth in the result encouraged Narabayashi to take this up as his life’s work. Following the Professor’s advice faithfully Narbayashi devoted his whole life to this field of stereotaxy. There

60

5

History of stereotactic surgery in Japan

is a moving anecdote reported in the story of his first pallidotomy. The patient was a relatively young telecommunist suffering from tremors and rigidity. The patient was a very generous man and agreed to undergo the new brain operation without any guarantee. The first and the second operations to caudate nucleus and putamen were in vain. When Narabayashi pleaded for a third attempt, the patient willingly accepted and Narabayashi’s enthusiasm was rewarded with triumph. Another important and unforgettable episode that centered around his first pallidotomy was, the now well known but controversial at that time, the bitter criticism of Spiegel and Wycis and the firm response from Narabayashi. However, Spiegel and Wycis soon recognized the hardwork of Narabayashi and a deep understanding of each other’s scientific attitude fostered a long lasting friendship between them. The letters they exchanged were published in the Archives of Neurology and Psychiatry in 1956 and were later reproduced in a tribute to Narabayashi [9,10]. For several years after the war, the Japanese people had a very difficult time with shortage of food, bare necessities, housing etc. Nevertheless, the Japanese people worked hard, doctors with poor medical facilities were no exception. In the Japanese University Clinic, Neurosurgery including stereotactic surgery was started in earnest. Stereotactic surgery attracted many young neurosrgeons and as a natural consequence Narabayashi became the central figure. He extended his stereotactic research working with young researchers of the Brain Research Institute, University of Tokyo, where Professor Toshihiko Tokizane chaired the Neurophysiology section. Narabayashi recognized from the very beginning, that cooperation with researchers in the basic sciences was essential for the development of stereotactic surgery; the author is one of those who worked with him. In 1956 he moved to Juntendo University as Associate Professor of Psychiatry, but in 1957 he founded his own private clinic devoted mainly to

stereotactic surgery; this became a world famous ‘‘Neurological Clinic’’ later on. As Narabayashi was very frank and open minded, many young doctors gathered in his clinic to see new operations and discuss freely with him. His clinic became a salon for discussions on not only stereotactic surgery, but all areas of Neuroscience. From this salon, in fact, many fledgling doctors of Neuroscience – neurologists, neurosurgeons, neurophysiologists, neurochemists, neuroanatomists etc. – have emerged as leaders in their fields. Narabayashi’s contribution in this sphere also has been remarkable.

Birth of the Japanese Society of Functional and Stereotactic Surgery Many Professors of Neurosurgery [11–13] gathered around Narabayashi, and the Japanese Society for Research in Stereoencephalotomy was founded in 1963, following the International Society, which was founded in 1961 The first Meeting was held in Kyoto, February, 1963, hosted by Professor Araki of Kyoto University. The second Meeting was held in June of the same year organized by Professor Takebayashi of Wakayama Medical School. After the third meeting in the following year, the Japanese Society has been meeting regularly every year, so the 47th Meeting will be held in 2008, organized by Professor Namba of Hamamatsu Medical School. We are very proud that the annual meetings have been conducted regularly without interruption. At one point of time there was an effort to establish the Japanese Society (a research group) but the research fund provided by the Japanese Ministry of Education for the purpose of ‘‘Stereotaxic deep brain surgery’’ was accepted by our Society. This governmental support was, in fact, the successor to the previous research fund for the ‘‘Study on the extrapyramidal movement disorder.’’ The name of the society was changed, following the name of the World Society to


‘‘the Japanese Society for Stereotactic and Functional Neurosurgery’’ in 1973. At the first meeting, some 50 interested doctors came from several centers and 13 papers from different University clinics were presented including my special lecture on neurophysiology of the basal ganglia symptoms. There were many lively discussions on this new operation. At present the society has 500 odd members and some 100 papers were presented during the two days of the meeting with lively discussions. We pride ourselves on the organization of these meetings and the free and active discussions. Narabayashi attended the first International Meeting of Neurology and Neurosurgery in Brussels in 1959. In the stereotactic section he was the first Japanese to attend an international meeting. He presented his work on pallidotomy, met most of the pioneering stereotactic surgeons from all over the world who shared their initial experiences of stereotactic surgery. Narabayashi’s work on pallidotomy was appreciated by many. It was very important that the Japanese development of stereotaxy was recognized even in the early stages. Sterotactic surgery was once prevalent all over Japan.Asin othercountries almost every Neurosurgical Department has a stereotactic unit. Following are the names the first generation of stereotactic surgeons in Japan with their areas of work: Ueki (Niigata) – Ultrasonic device, Sano (Tokyo) – hypothalamotomy for behavior disorder, Narabayashi,Nagao(Tokyo)–pallidotomy,Takebayashi, Komai (Wakayama) – Superior colliculus for Nystagmus, Araki, Handa (Kyoto) – Torticollis, Ozawa, Hori (Osaka) – Epilepsy, Jinnai, Nishimoto (Okayama) – Forel H, Hoshino (Hiroshima) – Radioisotope. Many young stereotactic neurosurgeons went abroad (USA, Canada, Germany, France, Swiss, Sweden, Italy, England etc.), to learn about the different aspects of stereotactic surgery. Sterotactic surgery has progressed rapidly in Japan and consequently, nowadays, several young stereotactic doctors of second and third generation attend

5

international meetings related to modern stereotactic research and treatment in the United States and in Europe.

The Ups and Downs of the Society Around the 1970s stereotactic surgery went through several ups and downs. One of the unhappy incidents was the world wide student violence toward the end of the 1960s, known as the Zengakuren in Japan. Some professors were attacked by the students, professors of Psychosurgery being their main focus. Unfortunately the hypothalamotomy of Sano and the amygdalotomy of Narabayashi had to be stopped. As a consequence, the stereotactic team in Tokyo University was almost disbanded, which resulted in a great loss for the progress of stereotactic surgery in Japan. Another critical event for the world of stereotactic surgery was the progress in the L-dopa drug therapy for Parkinson’s disease around the 1970s that sidelined the surgical option. The next 20 years were difficult times for stereotactic surgeons all over the world [14]. In Japan the surgeries were reduced and many Neurosurgical Departments closed their stereotactic section. But Narabayashi’s group continued with surgeries because of his reputation, and the microrecording technique learnt from the French group (Albe-Fessard and Guiot) by Ohye was established during this time. Thereafter, microrecording during stereotactic surgery became one of the prestigious features in Japan. It also gave us an opportunity to study the functional organization of the human thalamus (see ‘‘Thalamotomy’’ in this book). On the other hand, the use of L-Dopa for more than 10 years brought to light an unexpected new hazard in the treatment of Parkinson’s disease. As every one is aware, L-Dopa is quite effective for a few years initially but it has complicated side effects. This inevitably suggests revival of the surgical treatment.

61

62

5


Nevertheless, during this period, the Japanese Society continued to have its annual meetings as before. One aspect of the scientific activity of the Japanese Society, the number of presentations in each year’s meeting, is shown in > Figure 5-1 [15]. It is clear from the figure that the number of papers presented at these meetings has grown steadily from 20 papers in the earlier years to nearly 100 papers in recent times. As a result of this increase, since the 39th meeting (in Fukuoka, in 2000), the Society meets for two days. In addition, the Society, which usually held its meetings just before the Annual Meeting of the Japanese Society of Neurosurgery (usually one day earlier), has since the 43rd meeting in Nara become independent of the general neurosurgery meeting. Moreover we have two local functional neurosurgery study groups namely, the Kanto (East Japan) Functional Neurosurgeon Conference twice a year and the Yamaguchi and Kyusyu

(Western) Stereotactic Neurosurgery Seminar once a year. The former had its 25th conference in 2007, and the latter its 15th seminar in 2007. In effect, our Japanese stereotactic group continues to be very active.

New Age of Stereotactic Surgery During the ‘‘cold age’’ of stereotactic surgery, we had a few pleasant events. Narabayashi hosted the World Society Meeting in Tokyo in 1973, and Ohye the 10th Meeting in Maebashi in 1989. They both served as the President of the World Society following each World Society Meeting they hosted. Further, Narabayashi was awarded the Spiegel Wycis Medal, the highest honor of the Society, in 1989, and Ohye received it in 2001. The Japanese Society is very proud of them. The Japanese Society continued its activity into the new age. We now have several advanced

. Figure 5-1 Time sequential change of number of papers (ordinate) presented at the Annual Meeting of Japanese Society from the beginning (1st) to the present time (47th) (Abscissa). Results of 21st–35th meeting were omitted. Note that since 39th meeting, we had two days meeting, and the submitted pages increased from the next year of 2001. The 15th meeting was somewhat different from others. It was invited lectures by foreign guests and related discussions


techniques in surgery and fresh knowledge of the basal ganglia and hence, Parkinson’s disease. Computerized imaging system is now close to neurosurgeon’s hand to visualize deep brain structures in 3D fashion. Understanding Parkinson’s disease opens several new avenues that are still growing. On the other hand, advanced new surgical technology has given us new tools – chronic stimulation by implanted electrode initiated by A. Benabid in France, stereotactic radiosurgery using Gamma knife by L. Leksell, brain graft and etc. We are now in a new era of stereotaxy with many new possibilities. In fact, more than ten stereotactic centers are actively involved in stereotactic selective thalamotomy with microrecording, chronic stimulation of different targets including the thalamus, GPi, subthalamic nucleus, or stimulation of cerebral cortex (initiated by Tsubokawa), Although we do not yet know the optimal target for the treatment of different symptoms, we are quite optimistic about the future of stereotactic surgery, because the idea of stereotaxy is now widely applied in the field of general neurosurgery and therefore it stands at the center of Neurosurgery. Without stereotaxy, precise minimally invasive neurosurgery cannot be performed.

References 1. Kanazawa I. Interview with Professor Narabayashi. Tracing the idea and footstep of a neurologist. Prog Neurol 2001;45(3):512-22 (Japanese).

5

2. Jacob A. Die extrapyranuidale Erkrankungen. Berlin: Springer-Verlag; 1923. 3. Narabayashi H. (Stereotaxic apparatus (Model II) (Japanese). Psychiat Neurol Jpn 1953;54:669. 4. Uchimura Y, Narabayashi H. Stereotaxic apparatus. Psychiat et neurol 1951;52:265-70. 5. Spiegel EA, Wycis HT. Pallido-thalamotomy in Chorea. Arch Neurol Psychiat 1950;64:495-6. 6. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 7. Narabayashi H, Okuma R. Procaine oil blocking of the globus pallidus for the treatment of rigidity and tremor of parkinsonism (priminary report). Proc Jpn Acad 1953;29:134-7. 8. Narabayashi H, Okuma T, Shikiba S. Procaine oil blocking the globus pallidus. Arch Neurol Psychiatry 1956;75:36-48. 9. Ohye C, Fodstad H. Forty years with professor Narabayashi. Neurosrgery 2004;55:222-7. 10. Spiegel EA, Wycis HT. Procaine oil blocking the globus pallidus (letter). Arch Neurol Psychiatry 1956;76:263. 11. Narabayashi H. Stereo encephalotomy in Japan. Conf Neurol 1964;24:314-20. 12. Narabayashi H. Begging and development of sterotaxic surgery in Tokyo. Conf Neurol 1975;37:364-73. 13. Ohye C, Fodstad H. Prof. Hiro Narabayashi: in memoriam. Stereotact Funct Neurosurg 2001;76:125-8. 14. Ohye C. Stereotactic surgery of Parkinson’s disease over 30 years. In: Mizuno Y, Fisher A, Hanin I, editors. Mapping the progress of Alzeimer’s and Parkinson’s disease. New York: Kluwer Academic/Plenum Publishers, (2002) p. 429-34. 15. Matumoto K., Abstracts of the papers presented at the 1st–20th meeting of the Japanese society of stereotactic and functional neurosurgery. Neuron Tokushima, Tokyo; 1892.

63

13 History of Stereotactic Surgery in Korea S. S. Chung

In Korean history, the end of the nineteenth century is considered as the period of the Chosun Dynasty’s transition from medieval to modern. This period was marked by wide-ranging social and economic changes on the domestic front, and, externally, by the threat of foreign domination. It was during this period when modern medicine began, perhaps personified historically when Dr. Horace N. Allen (> Figure 13‐1), a Protestant missionary from the Presbyterian Missions in New York, stepped onto Korean soil. In December 1884, Allen was given the opportunity of saving the life of Queen Min’s nephew. In gratitude, at the behest of King Kojong, the Royal Hospital Kwanghyewon, ‘‘House of Extended Grace,’’ was founded on 10 April 1885, and on April 23rd the name was changed to Chejungwon (> Figure 13‐2), which means ‘‘Universal Helpfulness,’’ Kwanghyewon was the first modern hospital in Korea, which later became the Severance Hospital, an affiliate of the Yonsei University Medical School [1,2]. The hospital provided a legitimate venue for the first protestant missionaries to pursue their religious activities while conducting systematic research on diseases endemic to Korea. In March 1886, a year after the founding of Chejungwon, the hospital launched its Medical Department to educate future medical practitioners and professionals in modern medicine, aiming to treat the ailments of the Korean people. Since Allen had been appointed as head of the hospital, J. E. Heron, Charles C. Vinton, and Oliver R. Avison (> Figure 13‐3) successfully carried on his work as a director of the hospital. #


As the result of advances made possible by Avison’s fund-raising initiatives in the United States, in 1899 the Medical Department of Chejungwon was accredited as a full-fledged medical educational institution. An American entrepreneur, Louis H. Severance, deeply moved by Avison’s speech about his missionary activities in Chosun, decided to donate $10,000, a very large sum of money in 1900. With funding provided by Severance, a new hospital building was completed on 23 September 1904. It was named Severance Hospital (> Figure 13‐4) to commemorate the man who had made it possible. It was Korea’s first modern Western style hospital building. In 1908, the Chejungwon Medical School (later Yonsei University Medical School) celebrated its first graduation. While the Chejungwon Medical School was founded and established, several medical schools were also founded in sequence. In the early 1940s, a few pioneering general surgeons started to practice neurosurgical procedures in medical school accredited hospitals. The neurosurgical procedures included neurotraumas, epilepsy surgeries, and surgeries for mental disorders. Actually all the early stage neurosurgeons were functional neurosurgeons. In 1943, Dr. Chu Kul Lee performed corticectomy on a patient with posttraumatic epilepsy. Dr. Lee graduated from Daegu Medical College (later Kyungbook National University) in 1937 and continued to study neurosurgery in Nagoya University, Japan. He returned to Korea in 1942 and proceeded to perform cortical resection for epilepsy patients in Seoul’s Women’s Medical College (later Korea University Medical College).

172

13

History of stereotactic surgery in korea

Later on he conducted cortical coagulation to interrupt the epileptic impulse in patients with epilepsy. He was interested in epilepsy surgery throughout his neurosurgical career and . Figure 13‐1 Dr. Horace N. Allen; Presbyterian missionary from New York who founded the first Western style hospital, Chejungwon on 10 April 1885

published his experience in the first issue of the official Journal of the Korean Neurosurgical Society in 1972 [3]. He was a leading figure in the neurosurgical field and trained many neurosurgeons in 1960s. He was elected as the first president of the Korean Neurosurgical Society in 1961. Another pioneer was Dr. Ki Sup Lee. He graduated from Severance Medical College (later Yonsei University) and studied neurosurgery in Kyoto University, Japan. He performed frontal lobotomy for mental disorder, corticectomy for epilepsy, and sympathectomy for pain in Severance hospital in 1943 (Lee KS, 1996, personal communication). Dr. Si Chang Kim graduated from Kyungsung University (later Seoul National University) in 1936. At one stage, he worked in Seoul Women’s Medical College and returned to Kyungsung University Hospital in 1948 where he performed corticectomy and cortical vessel ligation for epilepsy sufferers. He was one of the handful number of active surgeons practicing neurosurgery at that time. Unfortunately, he was

. Figure 13‐2 Chejungwon in Seoul (1885); the first Western style hospital in Korea


abducted and taken to North Korea during the Korean War (Moon TJ, Forty years History of the Korean Neurosurgical Society (19612001), 2002, personal communication).

. Figure 13‐3 Oliver R. Avison (served 1893–1935); Canadian missionary who was director of Chejungwon hospital and later Dean of Severance Medical College

13

The Korean War broke out on the 25 June 1950. War-related trauma injuries provided great momentum to the development of neurosurgery in Korea. During the Korean War, Scandinavian countries sent the Danish hospital ship Jutlandia to Busan to care of casualties. The chief neurosurgeon on this ship was Professor Edward A. V. Busch from Copenhagen University, Denmark, and his assistant was Dr. K. Vaernet. They not only treated patients but also taught neurology and neurosurgery to Korean military surgeons. Many American military surgeons also came to Korea during the war, among them colonel Arnold M. Meirowsky and George J. Hayes contributed enormously to the development of neurosurgery in Korea. They served in the mobile army surgical hospitals (MASH) and also taught neurosurgery to Korean military surgeons [4]. Some Korean surgeons had several months of continuous neurosurgical training, which served them well in organizing neurosurgical teams. Casualties from the war included patients with cranio-spinal trauma, peripheral nerve injury, and causalgia. The neurosurgeons in the army

. Figure 13‐4 Severance Hospital (1904), the first Western style hospital building in Korea. Louis H. Severance, businessman from Cleveland, donated for the new hospital building and Chejungwon Hospital was renamed as Severance Hospital

173

174

13


performed many thoracic sympathectomy for patients with causalgia. It was a great opportunity for the Korean neurosurgeons to observe Western medicine and neurosurgical practices. After the war some of them continued to practice neurosurgery while others went abroad to study neurosurgery in Western countries [4]. In the late 1950s, neurosurgeons who went abroad came back to Korea after completing neurosurgical training. Dr. Tae Joon Moon had residency training at Thomas Jefferson University (USA), and became a qualified neurosurgeon by the time he came back in 1957. Upon his return, he practiced neurosurgery in Yonsei University and performed trigeminal ganglion block with hot saline or absolute alcohol in the treatment of trigeminal neuralgia and open thoracic cordotomy for intractable cancer pain. Dr. Bo Sung Sim graduated from Seoul National University in 1949. He served as an army neurosurgeon during the Korean War and studied neurosurgery in Minnesota University (USA). He performed hemispherectomy in patients with intractable epilepsy due to cerebral paragonimiasis in 1958. He also performed retrogasserian rhizotomy in the middle cranial fossa for trigeminal neuralgia [4]. Dr. Hun Jae Lee graduated from Severance Medical College in 1944 and finished his residency program in Michigan University Hospital (USA) to become a qualified neurosurgeon. He came to Seoul Women’s Medical College in 1959, after which he performed open cordotomy for intractable cancer pain and retrogasserian rhizotomy for trigeminal neuralgia, while also applying alcohol in gasserian ganglion percutaneously for trigeminal neuralgia [5]. Dr. Kon Huh graduated from Severance Medical College in 1948, and continued studying neurosurgery in Wisconsin University (USA). He came back to Severance Medical College in 1962 where he practiced pain surgery such as open cordotomy or retrogasserian rhizotomy in middle cranial fossa [4]. Dr. Jeong Wha Chu graduated from

Seoul National University in 1956, and studied neurosurgery in Minnesota University (USA). After coming back from Minnesota University in 1961, he practiced some ablative pain surgeries such as open cordotomy and retrogasserian rhizotomy in middle cranial fossa [6]. Stereotactic surgery in Korea began in the early 1960s. Dr. Hun Jae Lee performed chemothalamotomy in patients with Parkinson’s disease using Cooper’s frame in 1960. It was the first stereotactic surgery using stereotactic apparatus in Korea. He presented the results of chemothalamotomy in seven cases of Parkinson’s disease patients and four cases of dystonia patients. He reported on follow up results of those patients in 1963 (Chu JW, 2006, personal communication). Dr. Tae Joon Moon also performed thalamotomy using simple burr hole mounted Mackinie apparatus. Dr. Chul Woo Lee of Kyungbook University, who trained at the Saint Vincent Hospital in Wooster city near Boston (USA), made his own stereotactic frame in 1960. He performed thalamotomy for patients with dystonia and Parkinson’s disease. Dr. Jeong Wha Chu also performed thalamotomy for patients with Parkinson’s disease [6]. However, there were very few neurosurgeons who carried out stereotactic surgery for movement disorders at that time. In 1971, the radiofrequency lesion generator was introduced in to Korea. Sang Sup Chung of Yonsei University performed percutaneous radiofrequency cervical cordotomy, radiofrequency trigeminal thermocoagulation, and radiofrequency ventrolateral thalamotomy for Parkinson’s disease sufferers in 1972 [7,8]. In 1973, radiofrequency facial nerve neurotomy was performed for hemifacial spasm at the stylomastoid foramen [9]. It was the beginning of the subspecialty of stereotactic and functional neurosurgery in Korea. Introduction of radiofrequency lesion generator made it possible for many procedures to be conducted percutaneously, allowing for a more accurate,


safer, and simpler operation. In 1976, Computed tomography (CT) scanner was introduced in to Korea. The introduction of radiofrequency lesion generator and CT scanners were substantial moments for the development of stereotactic and functional neurosurgery in Korea. In 1975, Sang Sup Chung went to Edinburgh University, Britain, for further study where he practiced stereotactic and functional neurosurgery under professors F. John Gillingham and Edward R. Hitchcock. After returning to Yonsei University he became a fulltime stereotactic and functional neurosurgeon. In 1976, Chang Rak Choi studied functional neurosurgery under professor Umbach in Berlin, Germany, and came back to Catholic University where he continued to practice stereotactic and functional neurosurgery. In 1978, Dr. Kil Soo Choi and Dr. Hun Jae Lee performed microvascular decompression for hemifacial spasm and trigeminal neuralgia. Soon after, Sang Sung Chung followed the procedures [10]. In 1979, depth recording was performed using semi-micro electrode during thalamotomy. In the 1980s there was rapid progress in research activities and surgical techniques in Korea. Various stereotactic and functional surgeries were performed. In 1980, Sang Sup Chung performed stereotactic chemical or radiofrequency hypophysectomy [11], percutaneous spinal rhizotomy and percutaneous medullary trigeminal tractotomy in treating cancer pain. Dorsal root entry zone lesioning and facet denervation were performed for chronic intractable pain and for chronic low back pain. Centrum medianum and parafacicularis nucleus lesioning were done for chronic central pain. Hypothalamotomy was performed for aggressive psychosis and glycerol injection was introduced as a treatment modality for trigeminal neuralgia. Also in 1980, percutaneous spinal cord stimulation for chronic pain and intrathecal infusion pump were introduced for cancer pain. However, these devices were not covered by insurances and it was difficult to treat many patients in the early stages of diagnosis

13

because of economic problems. During this period, various newly developed functional neurosurgical procedures were attempted while a number of stereotactic apparatus such as ToddWells, Riechert-Mundinger, Guiot-Gillingham, BRW frame were introduced in to Korea. In the 1980s, CT compatible CRW, Hitchcock, and Leksell frame, which were introduced, enabled surgeons to perform image guided surgery. In 1984, stereotactic evacuation of intracerebral hematoma or brain biopsy was carried out by CT image guided surgery. In 1982, Moon Chan Kim came back to Catholic Medical College after studying in Birmingham under Professor Hitchcock. As a full time neurosurgeon, he practiced surgery for movement disorder, pain, and psychiatric illness. In 1988, linear accelerated based radiosurgery were performed by Moon Chan Kim and Sang Sup Chung. In 1988, Sang Sup Chung performed adrenal gland transplantation for Parkinson’s disease sufferers and reported on the 5 year follow-up results in 1993 [12]. Epilepsy surgery was conducted sporadically from its inception. However, a comprehensive epilepsy protocol (Yonsei Epilepsy Protocol) was established in 1989, after which epilepsy surgery became more standardized and several centers were built. During the 1990s, stereotactic and functional neurosurgery achieved substantial development owing to the great technological advancements in computer science, surgical softwares, engineering, neurophysiology, and various diagnostic tools. The resolution of MRI improved immensely and accurate MRI guided surgery became possible from 1995. In 1999, vagal nerve stimulation was done for intractable epilepsy and microdepth recording started during thalamotomy. Deep brain stimulation for Parkinson’s disease was introduced in 2000 by Jin Woo Chang and Sang Sup Chung. From late 1980s through to the 1990s, many young neurosurgeons were getting interested

175

176

13


in stereotactic and functional neurosurgery. Uhn Lee of Ghil Hospital, who graduated from Hanyang University, is an active neurosurgeon who treats patients with movement disorders. He conducted thalamotomy and pallidotomy in the late 1980s and now continues to practice DBS. Jae Hyoo Kim of Chonnam University is also an active functional neurosurgeon, who is carrying out pain surgery, DBS for movement disorders and sympathectomy for hyperhidrosis. Jin Woo Chang graduated from Yonsei University and studied at the University of Chicago (USA). He is also an active functional neurosurgeon doing DBS surgery for movement disorders, pain surgery and psychosurgery. Others are Sung Nam Hwang of Chungang University, Young Soo Kim of Hanyang University, Yong Tae Chung of Busan Baik Hospital, Kyung Jin Lee of Catholic University, Jung Yul Park of Korea University, Seong

Ho Kim of Yeungnam University, Young Hwan Ahn of Ajou University, Jeong Il Lee of Samsung Medical Center, Moo Seong Kim of Busan Baik Hospital and Ryoong Huh of Pochon Cha University. They are all active in the treatment of movement disorders or pain surgery. The Korean stereotactic and functional neurosurgery society was founded on February 24th 1990, and the first president and secretary were Sang Sup Chung and Chang Rok Choi, respectively (> Figure 13‐5). Approximately 150 members were registered with the society in 2006. In 1996, Sang Sup Chung was elected as the president of the Asian Society of Stereotactic, Functional and Computer assisted Neurosurgery. In 1999, he held the third Asian Society meeting in Seoul successfully (> Figure 13‐6). The first Gamma Knife unit was installed in 1990, and now there are 11 gamma knife units

. Figure 13‐5 The first meeting of the Korean Sterotactic and Functional Neurosurgery Society; Seoul, February 24, 1990


13

. Figure 13‐6 The third meeting of Asian Society of Sterotactic, Functional and Computer assisted Neurosurgery; Seoul, June 13, 1999

in Korea. In 2006, 2700 gamma knife radiosurgeries were conducted in Korea. The gamma knife radiosurgery society meeting was held on 15th of November 2002 and there have been also annual meetings thereafter. The neurosurgeons involved in radiosurgery are Yong Gou Park of Yonsei University, Dong Gyu Kim of Seoul National University, Young Jin Lim of Kyunghee University, Do Hoon Kwon of Seoul Asan Hospital, and Chang Wha Choi of Busan National University. In 2006, Dong Gyu Kim organized the 13th international meeting of Leksell gammma knife society in Seoul successfully. In the beginning of functional neurosurgery, our pioneers conducted epilepsy surgery, while epilepsy surgery according to a comprehensive proocol began in 1989. The Korean epilepsy

society was founded in 1996 and Sang Sup Chung was elected as the first president. Many active and functional neurosurgeons are participating members of the epilepsy society. Active epilepsy surgeons are Jung Kyo Lee of Seoul Asan Hospital of Ulsan University, Hyung IL Kim of Chunju Presbyterian Hospital, Eun Ik Son of Kyemyung University, Seung Chyul Hong of Samsung Hospital, Chun Kee Chung of Seoul National University, Ha Young Choi of Chonbuk National University and Jong Hee Chang of Yonsei University. Many of the institutes use Leksell stereotactic apparatus while 54 hospitals have Leksell apparatus. Many of the hospitals are using the apparatus simply to conduct brain biopsy or evacuation of hematoma. Nineteen hospitals performed 271 DBSs in 2006; 190 Parkinson’s

177

178

13


disease, 34 Essential tremors & 36 Dystonia. There were also clinical experiment of DBS for epilepsy and psychiatric disorders. With more than 60 years of achievement, stereotactic and functional neurosurgery has evolved and become one of important fields of neurosurgery in Korea.

References 1. Commemorating symposium of 122nd anniversary of Chejungwon. Evolution of modern medicine in Korea. Yonsei University Medical Center. 2007. 2. Photographs of 120 years of modern medicine in Korea. Yonsei University Medical Center. 2007. 3. Lee CK. Surgical Treatment of Epilepsy, Preoccipital Coagulation. J Kor Neurosurg. 1972;1:1-14. 4. Forty years History of the Korean Neurosurgical Society (1961~2001). The Korean Neurosurgical Society. 2002. 5. Woo CH, Chang NS. Anterolateral Cordotomy for Relief of Various Intractable Pain. Kor J Soci. 1963;5:121-5.

6. Woo CH. Stereoencephalotomy for Extra Pyramidal System Disorders. J Kor Medi Assoc. 1963;6:248-58. 7. Lee CW. A new method of stereotactic Encephalotomy for Dystonia and Dyskinesia. Modern Med. 1960;3:69-78. 8. Chung SS, Park TS, Kim CS, et al. Percutaneous radiofrequency rhizotomy for Trigeminal neuralgia. J Kor Neurosurg. 1975;4:323-9. 9. Kim SH, Lee KH, Chung SS, et al. Percutaneous cervical radiofrequency Cordotomy for Intractable pain. Yonsei Med J 1975;16:72-82. 10. Doh JW, Park JU, Chung SS, et al. Percutaneous Neurotomy for Clonic Facial Spasm, a case report. J Kor Neurosurg. 1975;4:331-4. 11. Kim SH, Chung SS, Lee HJ, et al. Neurovascular decompression in posterior fossa for Trigeminal Neuralgia. J Kor Neurosurg. 1981;10:469-75. 12. Chung SS, Lee HJ, Lee Ks, et al. Stereotatic radiofrequency hypophysectomy, for disseminated breast and prostate cancer: Transseptal trans sphenoidal approach. Yonsei Med J. 1981;22:53-7. 13. Chung SS, Park YG, Chang JW, Cho J. Long term follow-up results of Stereotactic adrenal medullary transplantation in Parkinson’s diseases. Stereotact Funct Neurosurg. 1994;62:141-7.

14 History of Stereotactic Surgery in Spain J. Guridi . M. Manrique

The history of stereotaxy in Spain dates back to the mid-1950s. Its early history was constrained by the negative impact of the Spanish Civil War on the development of medicine, as on other areas of Spanish society. However, during the 1960s, when various neurosurgeons returned to Spain, the speciality grew. Today it is difficult to find reports of surgical interventions performed by neurosurgeons during that time; in some cases, oral transmission provided the only way of finding out about the development of functional stereotactic surgery in Spain during the 1950s and 1960s. Historical concepts concerning movement disorders developed during the early years of the twentieth century. As a result of early pathological studies on Huntington’s or Wilson’s disease, encephalitis letargica and Parkinson’s disease (PD), which affected subcortical circuits such as the striatum, pallidum, subthalamic nucleus (STN), and substantia nigra, it was discovered that the basal ganglia were involved in motor function. The understanding of the extrapyramidal motor system and movement disorders constituted a turning point in the neurosurgical field. Meyers operated on a patient with postencephalitic parkinsonism in 1939, resecting two-thirds of the caudate head using a transcortical-transventricular approach, and achieving complete arrest of tremor [1]. This pioneering procedure was followed by other operations approaching the caudate nucleus, internal capsula and globus pallidus (GP) with the ansa lenticularis. At that time, the GP was the target of choice for PD and for hyperkinetic disorders; early surgery was performed by open #


procedures, and later with the stereotactic armamentarium. The first stereotactic operation in humans was performed in 1947 by Spiegel (a neurologist) and Wycis (a neurosurgeon) in Philadelphia, on a patient with a psychiatric disorder. They called stereoencephalotome to the first frame, and an account of this significant technical advance and its benefits for patients was published in Science [2]. During the first decades of the twentieth century, general surgeons performed brain operations in different hospitals in Spain. Initially, some of them were in close contact, and worked under the direction of a neurologist, because early neurosurgery was not well developed in the country. After the Spanish Civil War (1936–1939) and the Second World War (1939– 1945), the creation of the National Health System facilitated the development of specialties, which contributed to medical advances in the field of the neurosurgery in the 1950s. Dr Obrador was a pioneering neurosurgeon, who spearheaded and worked in specific areas in the field of surgical neurology. On one of their several visits to different neurosurgical departments, Obrador went to Oxford with Sherrington, where Liddel was using stereotactic methods in monkeys. Later, he visited Dr. Fulton in New Haven and Bailey in Chicago, where he came to know and study techniques relating to the implantation of electrodes producing localized deep cerebral lesions. The first Spanish experience in the functional field was with electrostimulation and the coagulation of the thalamus in a patient with localized myoclonic epilepsy and in whom a previous cortical excision had failed [3].

180

14

History of stereotactic surgery in Spain

1955–60 Dr Obrador performed GP lesions in parkinsonian patients in 1955; he described a simplified and open technique, approaching the pallidal region from a temporal burr hole and performing a mechanical lesion with a pallidotome without stereotactic armamentarium [4,5]. The instrument described was a canula with a wire loop at the lower end; the rotation of the instrument (the loop) severs the tissue. However, the wire loop may tear tissue and blood vessels resulting in an irregular and non-predictable lesion. Obrador, along with Dierssen, wrote a groundbreaking report describing 100 lesions, (95 in the pallidum and 5 in the thalamus) in 69 patients (55 parkinsonian and 14 hyperkinetic) [6]. The authors reported that 67 surgical procedures had been carried out, involving mechanical lesions. Nine lesions were placed using the mechanical-chemical procedure; five were made using the Cooper balloon technique; a further nine chemical lesions, with alcohol; and ten through stereotactic electrocoagulation. The results showed that, whatever the surgical technique used, more than half of the lesions (56%) led to a dramatic alleviation of rigidity, tremor and general patient mobility [6]. However, during the follow-up period, from months to 3 years, clinical signs returned in a high percentage of cases. The report’s authors concluded that only 32% of the parkinsonian patients who received surgical treatment could be regarded as improved after that time. The complications arising after open surgical procedures were also detailed: facial paresis appeared in 20 of the 79 lesions; 17 patients showed contralateral hemiparesia; there were speech disorders in 7 cases; and cognitive changes in 8 patients. One patient died after mechanical pallidotomy [6]. This interesting report on early neurosurgery in Spain showed a high degree of failure in the clinical signs of parkinsonian patients in lengthy follow-up periods, along with a high

percentage of complications caused by the open surgical procedure. Consequently, the open approach to deep structures such as basal ganglia without stereotactic apparatus was practically abandoned, because it had become obvious that lesion placement was only well controlled using stereotactic armamentarium. There was no stereotactic frame in the country at that time; thus, a number of Spanish neurosurgeons went abroad to learn how to use different stereotactic instruments. In those days, Irving Cooper was a renowned and charismatic neurosurgeon working in the field of movement disorders and, in particular, PD. G Bravo (1957–60), G Dierssen (1958–61) and F Isamat (1960–61) – all from Spain – worked with him at St Barnabas Hospital in New York. This period encompasses a significant moment in the history of neurosurgery, when the surgical target in parkinsonian patients was shifting from the pallidum to the ventrolateral (VL) of the thalamus. At St Barnabas, Cooper and Bravo developed a technique for lesioning in the GP and thalamus. They introduced a balloon on the target, inflating it with air or a liquid and inserting it into the lumen as a compressive test. If the patient improved, the lesion was later produced with alcohol [7–10]. On the basis of this early technique, in conjunction with chemopallidectomy or chemothalamectomy, rigidity was alleviated by 80% and tremor by 75% in parkinsonian candidates after a 4-year follow-up period [9,10]. Cooper and Bravo showed that if the lesion was performed accurately on the target, the improvement induced as a result held for a long follow-up period. Their surgical procedure produced some adverse effects, such as a mortality of 2.4% and 2% hemiplegia in patients treated [9]. Alcohol injections in the brain’s deep structures also had irregular diffusion in the tissue, lesioning cellular and fiber elements without affecting vessels. Cooper et al reintroduced alcohol through the chemopallidectomy approach,


preparing a mixture of ethyl cellulose and ethanol (etopalin) to prevent diffusion in the target. They reported, as Hassler had in Germany, that the interruption of fibers between GP and thalamus was very important for clinical alleviation in parkinsonian and hyperkinetic disorders [10,11]. The lesions in the GP provided good relief of rigidity but less consistent improvement of tremor [10]; they later reported that the alleviation of tremor and rigidity in parkinsonian patients was obtained after a lesion was placed in the VL of the thalamus. Dr Cooper and Bravo made a striking discovery. Cooper planned to perform a lesion in the posterior part of the GP (15 mm behind Monro foramen) in a tremoric patient; after surgery, the tremor disappeared. However, the patient committed suicide 6 months after surgery, and the brain was analyzed in the hospital. The autopsy, which was carried out by Bravo, revealed that the lesion was placed in the VL of the thalamus. Based on this information, Cooper and Bravo changed the surgical approach from the GP to the VL of the thalamus as “the most effective lesion for producing complete and enduring relief of parkinsonian tremor and rigidity” [7,8]. The surgical results with the GP lesion (pallidotomy) for the alleviation of rigidity and tremor in parkinsonian patients were disappointing because large lesions had to be made in the nucleus if improvement was to be obtained. This prompted Hassler to study the pallidal projections to the thalamus; he showed that most of the pallidal efferences projected in the ventral lateral nuclear region: in his own classification, ventralis oralis anterior (Voa). As a consequence, he proposed performing a lesion on the new target; nevertheless, despite the shift to the thalamus, the surgery did not effect total abolition of parkinsonian tremor [11,12]. Whether the shift was based on scientific discovery through an anatomic rationale or on serendipity, the VL of the thalamus replaced the GP as the target of interest [13]. Thus,

14

thalamotomy became the primary surgical procedure for the treatment of PD, starting in the mid-1950s and reaching a peak during the 1960s. Hyperkinetic patients were also operated on in the most anterior part of the VL thalamic nucleus with good results; thus, in practice, pallidotomy was discarded. Ventro lateral-thalamotomy was carried out on patients with PD or other movement disorders throughout the world. However, the accuracy of the surgical procedure at that time fell far short of what is currently being achieved, mainly because imaging analysis was not yet available [14,15]. A common complication arising in surgery was the induction of a movement disorder, hemichorea or hemiballism (Hc/Hb), especially after thalamotomy. Cooper and Bravo analyzed this complication in their patients: 21 patients developed involuntary movements in 850 consecutive BG operations [16]. The patients affected showed a lesion involving the corpus Luysii (subthalamic nucleus STN). After his return to Spain in 1966, Bravo reported that the incidence of Hc/Hb in patients after surgery could be attributed to the placement of the lesion below the intercomissural line. The duration of the disease and the age of patients also became significant factors determining the appearance of complications [17]. He found that no patient over 50 years of age with a disease-duration of less than 10 years developed Hc/Hb. In contrast, patients under 40 years old and with a diseaseduration of more than 10 years developed Hc/Hb after VL thalamotomy, even if the lesion had not reached the subthalamic area [17]. Dr G Dierssen, another Spanish neurosurgeon who trained with Cooper, carried out different studies in the same area, which showed that not all Hc/Hb complications after surgical procedures were due to STN involvement (> Figure 14-1). Dierssen et al first published an account of the case of a 35-year-old parkinsonian patient with severe tremor and rigidity, on whom a right thalamotomy was performed. The patient

181

182

14


. Figure 14-1 Dr Dierssen (front row left in the photograph) with Dr Cooper (second from left) in Brussels in 1958

developed dyskinesia (Hb) hours after surgery, which lasted for days. A second procedure was undertaken to undo the movement induced but the surgical intervention failed and the movement disorder persisted. The patient died of a pulmonary embolism 5 months after the second surgical procedure; an anatomopathological study showed that no lesions were placed in the STN. The first lesion destroyed the VL nucleus, involving the thalamic fasciculus; and the second was placed in the red nucleus and rubro tegmental fibers [18]. This case was also reported by Cooper in his book “The Vital Probe” as case number 1, and by Gioino as case number 6 [19,20].

1960–70 After his return to Spain, Dierssen continued to analyze Hc/Hb in patients with PD who had undergone surgical procedures, developing the diskynesia ipsilateral in thalamotomy [21] and an hemiballism as the primary manifestation of a chiasm-optic nerve glioma in a child [22]. However, the most interesting study, which is not well-known because it was written in Spanish, was the report on 116 Hb with anatomopathological study [23]. In the conclusion, he held that in a high percentage of cases (65%) with Hc/ Hb—not only surgical patients—multiple anatomical structures were affected and the movement


disorder was not always induced by STN involvement; he pointed out that a number of patients with subthalamic lesion showed no clinical sign. A lesion in the striatum or GP may also induce a movement disorder similar to the subthalamic affect [23]. He reported that the percentage of patients developing diskynesia after thalamotomy increased in cases of postencephalitic parkinsonism and after bilateral surgeries [23]. Dierssen returned to the Hospital La Paz with S Obrador and later (1975) moved to Santander, but during that time the number of surgical procedures carried out on patients with PD decreased through the introduction of levodopa as a treatment for the disease. Hundreds of patients with PD were operated on in two different centers in Madrid at that time—the early 1960s—and in Barcelona (Dr Isamat). The procedure was performed with the stereotactic frame (Cooper frame) and ventriculography with lypiodol. The intercommisural line (ICL) was measured and the target was selected in the medial third of the ICL and 11–14 mm from the midline. A leucotome (McKenney) was introduced and opened near the target; patient tremor and rigidity were evaluated intraoperatively. If tremor was arrested, a mixture of lypiodol and hot wax was injected into the target; if tremor persisted, the wire of the leucotome was opened more. Very few papers and reports about Spanish patients operated on during that period of time exist, but a doctoral thesis recorded 200 patients with bilateral thalamotomies (in a group of 1700 patients operated on by Bravo et al) during a 5-year or longer follow-up period in the 1960s and 1970s [24]. The study showed that tremor, rigidity and hypocinesia were stabilized after bilateral surgery, and the benefit persisted for years. Some of the patients presented speech disorders that correlated with the second lesion volume [24]. Stereotactic surgery spread to other Spanish regions when the neurosurgeons returned after their study-periods abroad. In 1969, V Arjona

14

returned from Newcastle (1962–69) (Dr Hankinson) and began to treat parkinsonian patients in Seville with other movement disorders and pain disorders using stereotactic procedures. He had previously spent time with Denise Albe-Fesard and Guiot in Paris, learning about microrecording for target localization. Recording during surgery was a control procedure to place the lesions more accurately and to obtain better results for a long follow-up period. They used a Leksell frame and performed two burr holes: one frontal, for ventriculography and lesioning; the other occipital, for recording. Dr Arjona favored the criogenic lesion because it was reversible, and temperature control was more accurate with this apparatus than in thermolitic procedures. At that time, the target was the ventralis intermedius (Vim) of the thalamus in parkinsonian tremor; the Voa-Vop (Ventralis oralis posterior) was also lesioned in patients with choreo-athetosis or other dyskinesias. Despite the fact that larger lesions were placed, patients with dystonia responded less satisfactorily to thalamic surgery [25]. The surgical target for pain was the Cm-Pf (Centromedianus-Parafascicular) of the thalamus, and the hypothalamus was the target in surgical candidates with aggressivity and oligophrenia erethica [26,27]. Dr JG Martin Rodriguez also performed surgical stereotactic formation with Hankinson in Newcastle (1967–71). He moved to Madrid in 1971, where he developed functional surgery at La Paz Hospital. He acquired extensive experience in thalamic criogenic lesions in patients with PD or other movement disorders. During surgical recording from the occipital cortex, the electrode crossed the internal capsule reaching the GP or VL of the thalamus. Surgical results showed that thalamo-capsular lesions improved tremor, and GP-capsular lesions rigidity [28,29]. During the 1970s, Dr Martin Rodriguez reported on a number of papers about the electrophysiology of the pulvinar nucleus and the role of this anatomical structure in the mechanism of pain

183

184

14


appreciation. He performed pulvinar lesions, combined with other thalamic nuclei (Vop, Vim), in patients with hypertonic-hyperkinetic syndromes; however, the surgical conclusion after patient evolution was not consistent for clinical alleviation [30–32]. In 1972, he and Dr Delgado placed the first electrode with transdermic stimulation in the septal nuclei of the thalamus of a patient with phantom limb pain, by braquial plexus avulsion (see below) [33]. In 1977, Dr Martin Rodriguez moved to the Hospital Santiago Ramoń y Cajal, but as a result of the positive response to levodopa in patients with PD the number of thalamic surgeries decreased and all stereotactic activity declined. The neurophysiologist, Dr J.M.R. Delgado (et al), developed the technique of implanting multilead electrodes in the brain for several days (> Figure 14-2). The electrodes were made of enameled stainless steel wires or of coated silver wires, which had the advantage of flexibility. Their diameter (0.1–0.2 mm) permitted the insertion of the active portion without isolate for a long period of time [34]. J.M.R. Delgado was . Figure 14-2 Dr Delgado circa 1965

a physiologist who moved to Yale University in 1950; he was to work with Dr J.F. Fulton there for 20 years. After his return to Madrid, he worked with Dr Obrador and Dr J.G. Martin Rodriguez; during the 1960s, they developed a device connected to the electrodes and controlled by radio waves which was able to record neuronal activity and could also enable the stimulation of internal structures. The device was called the “stinoceiver,” and it was tested in different animal species, including a bull. The device was finally implanted in a patient with a phantom limb in 1968 [35]. Electrodes were implanted in the Caudate head and septal nuclei (near accumbens nucleus), bilaterally, through two frontal burr holes. The transdermic stimulation was “on” for a number of hours daily, alleviating patient pain. This may have been the first implantation of a deep brain stimulation device in Europe.

1970s In the early 1970s, Dr G Bravo (> Figure 14-3) and Dr J Miravet, a neurophysiologist in the Clinica Puerta de Hierro in Madrid, designed a operating room (OR) for stereotactic surgery. This OR was built using the Talairach technique for epilepsy study, and their surgical treatment permitted the implantation of multicontact electrodes in stereotactic conditions to record electrical activity during the onset and propagation of seizures (estereoelectroencephalography: SEEG). The operating table could be rotated through 360 to perform ventriculographies with teleradiography, without magnification or distortion in the studies. The Talairach frame permitted their replacement in patients; this approach had the advantage of supplanting different radiological studies, such as angiography and ventriculography. A room with a glass wall next door to the OR was the neurophysiological laboratory, which had 41 electroencephalography


14

. Figure 14-3 Dr Bravo one of the spanish pioneers in the surgical treatment of Parkinson’s disease in his office in Madrid

channels for recording. The room as a whole was electrically insulated. During that period Dr M Manrique spent 2 years in the Delgado lab and later moved to Paris with Dr J Talairach to study electrode placement in patients with epilepsy [36–38]. He and Dr R Garcıá de Sola went on to develop the surgical treatment of refractory epilepsy in Madrid. These multiple radiological studies enabled the implantation of subdural electrodes, to define the epileptic zone, and their resection with the advantage of superimposed angiography. Later, M Manrique applied stereotactic procedures for epilepsy (stereoelectrocorticography), with reference to subdural electrodes in MRI (magnetic resonance image) in the Talairach gride, and after identification of the anterior and posterior commissures [36,39,40]. The stereotactic study referenced the position of the electrodes and the epileptogenic zone, the different cortical areas and vascular information in the frame space. The information was used during surgery to facilitate accurate resection. In 1977, in the Hospital Santiago Ramoń y Cajal, a second OR for stereotactic surgery, similar to the one described above, was designed using a Leksell

frame. Dr JG Martin Rodriguez with JC Bustos, F Figueiras in neurosurgery, and E Garcia Austt and W Bunõ in neurophysiology treated hundreds of patients during that time. Dr Eiras carried out his stereotactic surgical activity in Zaragoza. In 1966, he went to Freiburg in Germany—the Hassler school, with Riechert and Mundinger. These authors described other targets for tremoric parkinsonian patients, such as the zona incerta placed dorsal to the STN, near the fasciculus lenticularis (campotomies). After his return, using the Riechert-Mundinger frame, Dr Eiras performed campotomies in patients with PD, and also targeted the CM-Pf bilateral with asymmetric lesions in patients with pain disorders and Voa-Vop for dystonia. He described a post-traumatic case with action myoclonus that was alleviated through Vop-Vim thalamotomy [41]. Following the introduction of CT (computer tomography), he and his team pioneered the implantation of stereotactic biopsies with frame and the new image tool. They performed tumoral resections with the help of the stereotactic coordinates, using the Robert Wells (Radionics) equipment for these surgical interventions.

185

186

14


The significant clinical impact of the introduction of levodopa therapy for PD during the 1960s brought about a decline in the surgical treatment of movement disorders. In a large number of surgical departments, such surgery practically disappeared. A few neurosurgeons continued to perform the procedure on patients with levodopa intolerance. Sixteen neurosurgeons were questioned by Laitinen in 1985 regarding their preferred target for PD [42]. The responses reveal a variation in the chosen target (VL, zona incerta, Vim, Forel’s field); however, none of the neurosurgeons used GP as the target for PD during the 1980s [42]. During these years in Spain, only one or two teams carried out operations on parkinsonian patients who did not respond to treatment with levodopa. Nowadays, and over the last 15 years, surgery for advanced PD has been revived. The major reasons for this renaissance are the increased understanding of the pathophysiology of the

basal ganglia and the demonstration of surgical alleviation of experimental parkinsonism. In addition, a large proportion of patients suffer severe complications after years of chronic treatment. Moreover, the significant advances in neurosurgery brought about by the introduction of computers and image development, and deep brain stimulation (DBS) that enables low-risk bilateral surgery, are further factors in this regard. However, this new approach in PD surgery centers on a target that had never before been considered: the STN [43].

1970–80 JL Barcia-Salorio, another pioneer in Spanish neurosurgery, worked and carried out stereotaxy in Valencia (> Figure 14-4). He went to Freiburg in 1960, a renowned school at that time, given the participation of Riechert and Toennis;

. Figure 14-4 Dr J L Barcia-Salorio placing the stereotactic frame for a radiosurgical treatment


however, a few months later, he moved to Sweden to work with Prof Leksell in Stockholm and Dr Larsson in Uppsala. His primary intention was to learn about pallidal lesions in patients with movement disorders, but he was very impressed at the Karolinska Institute by the initial steps taken in radiosurgery, using the gamma-knife in different neurosurgical procedures. After Barcia returned to Valencia, he began to carry out stereotactic surgery on parkinsonian patients, above all, pallidotomies; however, he also inaugurated the use of radiosurgical treatment in Spain. Basing his development on the Leksell frame, he designed and built the Barcia frame, capable of supporting the heavy collimators; cobalt was used as the source of radiation in his groundbreaking approach to radiosurgical treatment. He calculated the dose due to a specific target mathematically, and developed a system based on the tomography in AP and lateral to localize a brain lesion using cartesian coordinates, before the technical innovation of CT and MRI. He localized the coordinate position and lesion size through this early procedure, creating a target volume to be treated [44]. In 1975, he performed the first radiosurgical treatment in Spain; this was reported to the ESSFN (European Society of Stereotactic and Functional Neurosurgery) in 1977. In 1979, he described the radiosurgical treatment of large acoustic schwanomas; later (1982), he performed the first treatment of carotid cavernous fistulas (ccf) using radiosurgery [45–48]. He also developed the introduction of low dosages of radiotherapy in patients with resistant epilepsy, who had previously been treated unsuccessfully with a temporal lobectomy [49]. Prof. Barcia’s most significant contribution to neurosurgery was his pioneering study of radiosurgery for ccf with low flow and the treatment of epilepsy. He established the Valencia school that lives on into the present in the work of Dr J Barbera´, Dr J Broseta and Dr P Roldań.

14

In 1990, the first LINAC for radiosurgery was installed in the Sanatorio San Francisco de Asis in Madrid, under the direction of Dr Sambla´s and Dr JC Bustos. Today, there are more than 20 LINAC units in Spain; there is also one gamma knife, in the Clinica Ruber in Madrid. As a young Spanish neurosurgeon, J. A. Burzaco (1932–2006) went to Cheshire (England) (1956–57), and later worked in the Walton Hospital in Liverpool as a registrar (1957–58); his contact with L. Leksell in the Serafirmerlassarettat in Stockholm (1961–62) had a decisive impact on his future development in the stereotactic field (> Figure 14-5). He returned to Spain in 1962, practiced surgery in a number of different hospitals in Madrid; he always had a great interest in the neurosurgical treatment of psychiatric diseases. On the basis of Egas Moniz’s description of such surgery in psychiatric patients,

. Figure 14-5 Dr Burzaco, a pioneer figure in psychiatric surgery in Spain also worked with the gamma knife

187

188

14


lobotomies were regularly carried out in Spain at that time. Burzaco combined both approaches: stereotactic surgery and the selection of psychiatric candidates [50]. He reported his experience with thermolesions in the anterior arm of internal capsula (capsulotomies), using the Leksell’s target in patients with obsessive neurosis and severe autoaggressive compulsions, and describing 108 capsulotomies in patients with obsessive compulsive disorders (OCD) [51,52]. Lesions were placed bilaterally and symmetrically, and the follow-up period was either one or 2 years. His results showed a 73% rate of great alleviation (symptom-free or an improvement in normal life); that is, in 62 of the 85 patients on whom he operated. His analysis revealed that improvement was greater among patients who were 30 years old or younger, as compared with patients who were 40 years old or older (83.3% vs. 67%) [51,52]. Dr Burzaco’s most significant contribution to the field of psychiatric surgery was thermolesion in the stria terminalis (fascicle from amygdala to septal nuclei in the hypothalamus) in complex patients experiencing severe character, aggressivity and behavior changes. He first described the technique in a young patient after an encephalitis, but the surgery was later performed on a selection of patients with violent aggressive behavior [53]. The target was placed 3 mm anterior to the anterior commissure, 5–6 mm lateral to the wall of the third ventricle, and at the level of the intercommissural line. However, the complexity of the cases treated led to the formation of combined lesions. The fundus of the stria terminalis lesion was placed in 90% of 116 patients with aggressivity, and a posteromedial hipothalamotomy, a capsulotomy or a cinculotomy was also performed as a complementary target [54]. A second operation was carried out on 27% of the patients, which he himself evaluated as showing a dramatic reduction in aggressive behavior, with no mortality and very few complications. Dr Burzaco was a

pioneer in the field of the surgical treatment of complex cases involving psychiatric disorders, and was renowned for his extensive experience throughout the country. His description of the pathophysiology of aggressivity showed that components of the limbic system (hippocampus and amygdala), as well as cortical areas, such as the orbital and prefrontal cortex, were implicated. His experience led him to conclude that major deafferentization should be carried out to alleviate the clinical condition of chronic psychiatric diseases. A bilateral striae terminalis lesion and—in a high percentage of cases—a bilateral cingulotomy were required to disrupt cortical and deep limbic anatomical pathways in patients with aggressive behavior [54,55].

Final Remarks Stereotactic localization within the brain, using CTor MRI, and the introduction of computers in neurosurgery have marked significant advances in this field. The target may now be visualized in tumors, and may also be recorded in the case of a functional anatomical deep structure. Following target selection, computers automatically introduce the coordinate information of the frame, reducing surgical time with a high degree of accuracy. During the 1970s and 1980s, different neurosurgical departments implemented these techniques for brain biopsies, and later for craneotomies with frame. Tumoral pathology advanced with the help of stereotaxy for years, until the introduction of neuronavigation in neurosurgery in current times. A survey carried out by the Stereotactic and Functional group of the Spanish Society of Neurosurgery (2002) in Spain showed that 50% of departments used stereotaxy and/or neuronavigation for tumoral pathologies; 15 of 34 used stereotaxy alone as a technique; only two groups used neuronavigation without stereotactic frame [56]. At that time, there were 16 Radionics


frames, 14 Leksell, two Barcia, and only one Laitinen and Riecher-Mundinger frame. In 1 year, 656 stereotactic biopsies were performed in thirty-four surgical departments, 125 craneotomies guided by stereotaxy, and 95 cases of cyst or abscess evacuation using frame help. Eleven departments performed functional stereotactic procedures on 115 previously operated PD patients and 10 patients with other movement disorder pathologies. In target planning, nine groups used only TAC, five used MRI, and ten groups fusion image; only one group performed ventriculography for intercommissural reference [56]. Stereotactic and functional neurosurgery is currently one of the major areas in neurosurgery, in Spain, as in other developed countries: its future is wide and open. Current applications and procedures include DBS techniques, ablative procedures, along with drug delivery, gene therapy and cell transplantation in degenerative disorders. The patients involved suffer from PD, essential tremor, dystonia, chorea, multiple sclerosis, pain, epilepsy and psychiatric disorders, such as depression and OCD. Surgical techniques have evolved over the course of many years, but patients continue to suffer the burden of their diseases in ways similar to patients in the past.

Acknowledgments We would like to acknowledge the information provided and opinions shared by M Arra´zola, F Aguilera, V, Arjona, K Buus, M Dierssen, J Eiras, R Figueiras, F Isamat, R Martinez, J Molet, and JG Martıń Rodriguez.

References 1. Meyers R. Surgical procedure for postencephalitic tremor, with notes on the physiology of the premotor fibers. Arch Neurol Psychiatry 1940;44:455-9.

14

2. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus on the human brain. Science 1947;106:349-50. 3. Obrador S. Personal recollections of the development of human stereotactic neurosurgery. Confinia Neurol 1975;37:378-83. 4. Obrador S, Dierssen G. Cirugıá de la regioń palidal en el sıńdrome de Parkinson. Tećnica personal y resultados inmediatos en los seis primeros casos operados. Revista Clıńica Espanõla 1956;61:229-37. 5. Obrador S. A simplified neurosurgical technique for approaching and damaging the region of the globus pallidus in Parkinson’s disease. J Neurol Neurosurg Psychiatr 1957;20:47-9. 6. Obrador S, Dierssen G. Results and complications following one hundred subcortical lesions performed in Parkinson’s disease and other hyperkinesias. Acta Neurochi 1959;7:206-13. 7. Cooper IS. Surgical treatment of parkinsonism. Ann Rev Med 1965;16:309-30. 8. Cooper IS, Bravo G. Chemopallidectomy and chemothalamectomy. J Neurosurg 1958;15:244-56. 9. Bravo GJ, Cooper IS. Chemopallidectomy: two recent technical additions. J Am Geriatr Soc 1957;5:651-5. 10. Bravo GJ, Cooper IS. A clinical and radiological correlation of the lesions produced by chemopallidectomy and chemothalamectomy. J Neurol Neurosurg Psychiatr 1959;22:1-10. 11. Hassler R. The pathological and pathophysiological basis of tremor and parkinsonism. Second International Congress of Neuropathology, Excerpta Medical Foundation 1955;1:29-40. 12. Hassler R, Riechert T. Indikationen und Lokalisationsmethode der gezielten Hirnoperationen. Nervenarzt 1954;25:441-7. 13. Guridi J, Lozano AM. A brief history of Pallidotomy. Neurosurgery 1997;41:1169-83. 14. Smith MC. Location of stereotactic lesions confirmed at necropsy. Br Med J 1962;1:900-6. 15. Cooper IS, Bergmann LL, Caracalos A. Anatomic verification of the lesion which abolishes parkinsonian tremor and rigidity. Neurology 1963;13:293-315. 16. Schachter JM, Bravo G, Cooper IS. Involuntary movement disorders following basal ganglia surgery in man. J Neuropathol Exp Neurol 1960;19:228-37. 17. Bravo G, Parera C, Seiquer G. Neurological side-effects in a series of operations of the basal ganglia. J Neurosurg 1966;24:640-7. 18. Dierssen G, Bergmann L, Gioino L, Cooper IS. Hemiballism following surgery for Parkinson’s disease. Arch Neurol 1961;5:627-37. 19. Cooper IS. The vital probe: my life as a brain surgeon. New York: Norton; 1981. p. 293-315. 20. Gioino GG, Dierssen G, Cooper IS. The effect of subcortical lesions on production and alleviation of hemiballic and hemichoreic movements. J Neurol Sci 1966;3:10-36.

189

190

14


21. Dierssen G, Gioino G, Cooper IS. Participation of ipsilateral hemisphere lesions in the pathology of hemichorea and hemiballismus. Neurology 1961;11:894-8. 22. Dierssen G. Sobre un caso de hemibalismo producido por un glioma de ta´lamo o´ptico. Rev Clinica Espanõla 1962;84:339-42. 23. Dierssen G, Gioino G. Correlacioń anatomica del hemibalismo. Rev Clin Esp 1961;82:283-305. 24. Herrero J. Resultados tardıós de la cirugıá en la enfermedad de Parkinson. Situacioń clıńica de 200 enfermos despue´s de cinco o ma´s anõs de haber sido operados bilateralmente. Doctoral Thesis, Universidad Autonoma de Madrid; 1980. 25. Burzaco J. Stereotactic pallidotomy in extrapyramidal disorders. Appl Neurophysiol 1985;48:283-7. 26. Arjona V, Hankinson J. Cirugıá tala´mica del dolor. Rev Clin Espa 1969;113:313-22. 27. Arjona VE, Martin-Rodriguez JG. Cryohypothalamotomy for the treatment of aggressive behavior in oligophrenia erethica. Neurological surgery. Amsterdam: Excerpta Medica; 1978. p. 320-4. 28. Gillingham FJ, Watson WS, Donaldson AA. The surgical treatment of parkinsonism. Brit Med J 1960;12 Nov:1395-402. 29. Gillingham FJ, Kalyanaraman S, Donaldson AA. Bilateral stereotaxic lesions in the management of parkinsonism and the dyskinesias. Brit Med J 1964;2:656-9. 30. McComas AJ, Wilson P, Martin-Rodriguez JG, Wallace AC, Hankinson J. Properties of somatosensory neurons in the human thalamus. J Neurol Neurosurg Psychiatr 1970;33:716-7. 31. Martin-Rodriguez JG, Bunõ W, Jr, Garcia-Austt E. Human pulvinar units, spontaneous activity and sensorymotor influences. Electroenceph Clin Neuropath 1982;54:388-98. 32. Martin-Rodriguez JG, Obrador S. Evaluations of stereotaxic pulvinar lesions. Conf Neurol 1975;37:56-62. 33. Obrador S, Delgado JMR, Martin-Rodriguez JG, SantoDomingo, Alonso A. Estimulacioń cerebral transdermica en miembro fantasma doloroso. Revista espanõla de OtoNeuro-Oftalmologıá y Neurocirugıá 1972;30:269-72. 34. Delgado JMR. Permanent implantation of multilead electrodes in the brain. Yale J Biol Med 1952;24:351-8. 35. Delgado JMR, Mark V, Sweet W, Erwin F, Weiss G, et al. Intracerebral radio stimulation and recording in completely free patients. J Nerv Ment Dis 1968;147:329-40. 36. Talairach J, Peragut JC, Farnarier P, Manrique M. The role of the stereotactic radiographic exploration in neurosurgical interventions. In: Salomon GG, editor. Advances in cerebral angiography. Berlin: Springer; 1976. p. 272-3. 37. Manrique M, Alborch E and Delgado JMR. Cerebral blood flow and behavior during brain stimulation in the goat. Am J Physiol 1977;232:495-9.

38. Bancaud J, Talairach J, Geier S, Bonis A, Trorrier S, Manrique M. Manifestations comportamentales induites par la stimulation electrique du gyrus cingulaire anterior chez l’homme. Rev Neurol (Paris) 1976;10:705-24. 39. Manrique M, Rodriguez Albarinõ M, De la Torre M, Blazquez MG. Stereocorticography in the surgical treatment of epilepsy. Stereot and Funct Neurosurg 1994;63 (1–4):101. 40. Manrique M, Parera C, Nombela L, Miravet J, Bravo G. Stereotaxic radiographic and computerized axial tomographic explorations in the surgical treatment of epilepsy. Presented at the third meeting of European Society for Stereotactic and Functional Neurosurgery (ESSFN). Freiburg 1977. 41. Eiras J. Sıńdrome mioclońico postrauma´tico. Efectividad de las lesiones tala´micas sobre las mioclonıás de accioń. Arch Neurobiol 1980;43:17-28. 42. Laitinen LV. Brain targets in surgery for Parkinson’s disease. Results of a survey of neurosurgeons. J Neurosurg 1985;62:349-51. 43. Guridi J, Luquin MR, Herrero MT, Obeso JA. The subthalamic nucleus: a possible target for stereotaxic surgery in Parkinson’s disease. Mov Disord 1993;8:421-9. 44. Barcia-Salorio JL, Barbera´ J, Broseta S, Soler F. Tomography in stereotaxis. A new stereoencephalotome designed for this propose. Acta Neurochir 1977;Suppl 24:77-83. 45. Barcia Salorio JL. Radiosurgical treatment of huge acoustic neurinomas. In: Szikla, editor. Proceedings of the inserm symposium No. 12, 1979. Oxford, UK: Elsevier; p. 245-9. 46. Barcia-Salorio JL, Hernandez G. Stereotactic radiosurgery in acoustic neuninomas. Acta Neurochir 1984;Suppl 33:373-6. 47. Barcia Salorio JL, Hernandez G, Broseta J. Radiosurgical treatment of carotid-cavernous fistula. Appl Neurophysiol 1982;45:520-2. 48. Barcia-Salorio JL, Roldan P, Hernandez G. Radiosurgical treatment of epilepsy. Appl Neurophysiol 1985;48:400-3. 49. Barcia Salorio JL, Soler F, Hernandez G. Radiosurgical treatment of low flow carotid-cavernous fistulae. Acta Neurochir 1991;Suppl 52:93-5. 50. Burzaco JA, Gutierrez Gomez D. Trastorno de conducta postencefalitico (su tratamiento por cirugia estereotaxica). Archivos de Neurobiologia 1968;31:69-77. 51. Lopez-Ibor JJ, Burzaco J. Stereotaxic anterior limb capsulotomy in selected psychiatric patients. In proceedings of the second international conference in psychosurgery, Copenhagen, Denmark, Springfield, IL: Charles Thomas Publishers; 1970. p. 391-9. 52. Lopez-Ibor JJ, Lopez-Ibor JA, Burzaco JJ, Duque del Rio M. Capsulotomia esterotaxica. Indicaciones y Resultados. Actas Luso Espanõlas de Neurologıá Psiquiatrica y Ciencias Afines 1974;3:219-24. 53. Burzaco JA. Fundus Striae terminalis, an optional target in sedative stereotactic surgery. Surgical Approaches in Psychiatry. In: Laitinen L, editor. Third international


congress of psychosurgery. Chapter 20, Lancaster: MTP; 1973. p. 135-7. 54. Burzaco J. The role of some limbic structures in aggressive behavior. In: Gris P, Struwe G, Jansson B, editors. Elsevier. North-Holland. Biomedical Press. Biological psychiatry; 1981. p. 1223–1126.

14

55. Burzaco JA. Stereotactic surgery in severe unresponsive aggressive syndromes. Biological Psychiatry Vol. 2 Ed Racagni. Elsevier Science Publishers BV; 1991. p. 241–244. 56. Guridi J. Stereotactic neurosurgery in Spain. Neurocirugıá 2002;13:406.

191

3 History of Stereotactic Surgery in US P. L. Gildenberg

The early use of functional neurosurgery was begun by Horsley [1] in England at the end of the nineteenth century, and the invention of stereotactic surgery in animals was begun by Horsley and Clarke [2] at the beginning of the twentieth century. Human stereotactic surgery began almost at mid-century, i.e., 1947, in the United States by Ernest A. Spiegel and Henry T. Wycis [3] at Temple Medical School in Philadelphia. Stereotactic surgery was accepted at so many institutions so quickly that it is difficult to identify individual ‘‘schools’’ of stereotactic surgery, as it is in other countries. Indeed, in the early 1960s many neurosurgeons, who did not publish, were practicing functional neurosurgery, but most quickly left the field when L-dopa was introduced near the end of that decade [4] and have not returned to the field [5]. In the first half of the twentieth century, prior to the introduction of human stereotaxis, several American functional neurosurgeons were dominant in the use of non-stereotactic techniques for movement disorders. Paul Bucy, in Chicago, recommended motor cortex extirpation or corticospinal tractotomy for the treatment of athetosis [6] and Parkinson’s disease [7]. Bucy [7,8] had declared well into the 1950s that it was necessary to damage the corticospinal system in order to achieve relief from involuntary movements. Dandy [9], based on observation of several patients with intracerebral hemorrhage, had averred that damaging the extrapyramidal system would result in permanent intractable coma. Earl Walker, of Johns Hopkins, reported a pedunculotomy wherein he sectioned the lateral #


two-thirds of the peduncle for relief of hemiballismus [10] or parkinsonian tremor [11]. It was in attempting the Walker procedure that Irving Cooper had his famous ‘‘surgical accident,’’ where he accidentally cut the anterior choroidal artery and found the patient much improved, which led him to advocate ligation of that vessel for treatment of movement disorders [12], as discussed below. The most important but often overlooked American functional neurosurgeon of the prestereotactic era is Russell Meyers of the University of Iowa. It was his pioneering work that proved both Paul Bucy and Walter Dandy wrong. In 1939, Meyers [13,14] performed a craniotomy and transventricular approach to resect the head of the caudate nucleus for successful treatment of Parkinsonian tremor. His observations were presented at a meeting of the Research Association for Nervous and Mental Disease the following year [14], where a number of senior members encouraged him to pursue these observations, which led to the publication of results in 39 patients operated with various open surgical procedures of the basal ganglia [15]. The results were good, but the mortality rate was so high that even Meyers recommended against such surgery. His observations, however, were critical to defining the first extrapyramidal targets, significant information that led directly to the first stereotactic surgery for a movement disorder, Huntington’s chorea, with the pallidum as the target [3]. Stereotactic surgery produced even better results and brought the mortality rate down to 1% within 3 years of its introduction [16]. In 1963, Meyers resigned his appointment

46

3

History of stereotactic surgery in US

at the University of Iowa, reportedly for personal reasons, and became Chief of Neurosurgery of the Appalachian Regional Hospital at Williamson, West Virginia. It was not uncommon thereafter for the question to be raised at each Stereotactic Society Meeting, ‘‘Whatever happened to Russell Meyers?’’ Ernest A. Spiegel (> Figure 3-1) graduated from the University of Vienna, where he remained on the faculty both as a neurologist and laboratory scientist. As anti-Semitism advanced in Austria during the 1930s, he found himself with fewer students and restricted access to his laboratory. A businessman who observed this on a trip to Vienna related Spiegel’s plight to Dr. William Parkinson, the Dean of Temple Medical School, who invited not only Ernest Spiegel but also his wife, Mona Spiegel-Adolph, PhD, Professor of Colloid Chemistry, to join the faculty in Temple, where they both worked until retirement. Spiegel was small in stature, with wirerimmed glasses, a somewhat unkempt look, and a thick accent which identified him as a professor

steeped in the European academic philosophy. He remembered everything, and as he wrote his manuscripts (usually with a pencil stub on the back of used EEG paper), he inserted the citations without the need to consult his library. His laboratory consisted of two small and cramped rooms – one lined with shelves that contained an uncounted number of cat brains in baby food jars and the other containing a Faraday cage for recording. Inside the cage was an animal stereotaxic apparatus on a small table. Although many considered Spiegel severe because of his professorial look, he had a wry sense of humor, which was a necessary trait to work with Henry T. Wycis, who was a practical joker. Henry Wycis (> Figure 3-2) was incorrigible. He came from a middle class family, and helped supplement his living expenses by playing semiprofessional baseball during college, which seemed remarkable, because he weighed over 300 pounds by the time I met him. He began to work with Spiegel when he was a medical student – although his academic record was unsurpassed for many years, he also held the record

. Figure 3-1 Ernest A. Spiegel

. Figure 3-2 Henry T. Wycis


for having missed the most classes, which may have accounted for his time in Spiegel’s laboratory. They continued to work together while Wycis was a neurosurgical resident at Temple and after he had joined the faculty. Just the opposite of Spiegel, Henry Wycis loved to be with a diverse group of people. He enjoyed doing card tricks for his friends. It was rumored that he had earned his way through Medical School by successfully playing poker, which may be because of his photographic memory or may explain his love of card tricks. After stereotactic techniques were introduced in 1947 [3], many neurosurgeons visited Spiegel and Wycis at Temple Medical School in Philadelphia to learn of this new field and return home to make their own apparatus, (since none were commercially available for approximately the next decade,) and become practitioners of this new discipline. Although many of the visitors were from Europe, many were also US neurosurgeons. Even though the field was born in the US, the stereotactic community was and remains truly an international community. Those most active in the field during the 1950s, probably fewer than 30 surgeons, met irregularly at various institutions. The information exchange was informal, open, and enthusiastic, since those scientists shared the excitement of participating in a new science. The first patient reported by Spiegel and Wycis in 1947 [3] had Huntington’s chorea. His involuntary movements became more severe when he was stressed. Consequently, two lesions were made, one in the globus pallidus for the involuntary movement and one in the dorsomedial nucleus of the thalamus to lessen the stress reaction. Lesions were made with alcohol injection, in hopes of sparing the fibers en passage. The next patients had lesions made with a direct current, which had been described in detail in Horsley and Clarke’s original animal stereotaxis paper [2]. Spiegel and Wycis concluded their paper by commenting, ‘‘the apparatus is being

3

used for psychosurgery. . . Lesions have been placed in the region of the medial thalamus. . .Further applications of the stereotaxic technique are under study, e.g., interruption of the spinothalamic tract in certain types of pain or phantom limb; production of pallidal lesions in involuntary movements; electrocoagulation of the Gasserian ganglion in trigeminal neuralgia; and withdrawal of fluids from pathological cavities, cystic tumors.’’ Spiegel was somewhat secretive about future plans, and I am certain that they had already done those procedures prior to the first publication. I first met Spiegel and Wycis during my freshman year at Temple Medical School. A new summer research program had been announced, and I was looking for someone in the neurological sciences to be my sponsor and guide my research. I was referred by the head of physiology, since there was no neurophysiologist in that department I walked into Spiegel’s laboratory to find him and Wycis reviewing pre- and postoperative 8-mm motion picture films of patients. They invited me to sit and watch, and 2 hours later I was accepted as Spiegel’s graduate student. This was in the spring of 1956, just 9 years since the field of stereotactic surgery had begun, when it was still in its infancy. Spiegel and Wycis then performed surgery on only one patient per week. Targeting required a pneumoencephalogram, which was performed with the apparatus in place on Tuesday. The patient was so sick from the study that the surgical part was delayed until Thursday, when the apparatus was re-applied and the procedure done. Every time an electrode was inserted into the brain of an awake patient was an opportunity to study human neurophysiology, which both helped our understanding of the human extrapyramidal nervous system and provided physiologic confirmation that the electrode was at the intended target. In 1968, the International Society for Research in Stereoencephalotomy was formed, along with the American and European

47

48

3


Branches. The officers of the American Branch, which included both Canada and the United States, included Ernest Spiegel as President, Lyle French as Vice President, and Henry Wycis as Secretary-Treasurer. Claude Bertrand of Montreal was the sole Canadian member of the Board, and served along with Blaine S. Nashold, Jr., Vernon Mark, Robert W. Rand, and Orlando Andy. At the next official meeting in 1970, those same neurosurgeons remained officers, an indication of how small the membership was, with French as the President, and the non-neurosurgeon Spiegel resigned from his office. At the next meeting in 1972, John Alksne was a new member of the Board, in place of Lyle French. The organization was revamped in 1973 as the World Society for Stereotactic and Functional Neurosurgery, and the American Society for Stereotactic and Functional Neurosurgery (as well as the European Society) were formed. It was by that time 25 years since the introduction of stereotactic surgery, and each of the founders had taught a number of trainees, so the field had been expanding rapidly. The officers of the American Society retained Blaine Nashold as President and Claude Bertrand as Vice-President. I took over as Secretary shortly after Henry Wycis’ death in 1972, and held that position in both the American Society and the World Society for the next 28 years, except for the years I was President of the American Society from 1977 to 1980 and President of the World Society for Stereotactic and Functional Neurosurgery from 1993 to 1997. The evolution of stereotactic and functional neurosurgery in the United States took a somewhat different course than in most of the rest of the world. Up until the last decade, most stereotactic surgeons performed such surgery as part of a general neurosurgical practice, or one which also featured other subspecialties. There were few ‘‘stereotactic centers,’’ even in universities, as was the norm in other countries where medicine

was directed by government or academic agencies. Both medical school and private neurosurgical practitioners devoted part of their practice to stereotaxis, but also performed other neurosurgery, as well. This is in contrast to the last decade, where more sophisticated technology requires a team, so that multispecialty stereotactic services are more common, usually centered around teaching programs, such as Pat Kelly at NYU, Ali Rezai at the Cleveland Clinic, and others. In the pre-dopa days, the largest functional neurosurgical practice in the US was that of Irving Cooper, who devoted his practice to functional neurosurgery after the mid-1950s [17]. He embarked on functional neurosurgery because of a ‘‘surgical accident’’ [12]. He was performing a Walker [11] pedunculotomy for Parkinson’s disease, when he accidentally cut the anterior choroidal artery. He aborted the procedure, but the patient awoke with marked improvement. That led Cooper to advocate ligation of that vessel for the treatment of Parkinson’s disease [12]. The distribution of that artery, however, is very variable, and so were the results of its ligation. By that time, the pallidum had become a common target for stereotactic treatment of parkinsonism [18], so Cooper advocated injecting alcohol into that structure in so-called chemopallidectomy [19]. Again the results were variable, since the insertion of the needle was free-hand and the alcohol spread in an uncontrolled fashion along the adjacent tracts. He tried using a thicker solution, Etopalin, and a cannula with a balloon at the end in hopes of making a cavity that would contain the injected solution [20]. Neither of these maneuvers produced a more predictable lesion [21,22]. Finally, he recruited a freelance engineer, Arnold StJ. Lee, who designed the cryoprobe, which used a controlled release of liquid nitrogen through a probe to freeze the tissue at the tip [23]. Although it had a large blunt tip that injured tissue on insertion, it produced a predictable lesion at the tip. It was purely coincidence


that the engineer who built the Cryoprobe was the same Lee who is the fourth author on the first stereotactic paper by Spiegel and Wycis [3] in 1947. Cooper used an aiming device that was not truly stereotactic, in that it was not based on a Cartesian system. He approached the pallidum by inserting a cannula ventral and medial through the temporal lobe. One of his patients, who had an excellent result, was killed in an accident. An autopsy involving the brain showed that the lesion was in the ventrolateral thalamus, a target that had already been described as preferable for tremor. Cooper changed his target to that structure and reported another large series of chemopallidectomy. Cooper brought functional neurosurgery, including stereotactic surgery, to the public through the mass media. One of his patients was Margaret Bourke-White, the famous Life magazine photographer who suffered from Parkinson’s disease. She insisted that her surgery be photographed by one of her colleagues, Alfred Eisenstadt, who was equally famous. Her procedure and result were excellent, and the pictures and article in a mass circulation magazine brought considerable attention to Dr. Cooper and to stereotactic and functional neurosurgery. There have been persistent stories about clashes between Cooper and Wycis at stereotactic meetings. Some of the more colorful stories had them coming to physical blows. Not only is that not true, as far as I can document, but that competition would have gone to Wycis. He was a bear of a man, who weighed more than 300 pounds, and was a semi-professional athlete in his college years. There certainly were verbal assaults, however, and a sense of one-upmanship when they disagreed. I was present at a stereotactic meeting at Temple Medical School in 1958, when such a competition occurred. Cooper brought one his successful patients to the meeting to show how well he could write on a blackboard after surgery. Not to be outdone,

3

Henry Wycis brought a patient to the meeting the following day to demonstrate how well he played the piano after a pallidotomy. We have lost much of the color of those early meetings. There has been a recent shift toward implanting deep brain stimulators, but story actually began in the late 1960s. The development of the majority of neuromodulation devices presently used occurred in the United States, mainly through Medtronic1, that remains the dominant supplier of implantable stimulators [24]. The introduction of the gate theory by Melzack and Wall [25] in 1965 led Norm Shealy, a neurosurgeon at Western Reserve Medical School (now Case Western Reserve) in Cleveland, to consider stimulating the dorsal columns of the spinal cord to ‘‘close the gate’’ for relief of chronic pain [26]. He worked with Tom Mortimer, who previously had spent time working at an American company, Medtronic1, that at time had several implantable stimulators on the market. Their Barostat stimulator that was used for stimulation of the carotid nerve for management of hypertension in 1963 and their Angiostat in 1965 to treat angina, were adapted to electrodes designed by Mortimer to stimulate the spinal cord, and the field of neuromodulation using commercially available implantable stimulators was born [24]. Shealy left neurosurgery soon thereafter to become a horse rancher in Wisconsin and write mystery books. It was in 1973 that Hosobuchi [27], then at the University of California at San Francisco, inserted an electrode into the somatosensory thalamus to treat denervation pain, and deep brain stimulation became a reality. Shortly after that, Don Richardson [28,29], of Tulane in New Orleans, stimulated the periaqueductal area for management of somatic pain. At around the end of the 1970s, the use of deep brain stimulation required approval by the FDA, but only one of the three companies manufacturing implantable stimulators performed the necessary studies to document its benefit in pain; the third company

49

50

3


was Avery, who obtained approval, but their founder, Roger Avery, retired just at that time, so the use of deep brain stimulators was deapproved. The use of deep brain stimulation ceased in the US. It was not until the use of implantable deep brain stimulators for motor control was documented by Alim Benabid [30] and later by Jean Siegfried [31], colleagues in Europe, that there has been a re-ignition of DBS activity in US centers. This led to the reapproval of DBS for movement disorders by the FDA in early 2002. Lars Leksell [32,33] introduced the Gamma Knife for stereotactic radiosurgery in Stockholm. The first unit installed in the United Staates was guided through the vast regulatory bureaucracy by Dade Lunsford [34] at the University of Pittsburgh, which became one of the most active Gamma Knife centers anywhere. Once the regulatory problems had been addressed by Lunsford, many Gamma Knives were imported into the US, sometimes directed by neurosurgeons who had previously worked in Sweden, such as Ladislau Steiner at the University of Virginia and Georg Noren at Rhode Island Hospital. Although linear accelerator stereotactic radiosurgery was invented by Leksell [32] in Europe, he used primarily his Gamma Knife. Later Betti [35] and Colombo [36] reintroduced the linac for radiosurgery. Several US neurosurgeons also became active in that time when the benefits of radiosurgery were becoming apparent, but there were few Gamma Knives in this country. Linac systems were developed at the University of Florida at Gainesville by William Friedman and Frank Bova [37]. Several neurosurgeons including Peter Heilbrun were involved in development of the Radionics XKnife, which was first used at the Joint Center in Boston by Jay Loeffler [38]. The use of proton beam therapy was reported by Ray Kjellberg [39] as early as 1962. During the past few years, a Proton Beam Center was opened at Loma Linda and more recently at the MD Anderson Cancer Center in Houston.

Pain management has always been a major interest of US neurosurgeons. An appreciation for the evolving philosophy of neurosurgical management of pain can be obtained by perusing the three volumes co-authored by William Sweet, who was the epitome of professorship. The first was by James White and Sweet [40] in 1955, which emphasized the interruption of the primary pain pathways. The second by those same authors appeared in 1969 [41], and provided a somewhat more conservative approach that included the extralemniscal pathways, as well. The third by Jan Gybles of Belgium and Sweet [42] in 1989 emphasized the complexities of pain perception as a guide to management. Anterolateral cordotomy was simplified by Sean Mullan [43], who in 1963 reported a technique of lesioning that part of the spinal cord by percutaneous insertion of a strontium needle at the C1–2 level for a measured duration. The technique was further modified in 1965 by Hu Rosomoff [44], who used a radiofrequency lesion to interrupt the lateral spinothalamic tract, making it accessible to most neurosurgeons. Paul Lin and I [45,46] introduced a technique that introduce the needle through a lower cervical disk, thus avoiding fibers concerned with respiration. It is more than coincidence that the two neurosurgeons with the largest series of percutaneous cervical cordotomies, Rosomoff and I, came to recommend a very conservative approach to surgery and favored comprehensive multidisciplinary management of chronic pain [47,48]. Both of our programs resembled the comprehensive pain management program pioneered at the University of Washington in Seattle, which was led by the anesthesiologist John Bonica [49], and the neurosurgeon John Loeser [50,51]. Interest in stereotactic and functional neurosurgery has been increasing in the United States, especially since the use of deep brain stimulation for motor disorders became available in 2002. The field is of interest to neurologists


and neurophysiologists, as well as neurosurgeons, which fostered a team approach to management of Parkinson’s disease and other movement disorders. This has developed to the point where the use of intraoperative microelectrode recording has become the norm. Not only has the field become more active, but the scientific basis of the diseases and techniques are being studied with increased intensity in order to assure further progress. Because of the increased complexity and sophistication of electrode implantation, many multidisciplinary centers have developed in the US, such as David Roberts at the DartmouthHitchcosk Medical Center, Michael Kaplitt at Weill Cornell Medical College in New York, Pat Kelly at New York University, Ali Rezai at the Cleveland Clinic, Ray Bakay in Atlanta and then Chicago, then Robert Gross in Atlanta, Michael Schulder in Manhasset, NY, Philip Starr and Nicholas Barbaro at the University of California in San Francisco, Jamie Henderson at Stanford, Tony DeSalles and Mike Apuzzo in Los Angeles, and Kim Burchiel in Portland Oregon, to name but a few. In 1987 I asked, ‘‘Whatever happened to stereotactic surgery?’’ [5] The answer is, ‘‘It is doing well and advancing at an unprecedented rate.’’

References 1. Horsley V, Taylor J, Colman WS. Remarks on the various surgical procedures devised for the relief or cure of trigeminal neuralgia (tic douloureaux). BMJ 1891;2:1139-43. 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, Marks M, Lee AS. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 4. Cotzias GC, VanWoert MH, Schiffer LM. Aromatic amino acids and modification of parkinsonism. N Eng J Med 1967;276:374-9. 5. Gildenberg PL. Whatever happened to stereotactic surgery? Neurosurgery 1987;20:983-7.

3

6. Bucy PC, Buchanan DN. Athetosis. Brain 1932;55:479-92. 7. Bucy PC, Case TJ. Tremor. Physiologic mechanism and abolition by surgical means. Arch Neurol Psychiat 1939;41:721-46. 8. Bucy PC. Cortical extirpation in the treatment of involuntary movements. In: Putnam TJ, editor. The Diseases of the Basal Ganglia. New York: Hafner Publishing Co.; 1966. p. 551-95. 9. Meyers R. Dandy’s striatal theory of the ‘‘center of consciousness’’; Surgical evidence and logical analysis indicating its improbability. Arch Neurol Psychiat 1951;65:659-71. 10. Walker AE. Cerebral pedunculotomy for the relief of involuntary movements. I. Hemiballismus. Acta Psychiatr Neurol Scand 1949;24:712-29. 11. Walker AE. Cerebral pedunculotomy for the relief of involuntary movements. Parkinsonian tremor. J Nerv Ment Dis 1952;116:766-75. 12. Cooper IS. Ligation of the anterior choroidal artery for involuntary movements of parkinsonism. Psychiat Quart 1953;27:317-9. 13. Meyers R. Surgical procedure for postencephalitic tremor, with notes on the physiology of premotor fibers. Arch Neurol Psychiat 1940;44:455-9. 14. Meyers R. The modification of alternating tremors, rigidity and festination by surgery of the basal ganglia. Res Publ Ass Res Nerv Ment Dis 1942;21:602-65. 15. Meyers R. Historical background and personal experiences in the surgical relief of hyperkinesia and hypertonus. In: Fields W, editor. Pathogenesis and treatment of parkinsonism. Springfield, IL: Charles C Thomas; 1958. p. 229-70. 16. Gildenberg PL. Neurosurgical treatment of movement disorders: history and update. In: Germano I, editor. Neurosurgical treatment of movement disorders. New York: Thieme; 2002. p. 139-47. 17. Cooper IS. Results of 1000 consecutive basal ganglia operations for parkinsonism. Ann Intern Med 1960;52:483-99. 18. Hassler R, Riechert T. Indikationen und Lokalisationsmethode der gezielten Hirnoperationen. Nervenarzt 1954;25:441-7. 19. Cooper IS. Chemopallidectomy. Science 1955;121:217. 20. Cooper IS, Bravo G, Riklan M, Davidson N, Gorek E. Chemopallidectomy and chemothalamectomy for parkinsonism. Geriatrics 1958;13:127-47. 21. Gildenberg PL. Studies in stereoencephalotomy. VIII. Comparison of the variability of subcortical lesions produced by various procedures (radio-frequency coagulation, electrolysis, alcohol injection). Confin Neurol 1957;17:299-309. 22. Gildenberg PL. Study of methods of producing subcortical lesions and of evaluating their effect upon the tremor of parkinsonism. MS Thesis (Experimental Neurology), Temple University, Philadelphia, PA; 1959. 23. Cooper IS, Lee A St.J. Cryostatic congelation. J Nerv Ment Dis 1961;133:259-63.

51

52

3


24. Gildenberg PL. Evolution of neuromodulation. Stereotact Funct Neurosurg 2005;83:71-9. 25. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971-9. 26. Shealy CN, Mortimer JT, Reswick JB. Electrical inhibition of pain by stimulation of the dorsal columns. Preliminary clinical report. Anesth Analg (Cleve) 1967;46:489-91. 27. Hosobuchi Y, Adams JE, Rutkins B. Chronic thalamic stimulation for the control of facial anesthesia dolorosas. Arch Neurol 1973;29:158-61. 28. Richardson DE, Akil H. Pain reduction by electrical brain stimulation in man. Part 1: acute administration in periaqueductal and periventricular sites. J Neurosurg 1977;47:178-83. 29. Richardson DE, Akil H. Pain reduction by electrical brain stimulation in man. Part 2: chronic self-administration in the periventricular gray matter. J Neurosurg 1977;47:184-94. 30. Benabid AL, Pollak P, Hommel M, Gaio JM, de Rougemont J, Perret J. Treatment of Parkinson tremor by chronic stimulation of the ventral intermediate nucleus of the thalamus. Rev Neurol (Paris) 1989;145:320-3. 31. Siegfried J. Effect of stimulation of the sensory nucleus of the thalamus on dyskinesia and spasticity. Rev Neurol (Paris) 1986;142:380-3. 32. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta chir scand 1951;102:316-9. 33. Leksell L. Stereotaxis and radiosurgery: an operative system. Springfield, IL: Charles C Thomas; 1971. 34. Lunsford LD, Maitz A, Lindner G. First United States 201 source cobalt-60 gamma unit for radiosurgery. Appl Neurophysiol 1987;50:253-6. 35. Betti O, Derechinsky V. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurochir Suppl 1984;33:385-90. 36. Avanzo RC, Chierego G, Marchetti C, Pozza F, Colombo F, Benedetti A, Zanardo A. Stereotaxic irradiation with a linear accelerator. Radiol Med (Torino) 1984; 70:124-9. 37. Friedman WA, Bova F. The University of Florida radiosurgery system. Surg Neurol 1989;32:334-42.

38. Loeffler JS, Alexander E, III, Siddon RL, Saunders WM, Coleman CN, Winston KR. Stereotactic radiosurgery for intracranial arteriovenous malformations using a standard linear accelerator. Int J Radiat Oncol Biol Phys 1989;17:673-7. 39. Kjellberg RN, Koehler AM, Preston WM, Sweet WH. Stereotaxic instrument for use with the Bragg peak of a proton beam. Conf Neurol 1962;22:183-9. 40. White JC, Sweet WH. Pain, its mechanism and neurosurgical control. Springfield, IL: Charles C Thomas; 1955. 41. White JC, Sweet WH. Pain and the neurosurgeon. A forty year experience. Springfield, IL: Charles C Thomas; 1969. 42. Gybels JM, Sweet WH. Neurosurgical treatment of persistent pain. Physiological and pathological mechanisms of human pain. Pain Headache 1989;11:1-402. 43. Mullan S, Harper PV, Hekmatpanah J, Torres H, and Dobben G. Percutaneous interruption of spinal pain tracts by means of a strontium-90 needle. J Neurosurg 1963;20:931-9. 44. Rosomoff HL, Carrol F, Brown J, Sheptak P. Percutaneous radiofrequency cervical cordotomy: technique. J Neurosurg 1965;23:639-44. 45. Gildenberg PL, Lin PM, Polakoff PP II, Flitter MA. Treating intractable pain with percutaneous cervical cordotomy. GP 1968;37:96-7. 46. Lin PM, Gildenberg PL, Polakoff PP. An anterior approach to percutaneous lower cervical cordotomy. J Neurosurg 1966;25:553-60. 47. Gildenberg PL, DeVaul RA. The chronic pain patient. Evaluation and management. Basel: Karger; 1985. 48. Rosomoff HL, Green C, Silbret M, Steele R. Pain and low back rehabilitation program at the University of Miami School of Medicine. NIDA Res Monogr 1981;36:92-111. 49. Bonica JJ. Organization and function of a pain clinic. In: Bonica JJ, editor. International symposium on pain. New York: Raven Press; 1974. 50. Bonica JJ, Loeser J. Management of pain. Philadelphia, PA: Lea & Febiger; 1990. 51. Loeser JD. Disability, pain, and suffering. Clin Neurosurg 1989;35:398-408.

2 History of the Stereotactic Societies P. L. Gildenberg . J. K. Krauss

The field of human stereotactic neurosurgery was born in 1947 when Spiegel and Wycis (> Figure 2-1) published their groundbreaking manuscript in Science [1]. In this article they describe their device, which was actually a animal apparatus suspended on the patient’s head with a plaster cap. The original Horsley-Clarke animal stereotaxic (which was the original spelling) apparatus that had been used in the laboratory was aligned with the skull by means of earplugs and orbital tabs, and the target was defined by its relationship to those bony landmarks [2]. However, there was too much variability between the boney landmarks and the human brain to use this system clinically. It was only possible when intraoperative encephalographic x-rays became available in the 1940s, that it became possible to align the human apparatus with internal cerebral landmarks. Because their human system was based on encephalographic landmarks, Spiegel and Wycis termed their procedure ‘‘stereoencephalotomy.’’ During the next decade, neurosurgeons who were interested in learning this new technique visited Spiegel and Wycis in Philadelphia, and later at each other’s institutions. Since there were no commercially available stereotactic apparatus, they first had to design and build their own device when they returned home before they could use this new procedure. The pioneer stereotactic neurosurgeons exchanged information informally and personally. They met in small groups at irregular times, usually at each other’s institutions.

#


The International Society for Research in Stereoencephalotomy Periodically, the growing number of stereotactic neurosurgeons met more formally. Many of the early meetings were hosted by Ernest A. Spiegel and Henry T. Wycis at Temple Medical School in Philadelphia. In 1961, at one of their meetings, they attempted to provide a more formal venue for the exchange of stereotactic information and experience and founded the first stereotactic society, which was called the International Society for Research in Stereoencephalotomy, a term that never really caught on. When only the American members met, the meeting was designated as the American Branch of the International Society for Research in Stereoencephalotomy, and when European members met as the European Branch. They met as groups at approximately two year intervals (reference). Consequently, when the International Society for Research in Stereoencephalotomy met in 1963 in Philadelphia at what was designated the First International Symposium on Stereoencephalotomy, the society was formally chartered (> Table 2-1). The Second International Symposium on Stereoencephalotomy was held in 1965, partly in Copenhagen and partly in Vienna, the latter coincident with the Meeting of the World Federation of Neurosurgical Societies (WFNS) (> Table 2-2). The Third Symposium on Stereoencephalotomy took place in 1967 in Madrid, hosted by Sixto Obrador.

36

2

History of the stereotactic societies

. Figure 2‐1 Spiegel and Wycis

. Table 2-2 Meetings of the WSSFN 1961 1965

1967 1968

1969

1970

1973

1977 . Table 2‐1 History of the Stereotactic Societies 1961 1963 1968

1970

1973

1980

International Society for Research in Stereoencephalotomy founded, Philadelphia International Society for Research in Stereoencephalotomy chartered, Philadelphia American Branch of the International Society for Research in Stereoencephalotomy founded, Atlantic City, NJ European Society for Stereotactic and Functional Neurosurgery founded, Freiburg, Germany World Society for Stereotactic and Functional Neurosurgery founded, Tokyo, Japan American Society for Stereotactic and Functional Neurosurgery founded as society affiliated with the World Society for Stereotactic and Functional Neurosurgery First Meeting of the American Society for Stereotactic and Functional Neurosurgery, Houston, TX

1981

1985

1989

1993

1997

2001

2005

2009

The first formal independent meeting of the American Branch of the International Society for Research in Stereoencephalotomy met in Atlantic City, New Jersey, in 1968. It focused mainly on Parkinson’s disease, and was the last meeting to do so, since the field became

First International Symposium on Stereoencephalotomy, Philadelphia, PA Second International Symposium on Stereoencephalotomy, Copenhagen and Vienna Third International Symposium on Stereoencephalotomy, Madrid, Spain First Meeting of the American Branch of the International Society for Research in Stereoencephalotomy, Atlantic City, NJ Fourth Symposium of the International Society for Research in Stereoencephalotomy, New York, NY Fifth Symposium of the International Society for Research in Stereoencephalotomy, Freiburg, Germany Sixth Symposium of the International Society for Research in Stereoencephalotomy Meeting of the World Society for Stereotactic and Functional Neurosurgery, Tokyo, Japan Seventh Meeting of the World Society for Stereotactic and Functional Neurosurgery, Sao˜ Paulo, Brazil Eighth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Zurich, Switzerland Ninth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Toronto, Canada Tenth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Maebashi, Japan Eleventh Meeting of the World Society for Stereotactic and Functional Neurosurgery, Ixtapa, Mexico Twelfth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Lyon, France Thirteenth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Adelaide, Australia Fourteenth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Rome, Italy Fifteenth Meeting of the World Society for Stereotactic and Functional Neurosurgery, Toronto, Canada

quiescent after L-dopa was introduced in 1968 [3]. Although the meetings were relatively small, advances in pain management, epilepsy, and other movement disorders were exchanged.


The core of dedicated stereotactic neurosurgeons remained interested and active in the field, which led to the Fourth International Symposium on Stereoencephalotomy in 1969 in New York, along with the meeting of the World Federation of Neurosurgical Societies (WFNS). In 1970 the Fifth International Symposium on Stereoencephalotomy was held in Freiburg, Germany, hosted by Traugott Riechert, which is also recognized as the founding and the first meeting of the ESSFN. At that time [4], both the American and European branches were considered components of the International Society.

. Figure 2-2 Logo of both the WSSFN and the ASSFN

The World Society for Stereotactic and Functional Neurosurgery It was well into the early 1970s before activity began to increase once again, which led also to organizational changes. At the Sixth Symposium of the International Society for Research in Stereoencephalotomy in Tokyo in 1973, hosted by Hirotaro Narabayashi, the name of the international organization was changed to the World Society for Stereotactic and Functional Neurosurgery (WSSFN) (> Figure 2‐2). The American branch became the American Society for Stereotactic and Functional Neurosurgery (ASSFN), and the European branch became the European Society for Stereotactic and Functional Neurosurgery (ESSFN) (> Figure 2‐3). This caused an interesting dilemma. Horsley and Clarke designated their new technique as ‘‘stereotaxic,’’ as did many of the neurosurgeons involved in this new field. However, some neurosurgeons, particularly in Europe, spelled the term ‘‘stereotactic’’ [4]. There needed to be consensus on how to spell the names of these new societies.

. Figure 2-3 Logo of the ESSFN

2

37

38

2


A vote was taken, and ‘‘stereotactic’’ won, even though it was a mongrel term with ‘‘stereo,’’ meaning three dimensional, derived from Greek and ‘‘tactic,’’ meaning to touch, derived from Latin. The original term ‘‘stereotaxic’’ had been derived from two Greek words, ‘‘stereo’’ and ‘‘taxic,’’ meaning an arrangement. It was felt to reflect what stereotactic surgery did to use the term ‘‘to touch’’ rather than just to observe. The first meeting of the WSSFN was actually the reorganizational meeting held in Tokyo in 1973 and hosted by Hirotoro Narabayashi. At that meeting it was decided that the WSSFN would meet each time just prior to the meeting of the World Federation of Neurological Societies (WFNS) that met every 4 years. Since the WFNS alternated the location of the meetings from one continent to another, it became customary for the WSSFN to meet somewhere near the WFNS so members might attend both meetings with a minimum of travel, but yet in a separate venue. The WSSFN met every 4 years, with the next meeting held in 1977 in Sao˜ Paulo, Brazil, hosted by Raul Marino [5]. However, advances in the field occurred ever more rapidly, so each of the continental societies met during the intervening years. Each four year cycle began with the WSSFN meeting, then the ESSFN the following year, the ASSFN the next year, and the ESSFN the next year again. With that schedule, the ESSFN Meetings were held at 2 year intervals the first and third year of the cycle, whereas the WSSFN and ASSFN each met once every 4 years. Because there was more new information than could be provided to the members every 4 years, during the past 4 years the ASSFN also began to meet at 2 year intervals coincident with the years the ESSFN meets. The atmosphere at each of the meetings has been fraternal and informal, and the gathering of international stereotacticians maintained a feeling of collegiality. A perusal of the topics at each of the meetings of the WSSFN and the ASSFN has

served as a good indicator of what basic and clinical advances were made and what procedures and technology in stereotactic and functional neurosurgery were developed. The Proceedings of each of the meetings was published as complete issues of volumes of the official journal, Stereotactic and Functional Neurosurgery from 1968 through 2001. When the Eighth Meeting of the WSSFN in Zurich in 1981 was hosted by Jean Siegfried [6], the ESSFN declared that it was a joint meeting between the ESSFN and the WSSFN, and the ESSFN was not a component society of the WSSFN, but rather an independent society. The Ninth Meeting of the WSSFN was held in Toronto in 1985, with Ronald Tasker [7] as host. The Tenth Meeting of the WSSFN was held in Maebashi, Japan, in 1989 with Chihiro Ohye [8] as host. The Eleventh Meeting of the WSSFN was held in Ixtapa, Mexico, in 1993, with Philip L. Gildenberg [9] as the host (> Figure 2‐4). The 1997 the Twelfth Meeting of the WSSFN was held in Lyon, France, with Marc Sindou [10] as the host. It was by far the largest meeting and included many European neurologists and many local neurosurgeons. Deep brain stimulation for movement disorders was advancing rapidly, as was reflected in the program. The Thirteenth Meeting of the WSSFN was held in Adelaide, Australia, in 2001, hosted by Brian Brophy, and became a memorable meeting for an unrelated event. It was during that meeting that the terrorist attack on the World Trade Center occurred in New York. In addition to concerns about being far from family, all flights to the United States were canceled, both stranding attendees from returning to the US and preventing US neurosurgeons from attending the meeting of the World Federation of Neurosurgical Societies in Sydney immediately following the WSSFN Meeting. Many of the WSSFN speakers found themselves filling in for


2

. Figure 2‐4 Presidents of the WSSFN prior to 1993, when this picture was taken. Gildenberg, Nashold, Tasker, Narabayashi, Siegfried, Ohye, and Gybels

neurosurgeons who were unable to travel to speak at the WFNS in Sydney. The most recent WSSFN Meeting was held in Rome in 2005 and hosted by Mario Meglio. The 2009 meeting will be hosted by the current President, Andres Lozano, in Toronto. Additional updated information may be obtained on the WSSFN web site at www.wssfn.org. At each international symposium from 1977 to 1985, (see > Table 2‐3) the WSSFN has presented an award to an outstanding stereotactic neurosurgeon. At the presentation of the first award, it was named the Spiegel-Wycis Award, and a gold medal was made by Prof. Manuel Velasco-Suarez to be presented to Lars Leksell (> Table 2‐4). In 1981 the award was presented to Traugott Riechert. It was felt that there were too many potential recipients for only one award to be given every 4 years, so in 1985 a gold award was presented to Jean Talairach and a silver award to Manuel Velasco Suarez. Since there was little distinction between the

requirements for the gold and silver awards, two identical gold awards have been presented since 1989. In July, 2006, the WSSFN sponsored and organized for the first time an interim meeting in Shanghai, China, which was hosted by Bomin Sun from the Shanghai Jiao Tong University, Rui Jin Hospital. The meeting covered surgical treatment for movement disorders, pain, psychiatric disease, and epilepsy. Andres M. Lozano was the meeting chairman, and Drs. Benabid, Kaplit, Krauss, Schulder, Taira, Delong, Chang, Aziz, Broggi, Sun and Zhang were among those who gave the 36 plenary lectures. It was the first time that a WSSFN meeting was held on mainland Asia. The meeting coincided with a satellite symposium on functional neurosurgery in China. Overall 250 attendees from mainland China, Taiwan and Hong Kong joined WSSFN members from 15 other countries throughout the world. The event was considered a tremendous success by those who attended.

39

40

2


The American Society for Stereotactic and Functional Neurosurgery

. Table 2‐3 Meetings of the ASSFN 1980

1983

1987

1991

1995

1999

2003

2006

2008

Meeting of the American Society for Stereotactic and Functional Neurosurgery, Houston, Texas Meeting of the American Society for Stereotactic and Functional Neurosurgery, Durham, North Carolina Meeting of the American Society for Stereotactic and Functional Neurosurgery Montreal, Canada Meeting of the American Society for Stereotactic and Functional Neurosurgery, Pittsburgh, PA Meeting of the American Society for Stereotactic and Functional Neurosurgery, Los Angeles, CA Meeting of the American Society for Stereotactic and Functional Neurosurgery, Snowbird, Utah Meeting of the American Society for Stereotactic and Functional Neurosurgery, New York, NY Meeting of the American Society for Stereotactic and Functional Neurosurgery, Boston, MA Meeting of the American Society for Stereotactic and Functional Neurosurgery, Vancouver, Canada

The first separate meeting of the ASSFN was held in Houston in 1980 and hosted by Philip L. Gildenberg [11]. The 27 papers included basic neurophysiology, movement disorders, epilepsy and pain. The meeting attracted an international audience, which has been the hallmark of all of both the continental and the world stereotactic society meetings. It is interesting that even at that early date, 11 papers concerned the use of the newly developing use of computers in neurosurgical guidance, both for addressing functional targets and in the new field of image-guided neurosurgery. The next independent ASSFN meeting took place in Durham, North Carolina, in 1983, with Blaine Nashold [12] as the host, as the field showed signs of increased activity and reawakening. New technology was of greatest interest, and half of the papers concerned image guidance as it impacted brain tumor management.

. Table 2‐4 Past Recipients of the Spiegel-Wycis Award Recipient

Location

Year

Lars Leksell Traugott Riechert Jean Talairach Manuel Velasco Suarez Hirataro Narabayashi Denise Albe-Fessard Ronald R. Tasker Blaine S. Nashold, Jr. Bjorn Meyerson Philip L Gildenberg Lauri V. Laitinen Chihiro Ohye Patrick Kelly Alim-Louis Benabid

Sao˜ Paolo Zurich, Switzerland Toronto, Canada Toronto, Canada Maebayashi, Japan Maebayashi, Japan Ixtapa, Mexico Ixtapa, Mexico Lyon, France Lyon, France Adelaide, Australia Adelaide, Australia Rome, Italy Rome, Italy

July, 1977 July, 1981 July, 1985 July, 1985 October, 1989 October, 1989 October, 1993 October, 1993 July, 1997 July, 1997 September, 2001 September, 2001 June, 2005 June, 2005


The 1987 meeting in Montreal, which was hosted by Andre´ Olivier, was the largest ASSFN meeting up to that time. Out of the 103 papers, 14 concerned the rapidly developing field of stereotactic radiosurgery and 41 concerned the use of computers in neurosurgery. Functional neurosurgery showed signs of significant awakening, as represented by 36 presentations. The 1991 meeting in Pittsburgh was hostedby L. Dade Lunsford [13], and represented another milestone. The stereotactic surgery meeting was followed by a two day meeting on stereotactic radiosurgery, which attracted nonneurosurgical colleagues in radiation physics and radiotherapy, as well as those stereotactic neurosurgeons who were using the Gamma Knife, the only stereotactic radiosurgical system available at that time. It was at that satellite meeting that the International Society for Stereotactic Radiosurgery was formed. In addition, this large meeting involved such new fields as tissue transplantation. It was also at that meeting that Lauri Laitinen [14,15] first presented his observations on ventroposterior pallidotomy, which signaled the return of Parkinson’s disease surgery to the functional neurosurgery arena. The ASSFN meeting was held in Los Angeles and hosted by Michael Apuzzo and David Roberts in 1995 [16,17]. Half of the papers were about classical functional neurosurgical topics of movement disorders, epilepsy and pain. There were eight papers about stereotactic radiosurgery and the rest about image guided surgery and brain tumor management. The 1999 ASSFN meeting was held in Snowbird, Utah, and hosted by Peter Heilbrun and Douglas Kondziolka. Two-thirds of the 55 papers were in functional neurosurgery, with emphasis on movement disorders, epilepsy, and pain. The rest were divided among imaging and computer guidance, stereotactic radiosurgery, and brain tumors. In 2003, the ASSFN met in New York, and was hosted by Patrick Kelly. The program

2

demonstrated the increasing use of computers in stereotactic and image guided surgery. Prior to that meeting, the FDA had provided approval of chronic deep brain stimulation, and much of the data related to that was presented. The meeting in 2006 was held in Boston and hosted by G. Rees Cosgrove. There was a significant shift toward interest in functional neurosurgery, involving both DBS and extending to epilepsy surgery. The current administrative structure includes officers elected for a term of 4 years, a board of directors and five Continental Vice Presidents. During the period from 2005 – 2009 the ESSFN leadership is represented by Andres M. Lozano (Toronto, Canada) as President, Joachim K. Krauss (Hannover, Germany) as Vice-President, Takaomi Taira (Tokyo, Japan) as Secretary-Treasurer, Michael Schulder (Newark, USA) as Assistant Secretary-Treasurer, and Philip L. Gildenberg (Houston, USA) as Historian. Honorary members include Blaine Nashold, Jean Siegfried, Ronald Tasker, Chihiro Ohye and Philip L. Gildenberg. Prior to 1975, there was no formal arrangement between the ASSFN and the American Association of Neurological Surgeons (AANS), although it was customary to have a representative of stereotactic surgery on the AANS program committee. In 1975, due to the efforts of then ASSFN President John Van Buren, the ASSFN amalgamated with the AANS and became the Section for Stereotactic and Functional Neurosurgery. This arrangement put the responsibility for the program of one and later two afternoon sessions in the hands of the ASSFN officers. In addition, members of the ASSFN automatically became members of the AANS Section of Stereotactic and Functional Neurosurgery. The ASSFN treasury, however, remained independent of the AANS, which provided the opportunity for the ASSFN to hold independent meetings every 4 years. The business meeting of the ASSFN is held at the conclusion of one of the days when the

41

42

2


stereotactic scientific program is held. Since 2004, the annual residents’ research award, which has been named the Gildenberg Award, has been presented at one of the Stereotactic Sections programs. In 1984, a similar administrative arrangement was made between the ASSFN and the Congress of Neurological Surgeons (CNS). The amalgamation was facilitated by the presence of the ASSFN President, George Sypert, on the CNS Board of Directors at that time. Current details of the ASSFN can be found on the web site at www.assfn.org.

The European Society for Stereotactic and Functional Neurosurgery The formation of the European Society for Stereotactic and Functional Neurosurgery (ESSFN) dates back as far as 1970 when the International Society for Research in Stereoencephalotomy met in Freiburg, Germany. The founding fathers at that time were Fritz Mundinger, Traugott Riechert and John Gillingham, and this meeting now is also considered as the founding meeting of the ESSFN. The society was registered as a taxexempted society in Freiburg, Germany, in 1971. It was founded primarily in order to represent the interests of European functional neurosurgeons and also to enhance communication and exchange between European countries. At that time, the European idea was blossoming, but it took still several years until traveling from one European country to the other was as comfortable as it is nowadays. The judicial seat of the society was moved to Toulouse, France, in 2002. Initially, the ESSFN had a relatively isolated position in the general neurosurgical community. With the rapid development of new imaging techniques and computer technology, however, stereotactic concepts and methods had a profound impact on the progress of European neurosurgery, in general. The ESSFN over the

decades thus served both the interests of those who were subspecialized in functional and stereotactic neurosurgery and those whose activities were embedded in general neurosurgery. The ESSFN has stated its principal objectives in a mission which is displayed on the ESSFN web site (www.essfn.org). Also, its Constitution and By-Laws are displayed on the web site. The logo of the ESSFN became popular during the 1990s (> Figure 2‐3). The board of officers is elected for a term of 4 years by the General Assembly which comes together during the congresses. It is supported by an Executive Committee which has representatives from 20 different European countries. During the period from 2006 – 2010 the ESSFN leadership is represented by Yves Lazorthes (Toulouse, France) as President, Giovanni Broggi (Milano, Italy) as Vice-President, Joachim K. Krauss (Hannover, Germany) as Secretary, Damianos Sakas (Athens, Greece) as Second Secretary, and Bart Nuttin (Leuven, Belgium) as Treasurer. The honorary presidents are the former ESSFN presidents F. John Gillingham, Fritz Mundinger, Gian Franco Rossi, Bjo¨rn Meyerson, Christoph B. Ostertag, David G. T. Thomas, and Andries Bosch. There are five different categories of membership: active, associate, resident, honorary and benefactor. Membership has increased steadily over the years, and in early 2008 there were more than 230 members. The majority of members come from France, Germany, Italy, Spain, The Netherlands, and the UK. There are members, however, from almost all European countries with increasing numbers in particular from Eastern European countries and Russia. Overall, the society has about 30 non-European members coming mainly from South Korea, Japan, Mexico and the United States. Since its inception in 1970, the ESSFN organized congresses at regular intervals in various locations all over Europe (> Table 2‐5). While the first meeting was still under the umbrella of the International Society for Research in


2

. Table 2‐5 Congresses of the ESSFN 1970 1972 1975 1977 1979 1981 1983 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

Founding Meeting of the European Society for Stereotactic and Functional Neurosurgery, Freiburg, Germany First Congress of the European Society for Stereotactic and Functional Neurosurgery, Edinburgh, United Kingdom Second Congress of the European Society for Stereotactic and Functional Neurosurgery, Madrid, Spain Third Congress of the European Society for Stereotactic and Functional Neurosurgery, Freiburg, Germany Fourth Congress of the European Society for Stereotactic and Functional Neurosurgery, Paris, France Fifth Congress of the European Society for Stereotactic and Functional Neurosurgery, Zurich, Switzerland Sixth Congress of the European Society for Stereotactic and Functional Neurosurgery, Rome, Italy Seventh Congress of the European Society for Stereotactic and Functional Neurosurgery, Birmingham, United Kingdom Eighth Congress of the European Society for Stereotactic and Functional Neurosurgery, Budapest, Hungary Ninth Congress of the European Society for Stereotactic and Functional Neurosurgery, Marbella, Spain Tenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Stockholm, Sweden Eleventh Congress of the European Society for Stereotactic and Functional Neurosurgery, Antalya, Turkey Twelfth Congress of the European Society for Stereotactic and Functional Neurosurgery, Milano, Italy Thirteenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Freiburg, Germany Fourteenth Congress of the European Society for Stereotactic and Functional Neurosurgery, London, United Kingdom Fifteenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Toulouse, France Sixteenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Vienna, Austria Seventeenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Monterux, Switzerland Eighteenth Congress of the European Society for Stereotactic and Functional Neurosurgery, Rimini, Italy

Stereoencephalotomy, the subsequent congresses were proper ESSFN congresses. The 1981 congress in Zu¨rich was considered a joint meeting of both the WSSFN and the ESSFN. While attendance was limited in the early years to those who practiced in a subspeciality frame, the scope of topics in recent years became much broader attracting also many neurosurgeons not devoted exclusively to functional neurosurgery. During the congress, awards are available for best oral presentations by young neurosurgeons and also prizes for best posters. The ESSFN also provides grants for research in stereotactic and functional neurosurgery considering both basic and clinical research. The grant recipients will report at the congresses about their results. In the years between the congresses the ESSFN administers hands-on courses which are open to ESSFN members or those who wish to apply for membership. The purpose of these courses is to provide education and training in

functional and stereotactic techniques. The first course was given in 2003 (Tolochenaz, Switzerland) on movement disorders surgery, followed by courses on pain surgery (Toulouse, France) in 2005, and on radiosurgery (Marseilles, France) in 2007. The topics for future courses will be adapted to include new horizons, but also to provide in-depth teaching about common standards. The web site has become an important medium to keep the membership posted with updated information. Application forms for membership can be downloaded from the web site. Furthermore, members are supplied with a Newsletter that is distributed about yearly detailing new developments and announcements for grants and courses. There are many close institutional and personal links between the ESSFN and the other societies for stereotactic and functional neurosurgery, in particular to the WSSFN. Nevertheless, there is no such a formal affiliation between the

43

44

2


ESSFN and the WSSFN as it has been established between the ASSFN and the WSSFN. The ESSFN also communicates with the national functional neurosurgery societies which have been established in almost all European countries. In addition, a subcommittee of the ESSFN cooperates with the UEMS (Union Europeene des Medecines Specialistes) to establish more uniform guidelines and standards for stereotactic and functional neurosurgery for all European countries.

Other International Stereotactic Surgical Societies In addition to the time-honored European and American functional neurosurgery societies, there are other international societies that organize meetings to enhance regional information exchange. The Sociedad Latinoamericana de Neurocirugia Funcional y Estereotaxia (SLANFE) is active since 1998 and covers the interests of functional neurosurgeons in Latin America. The Asian Society for Stereotactic, Functional and Computer Assisted Neurosurgery (ASSFCN) convenes meetings in Asia. In addition, several countries maintain their own Society for Stereotactic and Functional Neurosurgery, in order to accommodate their junior neurosurgeons who may not speak English or have funds to travel to the international meetings, such as Japan, Korea, Argentina, and China.

References 1. Spiegel EA, Wycis HT, Marks M, et al. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 2. Horsley V, Clarke RH. The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45-124.

3. Cotzias GC, VanWoert MH, Schiffer LM. Aromatic amino acids and modification of parkinsonism. New Engl J Med 1967;276:374-9. 4. Gildenberg PL. ‘‘Stereotaxic’’ versus ‘‘stereotactic’’. Neurosurgery 1993;32:965-6. 5. Gildenberg PL, Marino RJ. Seventh symposium of the international society for research in stereoencephalotomy. Conf Neurol 1978;41:1-250. 6. Gildenberg PL, Siegfried J, Gybels J, et al. Eighth meeting of the world society and the fifth meeting of the european society for stereotactic and functional neurosurgery. Appl Neurophysiol 1982;45:1-554. 7. Tasker RR, Turnbull IM, Gildenberg PL, Franklin PO, editors. Proceedings of the ninth meeting of the world society for stereotactic and functional neurosurgery. Appl Neurophysiol 1985;48(1–6):1‐498. 8. Ohye C, Gildenberg PL, Franklin PO. Proceedings of the tenth meeting of the world society for stereotactic and functional neurosurgery. Stereotac Funct Neurosurg 1990;54–55:1-564. 9. Gildenberg PL, Franklin PO, Escobedo FR, et al. Proceedings of the eleventh meeting of the world society for stereotactic and functional neurosurgery. Stereotac Funct Neurosurg 1994;63:1-301. 10. Sindou M, Martens F, Gildenberg PL, et al. Proceedings of the twelfth meeting of the world society for stereotactic and functional neurosurgery. Stereotac Funct Neurosurg 1997;68:1-318. 11. Gildenberg PL. Proceedings of the meeting of the American society for stereotactic and functional neurosurgery. Appl Neurophysiol 1980;43:89-266. 12. Nashold BS Jr, Gildenberg PL, Franklin PO. Proceedings of the American society for stereotactic and functional neurosurgery. Appl Neurophysiol 1983;46:1-252. 13. Lunsford LD, Gildenberg PL, Franklin PO. Proceedings of the meeting of the American society for stereotactic and functional neurosurgery, Part I. Stereotac Funct Neurosurg 1991;58:1-208. 14. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. 15. Laitinen LV, Bergenheim AT, Hariz MI. Ventroposterolateral pallidotomy can abolish all parkinsonian symptoms. Stereotact Funct Neurosurg 1992;58:14-21. 16. Roberts DW, Apuzzo MLJ, Gildenberg PL, et al. Proceedings of the meeting of the American society for stereotactic and functional neurosurgery, Part I. Stereotac Funct Neurosurg 1995;65:1-207. 17. Roberts DW, Apuzzo MLJ, Gildenberg PL, et al. Proceedings of the meeting of the American society for stereotactic and functional neurosurgery, Part II. Stereotac Funct Neurosurg 1996;66:1-156.

21 Angiography, MRA in Image Guided Neurosurgery P. Jabbour . S. Tjoumakaris . R. Rosenwasser

Introduction

Indication

Image-guided neurosurgery has become a standard practice in the last decade. This technology can assist the surgeon in different steps, including planning and executing the surgical procedure. So far advances in image guided surgery had little influence on vascular neurosurgery because of technical difficulties transferring 3-D vessels imaging to the neuronavigation system. The introduction of the 3-D rotational angiography has revolutionized the way we visualize and treat brain aneurysms. Sometimes it can be challenging to try to compare these views with the intraoperative microscopic view that the surgeon has. Incorporating the images of 3-D angiography in the image guided system during real time surgery is a recent technological innovation that enables the surgeon to navigate using 3-D angiogram pictures. Another innovation is fusing MRA or CTA images to the neuronavigation technique. These new techniques can be used for aneurysm, AVM and tumor surgery. One of the concerns for the neurosurgeon during tumor surgery is to avoid injury of blood vessels surrounding the tumor or encased by the tumor. Tumor surgery may be performed more safely using the intraopeartive navigated 3-D ultrasound angiography. The classical technique involves using an intraoperative ultrasound to correct the shift during surgery to keep the neuronavigation accurate. The same equipment has also Doppler capability with power Doppler to visualize vessels.

Angiography, CTA or MRA coupled to the image guided techniques is a new addition to the armamentarium for the surgical treatment of vascular lesions and tumors in the proximity of big vessels. It is essential to define exactly the 3-D relationship of the neck of an aneurysm with the surrounding vascular branches and perforating vessels in aneurysm surgery, or the exact drainage and feeding pattern of a complex arteriovenous malformation, or the exact relationship of a tumor with the surrounding vessels and its pattern of vascular invasion. All this makes this new technique essential in the surgical treatment of complex brain aneurysms, as well as distal aneurysms like pericallosal aneurysms or distal middle cerebral artery (MCA) aneurysm because it is sometimes challenging to anatomically locate these aneurysms due to the lack of big parent vessels close by. It is also used in the treatment of complex AVMs and it helps in identifying feeding and draining vessels and provides a better understanding of the AVM architecture and facilitates its resection. And finally it is also a useful technique in the surgical resection of skull base tumors located close to big vessels.

#


Techniques CTA and Neuronavigation A multislice thin cuts CT is performed with superficial fiducials on the skin with intravenous

300

21

Angiography, MRA in image guided neurosurgery

contrast injection. The data obtained is transferred directly or with a compact disc to the navigation station. A pre surgical planning can start at the navigation station on the 3-D model of the CT angiogram with special attention to the lesion and the surrounding vessels and structure. During surgery the 3-D vessel model is displayed on the navigation screen, and the screen display is matched with the intraoperative microscopic view of the surgeon to try to optimize the microdissection, for skull base pathology the bone can be substracted from the image displayed by the navigation system (> Figure 21-1). Chibbaro et al. [1] reported their experience using this technique, and they concluded that this technique can provide many advantages like a quick processing time of the raw data from the CT, a better understanding of the real position of the aneurysm with respect to the surgical position of the patient, it is also helpful in tailoring the craniotomy in case of small AVMs, sparing eloquent brain and fusing functional MRI with it, and they concluded that this technique avoids major problems, such as premature bleeding and or ischemic complications which often occur due to insufficient anatomical understanding of complex vascular lesions. Kim et al. [2], described their experience using the non formatted CTA coupled to neuronavigation in the clipping of 12 distal Anterior cerebral artery aneurysms (ACA) aneurysms, they concluded that this technique enables neurosurgeons to identify a distal ACA aneurysm and facilitates the safe exposure and clipping of the aneurysm. Rohde et al. [3,4] used the term ‘‘advanced neuronavigation.’’ They coupled the 3-D CTA images, MRA images and tractography to the neuronavigation system, they reported their series of 16 patients where they used this technique during surgery. Nine skull base meningiomas, one craniopharyngioma, one epidermoid, one giant carotid artery aneurysm, two basilar aneurysms, two brain stem cavernomas. This technique

facilitated the approach in four cases contributed to a tailored approach in two cases, helped to identify hidden vessels in nine cases. Rohde [4] later on described his series of 42 aneurysms in a prospective study. They used a 3-D reformatted CTA coupled to the neuronavigation system. Twelve aneurysms of the anterior communicating artery ACOM, 17 MCA aneurysms, six internal carotid artery aneurysms (ICA), three pericallosal artery aneurysms, two vertebro basilar aneurysms, and two superior cerebellar artery (SCA) aneurysms, all ruptured with subarachnoid hemorrhage. In 24 cases the neurosurgeon used this technique to localize the aneurysm, in 18 cases it was used to understand the branching anatomy, in 8 cases to visualize hidden structures, 5 to evaluate the projection of the dome, in 2 cases to tailor the approach, the author concluded that this technique has the potential to improve the operative results by reduction of the surgical trauma and avoidance of intraoperative complications. Coenen et al. [5], used the 3-D CTA coupled to neuronavigation for the surgical treatment of small AVMs with large hematomas in four patients, this technique allows feeding arteries to be distinguished from draining veins thereby allowing the nidus of the AVM to be identified despite the presence of intracerebral blood, and decreasing surgical morbidity.

MRA and Neuronavigation Coupled to Microvascular Doppler Sonography A multislice thin cuts MRI and MRA is performed with fiducials on the skin. The images are archived on a compact disc or transferred directly to the image guidance station at the operating room where the surgery is being performed. The 3-D ultrasound volumes are acquired within the same coordinate system as navigation is performed, when the craniotomy is performed a 3-D ultrasound scan is


21

. Figure 21-1 Digital subtraction angiography (upper left) and CT angiography (CTA) show a giant aneurysm of the right internal carotid artery (ICA) in a 71-year-old female patient, who was admitted after subarachnoid hemorrhage (Hunt/Hess grade IV). A 3-D reconstruction of the arterial vessel anatomy was created, rotated according to the surgeon’s perspective and displayed on the screen of the neuronavigational system during surgery (upper right). Additionally, the 2-D CT angiographic images were screen-displayed (lower left). For the surgeon, neuronavigation with CTA was especially helpful to tailor the extradural anterior clinoid process resection. Aneurysm clipping was successfully performed (lower right) and the patient made a good recovery despite her poor Hunt/Hess grade (with permission from Rohde et al. Advanced neuronavigation in skull base tumors and vascular lesions. Minim Invasive Neurosurg 2005;48(1):13-8)

performed before the dura is opened, the surgeon will be able to view reconstructed 2-D cross sectional images. Another 3-D ultrasound scan is obtained at the end of the resection and a comparison of the preoperative and postoperative 3-D ultrasounds

images is performed, this is mainly used in AVM surgery (> Figure 21-2). Mathiesen et al. [6] reported their experience using MRA with 3-D ultrasound angiography coupled to neuronavigation in the surgical

301

302

21


. Figure 21-2 Preoperative angiograms (a–d) and MRI scans (e and f), intraoperative ultrasound navigational 3-D (g, posterior view; h, frontolateral view) and postoperative angiograms (i–l) (with permission from Mathiesen et al. Neuronavigation for arteriovenous malformation surgery by intraoperative 3-D ultrasound angiography. Neurosurgery 2007;60(4 Suppl 2):345-50; discussion 341-50

treatment of AVMs. Nine patients with AVMs were treated, five had underwent pre operative embolization, six patients had pre operative MRIs, in six patients the stereoscopic display technique was used, it is a volume rendering technique that provides a true sense of 3-D vision when using special glasses the skin signal is subtracted and the surgeon is able to look at the MRA images of the AVM, allowing him to plan his approach and could navigate using the pointer and trying to identify feeders and draining vessels. The AVMs were totally removed in all nine patients. The ultrasound images corresponded to the intraoperative findings. This technique led to a

minimum exploration into the nidus, in two cases when the surgeon thought that he had a complete resection of the AVM, the 3-D ultrasound was able to demonstrate a residual nidus, allowing the surgeon to remove the residual AVM. The author concluded that AVM surgery was facilitated by navigation based on preoperative MRA coupled to intraoperative 3-D ultrasound and 3-D reconstruction. The AVMs were well outlined with better understanding of the feeders and draining vessels (> Figure 21-3). Unsgaard et al. [7], used also this technique in nine patients, seven patients had AVMs in eloquent brain with Spetzler-Martin grade II in


21

. Figure 21-3 Intraoperative photograph (a), biplanar angiogram (b), and 3-D-RA view (c) demonstrating a ruptured right middle cerebral artery aneurysm during the early phase of microsurgical preparation. The 3-D-RA view helps to identify the origin of the branching vessel and the localization of a large daughter bleb that was the site of rupture and was at high risk for intraoperative rerupture. The 3-D-RA model can also be viewed from behind for better identification of the temporal M2 branch that was covered by other temporal branches arising from the M3 and M4 segments (with permission from Raabe et al. 3-D rotational angiography guidance for aneurysm surgery. J Neurosurg. 2006;105 (3):406-11)

four patients, III in four patients and IV in one patient. The 9 AVMs had 28 feeders seen on the preoperative angiograms. Twenty-five of those feeders were clipped at the beginning of surgery based on the stereoscopic information from both 3-D MRA and ultrasound angiography. The other feeders were clipped in a later phase of the surgery, in one patient intraoperative ultrasound angiography revealed residual nidus that was immediately removed. Four of the seven patients with AVMs in an eloquent area had a temporary worsening of their neurological status, one patient had a permanent neurological deficit. One patient with a grade IV AVM had to be reoperated for a post operative hematoma due to a residual nidus. The authors concluded that navigated stereoscopic display of angiography offers a technology that can be used successfully to identify and clip AVM feeders in the initial phase of the operation.

Rotational Angiography and Neuronavigation 3-D rotational angiography is increasingly used to diagnose aneurysms and to better evaluate the branching vessels close to the aneurysm neck. A rotational 3-D angiography is performed. 3-D data is reconstructed, the images are archived on a compact disc or transferred directly through a network to the image guidance station. Conventional registration techniques can not be used without any CT or MRI, because the 3-D data contains only the region of interest of the angiogram, and no data representing surface matching with the patient. The registration is done with a technique that correlates the 3-D angiogram with the head positioning of the patient and this by using the angulation and the rotational coordinates from the angiogram images. A special reference head frame is used just before

303

304

21


the rotational angiogram is performed to record the position of the head with respect to the coordinates of the angiographic system. After reconstructing the rotational angiography images and optimizing the quality of the image, 3-D data is exported and loaded in the navigation system. The same three point headframe that was used during angiography is repositioned on the patient’s head. The position of the three point headframe in relation to the 3-D angiography data is already known from the fluoroscopic reference position image obtained before the rotational angiography procedure. Raabe et al. [8], used the previous technique on a series of 16 aneurysms, the procedure was successful in all cases. Using this technique added an average of 19 min to the procedure. There was a good correspondence with the intraopeartive vascular anatomy. It was determined by the surgeon, that this technique was helpful in eight cases in predicting the exact location of the aneurysm and the branching vessels covered by clots or brain parenchyma. The authors concluded that this technique provides useful information about the anatomical relationship between the aneurysm and the parent or branching vessel it helps minimizing the exposure and improves the quality of aneurysm surgery in selected cases. Willems et al. [9], used a different registration technique, not involving a headframe. Following the 3-D rotational angiogram acquisition, a navigated pointer is used to determine the position of the fiducials relative to the 3-D rotational angiography data set. The angiography room in this case is provided with navigation capabilities. The rotational angiography is coupled to an image guidance system by using a new software module, that enables the determination of the positional relation between the imaged volume and the tracker plate on the image intensifier, the positional relationship between the patient and the tracker plate is determined after the patient undergoes imaging, the data is transferred to

the operating room to the navigation system, the fiducials position are localized with a pointer and the patient to image registration is completed. They tried this technique on a phantom and on two cases and they concluded that there is a small error using this technique, but this does not prevent them from preserving the orientation of the vascular tree, and the goal of offering a 3-D rotational angiograms mimicking the surgical view was also achieved with technique.

Limitations One of the limitations is the failure to depict small arteries, another limitation would be the accuracy of the navigation system, and the brain shift after the bone is removed, the dura is open, and the cerebrospinal fluid CSF is drained. The registration process is still experimental especially in the image guided rotational 3-D angiography.

Conclusion Image guided vascular surgery using MRA, CTA or 3-D rotational angiography is a new technique developed to help surgeons identifying the topographical relationship between the vascular lesion and the related parent and branching vessels it can help to tailor the exposure and the dissection of the vascular tree, and therefore minimize the risk of intraoperative complications. It is still a new technique and it can benefit from a lot of improvement mainly in the registration step. Further studies and technical development are needed to try to make this new technique a standardized one.

References 1. Chibbaro S, Tacconi L. Image-guided microneurosurgical management of vascular lesions using navigated computed tomography angiography. An advanced IGS technology application. Int J Med Robot 2006;2(2):161-7.


2. Kim TS, Joo SP, Lee JK, et al. Neuronavigation-assisted surgery for distal anterior cerebral artery aneurysm. Minim Invasive Neurosurg 2007;50(3):140-4. 3. Rohde V, Spangenberg P, Mayfrank L, Reinges M, Gilsbach JM, Coenen VA. Advanced neuronavigation in skull base tumors and vascular lesions. Minim Invasive Neurosurg 2005;48(1):13-8. 4. Rohde V, Hans FJ, Mayfrank L, Dammert S, Gilsbach JM, Coenen VA. How useful is the 3-dimensional, surgeon’s perspective-adjusted visualisation of the vessel anatomy during aneurysm surgery? A prospective clinical trial. Neurosurg Rev 2007;30(3):209-16; discussion, 207-16. 5. Coenen VA, Dammert S, Reinges MH, Mull M, Gilsbach JM, Rohde V. Image-guided microneurosurgical management of small cerebral arteriovenous malformations: the value of navigated computed tomographic angiography. Neuroradiology 2005;47(1):66-72.

21

6. Mathiesen T, Peredo I, Edner G, et al. Neuronavigation for arteriovenous malformation surgery by intraoperative three-dimensional ultrasound angiography. Neurosurgery. 2007;60(4): Suppl 2:345-50; discussion 341-50. 7. Unsgaard G, Ommedal S, Rygh OM, Lindseth F. Operation of arteriovenous malformations assisted by stereoscopic navigation-controlled display of preoperative magnetic resonance angiography and intraoperative ultrasound angiography. Neurosurgery 2005;56 Suppl 2:281-90; discussion 281-90. 8. Raabe A, Beck J, Rohde S, Berkefeld J, Seifert V. Threedimensional rotational angiography guidance for aneurysm surgery. J Neurosurg 2006;105(3):406-11. 9. Willems PW, van Walsum T, Woerdeman PA, et al. Image-guided vascular neurosurgery based on threedimensional rotational angiography. Technical note. J Neurosurg 2007;106(3):501-6.

305

19 CT/MRI Safety in Functional Neurosurgery M. Schulder . A. Oubre´

Introduction Unique and ever-changing applications of imaging technologies have played an integral role in transforming the landscape of stereotactic and functional neurosurgery [1]. Both computer tomography (CT) and magnetic resonance imagery (MRI) continue to push forward the boundaries of image-guided neurological surgery. Despite the great benefits offered by CT and MRI, however, each technique poses certain risks to the safety of patients and, in some cases, of healthcare workers [2–4]. This chapter addresses pivotal considerations of the safe use of CT and MRI. The primary risks of CT scanning are associated with ionizing radiation and reactions to iodinated contrast media (ICM) [2,3,5] While MRI is often considered safer than CT because of the absence of ionizing radiation, MRI has raised its own set of safety issues. The use of gadolinium-based MR contrast agents (GBMCAs) has been linked with various types of adverse reactions, especially contrast-induced nephropathy in patients in advanced stages of renal disease [2]. In addition, MRI must be used with caution in patients with implanted devices [1,2].

DNA and generate free radicals. This is a safety concern because CT scanners usually deliver radiation doses that are often 100 times greater than those of conventional radiographic examinations, including chest X-rays or mammograms [3]. Scanner-based CT radiation carries a small but serious risk of causing cancer. Ionizing radiation can injure biologic material through several mechanisms, including formation of hydroxyl radicals that damage or break DNA doublestrands bases Concern over the risk of ionizing radiation for health problems, including malignancy, has reached a critical level in the current medical climate [5].

CT Scanning and Radiation Parameters The radiation dose for a specific study is determined by several scan parameters. These include the number of scans, tube current and scanning time in milliampseconds (mAs), size of the patient, axial scan range, scan pitch (the amount of overlap between adjacent CT slices), tube voltage in the kilovolt peaks (kVp), and the design features of the scanner used to deliver the dose [5].

Computed Tomography Safety of Ionizing Radiation

Estimations of CT Radiation-Related Cancer Risk

Background Computed tomography uses ionizing radiation: high-energy photons that are known to damage #


Estimates of radiation-induced cancer risk are based on epidemiologic follow-up studies of atomic-bomb survivors in Japan and other large

280

19

CT/MRI safety in functional neurosurgery

population investigations. One large-scale study examined 400,000 radiation workers in the nuclear industry receiving about 20 milliSieverts (mSv) (a representative organ dose from a single CT scan for an adult). These investigations yielded a significant association between radiation dose and cancer-related mortality in persons exposed to doses between 5 and 150 mSv [3]. The estimated lifetime risk of death from cancer resulting from a single CT scan of the head is determined by adding the estimated organ-specific cancer risks. Again, these risks are derived from estimations of organ-specific data for cancer incidence or mortality in atomic-bomb survivors. A CT study with 2 or 3 series produces a radiation dose in the range of 30–90 mSv. Findings from some investigations suggest that this dose range is associated with a statistically significant increase in the risk of cancer in adults. Approximately 0.4% of all cancers in the United States may be associated with CT scanner-based radiation. Current estimates of excess radiationrelated cancer rates range between 1.5 and 2.0%. The evidence for increased risk is most compelling for children. Compared to adults, they are more radiosensitive and have a longer remaining life span during which time a radiation-induced cancer could form. The lifetime cancer mortality risk associated with a single head CT protocol in a 1-year-old child was 0.07% [3]. The methodology for estimating the longrange cancer risk from CT radiation is in dispute over bias. Some investigators claim that the linear no-threshold model in this dose range may overestimate the risk. Excess cancer rates have not been reported in humans for doses below 100 mSv, One possible reason for this is that defense mechanisms that inhibit radiocarcinogenesis may be much more effective at low doses [6,7]. Yet, other evidence reveals that exposure to CT-related radiation exceeds low-level radiation doses, instead falling within the range of medium-level exposure. This is noteworthy

because increased cancer risk is related to midlevel radiation doses [5].

Strategies for Reducing Radiation Dose Strategies for radiation dose reduction include inplane bismuth shielding, minimizing multiphase scanning, and decreasing or eliminating multiple scans with contrast material. CT settings can be optimized by decreasing tube current (often via automatic tube currentmodulation(ATCM)), using a larger pitch, and limiting the range of coverage. The automatic exposure-control option on new scanners can be adjusted to decrease the radiation dose. However, there is almost always a tradeoff between lowering the level of radiation dose and producing the highest quality images. The cost of reducing radiation dose by, for example, decreasing gantry rotation time, is an increase in image noise [5,8]. Minimizing patient exposure to radiation remains a priority for healthcare workers in radiology. In some cases, magnetic resonance imaging (MRI) may be a preferred option to CT scans [9]. In the absence of updated CT protocols that reflect current scientific thought, neuroradiologists and neurosurgeons must collaborate to identify optimal techniques for radiation dose reduction during a CT diagnostic or interventional procedure. Since current radiation risk estimates remain ambiguous, CT scans should be performed in accordance with the ‘‘ALARA’’ principle: ‘‘As low as reasonably achievable’’ [5,9]. Nowhere is this more crucial than for the pediatric population. Guidelines established for CT imaging in children recommend adjusting scan parameters for smaller size in order to achieve lower-dose scanning for specific applications. Following CT guidelines protocols for using age-adjusted, relatively lower tube currents may help to reduce the radiation dose for pediatric CT of the brain [3,5].


Iodinated Contrast Media (ICM) Used in Enhanced CT Scans

19

Non-ionic ICM are the preferred agents in CT enhanced scans of the head [2].

Background Indications for the use of iodinated contrast media (ICM) in stereotactic neurosurgery mainly include targeting of mass lesions for stereotactic biopsy and radiosurgery [10]. While in much of the developed world contrast– enhanced MRI has supplanted CT for these uses, CT contrast may still be needed [11]. Patients with pacemakers in general cannot safely undergo MRI, and others who are obese or claustrophobic may not tolerate the small bore and prolonged acquisition times [4]. MRI may not always be available, and in much of the world CT is by far the more common imaging modality. It therefore behooves neurosurgeons to understand potential problems associated with the use of ICM.

Types of Radiographic Iodinated Contrast Media (ICM) Contrast materials consist of ionic (high osmolar), and organic non-ionic (low osmolar) water soluble agents. The higher osmolarity (600–2,100 mOsm/kg) in solution for ionic contrast agents accounts for some of their adverse effects. By contrast, nonionic agents have approximately half the osmolality of ionic substances, making them less likely to affect the blood-brain barrier. These materials exhibit fewer side effects because they do not ionize in solution. Yet, nonionic agents possess the same degree of radiopacity as ionic contrast materials. Both high and low osmolar iodinated contrast agents are used in current medicine, although nonionic ICM are more common [2]. In clinical practice, ICM are typically classified by osmolality. Low-osmolality ICM may be subcategorized further into (1) nonionic monomers, (2) ionic dimers, and (3) nonionic dimers.

Safety Studies of Nonionic Versus Ionic Iodinated Contrast Media Large population studies have demonstrated the relatively lower risk of nonionic ICM compared with ionic ICM. Comparative data from two older large-scale studies suggested that the incidence of mild adverse reactions to contrast media was 2.5% for ionic ICM, but only 0.58% for nonionic ICM. Severe reactions were reported in 0.4% of patients administered ionic ICM and 0% for severe reactions after administration of nonionic ICM [12,13]. Katayama et al. reported that in a series of 337,647 cases, the overall risk of an adverse drug reaction associated with ICM was 12.66% for ionic ICM and 3.13% for nonionic ICM. The risk of a very severe adverse drug reaction was 0.04% for ionic ICM and 0.004% for nonionic ICM [14]. In a meta-analysis of studies published during the 1980s, Caro, et al. documented risks of mortality and severe nonfatal reactions in high-osmolality ICM compared to nonionic ICM. These investigators calculated a rate of severe adverse drug reaction of 0.157% for high-osmolality ICM and 0.031% for nonionic ICM. The rate of a fatal adverse reaction was one death in 100,000 patients for both types of ICM [15].

Adverse Reactions to Iodinated Contrast Media Background Despite their poorer safety record, high-osmolality ICM are still used in current medicine, primarily because of their lower cost. These media should be used selectively. High-osmolality ICM have an increased risk for adverse contrast reactions, and a significantly higher risk for contrast-related severe

281

282

19


adverse events [2]. The overall safety of lowerosmolality nonionic ICM has been well established since the 1990s, but adverse events have been reported. Mild and moderate adverse reactions are generally uncommon. Severe and even fatal adverse effects are quite rare, but may occur unpredictably in some patients. Serious reactions may be preceded by a mild or moderate prodromal phase. A ‘‘test injection’’ administered before a contrast-enhanced CT may increase the risk for severe adverse events [2]. ICM may also impair kidney function in certain patients or exacerbate pre-existing renal insufficiency in persons with compromised kidney function. Although contrast-induced nephropathy is not an adverse allergic reaction, it is a serious adverse event that may have debilitating consequences for high-risk persons undergoing iodinated contrast enhanced CT [2].

Safety Issues: Magnetic Resonance Imaging MRI-Related Management of Metal Implants and Foreign Bodies Background Increasing use of technologically advanced MR systems during the past 20 years has introduced growing safety concerns over the MRI environment itself [16]. Compared to older machines, new MRI scanners have stronger static magnetic fields, faster and stronger gradient magnetic fields, and more powerful radiofrequency transmission coils. While there is no evidence that magnetic fields produce irreversible biologic effects, under certain conditions several features of high-field MRI equipment pose serious hazards for the body and for implanted metal devices [17]. Expanding clinical applications of deep brain stimulation

(DBS), in particular, require a new set of safety measures for performing MRI examinations in patients with implanted neurostimulation devices [16–18]. Metal implants warrant special consideration because they are typically located near, or contiguous with, brain structures or cerebral vasculature. As recent descriptions of several MR-related injuries and at least two fatalities illustrate, strict adherence to updated evidence-based safety guidelines on MRI technology is essential. Failure to follow the manufacturer’s guidelines when performing MRI on patients with a specific neuromodulation or other metal implant may have devastating consequences. In one reported case, the DBS electrode was heated during an MRI scan of the lumbar spine on a patient with Parkinson’s disease. The heating produced a radiofrequency lesion that led to permanent neurological damage [19]. This single case study further underscores the importance of literally complying with safety guidelines for performing MRI in persons with metallic implants. Patients may be subjected to severe injury if healthcare workers attempt to generalize about various conditions, positioning schemes, or other scanning scenarios stipulated for one neurostimulation system during MRI scanning, and then inadvertently apply these generalizations to the operation of other systems [16–18]. The primary hazards associated with MRI equipment in conjunction with implanted devices are categorized as follows [4,16,17,20].

Risks associated with the static magnetic field (Bo), including complications such as movement of ferromagnetic objects, twisting, heating, artifacts, and device malfunction produced by the static magnetic field. Risks associated with radiofrequency field (RF) effects, including complications arising from body coils and specific absorption rate (SAR).


Risks Associated with Static Magnetic Field Strength Projectile Effect The projectile effect, or the disturbing movement of ferromagnetic material, is a primary complication of metallic implants that may occur during an MRI. Also known as the missile effect, it is caused by interactions between the static magnetic field and MRI systems [21]. Magnetic translational and rotational forces that are exerted on a ferromagnetic object can move or dislodge the object from its implanted position. A magnet of high field strength can rapidly pull different types of ferromagnetic objects into the MRI scanner. The patient is subsequently at risk for injury by any number of objects, ranging from internal aneurysm clips and pins in joints to oxygen canisters and wheelchairs [22,23].

Heating The greatest risk for MRI scanning in a patient with a DBS implant is MRI-related heating of metallic objects, especially DBS leads. Heating is poorly tolerated in the central nervous system. When electrically conductive materials are introduced within the magnet and touch the bore of the MRI scanner, these materials may overheat. If a conductive object comes in contact with the patient’s tissue, it may burn his skin, possibly resulting in irreversible lesions [16,18,20]. In addition, conductive loops that come in contact with tattoos and eye-liners containing iron-oxides may cause burns [20]. A neurostimulation system used for DBS can generate variable levels of heating. Which levels are most likely to occur, and what factors are most likely to cause overheating depend upon the specific type of implanted device as well as various parameters used for a given MRI procedure [16,18,23].

19

Intrinsic factors that influence heating include the static magnetic field strength of the MRI system (which determines the transmitted RF used for the device operation); the electrical characteristics and configuration of the individual system (electrode, extension, length, orientation of the IPG); lengths and routing of the extension and leads; the impedance of the wires; and wire breakage [16]. Extrinsic parameters in the heating equation include the type of RF coil used (transmit/ receive body vs. transmit/receive head RF coils); the landmarking site; geometry of the RF coil and the quantity of the DBS lead present within this coil; SAR (amount of RF energy delivered); method for calculating the SAR based on a particular MR system; and quantity of RF energy (whole-body averaged SAR) required for imaging [16]. RF burns may occur if currents are induced into electrocardiographic leads, or into monitoring cables and coils that are placed on the patient’s skin surface [16]. The safety of MRI in patients with DBS may be increased by placing concentric loops of DBS electrode around the burr hole cap, by using a headonly receive coil, and by adhering to the vendor recommendations re the maximum SAR that can be tolerated [16]. For certain implants that have undergone empirical testing, clinically significant thermal changes may occur at 3.0-T but not at 1.5-T. Yet other data indicate that in some cases a particular implant may exhibit clinically significant levels of heating in seconds at 1.5-T but not at 3.0-T. The greatest risk, therefore, appears to be linked with the rate of temperature increase rather than the thermal change per se. According to one report, most heating occurred within the first minute of the MRI procedure and reached a steady-state within 15 min [16]. As noted previously, in order to mitigate risks of excessive heating, established product safety guidelines for MR scanning in patients with metallic implants must be diligently followed.

283

284

19


Even then, the guidelines should be used only insofar as they apply to the magnetic field strengths that have been evaluated and specified in the guidelines. MR scanning at either stronger and/or weaker magnetic field strengths than those indicated in the manufacturer’s guidelines may cause substantial heating. Unless extreme precautionary measures are taken, inadvertent heating may arise and ultimately produce severe injury in the patient [16–18].

Contrast Administration in MRI Scans: Safety Issues Indications for the use of gadolinium based magnetic contrast agents (GBCMAs) are similar to those noted above for CT scanning.

Allergic Reactions Allergic reactions have been linked to the use of GBMCAs in persons with impaired kidney function. Gadopentetate dimeglumine and gadoteridol have elicited adverse reactions such as anaphylaxis in patients with impaired renal function. The package insert for gadopentetate dimeglumine warns that a history of asthma or other allergic respiratory condition may increase the possibility of a reaction, including serious, fatal, life-threatening, anaphylactoid, cardiovascular reactions, or other idiosyncratic reactions [4].

Gadolinium-based ContrastInduced Nephrogenic Systemic Fibrosis (NSF) More disconcerting than the risk for allergic reactions is the mounting evidence for a linkage between GBMCAs administered to kidney disease patients and an emerging disease called

nephrogenic systemic fibrosis (NSF). Nephrogenic systemic fibrosis is a rare, progressive, and potentially fatal fibrosing disorder that affects patients with pre-existing renal dysfunction. It is closely associated with the use of GBMCAs [24]. The disorder was originally called nephrogenic fibrosing dermopathy (NFD) because of its involvement with the skin [18]. NSF is characterized as a systemic disease of connective tissue that targets skeletal muscle, skin, and tendons. The condition is definitively diagnosed by clinical evaluation and a deep skin biopsy of the dermis, subcutaneous fat, and fascia. The pathology involves increased deposits of collagen in connective tissues, resulting in a thickening and hardening of the skin of the extremities [24,25]. In severe illness, joints may become immobile or deformed. In extreme cases of limited motion, some patients may be confined to a wheelchair. NSF also may cause injury to the diaphragm, esophagus, heart, lung, pulmonary vasculature, and skeletal muscles [25]. The disease tends to develop slowly, but advances rapidly in about 5% of patients. At present, there is no consistently efficacious therapy [18]. The American College of Radiology (ACR) recommends that patients at risk for NSF from dialysis or chronic kidney disease be screened before receiving GBMCAs. Glomerular Filtration Rate (GFR) should be measured in patients older than 60 with a history of renal disease, hypertension, diabetes, and/or severe hepatic disease/liver transplant/pending liver transplant. Patients with hepatic dysfunction should undergo a GFR assessment as close as possible to the time at which the GBMCA is to be administered for the MRI examination [26]. GBMCAs should be avoided in patients with GFRs less than 30 mL/min/ 1.73m2 unless absolutely necessary. Persons unaware that they have kidney dysfunction may be identified through medical history. If a definitive diagnosis of kidney status is not known, immediate serum creatinine testing may be warranted in addition to a GFR assessment [26].


Conclusions The evidence presented in this chapter clearly reinforces current expert consensus that safety remains a paramount issue for patients undergoing either CT or MRI scanning. Radiographic safety measures must be properly implemented and followed at the level of the institution, neurosurgical team, and individual healthcare worker. It is incumbent upon clinicians to keep informed of the most recent information generated by the professional organizations that develop practice guidelines and issue advisories, including critical periodic updates.

References 1. Ross PJ, Ashamalla H, Rafla S. Advances in stereotactic radiosurgery and stereotactic radiation therapy. Radiation Therapist 2001;10(1):57-72. 2. Segal AJ, Ellis JH, Baumgartner BR. ACR manual on contrast media. 6th ed. Reston, VA: ACR; 2008. 3. Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med 2007;357 (22):2277-84. 4. Chung SM. Safety issues in magnetic resonance imaging. State of the art. J Neuro-Ophthalmol 2002;22(1):35-9. 5. Coursey CA, Frush DP. CT and radiation: what radiologists should know. Appl Radiol 2008;37(3):22-9. 6. Nagataki S. Comment on: computed tomography and radiation exposure. N Engl J Med 2007;357 (22):2277-84. N Engl J Med 2008;358(8):850‐1. 7. Tubiana M. Computed tomography and radiation exposure. N Engl J Med 2008;358(8):850; author reply 852–3. 8. Fricke BL, Donnelly LF, Frush DP, Yoshizumi T, Varchena V, Poe SA, Lucaya J. In-plane bismuth breast shields for pediatric CT: effects on radiation dose and image quality using experimental and clinical data. Am J Roentgenol 2003;180(2):407-11. 9. Semelka RC, Armao DM, Elias J, Jr, Huda W. Imaging strategies to reduce the risk of radiation in CT studies, including selective substitution with MRI. J Magn Reson Imaging 2007;25(5):900-9. 10. Bauman G, Wong E, McDermott M. Fractionated radiotherapy techniques. Neurosurg Clin N Am 2006;17 (2):99-110. 11. Valk J. The role of CT and NMRI in neurosurgical diagnosis. Neurosurg Rev 1986;9(1–2):43-7.

19

12. Wolf GL, Arenson RL, Cross AP. A prospective trial of ionic vs. nonionic contrast agents in routine clinical practice: comparison of adverse effect. Am J Roentgenol 1989;152:939-44. 13. Lasser EC, Berry CC, Talner LB, Santini LC, Lang EK, Gerber FH, Stolberg HO. Pretreatment with corticosteroids to alleviate reactions to intravenous contrast material. N Engl J Med 1987;317(14):845-9. 14. Katayama H, Yamaguchi K, Kozuka T, Takashima T, Seez P, Matsuura K. Adverse reactions to ionic and nonionic contrast media. A report from the Japanese committee on the safety of contrast media. Radiology 1990;175 (3):621-8. 15. Caro JJ, Trindade E, McGregor M. The risks of death and of severe nonfatal reactions with high- vs lowosmolality contrast media: a meta-analysis. Am J Roentgenol 1991;156(4):825-32. 16. Rezai AR, Baker K, Tkach J, Phillips M, Hrdlicka G, Sharan A, Nyenhuis J, Ruggieri P, Henderson J, Shellock FG. Is magnetic resonance imaging safe for patients with neurostimulation systems used for deep brain stimulation (DBS)? Neurosurgery 2005;57:1056-62. 17. Shellock FG, Crues JV. MR procedures: biologic effects, safety, and patient care. Radiology 2004;232(3):635-52. 18. Kanal E, Barkovich AJ, Bell C, Borgstede JP, Bradley WG, Jr, Froelich JW, Gilk T, Gimbel JR, Gosbee J, Kuhni-Kaminski E, Lester JW, Jr, Nyenhuis J, Parag Y, Schaefer DJ, Sebek-Scoumis EA, Weinreb J, Zaremba LA, Wilcox P, Lucey L, Sass N. ACR Blue Ribbon Panel on MR safety. ACR guidance document for safe MR practices: Am J Roentgenol 2007;188 (6):1447-74. 19. Henderson JM, Tkach J, Phillips M, Baker K, Shellock FG, Rezai AR. Permanent neurological deficit related to magnetic resonance imaging in a patient with implanted deep brain stimulation electrodes for Parkinson’s disease: case report. Neurosurgery 2005;57 (5):E1063. 20. Stecco A, Saponaro A, Carriero A. Patient safety issues in magnetic resonance imaging: state of the art [Article in English, Italian]. Radiol Med (Torino) 2007;112 (4):491-508. 21. Shellock F. Metallic neurosurgical implants: evaluation of magnetic field interactions, heating, and artifacts at 1.5-Tesla. J Magn Reson Imaging 2001;14 (3):295-9. 22. Joint Commission. Sentinel event MRI safety alert. Preventing accidents and injuries in the MRI suite. Issue 38, February 14, 2008. http://www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_38.htm Accessed 20 May 08. 23. Shellock FG, Crues JV. Commentary: MR safety and the American college of radiology white paper. Am J Roentgenol 2002;178:1349-52.

285

286

19


24. Chewning RH, Murphy KJ. Gadolinium-based contrast media and the development of nephrogenic systemic fibrosis in patients with renal insufficiency. J Vasc Interv Radiol 2007;18(3):331-3. 25. Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA. Gadodiamide-associated nephrogenic systemic

fibrosis: why radiologists should be concerned. AJR Am J Roentgenol 2007;188(2):586-92. 26. Weinreb JC. Improving gadolinium-based contrast safety. Imaging Biz.com 2008;3(2):1-2.

18 CT/MRI Technology: Basic Principles M. I. Hariz . L. Zrinzo

Introduction Ever since the beginning of human stereotactic neurosurgery in 1947 [1], the radiological study has been an absolute prerequisite for the very existence of this surgery. Furthermore, the radiological study has always constituted an integral part of the surgical procedure itself. It was indeed the limitations of conventional radiology (plain X-ray, pneumoencephalography, ventriculography, arteriography), which for a long time did put the limits for what could be achieved with stereotactic neurosurgery. In 1973, Hounsfield published the method of computerized transverse axial scanning [2], which rendered him a Nobel Price in 1979. This method, which came to be known as computed tomography (CT), has literally revolutionized all neurosurgery including stereotactic neurosurgery. To begin with, and due to the axial views provided by this new imaging technique, CT introduced a new coordinate language, which had to be adopted by stereotactic neurosurgeons [3]: the anteroposterior direction became Y instead of the previous X, the dorso-ventral became Z instead of Y, and the lateral direction became X instead of Z. The advent of MRI further stimulated developments in functional stereotaxis by providing much improved soft tissue contrast, and possibility to image the brain in any desired plan. MRI forced the development of new stereotactic frames, compatible with this new imaging modality. At the same time, concerns were raised about possible distortions in the images due to magnetic susceptibilities and magnetic field inhomogeneities. Lately, #


concern arose also about safety of MRI scanning in patients who are implanted with deep brain stimulation (DBS) devices [4,5]. A prerequisite for accurate CT/MRI-guided functional stereotactic neurosurgery is the adaptation of the scanning method and the stereotactic frame to each other. Additionally, for stereotactic MRI scanning, accounting for the eventual geometric distortions and the adaptation of the imaging sequence to the functional brain target aimed at, are of paramount importance. In this chapter the general principles for performing stereotactic CT/MRI scanning in the practice of functional neurosurgery are detailed. A brief description of various imaging protocols in relation to brain structure being targeted is presented.

General Principles Although routine CT/MRI examinations are performed for diagnostic purposes, the stereotactic CT/MRIstudyisalocalizationprocedureandrepresents a crucial step of the functional neurosurgical procedure. The localization procedure delineates and defines a deep-seated structure (nucleus, part of a nucleus, or pathway) in relation to a coordinate system such that the structure may be surgically targeted. This intracranial structure may be readily ‘‘visualized’’ on MRI, provided a suitable imaging sequence, as in the case of the subthalamic nucleus (STN) and the globus pallidus internus (GPi). The structure may also be a particular area of the brain, which has to be defined in relation to visualized ventricular landmarks as in the case of the

270

18

CT/MRI technology: basic principles

ventral intermediate nucleus (Vim) of the thalamus, which cannot be visualized as such on either the CT or the MRI scans. Hence, it is the geometrical accuracy in performing the scanning and the accuracy of definition of brain targets and/ or ventricular landmarks that are the hallmarks of a stereotactic CT/MRI study. Therefore, all possible sources of errors that may interfere with the requirements of accuracy of a stereotactic CT/ MRI study, should be accounted for, or preferably eliminated. The first issue in that respect is that of the geometry and image-compatibility of the stereotactic frame.

Stereotactic Frame The general requirement of any head-containing stereotactic system is to provide compatible, wellvisualized, artifact-free, fiducials or landmarks on every relevant CT/MRI slice of the stereotactic imaging study. Besides, the relationship of these external landmarks to the head must be constant during the whole scanning and also between imaging study and the subsequent surgery. The number of visible landmarks on each CT/MRI slice varies according to which frame system is used; however, a minimum of three is required to define a zero origin for the anteroposterior Y and lateral X coordinates on the slice. The accuracy of measuring the height Z coordinate is much increased if the stereotactic frame contains a fiducial permitting the assessment of the dorso-ventral position of the axial slice containing the brain target, independently from the accuracy of movement of the CT couch (if a CT study is performed, i.e.), and hence, without relying on the accuracy of the Scoutview. In case MRI is used, it is an advantage for minimizing the measurement error, and minimizing the distortion at the periphery of the MR image, where the fiducials of the frame are as close as possible to the head, that is, as close as possible to the intracranial structure, the coordinates of which are to be measured.

Immobilization of the Patient Unlike conventional radiography, a CT study, and even more so, an MRI study lasts for a certain amount of time. In routine CT/MRI scanning, the immobility of the head is important to avoid movement artifacts on the picture. In stereotactic CT/MRI scanning, the immobility of the head in relation to the frame, and to the gantry or head coil is absolutely mandatory because the scanning is performed usually with very thin slices and because at least one of the target coordinates, that is, the height coordinate Z, is sometimes based on the primary position of the head at the beginning of the scanning. To achieve a strict immobility of the head of the nonanesthetized patient during the stereotactic CT/MRI scanning, there is only one solution: it is to secure the head to a frame with skull screws and to fix rigidly the frame to the supporting couch. The discomfort for the patient should be kept to a minimum and the patient should be able to tolerate the frame. Nonrigid fixation using noninvasive interfaces [6–8] or fiducials [9] have also been used, but they assume comprehensive cooperation of the patient during the imaging study, unless general anesthesia is used. A careful explanation of the stereotactic imaging procedure, combined if needed with slight sedation, will contribute to motivate most patients to achieve an acceptable immobility of the head. Otherwise, and especially if the imaging study requires long acquisition time, the patient, who in most cases suffers from a movement disorder, must have general anesthesia during the scanning [10,11].

Scanning Plane and Alignment Conventional diagnostic axial CT images are most often acquired with a scanning plane parallel to Reid’s baseline, the orbito-metal line, or the skull base in general [12,13]. In stereotactic CT/MRI scanning, it is in many stereotactic


systems the geometry of the stereotactic frame that dictates the axial scanning plane. Since in many centers, most functional brain targets are still related to the atlas, and thus, to the anterior commissure-posterior commissure (AC–PC) line of the third ventricle, the stereotactic frame is mounted to the head in such a way that its base ring will be parallel to the average orientation of the AC–PC line. [14–16]. Some stereotactic systems that employ various N-shaped, Z-shaped, or V-shaped localizers do not require any specific alignment of the head and frame in relation to the CT gantry [17–23]. However, a dedicated software is then required to enable coordinate calculation. Frameless stereotactic techniques do not require any specific orientation of the head within the CT gantry or the MRI head coil. These methods rely completely on dedicated software that allows reformatting of the images according to AC–PC orientation and fusion between different imaging modalities.

Scan Thickness Even though an axial stereotactic CT/MRI scan containing the target does generally also provide the X and Y coordinates of that target in relation to any external reference fiducials, the dorsoventral Z coordinate is far from being easy to obtain accurately: on axial scans; the Z coordinate depends mainly on the thickness of the CT/ MRI slice. The volume of the voxel, that is, the impact of the partial volume effect on the boundaries of the brain structure, can be reduced by examining thinner slices. However, the spatial resolution of the image is known to decrease with thinner scans because thinner sections have a more unfavorable signal-to-noise ratio than thicker slices. This can be improved by increasing the degree of contrast on the image. The contrast resolution on the image can also be improved by longer scan time, which unfortunately prolongs the duration of the scanning and, in case of CT,

18

increases the radiation dose to the patient. Typically, for functional stereotactic purposes, it is recommended to perform the CT/MRI scanning with contiguous scans not thicker than 2 mm. Furthermore, should a sagittal or coronal reformatting of the axial CT image be desirable, the image resolution of such a reformatted image is greatly enhanced if the axial scanning has been performed with thin contiguous slices. However, it must be stressed that a reformatted or reconstructed CT image almost never reaches the degree of resolution of the axial one, and it does always carry an increased risk of measurement error. MRI readily allows for coronal and sagittal scans, although, in analogy with CT scanning, it is typically the axial scans that are used for target calculation, and most dedicated functional stereotactic softwares are based on evaluation of, and targeting on, axial scans. The distortion on the image is less on axial than on coronal and sagittal views. In all other aspects, the same rules apply for stereotactic MRI scanning (thickness of slices, parallelity with AC-PC plan, immobilization of head and frame, etc.) as for stereotactic CT scanning. The advantage of an MRI coronal scan is that it allows more readily to assess the depth of the target in relation to other visible structures, such as the relation of GPi to optic tract or the relation of STN to Substantia Nigra. Furthermore, MRI is exquisite, when used together with a dedicated software, to rehears the trajectory of the probe from cortex to target, and to allow a virtual ‘‘dry-run’’ through the structures traversed by the probe on its way to the target. Finally, on postoperative stereotactic images, MRI allows very precise evaluation of the location of DBS electrode contacts in relation to targeted structure and its surroundings, provided thin scans are obtained. It must be kept in mind that eventual geometrical distortions on MRI, which are due to the ‘‘potato ship’’ effect at the periphery of the image, are less—all other parameters equal—when the fiducials of the stereotactic frame are as close as possible to the head such as

271

272

18


is always the case when using the Laitinen system [24] and quite often the case—depending on the size of the head—with the Leksell system.

Measurements and Calculations For details concerning the various methods to calculate target coordinates proper to each stereotactic system, it is referred to the chapters of this book describing the respective systems. Generally speaking, there are three ways to perform measurements and calculations of the CT/ MRI coordinates of a brain target: One may use either manual measurements on enlarged hard copies, eventually with the help of matching mm-grids, or the inbuilt software of the scanner, or a specifically inbuilt or separately provided software dedicated to stereotactic measurements of a specific system. In some stereotactic systems, all three ways may be available in conjunction. However, it must be kept in mind that although a given CT/MRI machine is proven reliable as far as calibrations, accuracy, and software properties are concerned, and although a given stereotactic frame is proven mechanically accurate and CT/MRI-compatible with minimal image artifacts, the combination of the two during a stereotactic scanning on a ‘‘real’’ patient may result in measurement errors greater than what would be attributable to either of them separately. In 1992, Maciunas published an extensive application accuracy evaluation of the four most commonly used stereotactic frame systems [25]. The study was done on test phantoms scanned stereotactically with a modern CT machine with slice thicknesses from 8 mm down to 1 mm. Even with the 1-mm thick scanning, a mean target error of between 1 and 1.9 mm and a maximal error between 3.1 and 5.0 mm, dependent on the frame used, were obtained! The errors were considered greater than the mechanical accuracy of the frames and greater than the error attributable to the imaging procedure.

Clinical Applications in Functional Stereotaxis A substantial proportion of stereotactic CT/MRI studies is performed in view of a functional stereotactic procedure, that is, a stereotactic surgery for pain, movement disorder, or psychiatric disorder, where an ablative lesion or an implantation of a chronic electrode will be made in an anatomically ‘‘normal’’ structure of the brain. In these cases, the specific sub-nuclei to be targeted may not be readily seen on the CT/MRI image, and their position has to be determined in relation to visible internal reference structures, most commonly the AC-PC line.

CT Scanning for Functional Stereotaxis Computed tomography is both user- and patientfriendly. In CT-guided functional stereotaxis, the brain target cannot be visualized as such even by the most sophisticated CT scanner. Instead, the anatomical position of these structures is always defined in relation to ventricular landmarks, as has been the case during conventional ventriculography. Whereas ventriculography provides lateral and anteroposterior views of the third ventricle, CT provides an axial view. Therefore, the stereotactic CT scanning should be done not only with thin scans, but also the scanning plane should be as parallel as possible to the AC-PC line. One main difficulty has been that the ACPC line does not readily ‘‘show up’’ on the CT Scoutview. Both its exact position within the brain and its inclination in relation to any bony landmarks are subject to individual variations [26]. It has been suggested that the scanning plane which is the most ‘‘parallel’’ to the AC-PC line would be the Glabella-inion plane [27,28]. The Twining line, i.e., a line between the tuberculum sellae and the protuberantia occipitalis interna has also been advocated [29,30]. Using these bony landmarks,


an initial scanning of the area of the third ventricle is performed, and if it appears that the AC and PC are visualized on different slices, the gantry of the CT machine is re-angulated accordingly, and the scanning partly repeated. This technique may be time consuming and is only possible when gantryindependent CT-localizers are used. However, this technique has become rather obsolete with the advent of independent dedicated imaging softwares enabling reformatting of CT scans into axial images parallel to the AC-PC line. In other methods in which the parallelity of the scanning plane and AC-PC line can be averaged thanks to a special design of the used frame, a superimposition of the CT slice containing one of the commissures on the CT slice containing the other commissure may be sufficient to measure the length and define the level of the AC–PC line, and to assess its inclination in relation to the stereotactic frame [24,31]. This presupposes that the nonparallelity between the plane of scanning and the plane of the AC-PC line is not too exaggerated. One of the major pitfalls of stereotactic CT scanning for functional stereotactic procedures is the uncertainty in determining the height Z coordinate of the brain target. This uncertainty is not only due to the thickness of the CT slice, to eventual movement inaccuracies of the CT couch, to the geometry of the actual stereotactic frame, and/or to the nonparallelity between scanning plane and plane of AC-PC line, as has been discussed previously. The source of error in defining a functional brain target coordinate may also lie in the uncertainty in defining on CT the ventricular commissures in relation to which this very brain target is to be defined. Spiegel et al. had shown already in 1952 that the dorso-ventral thickness of the AC ranged from 1.5 to 4 mm with a mean of 3 mm [32]. Therefore, even if the CT scanning is performed with contiguous 1.5-mm thin slices, there may still be a risk of error in determining the dorso-ventral position of the AC. This risk of error is theoretically the same for the determination of the PC. Since the

18

height level of the brain target is defined in relation to the height level of the line joining the AC and PC, a repercussion of these errors on the target coordinates in the sagittal plane would ensue. Therefore some authors still prefer to perform ventriculography in addition to a stereotactic CT study (and even in addition to MRI) for the determination of a functional brain target [33–36]. Others do indeed rely solely on the stereotactic CT study and define consequently and always the AC on the CT slice lying 4 mm below the one depicting the ventral-most part of the foramen of Monro, and the PC on the slice immediately above the one showing the beginning of the aqueduct [6,24,37].

MRI Scanning for Functional Stereotaxis General Requirements Two aspects have to be considered for an MR image to be suitable as a source for direct determination of coordinates in functional neurosurgery: First, its ability to provide a good discrimination of the brain structures to be targeted, thus allowing the images to act as the patient’s individualized atlas; and second, the geometrical accuracy of the MR images has to be validated. General requirements to perform stereotactic imaging may include the following: (1) an MRI compatible stereotactic frame than can be attached to the head support of the head coil MRI; (2) a localizer with fiducials as close as possible to the head; (3) an MRI sequence adapted to the brain structure being targeted; (4) a field of view that fits the stereotactic frame; (5) a slice thickness allowing good compromise between resolution and signal-noise-ratio: with current technology, 1.5–2.0 mm slice thickness is a good compromise; (6) the pixel size in the axial plane (x, y) has to be isometric, i.e., based on a square matrix; (7) enough slices to cover with good margins the whole brain target

273

274

18


area, without gap between slices, and without overlapping slices; (8) the use of a transmitter– receiver head coil to ensure maximal signal, and to ensure safe postoperative imaging in patients with DBS implants; (9) regular checks of the gradient fields to minimize distortions; (10) avoiding patient movements during scanning. With few exceptions [34–36], most functional stereotactic surgeons today do use a stereotactic MRI study to determine target coordinates; however, it seems that many workers still relate the target to its atlas-defined position in relation to visualized ventricular landmarks on the MRI study. This approach ignores the main advantage of MRI: unlike ventriculography and CTscanning, MRI is not a homogeneous imaging method. Depending on the parameters of imaging, MRI studies can visualize differently and unequally well, various structures in the brain.

Thalamic Subnuclei The only commonly used brain targets in functional stereotactic neurosurgery, which still cannot be visualized as such on stereotactic thin slice MRI, are the subnulei of the thalamus. Stereotactic MRI cannot convincingly visualize several thalamic targets used in the treatment of a range of functional brain disorders: the ventroinetermediate (Vim) nucleus, targeted for treatment of tremor, the ventral oral posterior (Vop) and ventral oral anterior (Voa) nuclei, targeted, although rarely, in the treatment of dystonia and Parkinson’s, the ventroposterolateral (Vpl) and ventroposteromedial (Vpm) targeted for treatment of chronic pain, the centre median, -parafascicular (CMpf) nuclei, and ventral oral internus (Voi) targeted for the treatment of Tourette disorder. All these thalamic targets still have to be defined based on atlas coordinates in relation to AC– PC. Nonetheless, MRI, especially a T2-weighted MRI sequence can delineate the thalamo-capsular border, making it easier to determine at least the laterality of the Vim/Vop/Voa targets.

Posteroventral Pallidum The subdivisions of the globus pallidus (Globus pallidus pars interna, GPi, laminae medullaris interna and externa, Globus pallidus pars externa, GPe) and their surrounding structures (putamen, internal capsule, optic tract) can be visualized stereotactically on thin slice axial and coronal MRI using various sequences. One such sequence [38] is a nonvolumetric proton density sequence (TR/ TE 4000/15, echo-train 7, field of view 250 mm, slice thickness 2 mm, gap 0, matrix 210 256, excitations 3, imaging time 6 min and 5 s), which depicts exquisitely the details of the pallidal area (> Figure 18-1). Another sequence that has been validated by Vayssie`re et al. in Montpellier use volumetric T1 sequences [39,40]. Other MRI scanning methods, based on inversion recovery sequences, have been described by Starr et al. [41]. In all these cases, since the target itself is readily visualized, there is no need to refer to the landmarks of the third ventricle and to an atlas to obtain the location of the target in the individual patient, especially since it has been shown that the individual target location may vary substantially between patients, and also between the two hemispheres in the same patient [38,41,42].

Subthalamic Nucleus (STN) In surgery on STN, most workers who rely on MRI for targeting this structure determine its position on T1-weighted images, in relation to third ventricle landmarks and brain atlases. The few publications reporting on the use of stereotactic MRI for direct visualization and targeting of STN describe volumetric T2 weighted sequences with a rather long acquisition time [10,11,43–45], sometimes requiring reformatting of images and/or additional T1-weighted sequences that are used for targeting, and often necessitating general anesthesia during imaging. These sequences do allow exquisite visualization


18

. Figure 18-1 Preoperative, 2-mm thin, axial proton density stereotactic MRI scans, in two patients: (a) patient with atrophic pallidii and enlarged putaminae and (b) patient with large pallidii

of STN. The present authors, together with others [46], implemented a nonvolumetric T2-weighted MRI sequence (TR 3000–4000, TE 80–100) allowing individual visualization of the STN with fast acquisition sequences that allow imaging without general anesthesia in most patients (between 3 min 5 s and 7 min 48 s, depending on the MRI machine) (> Figure 18-2). Here also, direct visualization will make it possible to target the center of the visualized STN at surgery, without need for an indirect localization based on the atlas and the AC–PC landmarks. In all these imaging procedures, the present authors have used the Laitinen stereotactic apparatus [24,46] or the Leksell stereotactic system [47], together with MRI machines of various makings (Siemens, Philips, General Electric), regularly calibrated and assessed for field inhomogeneities and other sources of distortion.

Postoperative Stereotactic Imaging In functional stereotactic neurosurgery, stereotactic postoperative MRI is mandatory: It will unequivocally demonstrate the exact location of the stereotactic lesion or implanted DBS electrode.

. Figure 18-2 Preoperative, 2-mm thin, axial T2 weighted stereotactic MRI scan, at a 4 mm level below the AC–PC line, showing the subthalamic nuclei

This imaging should ideally be obtained with the same sequence parameters as the preoperative one. Concern about the risks of performing an MRI study on patients with implanted DBS systems has been raised [5,48]. However, a number of safety measures can be adopted to address these concerns such as use of a transmitter–receiver head coil, and ensuring the average specific absorption rate (SAR) is no more than 0.4 W/kg [4]. Recently the group of Philip Starr in San Francisco reviewed their experience over 7 years: 405 patients with 746 implanted DBS systems were imaged using a

275

276

18


variety of scanning techniques and on various 1.5-Tesla MRI machines, for a total of 1,071 imaging studies. They reported no adverse event, even with a SAR of up to 3 W/kg in few patients [49]. In patients with DBS, it is an advantage if postoperative imaging is performed immediately after surgery, while the frame is still on the head. If the neuro-pacemaker has already been implanted, MR images can still be obtained provided the voltage of the pacemaker is set to zero, and the output is switched off, before the patient enters the MRI room. One should be able to assess the location of the lead within the visible target and if necessary, return the patient to the operating room to relocate the lead should it become apparent that it is misplaced. The present authors consider that a DBS implant procedure is not complete until electrode localization in the intended target is verified by means of a stereotactic MRI (> Figures 18-3– > 18-4), or a stereotactic CT with image fusion to the preoperative stereotactic images.

Conclusions Notwithstanding the imaging method used, the stereotactic CT or MRI study constitutes an integral part of the stereotactic surgical procedure. It is sometimes the most difficult part of the surgery. Taking into consideration its limits and potential errors, and the fact that a measurement or calculation error at this stage may have harmful repercussions on the results of surgery and on the patient, responsibility for the stereotactic CT/MRI study lies with the neurosurgeon. Therefore, the surgeon must be well acquainted with the scanning technique, its potential pitfalls, and with target coordinate calculation, as well as being acquainted with the stereotactic frame being used. It might be wise to keep in mind the following statement made in 1985 by Lars Leksell in a paper entitled ‘‘Stereotaxis and nuclear magnetic resonance’’ [50]: ‘‘In clinical practice brain imaging can now be divided in two parts: the diagnostic

. Figure 18-3 Postoperative, 2-mm thin, axial proton density stereotactic MRI scan, at the level of AC–PC, showing artifacts of DBS electrodes in the posteroventral pallidum

. Figure 18-4 Postoperative, 2-mm thin, axial T2 weighted stereotactic MRI scan, at a 4 mm level below the AC–PC line, showing artifacts of DBS electrodes in the subthalamic nuclei

neuroradiology and the preoperative stereotactic localization procedure. The latter is part of the therapeutic procedure. It is the surgeon’s responsibility and should be closely integrated with the operation.’’


References 1. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 2. Hounsfield GN.Computerized transverse axial scanning (tomography): Part I. Description of system. Br J Radiol 1973;46:1016-22. 3. Goerss S, Kelly PJ, Kall B, Alker GJ. A computed tomographic stereotactic adaptation system. Neurosurgery 1982;10:375-9. 4. Rezai AR, Phillips M, Baker KB, Sharan AD, Nyenhuis J, Tkach J, et al. Neurostimulation system used for deep brain stimulation (DBS): MR safety issues and implications of failing to follow safety recommendations. Invest Radiol 2004;39:300-3. 5. Henderson JM, Tkach J, Phillips M, Baker K, Shellock FG, Rezai AR. Permanent neurological deficit related to magnetic resonance imaging in a patient with implanted deep brain stimulation electrodes for Parkinson’s disease: case report. Neurosurgery 2005;57:E1063. 6. Laitinen LV, Liliequist B, Fagerlund M, Eriksson AT. An adapter for computed tomography-guided stereotaxis. Surg Neurol 1985;23:559-66. 7. Hitchcock E. Stereotactic-computerized tomography interface device. Appl Neurophysiol 1987;50:63-7. 8. Bergstro¨m M, Greitz T. Stereotaxic computed tomography. A J Roentgenol 1976;127:167-70. 9. Bjartmarz H, Rehncrona S. Comparison of accuracy and precision between frame-based and frameless stereotactic navigation for deep brain stimulation electrode implantation. Stereotact Funct Neurosurg. 2007;85:235-42. 10. Bejjani BP, Dormont D, Pidoux B, Yelnik J, Damier P, Arnulf I, Bonnet AM, et al. Bilateral subthalamic stimulation for Parkinson’s disease by using three-dimensional stereotactic magnetic resonance imaging and electrophysiological guidance. J Neurosurg 2000;92:615-25. 11. Patel NK, Heywood P, O’Sullivan K, Love S, Gill SS. MRI-directed subthalamic nucleus surgery for Parkinson’s disease. Stereotact Funct Neurosurg 2002;78:132-45. 12. Lang J, Schlehahn F, Jensen HP, Lemke J, Klinge H, Muhtaroglu U. Cranio-cerebral Topography as a Basis for Interpreting Computed Topograms. In Lanksch W, Kazner E, editors. Cranial computerized tomography. Berlin: Springer, 1976. p. 24-36. 13. Lange S, Golde G. Resolution characteristics of computerized tomography and their impact on quantitative brain diagnosis. In: Lanksch W, Kazner E, editors. Cranial computerized tomography. Berlin: Springer; 1976. p. 52-5. 14. Laitinen LV. The Laitinen system. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Nijhoff M; 1988. p. 99-116. 15. Lunsford LD, Leksell D. The Leksell system. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Nijhoff M; 1988. p. 27-46

18

16. Mundinger F, Birg W. The imaging-compatible riechertmundinger system. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Nijhoff M; 1988. p. 13-25. 17. Goerss S, Kelly PJ, Kall B, Alker GJ. A computed tomographic stereotactic adaptation system. Neurosurgery. 1982;10:375-9. 18. Kelly PJ. Contempory stereotactic ventralis lateral thalamotomy in the treatment of parkinsonian tremor and other movement disorders. In Heilbrun MP, editor. Stereotactic neurosurgery, Vol 2: Concepts in neurosurgery. Baltimore: Williams and Wilkins; 1988. p. 133-47. 19. Heilbrun MP. Computed tomography-guided stereotactic systems. Clin Neurosurg. 1982;31:564-81. 20. Kelly PJ, Goerss SJ, Kall BA. Evolution of contemporary instrumentation for computer-assisted stereotactic surgery. Surg Neurol. 1988;30:204-15. 21. Rosenfeld JV, Barnett GH, Palmer J. Computed tomography guided stereotactic thalamotomy using the BrownRoberts-Wells system for non-Parkinsonian movement disorders. Technical note. Stereotact Funct Neurosurg. 1991;56:184-92. 22. Brown RA. A computerized tomography-computer graphics approach to stereotaxic localization. J Neurosurg. 1979;50:715-20. 23. Couldwell WT, Apuzzo MLJ. Initial experience related to the use of the Cosman-Roberts-Wells stereotactic instrument. Technical note. J Neurosurg. 1990;72:145-8. 24. Hariz MI, Laitinen LV. The Laitinen apparatus. In: Gildenberg PL and Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: Mc Graw Hill; 1997. p. 87-94. 25. Maciunas RJ, Galloway RL Jr, Latimer J, Cobb C, Zaccharias E, Moore A, et al. An independant application accuracy evaluation of stereotactic frame systems. Stereotact Funct Neurosurg. 1992;58:103-7. 26. Talairach J, David M, Tournoux P, Corredor H, Kvasina T. Atlas d? Anatomie Ste´reótaxique. Paris, Masson, 1957. 27. Tokunaga A, Takase M, Otani K. The glabella-inion line as a baseline for CT scanning of the brain. Neuroradiology. 1977;14:67-71. 28. Takase M, Tokunaga A, Otani K, Hori T. Atlas of the human brain for computed tomography based on the glabella-inion line. Neuroradiology. 1977;14:73-9. 29. Ohye C, Kawashima Y, Hirato M, Wada H, Nakajima H. Stereotactic CT scan applied to stereotactic thalamotomy and biopsy. Acta Neurochir. 1984;71:55-68. 30. Spiegelmann R, Friedman WA. Rapid determination of Thalamic CT-stereotactic coordinates: A method. Acta Neurochir (Wien). 1991;110:77-81. 31. Gouda KI, Freidberg SR, Larsen CR, Baker RA, Silverman ML. Modification of the Gouda frame to allow stereotactic biopsy of the brain using the GE 8800 computed tomographic scanner. Neurosurgery. 1983;13:176-81.

277

278

18


32. Spiegel EA, Wycis HT, Baird HW. Studies in Stereoencephalotomy. I. Topical relationships of subcortical structures to the posterior commissure. Confin Neurol. 1952;12:121-33. 33. Fox MW, Ahlskog JE, Kelly PJ. Stereotactic ventrolateralis thalamotomy for medically refractory tremor in postlevodopa era Parkinson’s disease patients. J Neurosurg. 1991;75:723-30. 34. Pinto S, Le Bas JF, Castana L, Krack P, Pollak P, Benabid AL.Comparison of two techniques to postoperatively localize the electrode contacts used for subthalamic nucleus stimulation. Neurosurgery. 2007; Suppl 2:285-92. 35. Breit S, LeBas JF, Koudsie A, Schulz J, Benazzouz A, Pollak P, et al. Pretargeting for the implantation of stimulation electrodes into the subthalamic nucleus: a comparative study of magnetic resonance imaging and ventriculography. Neurosurgery. 2006;58 Suppl 1: ONS83-95. 36. Merello M, Cammarota A, Cerquetti D, Leiguarda RC. Mismatch between electrophysiologically defined and ventriculography based theoretical targets for posteroventral pallidotomy in Parkinson’s Disease. J Neurol Neurosurg Psychiatry. 2000;69:787-91. 37. Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computed tomography-guided determinations of brain targets in functional stereotaxis. J Neurosurg. 1990;73:565-71. 38. Hirabayashi H, Tengvar M, Hariz MI: Imaging of the pallidal target. Mov Disord. 2002;17 Suppl 3: S162-6. 39. Vayssiere N, Hemm S, Zanca M, Picot MC, Bonafe A, Cif L, et al. Magnetic resonance imaging stereotactic target localization for deep brain stimulation in dystonic children. J Neurosurg. 2000;93:784-90. 40. Vayssiere N, Hemm S, Cif L, Picot MC, Diakonova N, El Fertit H, et al. Comparison of atlas- and magnetic resonance imaging–based stereotactic targeting of the globus pallidus internus in the performance of deep brain stimulation for treatment of dystonia. J Neurosurg. 2002;96:673-9.

41. Starr PA, Vitek JL, DeLong M, Bakay RAE. Magnetic resonance imaging-based stereotactic localization of the globus pallidus and subthalamic nucleus. Neurosurgery. 1999;44:303-14. 42. Vayssiere N, Gaag van der N, Cif L, Hemm S, Verdier R, Frerebeau P, et al. Deep brain stimulation for dystonia confirming a somatotopic organization in the globus pallidus internus. J Neurosurg. 2004;101:181-8. 43. Zhu XL, Hamel W, Schrader B, Weinert D, Hedderich J, Herzog J, et al. Magnetic Resonance Imaging-Based Morphometry and Landmark Correlation of Basal Ganglia Nuclei. Acta Neurochir. 2002;144 959-69. 44. Schrader B, Hamel W, Weinert D, Mehdorn HM. Documentation of electrode localization. Mov Disord. 2002;17 Suppl 3:S167-74. 45. Starr PA, Christine CW, Theodosopoulos PV, Lindsey N, Byrd D, Mosley A, et al. Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imaging–verified lead locations. J Neurosurg. 2002;97:370-87. 46. Hariz MI, Krack P, Melvill R, Jorgensen JV, Hamel W, Hirabayashi H, et al. A quick, and universal method for stereotactic visualization of the subthalamic nucleus before and after implantation of deep brain stimulation electrodes. Stereotact Funct Neurosurg. 2003;80:96-101. 47. Chen CC, Pogosyan A, Zrinzo LU, Tisch S, Limousin P, Ashkan K, et al. Intra-operative recordings of local field potentials can help localize the subthalamic nucleus in Parkinson’s disease surgery. Exp Neurol. 2006;198:214-21. 48. Baker KB, Tkach JA, Phillips MD, Rezai AR. Variability in RF-induced heating of a deep brain stimulation implant across MR systems. J Magn Reson Imaging. 2006;24:1236-42. 49. Larson PS, Richardson RM, Starr PA, Martin AJ. Magnetic resonance imaging of implanted deep brain stimulators: experience in a large series. Stereotact Funct Neurosurg. 2008;86:92-100. 50. Leksell L, Leksell D, Schwebel J. Stereotaxis and nuclear magnetic resonance. J Neurol Neurosurg Psychiatry. 1985;48:14-18.

22 Diagnostic PET in Image Guided Neurosurgery B. Ballanger . T. van Eimeren . A. P. Strafella

Positron emission tomography (PET) is one of the most popular imaging techniques in current neuroscience research. Due to ongoing innovations, PET continues to flourish as an interesting diagnostic tool in clinical neurology. While some imaging techniques, such as computed tomography (CT) and magnetic resonance imaging (MRI), are able to identify structural changes in the body, PET imaging is capable of detecting areas of functional changes even prior anatomical abnormalities are observed. The PET scanner does this via the use of radiolabeled molecular probes that have different rates of uptake, depending on the type and function of tissue involved. The changing of regional blood flow in various anatomical structures (as a measure of the injected positron emitter) can be visualized and relatively quantified with a PET scan. The first section of this chapter will trace the origin of PET and describe the instruments and radiopharmaceuticals. The second section will highlight studies which have application to stereotactic and/or functional neurosurgery. Whenever applicable, the current role of functional magnetic resonance imaging (fMRI) in preoperative assessment of discrete brain functions will be covered as well.

Background What is PET? PET is a nuclear medical imaging technique first developed by Michel Ter-Pogossian, Michael E. Phelps and others (1975) at the Washington #


University School of Medicine [1,2], which produces a three-dimensional image of functional processes in the body with excellent sensitivity and moderate anatomic resolution (4–7 mm). PET may be used to investigate subjects in the resting state or in relationship to an event such as the occurrence of a seizure or the induction of pain. PET enables to study cerebral energy metabolism and receptor function. PET provides in vivo measurements of injected biologically active substances that have been radioactively labeled (radioligands or tracers). These radionuclides are incorporated into compounds normally used by the body such as glucose or water and then injected into the body to trace where they become distributed. Such labeled compounds are known as radiotracers. To conduct the scan, a short-lived radioactive tracer isotope is injected into the body. The radioligand decays by emitting a positron. Then the positron encounters and annihilates with an electron, producing two 511 keV gamma photons radiating at 180 from each other. PET scanners have detectors placed on opposite sides of the region from where the photons are emitted (within the patient), and the detectors register an event only if both detectors record the photon emission at the same time. As the photons are always emitted 180 from each other, this serves to both localize and quantify these events, and hence register the amount of metabolic activity. The technique depends on simultaneous or coincident detection of the pair of photons; photons which do not arrive in pairs (i.e., within a few nanoseconds) are ignored. This provides PET with a unique ability to detect and quantify

308

22

Diagnostic PET in image guided neurosurgery

physiologic and receptor processes in the body, especially in the cancer cells, that is not possible by any other imaging technique. The data collected by the PET scanner is mathematically reconstructed to produce tomographic images of tissue radioactivity concentration. Glucose is the main substrate for energy supply of the brain by oxidation. In fact, brain cells use glucose as fuel, and the more active the brain cells are, the more they will consume radioactive glucose. Using an imaging tracer that is like glucose, the PET scan is able to quantify the activity of brain tissue. Areas of less activity will use less energy, and areas with increased activity will use more energy.

Radiopharmaceuticals The key to molecular imaging in nuclear medicine is radiotracers. Radiotracers allow in vivo evaluation of different functions in the brain, namely cerebral blood flow, glucose metabolism, protein synthesis, neurotransmission, and neuroreceptor density. The application mainly depends on their chemical properties.

The standard tracer for measurement of the cerebral metabolic rate of glucose is 18F-2-fluoro-2deoxy-D-glucose (18F-FDG) [6], a sugar, for which the acquisition period is typically an hour. FDG is transported into tissue and phosphorylated to FDG phosphate, like glucose, but does not undergo significant further metabolism. Thus, it accumulates in brain in proportion to local cerebral metabolic rate of glucose. Although FDG-PET is a very powerful diagnostic tool because it gives a comprehensive image of synaptic function, it lacks specificity with regard to individual transmitter systems. There are neurodegenerative diseases involving specific neurotransmitters that do not have a distinct appearance on FDG-PET scans, probably because the cells synthesizing and releasing these transmitters are too few or too dispersed to have a local impact on energy consumption. The most evident example is Parkinson’s disease (PD), where the substantia nigra pars compacta with its profound degeneration of dopaminergic neurons is too small and metabolically too similar to the rest of the midbrain to be easily recognized in FDG-PET scans.

Specific Neurotransmitter System Unspecific Brain Activity Water

H2O15 radioactive water is administrated by bolus intravenous injection to obtain blood flow images, with a data acquisition time of about 2 min. Injections and data acquisition can be repeated at intervals equals or superior to five times the radioactive half-live of 15O (10 min). This technique is well adapted for comparing a ‘‘resting’’ condition with a condition of sensory, motor or cognitive activation. In activated brain regions, the increase in regional cerebral blood flow (rCBF) leads to a local increase in the tissue radioactive water content detected [3–5]. Glucose

In these instances, we need tracers that image specifically that particular neurotransmitter system. Several radiotracers have been developed for PET that are ligands for specific neuroreceptor subtypes (e.g., dopamine D2, serotonin 5-HT1A, etc.), transporters (e.g., [11C]McN5652, [11C] DASB, [11C]MP), or enzyme substrates (e.g., 6-FDopa). The development of PET radiolabeled receptor ligands for brain imaging holds great promise for improved specificity and sensitivity in cerebral functional imaging. Dopaminergic System

To investigate the function and integrity of presynaptic dopaminergic terminals, the most widely used tracer is the 18F-Fluorodopa (FDopa). This tracer is a substrate to DOPA decarboxylase


which is expressed in abundance by dopaminergic neurons. The product, 18F-fluorodopamine, accumulates in proportion to decarboxylase activity which in turn reflects the amount of viable dopaminergic cells. Therefore, the amount of radioactivity resulting from 18F-FDopa in a region-of-interest (ROI) will reflect the number of functionally intact dopaminergic neurons within this particular region, as well as presynaptic dopamine uptake, decarboxylation to dopamine and storage [7]. The adequate modeling and interpretation of data from FDopa PET studies are far from simple and different techniques have been developed, e.g., the calculation of the influx constant Ki according to Patlak for the quantification of FDopa accumulation in the striatum [8,9]. Postsynaptically, dopamine exerts actions through several subtypes of the dopamine receptor. The dopamine receptor family consists of 5 subtypes D1-D5. In order to investigate the role of each receptor subtype, selective and high-affinity PET radioligands are required. For the dopamine D1-subtype the most commonly used ligand is [11C]-Schering 23390 ([11C]SCH) or [11C]-NNC 112, whereas for the D2/D3subtype [11C]-raclopride is a common tracer [10,11]. [18F]-fallypride is a suitable PET tracer for the investigation of extrastriatal D2 receptors [12]. For the other subtypes no suitable radioligands have been developed yet. Other tracers like [11C]-methylphenidate ([11C]MP), [11C]-dihydrotetrabenazine ([11C] DTBZ), provide complementary information on the integrity of the dopaminergic system. Methylphenidate inhibits dopamine reuptake and enhances synaptic dopamine levels. One of its isomers, dl-threo-methylphenidate has been labeled with Carbon-11 for PET [13]. Its binding in the human brain is reversible, highly reproducible and saturable and thus [11C]MP is deemed an appropriate PET ligand to measure dopamine transporter (DAT) availability [14]. The dopamine transporter (DAT) is

22

located on the plasma membrane of nerve terminals in a small number of neurons in the brain especially in the striatum and nucleus accumbens, but also in the globus pallidus, cingulate cortex, olfactory tubercle, amygdala and midbrain [15]. DAT regulates the dopamine concentration in the synaptic cleft through reuptake of dopamine into presynaptic neurons; it plays a central role in the spatial and temporal buffering of the released dopamine. Accordingly, DAT provides a good site for monitoring the function and the integrity of the dopaminergic neurons. Several DAT agents have been developed for diagnosis PD and monitoring the treatment of PD patients, based on DAT antagonists such as [11C]MP [14,16], [11C]cocaine [17,18] and 18 F-2-b-carbomethoxy-3-b-(4-fluorophenyl)tropane, ([18F]CFT) [19]. [11C]DTBZ is utilized as a tracer for in vivo imaging of the vesicular monoamine transporter (VMAT2) system [20,21]. Serotoninergic System

5-Hydroxy-tryptamine (5-HT) or serotonin is a monoamine transmitter produced in brainstem raphe nuclei and released at cortical level through widely distributed ascending pathway. Serotonin function seems to be altered in many neurologic and psychiatric disorders, in particular in depression, obsessive compulsive disorders, Alzheimer’s disease (AD) and schizophrenia. Currently at least seven major serotonin receptor classes have been identified, some of them consisting of different subtypes [22]. However, the available radioligands permit the investigation of 5-HT1A and 5-HT2A receptors only. Recently, two antagonists ligands of 5-HT1A receptors have been developed for PET studies. The first one is 18F-trans-4-fluoro-N-2-[4-(2methoxyphenyl)-1-piperazinyl]ethyl]-N-(2pirydyl)cyclohexanecarboxamide known as [18F] FCWAY which presents a much higher affinity than endogenous serotonin for 5-HT1A receptors, comparable to that of the original WAY-100635 labeled with 11C [23]. The second one is the

309

310

22


4-(20 -methoxyphenyl)-1-(20 -(N-20 -pirydynyl)p-18F-fluoro-benzamido)ethylpiperazine known as [18F]MPPF which has an affinity for the 5HT1A receptor close to that of endogenous serotonin [24]. Using these two ligands, a high level of tracer uptake has been observed in high density in the hippocampus, amygdale, parahippocampal gyrus, hypothalamus, temporal pole, insula, anterior and posterior cingulated gyri [25]. Conversely, 5-HT2A receptors are present in all neocortical regions, with lower densities in hippocampus, basal ganglia and thalamus. The cerebellum and striatum are virtually devoid of 5-HT2A receptors. A quantization of 5-HT2A receptors has been possible with 18F-altanserin and 18F-setoperone [26,27]. These ligands are selective 5-HT2A antagonists and they show good reproducibility. The ligand [11C]MDL 100907 has previously been introduced to image the 5-HT2A receptor in human brain [28]. a-11C-methyl-L-tryptophan ([11C]AMT) is used as a PET marker of brain serotonin synthesis [29]. The a-methyl-L-tryptophan is converted to a-methylserotonin, which is not a substrate for monoamine oxidase and therefore accumulates in the brain. Cholinergic System

Acetylcholine (ACh) is synthesized in certain neurons by the enzyme choline acetyltransferase from the compounds choline and acetyl-CoA. In the synaptic cleft, the enzyme acetylcholinesterase (AChE) converts acetylcholine into the inactive metabolites choline and acetate. Nicotinic ACh receptors have been implicated in many psychiatric and neurologic diseases, including depression and cognitive and memory disorders, such as Alzheimer’s and Parkinson’s disease. Thus 11C-labeled nicotine was used to visualize and quantify nicotinic receptors in the brain. In recent years, the piperidine analogs C-11labeled N-methyl-4-piperidyl-acetate (MP4A) [30] and N-methyl-4-piperidyl-propionate [31] have been developed for in vivo imaging of cere-

bral AChE with PET [32,33]. As a substrate of AChE these tracers are hydrolyzed and accumulate depending on enzyme activity. AChE in human cortex is mainly expressed in the cholinergic axons, and to a lesser extent also in some cholinoceptive neurons. With impaired function and neurodegeneration of these cholinergic axons, the amount of cortical AChE is reduced. GABAergic and Glutamatergic System

GABA (Gamma Aminobutyric acid) is the most important inhibitory neurotransmitter. Its transmission is altered in epilepsy and other psychiatric disorders. Because the GABA receptor is abundant in the cortex and is very sensitive to damage, it represents a reliable marker of neuronal integrity. The tracer most widely used for central GABA binding sites is flumazenil (FMZ) labeled with [11C] [34]. The highest degree of binding is observed in the medial occipital cortex, followed by other cortical areas, the cerebellum, thalamus, striatum and pons with very low binding in the white matter. FMZ is a biochemical marker of epileptogenicity and neuronal loss; benzodiazepine receptor-density changes are more sensitive than 18F-FDG in detecting hippocampal sclerosis and benzodiazepine receptor studies were useful in the selection of patients for targeted surgery and for predicting outcome of these procedures [35,36]. As an antagonist of GABA, glutamate is the main excitatory neurotransmitter in the cortex, and alterations of glutamatergic neurotransmission are associated with many neurologic diseases. In the post-synaptic cell, glutamate receptors, such as the NMDA (N-methyl-D-aspartate) receptor, bind glutamate and are activated. Because of its role in synaptic plasticity, it is believed that glutamic acid is involved in cognitive functions like learning and memory in the brain. The tracers used to study this system ([11C]MK 801, [11C]-ketamine, [18F]-fluoroethyl-TCP and [18F]-memantine) have only


a low specificity to the NMDA receptor, and relevance for clinical studies has not been established [37].

Limits of PET Radionuclides used in PET scanning are typically isotopes with short half lives such as 11C (20 min), 13N (10 min), 15O (2 min), and 18 F (110 min). Due to their short half lives, the use of PET is limited by the need for an on-site cyclotron. Therefore, PET differs from other imaging techniques in requiring more expensive equipment and highly specialized personnel, not only for scanning but also for production of the radiotracers. Unlike CT or MRI, few hospitals and universities are capable of maintaining such imaging systems.

Functional MRI fMRI is the use of MRI to measure the hemodynamic response related to neural activity in the brain. It is one of the most recently developed forms of neuroimaging. It can noninvasively record brain signals without risks of radiation inherent in other scanning methods, such as PET scans. It can record on a spatial resolution in the region of 3–6 mm with a relatively good temporal resolution (in the order of seconds) compared with techniques such as PET. Since neurosurgery relies on a precise delineation of the structural and functional aspects of brain, the role for fMRI in neurosurgical planning can be significant, especially when the presence of a tumor alters the expected location of a function, or when the location of the tumor is in an area with an uncertain function such as association cortices or language-related processes. An emerging group of investigators have reported fMRI results that are consistent with

22

electrophysiology, PET, cortical stimulation, and magneto-encephalography and serve to document that fMRI does provide a source of precise functional and structural information for neurosurgery [38–45]. Further, the potential role of fMRI in directing decisions about surgical and diagnostic procedures has also been demonstrated [46,47]. fMRI can be useful in the selection of patients for whom a surgical resection is attempted and could aid in the decision-making whether to operate on a patient who has been previously considered inoperable. fMRI is a useful tool in the decisional scheme of treatment of low-grade astrocytomas or arteriovenous malformations (AVM) in the rolandic area in intact or slightly impaired patients. fMRI can be repeated in selected patients with slow growing brain tumors or congenital lesions such as AVM to study cortical reorganization phenomena. In many neurosurgical centers, the Wada-test (barbiturate injection into one of the internal carotid arteries) was substituted by fMRI to determine the hemispheric dominance of language. After the dura had been opened and/or part of the tumor had been removed, the functional tissue that the surgeon wants to preserve might have shifted. As a future potential, real-time fMRI could identify this functional tissue intraoperatively, comparable to an inverse intraoperative frozen section diagnosis.

PET as a Differential Diagnostic Tool Since the development of PET, in vivo imaging has become the method of preference to assess a variety of neurodegenerative and neuropsychiatric disorders [48,49].

311

312

22


Parkinson’s Disease Advances in Understanding Parkinson’s Disease In the last two decades, measurements of rCBF (with methods like PET or fMRI) as an indicator of neuronal activity have greatly advanced our understanding of the neuronal processes underlying motor and cognitive deficits in patients with idiopathic Parkinson’s disease (IPD). In initial H2O15 PET activation studies on motor planning and execution in IPD, the hypothesis of an excessive cortical inhibition from corticostriatal loops had been strengthened [50,51]. The separate contributions of dopamine loss in the striatum and direct cortical involvement in cognitive impairment in IPD had been the subject of several studies in recent years (see [52,53] for reviews). rCBF PET and fMRI have also been successfully used to investigate the functional impact of pharmacological or surgical interventions in IPD (see [54,55] for reviews). Currently, the identification of genetically defined at-risk populations yields great potential to identify adaptive neuronal mechanisms in a preclinical state of parkinsonism with neuroimaging techniques (see [55] for review).

Differentiate Between Idiopathic Parkinson’s Disease and Atypical Parkinsonism The clinical diagnosis of IPD seems fairly straightforward in most cases. Conversely, clinical pathology studies demonstrated that up to 25% of the patients clinically diagnosed with IPD had the post mortem diagnosis of progressive supranuclear palsy (PSP), multiple system atrophy (MSA), vascular parkinsonism or AD [56]. Yet, establishment of an early and accurate diagnosis impacts on management, helps to avoid inappropriate treatment and assists in the

evaluation of novel drugs. A wide range of objective neuroimaging methods currently contribute to establish diagnosis. Especially functional imaging techniques like PET and single photon emission computed tomography (SPECT) can be used with various radioligands to provide quantitative assessment of dopamine functioning in- and outside of the brain. In terms of a differential diagnostic tool, the degeneration of dopaminergic neurons with reduction of their respective PET markers has been described in all diseases that cause parkinsonism, but the relative involvement of the pre- and postsynaptic metabolism in the striatum and the rostral and caudal parts of the striatum provides some distinction between diseases. PET can demonstrate the disturbance of dopamine synthesis that is the hallmark of PD. The most widely used tracer in PD is FDopa. In Parkinson’s disease, the PET scan shows a characteristic pattern of reduced striatal uptake for FDopa, and it begins to appear very early in the course of the disease [57] most strongly on the side opposite to the major motor signs [58,59] and predominantly in the posterior part of the putamen, indicating loss of more than 50% of dopaminergic neurons projecting to this part of the striatum [60]. This typical differential intrastriatal distribution of reduced uptake is often referred to as the ‘‘rostrocaudal gradient’’ [61]. Although the ability to demonstrate reductions in FDopa uptake in the putamen can help in the diagnosis of IPD, the differentiation between IPD, PSP, MSA, or corticobasal degeneration (CBD) is more difficult. It had been indicated that in PSP and MSA, nigral projections to the caudate nucleus become involved earlier in the course of disease resulting in more equally decreased FDopa uptake in terms of rostrocaudal distribution [62]. However, due to a significant overlap, FDopa PET alone seems not to be able to distinguish between different forms of parkinsonism [63]. An alternative measure of presynaptic dopamine terminal integrity is PET


imaging with radiotracers that bind either to DAT ([11C]MP) or the VMAT2 ([11C]DTBZ). These methods show similar findings in IPD to those seen with FDopa PET and the ability to differentiate different forms of parkinsonism is similarly low [64]. The distinction is much better (80–90%, according to Brooks 2002 [65]) for indicators of striatal postsynaptic D2 receptors ([11C]raclopride) or glucose metabolism ([18F]FDG) [66] (see > Figure 22-1). This distinction between PSP and MSA on the one side and IPD on the other side rests mainly on the fact that in PSP and MSA, dopaminergic neurodegeneration affects both pre- and postsynaptic nerve fibers, whereas in IPD there is a presynaptic dopaminergic deficit in the striatum while postsynaptic striatal neurons are fairly intact and their D2 receptors binding and glucose metabolism have been shown to be normal or even increased in

22

untreated patients [67,68]. In contrast, striatal glucose metabolism is reduced in 80–100% of probable atypical parkinsonism patients [65]. With FDG, CBD differs from PD by a metabolic decrease in premotor, primary motor, supplementary motor, primary sensory and parietal associative cortices but also in caudate and thalamus [69]. Based on FDG or D2 receptors studies, however, distinction among MSA, PSP and CBD is barely possible, limiting the sensitivity of this investigation in the differential diagnosis of parkinsonism [65,70,71]. In many neuroimaging facilities only SPECT might be available. Generally speaking, in terms of differential diagnosis of parkinsonism, the same strengths and limitations apply to SPECT imaging of the presynaptic dopamine transporter (123I-b-CIT or 123 FP-CIT) and postsynaptic dopamine receptor (123I-IBZM).

. Figure 22-1 Representative PET images of FDOPA, RACLO and FDG at the mid-striatal level from one patient with Parkinson’s disease (PD) (top row) and one patient with multiple system atrophy (MSA) (bottom row). Each FDOPA, RACLO and FDG image is scaled relative to common maximum and background levels. At the putamen level, note the marked decrease of dopamine D2 receptor binding (RACLO) and glucose consumption (FDG) in MSA patient which cannot be found in PD patient. Reduction of putaminal DOPA influx constants are similarly visible in PD and MSA (from Antonini et al., Brain 1997;120(12):2187–95)

313

314

22


Dementias Dementia is a clinical syndrome most commonly characterized by impaired short- and long-term memory and associated with deficits in many cortical functions which interfere significantly with the activities of daily life. Dementia is a common disease and prevalence continues to grow with increased human life expectancy. The most frequent cause is Alzheimer’s disease, accounting for 60–70% of all dementias in the elderly, followed by Lewy body dementia (LBD) (10–25%), vascular dementia (10–15%) and frontotemporal dementia (FTD) (5–10%). Synaptic dysfunction or loss – a hallmark of different types of dementia – entails at the molecular level several mechanisms that finally result in decreased energy demand. Therefore, assessment of glucose metabolism with FDG-PET is a valid tool for imaging the energy metabolism of the brain and its typical changes in dementia [72,73] as it can differentiate Alzheimer’s from other confounding types of dementia [74–76]. A consistent finding that has been noted since the earliest PET studies in AD is the hypometabolism affecting the temporal and parietal association cortex [77,78]. Recently, some studies using voxel-based comparison against normal reference data show that the posterior cingulate gyrus and the precuneus are also impaired early in the course [79]. In the parieto-temporal association cortices of AD patients, the reduction in glucose metabolism is greater than the reductions in blood flow and oxygen metabolism [78]. In contrast to other dementia types, glucose metabolism in the basal ganglia, primary motor and visual cortex, and cerebellum is usually well preserved [72–75]. Indeed FDG-PET in dementia with Lewy bodies reveals changes similar to those seen in AD plus additional hypometabolism in primary and associative visual cortices [80–83]. Occipital hypometabolism is the feature of DLB that discriminates it from AD. FDG-PET

scanning in FTD is associated with hypometabolism in the frontal, anterior and medial temporal cortices [84,85]. Low striatal DAT activity (e.g., indexed by [11C]MP) occurs in DLB but is normal in AD, making DAT scanning particularly useful in distinguishing between the two disorders [86]. FTD is a syndrome that can be clinically difficult to distinguish from AD, but in FTD amyloid deposition is not a characteristic pathological finding. Recently, it has been shown that N-methyl[11C]2-(40 methylaminophenyl)6-hydroxy-benzothiazole (PIB, a PET tracer with amyloid binding properties) could potentially aid in differentiating between FTD and AD [87,88] (see > Figure 22-2). The assessment of specific neurotransmitter systems with PET is likely to contribute substantially to clinical distinction between different neurodegenerative diseases that may lead to dementia. Receptor ligands for the cholinergic, dopaminergic, and serotoninergic system and newly developed tracers that label amyloid plaques are likely to play an important role. The full clinical relevance of these developments probably will turn up when more specific and also coursemodifying drugs for dementia treatment become available.

PET as a Preoperative Assessment Tool Brain Tumors Diagnosis and Staging PET has emerged as a powerful diagnostic tool in differentiating malignant from benign tumors. Indeed, the uncontrolled cellular proliferation is the hallmark of malignant transformation and offers the perfect target for diagnosis and evaluation with functional imaging. To this end, a


22

. Figure 22-2 PIB standardized uptake values (SUV) of four representative subjects. FTD PIB negative: a 65-year-old patient with a clinical diagnosis of frontotemporal dementia (FTD; MMSE = 29, CT scan normal) with a PIB retention similar to healthy control. FTD PIB positive: a 75-year-old patient with a clinical diagnosis of FTD (MMSE = 27, CT scan normal) with a PIB retention similar to that found in AD patients. AD: scan of a typical patient with Alzheimer disease. SUV were obtained using the time interval 40–60 min. The results of this study indicate that the majority of FTD patients displayed no PIB retention (FTD negative) in line with the assumption that amyloid deposition is a characteristic neuropathological feature of AD, but not FTD. However, two of the ten patients with clinical diagnosis of FTD tested in this study (FTD positive) had a PIB retention similar to the AD patients suggesting that their true diagnosis might be AD. Therefore, PIB could potentially aid to differentiate between FTD and AD (from Engler et al., Eur J Nucl Med Mol Imaging 2008;35:100–6)

number of radiotracers – targeting changes in glucose metabolism (FDG), protein (e.g., [11C]methionine) or DNA (e.g., [11C]-thymidine) synthesis – have been developed to exploit these differences between malignant and normal cells. FDG-PET

Currently, the evaluation of brain tumors with FDG-PET is widely used in clinical oncology as the rate of glucose utilization is directly proportional to the degree of malignancy [89]. FDG uptake in low-grade gliomas (which are mostly grade II in adults) is usually similar to that of normal white matter, whereas most grade III anaplastic gliomas have a FDG uptake similar to or exceeding that of normal gray matter. Untreated glioblastomas, the most malignant

gliomas (grade IV), also have increased uptake of FDG, which might be heterogeneous throughout the tumor owing to the microscopic and macroscopic necroses that are typical in this tumor type [90]. Accordingly, FDG-PET is a valuable tool for accurate differentiation of lowand high-grade gliomas (> Figure 22-3). Across oncologic applications, the sensitivity and specificity of FDG-PET ranged from 84 to 87% and 88 to 93%, respectively. However, the main limitation of FDG for clinical studies of brain tumors is the high glucose consumption of normal gray matter (45 mmol/100 g/min) that may be in the same range as malignant tumors. Thus, even malignant tumors may be missed if surrounded by intact gray matter. Ac-

315

316

22


. Figure 22-3 Co-registered MRI and FDG and 11C-methioinine PET images in glioblastoma. The contour delineates the areas of contrast enhancement (top row, middle) and is projected onto all other images. On the FDG images, only part of the contrast-enhancing area shows uptake comparable to normal gray matter, indicating the most aggressive part of the tumor, whereas methionine uptake exceeds the contrast enhancing area, which includes low-grade and tumor infiltration zones (from Herholz et al., Lancet Neurol 2007;6(8):711–24)

cordingly, evaluation of glucose consumption in brain tumors can only be done reliably if the location of the tumor is accurately known, best by digital image coregistration with MRI or CT. Amino Acid Uptake

Most brain tumors show an increased uptake of amino acids that is probably due to increased carrier-mediated transport at the blood–brain barrier (BBB). Increased uptake is also seen in most low-grade gliomas in the absence of BBB damage, which is a substantial advantage over CT, MRI, and FDG-PET [91–93]. Other 18 F-tagged radioisotopes (e.g., tyrosine) have also been found useful to improve diagnostic assessment of cerebral gliomas [94]. The most commonly used radiolabeled amino acid is 11C-L-Methionine (MET). Many studies have reported the use of MET to investigate malignancy, extent of tumor spread, effec-

tiveness of therapy, and prognosis of brain tumors [95–99]. Although MET PET has a limited ability to grade gliomas, it will provide reliable information about the extent of tumor infiltration.

Prognosis PET helps to assess tumor localization, extension and degree of malignancy, but histopathological examination of the tissue still is indispensable for definitive diagnosis and prognosis. Stereotactic biopsy is the least invasive way to obtain a specimen for histopathological classification of brain lesions. To this end, FDG-PET is used as stereotactic PET for directing biopsies accurately in the abnormal foci of brain tumors [100–103]. Many malignant gliomas are heterogeneous, but grad-


ing needs to be done on the most malignant parts, which are commonly difficult or impossible to identify with standard structural imaging. The most metabolically active tumor part on PET (FDG or amino acid) indicates the most informative location for taking a biopsy [103–105]. Furthermore, in biopsy-proven low-grade gliomas, tumoral FDG uptake correlates well with the risk of malignant transformation [106,107].

Differentiation Between Recurrent Tumor and Radiation Necrosis Detection of recurrent tumor is an important issue because growth of recurrent tumor will lead to increase of symptoms and ultimate death of the patient. FDG-PET has been used successfully for that purpose in high-grade tumors [108], for detection of malignant progression in low-grade gliomas [109] and to distinguish radiation necrosis from recurrent tumor. Classically recurrent tumor is ‘‘hot’’ on FDG-PET studies and radiation necrosis is ‘‘cold’’; however, these two processes are often interleaved, so the reported sensitivity and specificity is quite low [110]. But there is increasing evidence that amino acid tracers can provide this discrimination [111].

Epilepsy Epilepsy is one of the most common neurological conditions. Almost 60% of patients respond to the first tried antiepileptic drug. However, 20% of people with epilepsy continue to have seizures despite adequate anti-epileptic drug treatment. This failure of drug treatment has led to an increasing interest in neurosurgery for epilepsy, particularly surgical approaches which aim to remove the ‘epileptic focus’ and, therefore offer the opportunity for these patients to be-

22

come free of seizures. One of the main objectives of presurgical investigation in patients with medically refractory epilepsy is to define the boundaries of the epileptogenic region to be resected. The epileptogenic zone refers to the region of cerebral cortex that is both necessary and sufficient to generate epileptic seizures, hence its entire removal is required for a successful outcome. Toward this goal, chronic intracranial EEG evaluation remains the gold standard. However, because this method carries a risk of morbidity and possible mortality, it is appropriate only when reliable conclusions cannot be obtained by less invasive methods. In parallel, a wide range of imaging techniques is valuable for imaging the epileptogenic zone, including highresolution T1 MRI, T2 signal quantitation, MR spectroscopy, diffusion imaging, PET, SPECT and simultaneous EEG-fMRI. MRI is often able to identify the source of seizure in patients with focal epilepsy. However, 20 to 30% of potential surgical candidates with focal epilepsy have normal MRI [112]. Interictal FDG-PET has been shown to be more sensitive than MRI in the identification of seizure of temporal lobe origin [113] and a valuable tool in patients with intractable epilepsy without a structural lesion [114–116]. The characteristic finding is a regional reduction in glucose uptake during the interictal state [117–119]. However, when investigating partial epilepsy, the benzodiazepine ligand [11C]FMZ may also be used which is a selective antagonist of GABAA– BZD receptors. The reduction of FMZ binding is much more focally restricted than reductions of FDG uptake [120] and the area of focal reduction of FMZ binding is probably also a better indicator of the epileptogenic zone that needs to be resected to become seizure free [121,122]. Therefore FMZ-PET can be more useful in preoperative planning than [18F]FDG alone. Although interictal PET provides useful information in temporal and extratemporal lobe epilepsy, its

317

318

22


role is more important in patients with normal structural imaging.

Functional Brain Mapping with PET and fMRI The neurosurgeon often has to balance the benefits of brain tissue removal (e.g., in focal epilepsy or mass lesions) and the potentially devastating iatrogenic disruption of brain functions. Therefore a precise delineation of the structural and functional aspects of brain can prove very helpful in the preoperative decision making process, especially when the presence of a tumor alters the expected location of a function. Two of the most widely used approaches include PET and fMRI. Advantages of fMRI are: higher spatial and temporal resolution, more and different functional runs, shorter examination time, wider availability, longitudinal examinations, non-invasiveness and cost-effectiveness, easy registration to anatomical images. Advantages of PET are: higher signalto-noise ratio, lesser susceptibility to artifacts (motion, draining veins). Moreover, in deep brain stimulation (DBS), functional imaging has been used to investigate the impact of the procedure, but also identified potential stimulation areas.

Motor Mapping H2O15 PET and fMRI both are able to delineate brain regions activated by volitional movements, namely the primary sensorimotor cortex, the supplementary motor area, the lateral premotor cortex and the superior parietal lobule. Preoperative motor mapping with fMRI has been described [123–125] and validated against (the more invasive) electric cortical stimulation (ECS). Majos and colleagues studied 33 patients with brain tumors in the rolandic area and found an agreement of both methods of 98%

for combined motor and sensory representations [124]. Using 3T fMRI and ECS, Roessler et al. studied 22 patients with gliomas involving the primary motor cortex [125]. Motor foci were successfully detected with fMRI in all patients, but a successful intraoperative stimulation of the primary motor cortex was possible only in 77% of the patients. Moreover, in this subgroup, the motor focus in ECS and fMRI was identical within 1 cm. These results point to the direction, that motor mapping with fMRI or PET are safe and reliable techniques to assess the risk of a motor deficit following surgical procedure.

Language Mapping The lateralization of language is of special interest in patients with medial temporal lobe epilepsy and with tumors of the ventral frontal and temporal lobe. To this end, the selective intracarotid amobarbital application (Wada test) had been the gold standard for almost 50 years. Recently however, the Wada test gets more and more replaced by less invasive procedures such as fMRI. Klo¨ppel and Bu¨chel reviewed four studies comparing the Wada test with fMRI based language lateralization [126]. The authors summarize an agreement of about 90% of the two methods and note the advantage of fMRI to additionally provide a precise localization of language functions. They predicted that fMRI will be most widely used to assess language lateralization. But there might be an essential advantage of PET on the clinical application of language activation studies that challenges a predominant role of fMRI. Active speaking during language production tasks does not induce technical artifacts (as it is common with fMRI), and therefore, direct monitoring of task performance is possible even in functionally impaired subjects [127–129].


Deep Brain Stimulation Pre- and post-operative imaging of H2O15 PET has been used to define the effects of deep brain stimulation in cases of Parkinson’s disease, depression and pain syndromes. For instance, concerning subthalamic nucleus (STN) stimulation in PD, three H2O15 PET studies found an enhanced movement-related activity in the dorsolateral and mesial prefrontal cortex, two areas known to be underactive in unmedicated PD [130–132]. For the treatment chronic pain syndromes, Davis and associates report on the changes on pre- and postoperative PET scanning in patients undergoing placement of thalamic stimulators. The investigators found that DBS caused activation of the contralateral anterior cingulate cortex, a known centre involved in pain and analgesia [133]. In parallel there has been some evidence that chronic electrical stimulation of the primary motor cortex (MCS) may relieve motor symptoms of PD. This surgical technique has been proposed as an alternative for selected PD patients who are considered poor candidates for DBS of the STN. However, to date the MCS technique remains a controversial procedure. Recently, using PET, our group [134] suggest that while unilateral MCS is probably a simpler and safer surgical procedure than DBS of STN, it did not improve motor performance nor significantly modify the activation pattern of movement-related rCBF in patients with advanced PD. These observations along with recent negative clinical experiences in PD patients [135] raise some controversy about the efficacy of this therapeutic modality in PD. Sometimes, functional imaging studies even identify new DBS targets. Based on the converging PET findings that the subgenual cingulate region (Brodmann area, BA, 25) is overactive in treatment resistant depression, Mayberg and colleagues tested DBS of adjacent white matter tracts to modulate BA25. They found a sustained

22

remission of depression in four of six patients. This clinical improvement was associated with a marked reduction in local cerebral blood flow as well as changes in downstream limbic and cortical sites [136]. Using PET, a possible cerebral origin of cluster headache has been visualized in the hypothalamic gray matter [137] prompting the successful use of DBS in the inferior posterior hypothalamus in intractable chronic cluster headache patients [138–140].

References 1. Phelps ME, et al. Application of annihilation coincidence detection to transaxial reconstruction tomography. J Nucl Med 1975;16(3):210-24. 2. Ter-Pogossian MM, et al. A positron-emission transaxial tomograph for nuclear imaging (PETT). Radiology 1975;114(1):89-98. 3. Posner MI, et al. Localization of cognitive operations in the human brain. Science 1988;240(4859):1627-31. 4. Fox PT, et al. Mapping human visual cortex with positron emission tomography. Nature 1986;323(6091):806-9. 5. Raichle ME, et al. Practice-related changes in human brain functional anatomy during nonmotor learning. Cereb Cortex 1994;4(1):8-26. 6. Reivich M, et al. The [18F]fluorodeoxyglucose method for the measurement of local cerebral glucose utilization in man. Circ Res 1979;44(1):127-37. 7. Snow BJ, et al. Human positron emission tomographic [18F]fluorodopa studies correlate with dopamine cell counts and levels. Ann Neurol 1993;34(3):324-30. 8. Firnau G, et al. Cerebral metabolism of 6-[18F]fluoro-L-3, 4-dihydroxyphenylalanine in the primate. J Neurochem 1987;48(4):1077-82. 9. Martin WR, et al. Nigrostriatal function in humans studied with positron emission tomography. Ann Neurol 1989;26(4):535-42. 10. Farde L, et al. Stereoselective binding of 11C-raclopride in living human brain – a search for extrastriatal central D2-dopamine receptors by PET. Psychopharmacology (Berl) 1988;94(4):471-8. 11. Rinne JO, et al. Increased density of dopamine D2 receptors in the putamen, but not in the caudate nucleus in early Parkinson’s disease: a PET study with [11C]raclopride. J Neurol Sci 1995;132(2):156-61. 12. Mukherjee J, et al. Brain imaging of 18F-fallypride in normal volunteers: blood analysis, distribution, testretest studies, and preliminary assessment of sensitivity to aging effects on dopamine D-2/D-3 receptors. Synapse 2002;46(3):170-88.

319

320

22


13. Ding YS, et al. Pharmacokinetics and in vivo specificity of [11C]dl-threo-methylphenidate for the presynaptic dopaminergic neuron. Synapse 1994;18(2):152-60. 14. Volkow ND, et al. A new PET ligand for the dopamine transporter: studies in the human brain. J Nucl Med 1995;36(12):2162-8. 15. Ciliax BJ, et al. The dopamine transporter: immunochemical characterization and localization in brain. J Neurosci 1995;15(3 Pt 1):1714-23. 16. Lee CS, et al. In vivo positron emission tomographic evidence for compensatory changes in presynaptic dopaminergic nerve terminals in Parkinson’s disease. Ann Neurol 2000;47(4):493-503. 17. Logan J, et al. Graphical analysis of reversible radioligand binding from time-activity measurements applied to [N-11C-methyl]-(-)-cocaine PET studies in human subjects. J Cereb Blood Flow Metab 1990;10(5):740-7. 18. Telang FW, et al. Distribution of tracer levels of cocaine in the human brain as assessed with averaged [11C]cocaine images. Synapse 1999;31(4):290-6. 19. Haaparanta, M., et al. [18F]CFT ([18F]WIN 35,428), a radioligand to study the dopamine transporter with PET: biodistribution in rats. Synapse 1996;23(4):321-7. 20. Koeppe RA, et al. Kinetic evaluation of [11C]dihydrotetrabenazine by dynamic PET: measurement of vesicular monoamine transporter. J Cereb Blood Flow Metab 1996;16(6):1288-99. 21. Chan GL, et al. Reproducibility studies with 11C-DTBZ, a monoamine vesicular transporter inhibitor in healthy human subjects. J Nucl Med 1999;40(2):283-9. 22. Mansour A, et al. Biochemical anatomy: Insights into the cell biology and pharmacology of dopamine and serotonin systems in the brain. In Schatzberg AF, Nemeroff CB, editors. Textbook of psychopharmacology. Washington, DC: American Psychiatric Press; 1998. p. 55-73. 23. Gunn RN, et al. Tracer kinetic modeling of the 5-HT1A receptor ligand [carbonyl-11C]WAY-100635 for PET. Neuroimage 1998;8(4):426-40. 24. Costes, N., et al. A 18F-MPPF PET normative database of 5-HT1A receptor binding in men and women over aging. J Nucl Med 2005;46(12):1980-9. 25. Costes, N., et al. Modeling [18F]MPPF positron emission tomography kinetics for the determination of 5-hydroxytryptamine(1A) receptor concentration with multiinjection. J Cereb Blood Flow Metab 2002;22 (6):753-65. 26. Blin, J., et al. [18F]setoperone: a new high-affinity ligand for positron emission tomography study of the serotonin-2 receptors in baboon brain in vivo. Eur J Pharmacol 1988;147(1):73-82. 27. Kapur, S., et al. Reliability of a simple non-invasive method for the evaluation of 5-HT2 receptors using [18F]-setoperone PET imaging. Nucl Med Commun 1997;18(5):395-9.

28. Hinz, R., et al. Validation of a tracer kinetic model for the quantification of 5-HT(2A) receptors in human brain with [11C]MDL 100,907. J Cereb Blood Flow Metab 2007;27(1):161-72. 29. Rosa-Neto, P., et al. Stability of a-[11C]methyl-L-tryptophan brain trapping in healthy male volunteers. Eur J Nucl Med Mol Imaging 2005;32(10):199-204. 30. Irie, T., et al. , Brain acetylcholinesterase activity: validation of a PET tracer in a rat model of Alzheimer’s disease. J Nucl Med 1996;37(4):649-55. 31. Kilbourn MR, et al. In vivo studies of acetylcholinesterase activity using a labeled substrate, N-[11C]methylpiperdin-4-yl propionate ([11C]PMP). Synapse 1996;22 (2):123-31. 32. Namba H, et al. In vivo measurement of acetylcholinesterase activity in the brain with a radioactive acetylcholine analog. Brain Res 1994;667(2):278-82. 33. Snyder SE, et al. Synthesis of 1-[11C]methylpiperidin-4-yl propionate ([11C]PMP) for in vivo measurements of acetylcholinesterase activity. Nucl Med Biol 1998;25 (8):751-4. 34. Savic I, et al. In-vivo demonstration of reduced benzodiazepine receptor binding in human epileptic foci. Lancet 1988;2(8616):863-6. 35. Richardson MP, et al. Benzodiazepine receptors in focal epilepsy with cortical dysgenesis: an 11C-flumazenil PET study. Ann Neurol 1996;40(2):188-98. 36. Juhasz C, et al. Neuroradiological assessment of brain structure and function and its implication in the pathogenesis of West syndrome. Brain Dev 2001;23(7):488-95. 37. Bressan RA, Pilowsky, LS. Imaging the glutamatergic system in vivo – relevance to schizophrenia. Eur J Nucl Med 2000;27(11):1723-31. 38. Burgess RC. Functional localization by neurophysiologic and neuroimaging techniques. J Clin Neurophysiol 1995;12(5):405. 39. George JS, et al. Mapping function in the human brain with magnetoencephalography, anatomical magnetic resonance imaging, and functional magnetic resonance imaging. J Clin Neurophysiol 1995;12(5):406-31. 40. Simpson GV, et al. Dynamic neuroimaging of brain function. J Clin Neurophysiol 1995;12(5):432-49. 41. Puce A. Comparative assessment of sensorimotor function using functional magnetic resonance imaging and electrophysiological methods. J Clin Neurophysiol 1995;12(5):450-9. 42. Puce A, et al. Functional magnetic resonance imaging of sensory and motor cortex: comparison with electrophysiological localization. J Neurosurg 1995;83(2):262-70. 43. Fried I, et al. Functional MR and PET imaging of rolandic and visual cortices for neurosurgical planning. J Neurosurg 1995;83(5):854-61. 44. Detre JA, et al. Localization of subclinical ictal activity by functional magnetic resonance imaging: correlation with invasive monitoring. Ann Neurol 1995;38(4):618-24.


45. Ives J.R, et al. Monitoring the patient’s EEG during echo planar MRI. Electroencephalogr Clin Neurophysiol 1993;87(6):417-20. 46. Atlas S.W, et al. Functional magnetic resonance imaging of regional brain activity in patients with intracerebral gliomas: findings and implications for clinical management. Neurosurgery 1996;38(2):329-38. 47. Archip N, et al. Non-rigid alignment of pre-operative MRI, fMRI, and DT-MRI with intra-operative MRI for enhanced visualization and navigation in image-guided neurosurgery. Neuroimage 2007;35(2):609-24. 48. Jones T. The role of positron emission tomography within the spectrum of medical imaging. Eur J Nucl Med 1996;23(2):207-11. 49. Jones T. The imaging science of positron emission tomography. Eur J Nucl Med 1996;23(7):807-13. 50. Jahanshahi M, et al. Self-initiated versus externally triggered movements. I. An investigation using measurement of regional cerebral blood flow with PET and movementrelated potentials in normal and Parkinson’s disease subjects. Brain 1995;118 (Pt 4):913-33. 51. Jenkins IH, et al. Impaired activation of the supplementary motor area in Parkinson’s disease is reversed when akinesia is treated with apomorphine. Ann Neurol 1992;32(6):749-57. 52. Dagher A, Nagano-Saito A. Functional and anatomical magnetic resonance imaging in Parkinson’s disease. Mol Imaging Biol 2007;9(4):234-42. 53. Grafton ST. Contributions of functional imaging to understanding parkinsonian symptoms. Curr Opin Neurobiol 2004;14(6):715-9. 54. Ceballos-Baumann AO. Functional imaging in Parkinson’s disease: activation studies with PET, fMRI and SPECT. J Neurol 2003;250 Suppl 1:I15-I23. 55. van Eimeren T, Siebner HR. An update on functional neuroimaging of parkinsonism and dystonia. Curr Opin Neurol 2006;19(4):412-9. 56. Hughes AJ, et al. Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55 (3):181-4. 57. Heiss WD, Hilker R. The sensitivity of 18-fluorodopa positron emission tomography and magnetic resonance imaging in Parkinson’s disease. Eur J Neurol 2004;11(1):5-12. 58. Leenders KL, et al. Brain dopamine metabolism in patients with Parkinson’s disease measured with positron emission tomography. J Neurol Neurosurg Psychiatry 1986;49(8):853-60. 59. Nahmias C, et al. Striatal dopamine distribution in parkinsonian patients during life. J Neurol Sci 1985;69(3):223-30. 60. Damier P, et al. The substantia nigra of the human brain. II. Patterns of loss of dopamine-containing neurons in Parkinson’s disease. Brain 1999;122 (Pt 8):1437-48.

22

61. Garnett ES, et al. A rostrocaudal gradient for aromatic acid decarboxylase in the human striatum. Can J Neurol Sci 1987;14 Suppl 3:444-7. 62. Daniel SE. de Bruin VM, Lees AJ. The clinical and pathological spectrum of Steele-Richardson-Olszewski syndrome (progressive supranuclear palsy): a reappraisal. Brain 1995;118(Pt 3):759-70. 63. Ghaemi, M, et al. Differentiating multiple system atrophy from Parkinson’s disease: contribution of striatal and midbrain MRI volumetry and multi-tracer PET imaging. J Neurol Neurosurg Psychiatry 2002;73 (5):517-23. 64. Frost JJ, et al. Positron emission tomographic imaging of the dopamine transporter with 11C-WIN 35,428 reveals marked declines in mild Parkinson’s disease. Ann Neurol 1993;34(3):423-31. 65. Brooks DJ. Diagnosis and management of atypical parkinsonian syndromes. J Neurol Neurosurg Psychiatry 2002;72(90001):10i-16i. 66. Antonini A, et al. Complementary PET studies of striatal neuronal function in the differential diagnosis between multiple system atrophy and Parkinson’s disease. Brain 1997;120(Pt 12):2187-95. 67. Otsuka M, et al. Differentiating between multiple system atrophy and Parkinson’s disease by positron emission tomography with 18F-dopa and 18F-FDG. Ann Nucl Med 1997;11(3):251-7. 68. Antonini A, et al. Differential diagnosis of parkinsonism with [18F]fluorodeoxyglucose and PET. Mov Disord 1998;13(2):268-74. 69. Laureys S, et al. Fluorodopa uptake and glucose metabolism in early stages of corticobasal degeneration. J Neurol 1999;246(12):1151-8. 70. Booij J, et al. Imaging of the dopaminergic neurotransmission system using single-photon emission tomography and positron emission tomography in patients with parkinsonism. Eur J Nucl Med 1999;26(2):171-82. 71. Piccini P, Whone A. Functional brain imaging in the differential diagnosis of Parkinson’s disease. Lancet Neurol 2004;3(5):284-90. 72. Herholz K. PET studies in dementia. Ann Nucl Med 2003;17(2):79-89. 73. Silverman DH. Brain 18F-FDG PET in the diagnosis of neurodegenerative dementias: comparison with perfusion SPECT and with clinical evaluations lacking nuclear imaging. J Nucl Med 2004;45(4):594-607. 74. Devous MD Sr. Functional brain imaging in the dementias: role in early detection, differential diagnosis, and longitudinal studies. Eur J Nucl Med Mol Imaging 2002;29(12):1685-96. 75. Van Heertum RL, Tikofsky RS. Positron emission tomography and single-photon emission computed tomography brain imaging in the evaluation of dementia. Semin Nucl Med 2003;33(1):77-85.

321

322

22


76. Ishii K, Minoshima S. PET is better than perfusion SPECT for early diagnosis of Alzheimer’s disease. Eur J Nucl Med Mol Imaging 2005;32(12):1463-5. 77. Jagust WJ, et al. The cortical topography of temporal lobe hypometabolism in early Alzheimer’s disease. Brain Res 1993;629(2):189-98. 78. Fukuyama H, et al. Altered cerebral energy metabolism in Alzheimer’s disease: a PET study. J Nucl Med 1994;35(1):1-6. 79. Minoshima S, et al. Metabolic reduction in the posterior cingulate cortex in very early Alzheimer’s disease. Ann Neurol 1997;42(1):85-94. 80. Albin RL, et al. Fluoro-deoxyglucose positron emission tomography in diffuse Lewy body disease. Neurology 1996;47(2):462-6. 81. Imamura T, et al. Regional cerebral glucose metabolism in dementia with Lewy bodies and Alzheimer’s disease: a comparative study using positron emission tomography. Neurosci Lett 1997;235(1–2):49-52. 82. Ishii K, et al. Regional cerebral glucose metabolism in dementia with Lewy bodies and Alzheimer’s disease. Neurology 1998;51(1):125-30. 83. Higuchi M, et al. Glucose hypometabolism and neuropathological correlates in brains of dementia with Lewy bodies. Exp Neurol 2000;162(2):247-56. 84. Friedland RP, et al. Functional imaging, the frontal lobes, and dementia. Dementia 1993;4(3–4):192-203. 85. Herholz K. FDG PET and differential diagnosis of dementia. Alzheimer Dis Assoc Disord 1995;9(1):6-16. 86. McKeith IG, et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65(12):1863-72. 87. Rowe CC, et al. Imaging b-amyloid burden in aging and dementia. Neurology 2007;68(20):1718-25. 88. Engler H, et al. In vivo amyloid imaging with PET in frontotemporal dementia. Eur J Nucl Med Mol Imaging 2008;35(1):100-6. 89. Di Chiro G, Brooks RA. PET-FDG of untreated and treated cerebral gliomas. J Nucl Med 1988;29(3):421-3. 90. Delbeke D, et al. Optimal cutoff levels of F-18 fluorodeoxyglucose uptake in the differentiation of low-grade from high-grade brain tumors with PET. Radiology 1995;195(1):47-52. 91. Ogawa T, et al. Quantitative evaluation of neutral amino acid transport in cerebral gliomas using positron emission tomography and fluorine-18 fluorophenylalanine. Eur J Nucl Med 1996;23(8):889-95. 92. Kaschten B, et al. Preoperative evaluation of 54 gliomas by PET with fluorine-18-fluorodeoxyglucose and/or carbon-11-methionine. J Nucl Med 1998;39(5):778-85. 93. Chung JK, et al. Usefulness of 11C-methionine PET in the evaluation of brain lesions that are hypo- or isometabolic on 18F-FDG PET. Eur J Nucl Med Mol Imaging 2002;29(2):176-82.

94. Pauleit D, et al. O-(2-[18F]fluoroethyl)-L-tyrosine PET combined with MRI improves the diagnostic assessment of cerebral gliomas. Brain 2005;128(Pt 3):678-87. 95. Derlon JM, et al. [11C]L-methionine uptake in gliomas. Neurosurgery 1989;25(5):720-8. 96. Ogawa T, et al. Cerebral glioma: evaluation with methionine PET. Radiology 1993;186(1):45-53. 97. Nuutinen, J., et al. Radiotherapy treatment planning and long-term follow-up with [11C]methionine PET in patients with low-grade astrocytoma. Int J Radiat Oncol Biol Phys 2000;48(1):43-52. 98. De Witte O, et al. Positron emission tomography with injection of methionine as a prognostic factor in glioma. J Neurosurg 2001;95(5):746-50. 99. Ribom D, et al. Positron emission tomography 11C-methionine and survival in patients with low-grade gliomas. Cancer 2001;92(6):1541-9. 100. Levivier M, et al. Positron emission tomographyguided stereotactic brain biopsy. Neurosurgery 1992; 31(4):792-7; 101. Pirotte B, et al. PET in stereotactic conditions increases the diagnostic yield of brain biopsy. Stereotact Funct Neurosurg 1994;63(1–4):144-9. 102. Pirotte B, et al. Use of positron emission tomography (PET) in stereotactic conditions for brain biopsy. Acta Neurochir (Wien) 1995;134(1–2):79-82. 103. Levivier M, et al. Diagnostic yield of stereotactic brain biopsy guided by positron emission tomography with [18F]fluorodeoxyglucose. J Neurosurg 1995; 82(3):445-52. 104. Goldman S, et al. Regional methionine and glucose uptake in high-grade gliomas: a comparative study on PET-guided stereotactic biopsy. J Nucl Med 1997; 38(9):1459-62. 105. Pirotte B, et al. Combined positron emission tomography and magnetic resonance imaging for the planning of stereotactic brain biopsies in children: experience in 9 cases. Pediatr Neurosurg 2003;38(3):146-55. 106. Spence AM, Mankoff DA, Muzi M. Positron emission tomography imaging of brain tumors. Neuroimaging Clin N Am 2003;13(4):717-39. 107. Schaller B. Usefulness of positron emission tomography in diagnosis and treatment follow-up of brain tumors. Neurobiol Dis 2004;15(3):437-48. 108. Holzer T, et al. FDG-PET as a prognostic indicator in radiochemotherapy of glioblastoma. J Comput Assist Tomogr 1993;17(5):681-7. 109. Francavilla TL, et al. Positron emission tomography in the detection of malignant degeneration of low-grade gliomas. Neurosurgery 1989;24(1):1-5. 110. Ricci PE, et al. Differentiating recurrent tumor from radiation necrosis: time for re-evaluation of positron emission tomography? AJNR Am J Neuroradiol 1998;19(3):407-13.


111. Tsuyuguchi N, et al. Methionine positron emission tomography for differentiation of recurrent brain tumor and radiation necrosis after stereotactic radiosurgery in malignant glioma. Ann Nucl Med 2004; 18(4):291-6. 112. Carne RP, et al. MRI-negative PET-positive temporal lobe epilepsy: a distinct surgically remediable syndrome. Brain 2004;127(10):2276-2285. 113. Casse R, et al. Positron emission tomography and epilepsy. Mol Imaging Biol 2002;4(5):338-51. 114. Juhasz C, et al. Glucose and [11C]flumazenil positron emission tomography abnormalities of thalamic nuclei in temporal lobe epilepsy. Neurology 1999; 53(9):2037-45. 115. Koepp M.J, et al. 11C-flumazenil PET in patients with refractory temporal lobe epilepsy and normal MRI. Neurology 2000;54(2):332-9. 116. O’Brien TJ, et al. The utility of a 3-dimensional, largefield-of-view, sodium iodide crystal-based PET scanner in the presurgical evaluation of partial epilepsy. J Nucl Med 2001;42(8):1158-65. 117. Kuhl DE, et al. Epileptic patterns of local cerebral metabolism and perfusion in humans determined by emission computed tomography of 18FDG and 13NH3. Ann Neurol 1980;8(4):348-60. 118. Ollenberger GP, et al. Assessment of the role of FDG PET in the diagnosis and management of children with refractory epilepsy. Eur J Nucl Med Mol Imaging 2005;32(11):1311-6. 119. Koepp MJ, Woermann FG. Imaging structure and function in refractory focal epilepsy. Lancet Neurol 2005;4(1):42-53. 120. Savic I, Ingvar M, Stone-Elander S. Comparison of [11C]flumazenil and [18F]FDG as PET markers of epileptic foci. J Neurol Neurosurg Psychiatry 1993;56 (6):615-21. 121. Arnold S, et al. Reduction of benzodiazepine receptor binding is related to the seizure onset zone in extratemporal focal cortical dysplasia. Epilepsia 2000; 41(7):818-24. 122. Muzik O, et al. Intracranial EEG versus flumazenil and glucose PET in children with extratemporal lobe epilepsy. Neurology 2000;54(1):171-9. 123. Yousry TA, et al. Topography of the cortical motor hand area: prospective study with functional MR imaging and direct motor mapping at surgery. Radiology 1995; 195(1):23-9. 124. Majos, A, et al. Cortical mapping by functional magnetic resonance imaging in patients with brain tumors. Eur Radiol 2005;15(6):1148-58. 125. Roessler, K, et al. Evaluation of preoperative high magnetic field motor functional MRI (3 Tesla) in glioma patients by navigated electrocortical stimulation and postoperative outcome. J Neurol Neurosurg Psychiatry 2005;76(8):1152-7.

22

126. Kloppel S, Buchel C. Alternatives to the Wada test: a critical view of functional magnetic resonance imaging in preoperative use. Curr Opin Neurol 2005;18(4):418-23. 127. Kaplan AM, et al. Functional brain mapping using positron emission tomography scanning in preoperative neurosurgical planning for pediatric brain tumors. J Neurosurg 1999;91(5):797-803. 128. Meyer PT, et al. Preoperative mapping of cortical language areas in adult brain tumour patients using PET and individual non-normalised SPM analyses. Eur J Nucl Med Mol Imaging 2003;30(7):951-60. 129. Vanlancker-Sidtis D. McIntosh AR, Grafton S. PET activation studies comparing two speech tasks widely used in surgical mapping. Brain Lang 2003;85 (2):245-61. 130. Limousin P, et al. Changes in cerebral activity pattern due to subthalamic nucleus or internal pallidum stimulation in Parkinson’s disease. Ann Neurol 1997;42 (3):283-91. 131. Ceballos-Baumann AO, et al. A positron emission tomographic study of subthalamic nucleus stimulation in Parkinson disease: enhanced movement-related activity of motor-association cortex and decreased motor cortex resting activity. Arch Neurol 1999;56(8):997-1003. 132. Strafella AP, Dagher A, Sadikot AF. Cerebral blood flow changes induced by subthalamic stimulation in Parkinson’s disease. Neurology 2003;60(6):1039-42. 133. Davis KD, et al. Activation of the anterior cingulate cortex by thalamic stimulation in patients with chronic pain: a positron emission tomography study. J Neurosurg 2000;92(1):64-9. 134. Strafella AP, et al. Subdural motor cortex stimulation in Parkinson’s disease does not modify movement-related rCBF pattern. Mov Disord 2007;22(14):2113-6. 135. Roberto Cilia ALFVESGPAA. Extradural motor cortex stimulation in Parkinson’s disease. Mov Disord 2007; 22(1):111-114. 136. Mayberg HS, et al. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45(5):651-60. 137. May A, et al. Hypothalamic activation in cluster headache attacks. Lancet 1998;352(9124):275-8. 138. Leone M, Franzini A, Bussone G. Stereotactic stimulation of posterior hypothalamic gray matter in a patient with intractable cluster headache. N Engl J Med 2001;345(19):1428-9. 139. Leone M, et al. Hypothalamic deep brain stimulation for intractable chronic cluster headache: a 3-year follow-up. Neurol Sci 2003;24 Suppl 2:S143-S145. 140. Franzini A, et al. Stimulation of the posterior hypothalamus for treatment of chronic intractable cluster headaches: first reported series. Neurosurgery 2003;52 (5):1095-9; 141. Herholz K, Coope D, Jackson A. Metabolic and molecular imaging in neuro-oncology. Lancet Neurol 2007; 6(8):711-24.

323

20 Functional MRI in Image Guided Neurosurgery T. Sankar . G. R. Cosgrove

Introduction The principal goal of neurosurgical procedures on the cerebral cortex is to accurately localize and maximally resect lesions or abnormal tissue. An intimate knowledge of eloquent cortical regions which may be damaged during surgery is therefore a prerequisite for safe cortical neurosurgery. Identifying functionally important areas of the cortex during surgery can be aided by a priori knowledge of cortical functional organization, but is subject to inaccuracy caused by individual variations in anatomy and displacement of normal structures by pathological lesions. Consequently, several cortical mapping techniques have been established to facilitate safe and effective cortical resections. The gold standard for cortical mapping is direct electrocortical stimulation (ECS) of the exposed cortex at surgery [1–3]. ECS is limited, however, because it is an invasive technique which can usually only be performed in an awake patient [3]. Awake craniotomy is not always feasible and presents unique anesthetic and analgesic challenges. In addition, as a surface mapping technique, ECS cannot provide information about function buried within the sulcal depths which represent a substantial proportion of the cortical volume [4]. Perhaps most importantly, ECS is a purely intra-operative technique which cannot be used preoperatively to guide surgical planning [3]. The past two decades have seen the development and implementation of several noninvasive cortical mapping techniques, including positron emission tomography (PET), #


magnetoencephalography (MEG), and transcranial magnetic stimulation (TMS) [3]. Functional magnetic resonance imaging (fMRI), which assesses regional variations in cerebral blood flow during various tasks of cortical activation, is one such technique. fMRI makes use of readily available conventional MRI hardware to produce maps of functional cortical activation, which can be superimposed onto anatomical images of the brain. The resulting structural and functional model of the brain – which is obtained prior to surgery – can then be used both for preoperative surgical planning and intra-operative decisionmaking. Recent clinical experience in patients with lesions involving eloquent cortex has demonstrated the utility of fMRI in the neurosurgical armamentarium.

Principles of fMRI Acquisition Physiological Basis of fMRI Basic neurophysiologic studies have long shown that cerebral activity is tightly coupled to cerebral metabolism, which is in turn related to cerebral blood flow [5]. As a result, regionalhemodynamic measurements can serve as a surrogate marker of neuronal activity. This principle forms the basis of metabolic/hemodynamic mapping techniques such as PET [6] and fMRI [7]. In 1991, Belliveau et al. [8] first reported on the use of this principle in MR imaging. After intravenous bolus injection of a paramagnetic contrast agent, they rapidly acquired MR images and demonstrated signal change in the visual

288

20

Functional MRI in image guided neurosurgery

cortex due to activation with photic stimulation. This signal change was presumed to be due to increased concentration of contrast agent accompanying increased regional blood flow to the visual cortex, and became known as the ‘‘contrast bolus tracking’’ technique. fMRI based on contrast bolus tracking is seldom used today. The most commonly used fMRI acquisition method measures blood oxygen level-dependent (BOLD) changes in the magnetic resonance (MR) signal, first described by Ogawa et al. [9,10]. BOLD imaging has become the fMRI acquisition technique of choice because it does not require the administration of an exogenous contrast agent. An increase in neuronal activity in a particular region of the brain – caused by, say, performing a particular task – initially leads to a transient increase in oxygen extraction from the blood supply to that region. This is quickly followed by compensatory vasodilation, which produces a net increase in the overall concentration of oxy-hemoglobin (oxy-Hb) relative to deoxy-hemoglobin (deoxy-Hb) [7,11]. The physiologic mechanisms underpinning this vasodilatory response are not yet fully understood, but various locally-generated chemical mediators have been implicated, including serotonin, acetylcholine, neuroactive peptides, and nitric oxide produced by cortical neurons [12]. Since the iron in deoxy-Hb is paramagnetic, it reduces T2 MR signal through spin-dephasing effects. The relative decrease in deoxy-Hb concentration in a region of increased neuronal activity, then, produces a corresponding increase in T2 signal, forming the basis of BOLD imaging [9]. This T2 signal change is primarily thought to come from the microvasculature consisting of the capillary system and small venules, and is of low magnitude (on the order of 0.5–5%) [13]. Studies in both humans and primates have demonstrated that the BOLD signal is proportional to the neuronal firing rate, though the latency of the observed change is on the order of several seconds [14,15].

Technological Requirements for fMRI Acquisition Given the small magnitude of changes in the BOLD signal accompanying neuronal activation, signal-to-noise ratio is of critical importance during fMRI acquisition. Initial fMRI studies were carried out using high field strength magnets (2–4 T) using standard gradient-echo acquisitions. Subsequent improvements in gradient coil shielding and surface coil design allowed conventional 1.5 T scanners to be used, but at the expense of prolonged and frequently impractical acquisition times [16]. The recent development and application of echo-planar imaging gradients has revolutionized fMRI acquisition, allowing entire volumes to be imaged in seconds, with improved spatial resolution and signal-tonoise [17]. Most current fMRI studies are carried out on conventional 1.5 T MRI scanners with modifications to allow for echo-planar imaging, and involve multiple iterations of image acquisition and task repetition [7]. Following acquisition, fMRI data must be processed and displayed in a format suitable for clinical use. Typically, a conventional desktop computer workstation equipped with imageprocessing software is used to average multiple fMRI acquisitions. Next, voxel-by-voxel analysis is performed to identify those cortical regions which have an increased BOLD signal relative to baseline during performed tasks. The threshold used to classify increased signal as true cortical activation is arbitrary; normally, statistical methods are used to select a threshold coefficient 2–3 standard deviations from baseline. fMRI images are then co-registered, fused, and volume rendered with anatomical MRI images using both surface- and volume-based matching techniques [18]. Images of regions of cortical activation can then be superimposed onto anatomical images for display purposes or incorporated into neuronavigation systems for intra operative surgical guidance [19] (> Figure 20‐1).


20

. Figure 20‐1 fMRI of sensorimotor hand and foot activation superimposed on axial T2 weighted images demonstrating a tumor in the left posterior superior frontal gyrus just anterior to the motor strip. Note that the right foot activation is less intense and pushed posteriorly by the tumor

Task Design and Selection For neurosurgical mapping purposes, fMRI data are generated using a block design paradigm. Simply, block design involves multiple trials of a task or stimulus presentation – a so-called ‘‘task block’’ – alternating with multiple trials of a control task (‘‘control block’’) [3,20,21]. The resulting BOLD signals from task and control blocks are compared, allowing for the detection of subtle changes accompanying task-related cerebral activation. There is no standardization of the length of each task block or the number of alternating cycles between task and control blocks. Typically, task blocks last from 15 to 30 s; this ensures that they are of sufficient duration to allow the occurrence of hemodynamic changes underlying the fMRI signal [7,22,23]. Cycling multiple trials of a task block with control blocks improves signal-tonoise ratio and the overall statistical accuracy of inferred cortical activations [24] (> Figure 20‐2). Several task variants have been developed to assess cortical activation in different functional regions of the brain, including the sensorimotor cortex, the speech areas of Broca and Wernicke, and the primary visual cortex. The classical motor

activation task is finger-thumb tapping, though repetitive clenching of the fingers or sponge squeezing have also been employed. It is important to realize that this motor task includes stimulation of both cutaneous and proprioceptive sensory inputs and therefore is more accurately a combined sensorimotor task than a pure motor task. Sensory stimulation is frequently achieved by brushing of the palmar surface of the hand [7]. Language is assessed by picture naming, verb generation, or having the patient classify related nouns into categories [25–27]. Because of the motion artifact created by speech, most centers favor of using silent instead of vocalized speech for language mapping [27]. Visual tasks frequently involve intermittent photic stimulation with patterned stimuli presented via binoculars or on a screen [22].

Application of fMRI to Cortical Resections Illustrative Cases Case # 1 – This 17-year-old male had experienced intractable seizures for nearly 10 years which

289

290

20


. Figure 20‐2 Demonstration of the signal changes on fMRI during activation task versus resting state cycle

were characterized by the sudden onset of dystonic posturing and elevation of his right upper extremity followed by rapid secondary generalization. Video EEG recordings suggested a left frontal or bifrontal onset and MRI demonstrated a non-enhancing lesion in the left frontal parasaggital cortex. Functional MRI was carried out for sensorimotor testing of the right hand and the results mapped to the patients structural MRI. At surgery, direct ECS of the exposed cortex confirmed the location of primary sensorimotor cortex of the hand region in the pre and post central gyri. A 4 contact depth electrode was then placed in the region of the presumed supplementary motor area (SMA) and stimulation through the deepest and middle contacts elicited tonic contraction and elevation of the entire right upper extremity consistent with a SMA type seizure (> Figure 20‐3). Careful resection of the cortex just anterior to the SMA revealed a ganglioglioma and the patient has remained seizure free for nearly a decade. Case # 2 – This 24-year-old right-handed man had noticed episodes of sudden onset of a ‘‘tickling’’ in his throat followed by reflexive coughing and some slight difficulty speaking within the past year. These episodes would last for just a few minutes and occur irregularly. After

one such episode, he experienced a generalized tonic conic convulsion and he was taken to a local emergency room where an enhancing lesion was discovered in his dominant subcentral gyrus. A functional MRI was performed for sensorimotor tongue activation and for language mapping using a visual verb generation task (> Figure 20‐4a). At surgery under local anesthesia, cortical mapping of tongue sensorimotor and language areas corresponded closely to that predicted by fMRI. The location of the tumor could be easily determined by the gyral anatomy and cortical vessels and was resected without deficit (> Figure 20‐4b). Pathology demonstrated an xanthoastrocytoma and the patient has been seizure free since surgery. Case #3 – This 47-year-old bilingual, righthanded Greek woman had previously undergone resection of an oligodendroglioma 5 years before presenting with a recurrent tumor. Functional MRI was performed using visual and auditory verb generation tasks in both English and Greek (> Figure 20‐5a). These results were mapped to a structural MRI of her brain and at surgery under local anesthesia correlated closely with language mapping in both languages with ECS enabling safe resection of the recurrent tumor (> Figure 20‐5b).


20

. Figure 20‐3 (a) Axial and (b) surface rendered MRI images with superimposed fMRI activations demonstrating a tumor in the superior left frontal gyrus and both primary and supplementary motor area activations during hand movement (c) cropped close up view of cortical surface as predicted by fMRI as compared to (d) intraoperative view of cortical surface with green tags indicating motor responses in the hand, white tags motor responses in the proximal upper extremity and red tag the area where depth electrode stimulation elicited SMA type seizure. Asterisks are for localization purposes but gyral anatomy and cortical veins are also useful

Discussion Sensitivity and Interpretation of fMRI Results According to Kim and Singh [7], the tasks used for fMRI data generation can be considered either simple or complex, based on the area of cortex they consistently activate. For example, finger-thumb tapping is considered a simple motor task, which typically activates a welldefined focal area of motor cortex. However, even so called simple motor tasks are actually a combination of both motor and sensory activations because of the activation of cutaneous and proprioceptive input. Complex tasks such as

language function require simultaneous activation of several regions of the brain, each with multiple sensory inputs and functional outputs. However, even simple tasks may require supplementary areas whose interaction with the primary cortical activation area is poorly understood. Furthermore, in order to increase the overall sensitivity of fMRI, some centers assess cortical activation during tasks completed in both active and passive modes (e.g., spoken and silent speech), because redundancy is thought to exist in functional areas of cortical activation for related tasks [7]. As a consequence of these factors, preoperative fMRI imaging maps frequently demonstrate a greater spread and area of activation than do

291

292

20


. Figure 20‐4 (a) Axial MRI images of enhancing nodule in left central region and the surface rendered fMRI of the whole brain with tumor in yellow, tongue activation in pink and verb generation task activation in green (b) images of the cortical surface at surgery and that predicted by fMRI cropped for comparison. Note the excellent correlation between ECS and fMRI. The subcortical location of the tumor is easily deduced from the surface anatomy

invasive intraoperative mapping techniques such as ECS, and it is occasionally unclear which cortical regions are essential to the completion of a given task. Optimum interpretation of preoperative fMRI is predicated, then, on a reasoned consideration of the balance between increased sensitivity generated by task complexity or redundancy on the one hand, and a priori knowledge of functional neuroanatomy on the other. Keeping these principles in mind, the overall sensitivity of fMRI is usually reported as excellent. In a widely cited report, Hirsch et al. [28] were able to demonstrate 100% sensitivity for identifying language cortex in the superior temporal gyrus, motor

function in the precentral gyrus, and visual function in occipital cortex. They also showed 93% sensitivity for Broca’s area.

Comparison of Functional Cortical Localization Between fMRI and ECS Several studies have attempted to validate fMRI data by comparing it against standard cortical localization by ECS. The definition of adequate correlation between the two techniques varies from study to study, but most authors consider


20

. Figure 20‐5 (a) fMRI images of cortical activation during language tasks in both English and Greek. Note that the auditory verb generation task activates primary auditory cortex in addition to anterior language areas (b) images of the cortical surface cropped for comparison to that predicted by fMRI. Both the structural and functional correlation is excellent

correlation to be successful when both techniques demonstrate functional localization to within some arbitrary distance of one another, usually 10 or 20 mm. Most existing studies have focused on sensorimotor mapping, and have almost uniformly reported excellent correlation between the two techniques [28–39]. In particular, fMRI appears to be particularly successful at correctly identifying the central sulcus (CS): Majos et al. [35] reported a 98% success rate in mapping

the CS, and Lehericy et al. [32] showed a similar concordance rate of 92% in a large study of 60 patients. Furthermore, brain distortion as a result of edema and deformation due to tumors or other pathological entities does not seem to impact this concordance. Hirsch et al. [28] reported that overall localization of the CS was possible in 97% of patients being surgically treated for lesions of the central region. Excellent sensorimotor functional correlation has also been observed in the hand

293

294

20


area located within the pli de passage moyen (PPM) of the precentral gyrus [37], even in the presence of infiltrative glioma [36]. Interestingly, Boling et al. [40] recently used fMRI to demonstrate that there may be a distinct whole hand sensory and motor area within the PPM. To date, agreement between fMRI and ECS has been less robust for language localization. Some studies have indeed reported excellent concordance with word-generation tasks for mapping Broca’s area [28,29,41], as well as for language function in general using a battery of tasks [42]. The most recent and systematic study directly comparing language localization by both methods was published by Roux et al. [43], who examined a series of 14 right-handed patients – all with tumors in the left hemisphere – with preoperative fMRI obtained during naming and verb generation tasks. During operative tumor resection, they also assessed 426 distinct cortical sites by ECS across the patient cohort. In total, 22 sites were found to be ‘‘positive’’ during naming or verb generation (i.e., elicited speech arrest, anomia, hesitations, paraphasic errors, or delayed responses when stimulated). Of these, only 13 (59%) were concordant (i.e., within 1 cm of) with fMRI signals. The authors concluded from these data that fMRI itself is insufficient to make critical surgical decisions in essential cortical language areas. However, the study suffered from some methodological limitations which weakened its dismissal of fMRI for language mapping. Specifically, the study did not use a silent speech generation task during fMRI acquisition, thereby ignoring the principle of task redundancy and increasing susceptibility to motion artifact. In addition, as noted by McKhann II and Hirsch [44] in their comment on the study, language disruption sites within dominant face motor cortex were included among the 22 positive stimulation sites. Normally, fMRI obtained with silent speech tasks does not activate motor cortex, suggesting that the methods may have been inherently biased against the accuracy of fMRI for language. That being

said, recent work has suggested that the weakness of silent speech tasks may lie in their inability to activate contributory cortical sites within the precentral gyrus [45]. Clearly, further work needs to be done to elucidate the optimal testing paradigms which generate the most accurate preoperative fMRI maps of cortical language localization across a varied population of patients. Such work may solidify the value of fMRI in guiding cortical resections in language-specific regions. Currently, all fMRI results for language should be interpreted with caution and must still be validated intra-operatively with ECS in the awake patient.

Advantages of fMRI as a Cortical Mapping Technique in Neurosurgery fMRI has as its principal advantage the fact that it is non-invasive, which significantly increases its safety and repeatability compared to direct ECS. At best, intraoperative testing requires a larger craniotomy, increases operative time, and is associated with significant patient stress and discomfort accompanying an awake procedure under local anesthetic. At worst, ECS may be associated with intraoperative seizures which may force premature termination of mapping or require the procedure to be aborted altogether [3,46]. Furthermore, awake craniotomy generally requires dedicated neuroanesthesia support and sufficient experience which may preclude its use in smaller centers. fMRI has good spatial resolution, estimated to be on the order of 2–5 mm [47]; this is well within the 10 mm spread of electrical current to adjacent cortical areas when ECS is used [7]. fMRI also has the advantage of generating data over the entire extent of the brain, producing global maps of cortical activation. This data can be important even if we concede that the accuracy of fMRI for precise functional localization may


be uncertain [22]. Invasive mapping is typically performed if a proposed resection is believed, preoperatively, to place eloquent cortex at risk. fMRI can hence be used prior to surgery to identify patients in whom invasive intraoperative mapping will be required, avoiding in others the significant risks and increased operative time required by such mapping [48]. This complementary application of fMRI confirms its worth even as work continues to establish it as a cortical mapping technique independent of other imaging or stimulatory modalities. Perhaps the most appealing feature of fMRI is that it can demonstrate cortical activation within the sulcal depths, which have been reported to represent as much as two-thirds of functionally eloquent cortex [4]. This is a clear advantage over ECS, which can only assess surface cortex. The ability of fMRI to guide preservation of cortical grey matter in the deep sulci during surgery may be of particular relevance to avoiding damage to cognitive and memory functions, whose underlying neuroanatomical basis is poorly understood [3,7].

Disadvantages of fMRI in Neurosurgical Applications While fMRI can localize distinct cortical activity, it does not localize underlying white matter tracts or the important connections between functional areas. The risks of incurring neurological deficits are often greatest during subcortical resections in the white matter and therefore careful consideration of the location of these pathways must always be paramount. Only surgery performed under local anesthesia with constant testing of the patient’s function during resection can prevent deficits and therefore fMRI will never completely replace standard ECS. Most shortcomings of fMRI are technical. As with any preoperative imaging modality, fMRI is susceptible to inaccuracies because of the brain

20

shift accompanying craniotomy, dural opening, and CSF drainage [49]. Sensitivity to motion artifact, which is principally caused by a patient’s inability to hold the head still, or occasionally by involuntary movements related to breathing or heartbeat, can also be problematic [3,22]. As mentioned previously, during tasks involving overt speech, word vocalization may alone produce sufficient motion to degrade image acquisition and quality. The rapid image acquisition achieved with echoplanar technology partially reduces the impact of motion on fMRI studies. Firmly securing the head and providing a visual fixation crosshair may also help [28,42], but occasionally – particularly in cognitively impaired or uncooperative patients – an fMRI examination simply cannot be completed. Venous effects may also confuse fMRI interpretation. While fMRI is based on the assumption that the BOLD signal is confined to the microvasculature, in reality there is signal spread to veins draining blood away from activated cortical tissue and which may be located several millimeters from the actual focus of neuronal activation. This signal in turn can produce images depicting false-positive activation in these adjacent regions [50]. Obtaining highquality three dimensional anatomic MRI images showing the positions of draining veins can assist the neurosurgeon in anticipating areas of falsepositive signal [4]. In addition, venous signal appears to be less of an issue at higher magnetic field strength [50]. Related to aberrant venous signal is the theoretical possibility that various neurosurgical pathologies may alter normal blood flow in such a way as to disrupt fMRI interpretation. One such example of this phenomenon is the effect of tumor angiogenesis, whose impact on the BOLD signal is as yet unknown. Vascular steal related to arteriovenous malformations near eloquent cortex has similarly been cited as another potential pitfall [7]. Even atherosclerotic disease may potentially impact regional variations

295

296

20


in blood flow from individual to individual, and within the same individual at different times. Further study is necessary to assess the BOLD signal in these clinical situations to appropriately interpret fMRI data prior to surgical resections. As mentioned previously, perhaps the greatest current limitation of fMRI for guiding cortical resections relates to the – as yet – uncertain neuroscientific implications of the data it generates. At present, we cannot use fMRI to accurately distinguish between essential functional cortical areas and those which are participatory but nonessential to function [7]. This is certainly true for language, but even more so for memory, whose neuroanatomic basis is still incompletely understood [51]. Fortunately, fMRI data are being analyzed in the setting of several neuropsychological and cognitive functions, and the conclusions reached from this basic work will likely make their way into neurosurgical applications. Additionally, the ongoing development of increasingly standardized fMRI task protocols should help to improve the interpretation and reproducibility of cortical mapping data between different patients and different centers.

Conclusions fMRI is a powerful non-invasive neuroimaging technique that can create a unique structural and functional model of an individual patient’s brain. The widespread availability of MRI scanners allows fMRI exams to be acquired in appropriate patients at most modern neurosurgical centers. fMRI data are generated preoperatively and can be used to determine the feasibility of open cortical resections by delineating those areas of eloquent cortex which may be at risk in a planned surgical procedure. Such data can be of critical importance in assessing surgical risk, optimizing surgical exposure, as well as guiding surgical approaches to and the resection of cortical

lesions. fMRI has several advantages over ECS, including its non-invasiveness, its higher spatial resolution, and its ability to assess function within the sulcal depths. fMRI acquisitions can be impaired, however, by technical factors such as motion artifact, infiltrative tumors and venous effects, and fMRI data are not perfectly correlated with ECS, particularly for language localization. Future investigations will likely establish the appropriate task paradigms required to maximize the accuracy of functional cortical maps generated by fMRI. Meanwhile, fMRI has already come into widespread use in neurosurgery, either independently or as an adjunct to ECS. The future of cortical neurosurgery is one in which all patients will routinely undergo fMRI as part of their presurgical evaluation.

References 1. Penfield W. The cerebral cortex and consciousness, in the Harvey lectures. Baltimore: Williams & Wilkins Co.; 1937. p. 35-69. 2. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in left, dominant hemisphere. An electrical stimulation mapping investigation in 117 patients. J Neurosurg 1989;71:31-326. 3. Tharin S, Golby A. Functional brain mapping and its applications to neurosurgery. Neurosurgery 2007;60(4 Suppl 2):185-201. 4. Cosgrove GR, Buchbinder BR, Jiang H. Functional magnetic resonance imaging for intracranial navigation. Neurosurg Clin N Am 1996;7:313-22. 5. Colebatch JG, Deiber M-P, Passingham RE, Friston KJ, Frackowiak KS. Regional cerebral blood flow during voluntary arm and hand movements in human subjects. J Neurophysiol 1991;65:1392-401. 6. Fox PT, Burton H, Raichle M. Mapping human somatosensory cortex with positron emission tomography. J Neurosurg 1997;67:34-43. 7. Kim PE, Singh M. Functional magnetic resonance imaging for brain mapping in neurosurgery. Neurosurg Focus 2003;15:1-7. 8. Belliveau JW, Kennedy DN, Jr, McKinstry RC, Buchbinder BR, Weisskoff RM, Cohen MS, et al. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 1991;254:716-19. 9. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci USA 1990;87:9868-72.


10. Ogawa S, Tank DW, Menon R, Ellermann JM, Kim SG, Merkle H, et al. Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci USA 1992;89:5951-5. 11. Fox PT, Raichle ME. Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc Natl Acad Sci USA 1986;83:1140-4. 12. Iadecola C. Regulation of the cerebral microcirculation during neural activity: is nitric oxide the missing link? Trends Neurosci 1993;16:206-14. 13. Bandettini PA, Wong C, Hinks RS, Tifosky RS, Hyde JS. Time course EPI of human brain function during task activation. Magn Reson Med 1992;25:390-7. 14. Rees G, Friston K, Koch C. A direct quantitative relationship between the functional properties of human and macaque V5. Nat Neurosci 2000;3:716-23. 15. Heeger DJ, Huk AC, Geisler WS, Albrecht DG. Spikes versus BOLD: what does neuroimaging tell us about neuronal activity? Nat Neurosci 2000;3:631-3. 16. Connelly A, Jackson GD, Frackowiak RS, Belliveau JW, Vargha-Khadem F, Gadian DG. Functional mapping of activated human primary cortex with a clinical MR imaging system. Radiology 1993;188:125-30. 17. Stehling MK, Turner R, Mansfield P. Echo-planar imaging: magnetic resonance imaging in a fraction of a second. Science 1991;254:43-50. 18. Cosgrove GR, Buchbinder BR. Functional MRI for planning cortical resections. In: Alexander E, Maciunas R, editors. Advanced neurosurgical navigation. New York: Thieme; 1997. 19. Roux FE, Ibarrrola D, Tremoulet M, Lazorthes Y, Henry P, Sol JC, et al. Technical and methodological issues for integrating functional MRI in a neuronavigational system. Neurosurgery 2001;49:1145-57. 20. Gore JC. Principles and practice of functional MRI of the human brain. J Clin Invest 2003;112:4-9. 21. Detre JA. fMRI: applications in epilepsy. Epilepsia2004;45 Suppl 4:26-31. 22. Lee CC, Ward HA, Sharbrough FW, Meyer FB, Marsh WR, Raffel C, et al. Assessment of functional MR imaging in neurosurgical planning. Am J Neuroradiol 1999;20: 1511-19. 23. Wilkinson ID, Romanowski CA, Jellinek DA, Morris J, Griffiths PD. Motor functional MRI for pre-operative and intraoperative neurosurgical guidance. Br J Radiol 2003;76:98-103. 24. Elster AD, Burdette JH. Questions and answers in magnetic resonance imaging. 2nd ed. St. Louis: Mosby; 2001. p. 273-80. 25. Binder JR, Swanson SJ, Hammeke TA, Morris GL, Mueller WM, Fischer M, et al. Determination of language dominance using functional MRI: a comparison with the Wada test. Neurology 1996;46:978-84.

20

26. Herholz K, Reulen HJ, von Stockhausen HM, Thiel A, Ilmberger J, Kessler J, et al. Preoperative activation and intraoperative stimulation of language-related areas in patients with glioma. Neurosurgery 1997;41:1253-60. 27. Hinke RM, Yu X, Stillman AE, Kim SG, Merkle H, Salmi R, et al. Functional magnetic resonance imaging of Broca’s area during internal speech. Neuroreport 1993;4:675-8. 28. Hirsch J, Ruge MI, Kim KH, Correa DD, Victor JD, Relkin NR, et al. An integrated functional magnetic resonance imaging procedure for preoperative mapping of cortical areas associated with tactile, motor, language, and visual functions. Neurosurgery 2000;47:711-21. 29. Ruge MI, Victor J, Hosain S, Correa DD, Relkin NR, Tabar V, et al. Concordance between functional magnetic resonance imaging and intraoperative language mapping. Stereotact Funct Neurosurg 1999;72:95-102. 30. Fandino J, Kollias SS, Wieser HG, Valavanis A, Yonekawa Y. Intraoperative validation of functional magnetic resonance imaging and cortical reorganization patterns in patients with brain tumors involving the primary motor cortex. J Neurosurg 1999;91:238-50. 31. Jack CR, Jr, Thompson RM, Butts RK, Sharbrough FW, Kelly PJ, Hanson DP, et al. Sensory motor cortex: correlation of presurgical mapping with functional MR imaging and invasive cortical mapping. Radiology 1994;190:85-92. 32. Lehe´ricy S, Duffau H, Cornu P, Capelle L, Pidoux B, Carpentier A, et al. Correspondence between functional magnetic resonance imaging somatotopy and individual brain anatomy of the central region: comparison with intraoperative stimulation in patients with brain tumors. J Neurosurg 2000;92:589-98. 33. Puce A, Constable RT, Luby ML, McCarthy G, Nobre AC, Spencer DD, et al. Functional magnetic resonance imaging of sensory and motor cortex: comparison with electrophysiological localization. J Neurosurg 1995;83: 262-70. 34. Yetkin FZ, Mueller WM, Morris GL, McAuliffe TL, Ulmer JL, Cox RW, et al. Functional MR activation correlated with intraoperative cortical mapping. Am J Neuroradiol 1997;18:1311-15. 35. Majos A, Tybor K, Stefanćzyk L, Go´raj B. Cortical mapping by functional magnetic resonance imaging in patients with brain tumors. Eur Radiol 2005;15:1148-58. 36. Roux FE, Boulanouar K, Ranjeva JP, Tremoulet M, Henry P, Manelfe C, et al. Usefulness of motor functional MRI correlated to cortical mapping in Rolandic low-grade astrocytomas. Acta Neurochir (Wien) 1999;141:71-9. 37. Yousry TA, Schmid UD, Jassoy AG, Schmidt D, Eisner WE, Reulen HJ, et al. Topography of the cortical motor hand area: prospective study with functional MR imaging and direct motor mapping at surgery. Radiology 1995;195:23-9. 38. Mueller WM, Yetkin FZ, Hammeke TA, Morris GL, III, Swanson SJ, Reichert K, et al. Functional magnetic

297

298

20 39.

40.

41.

42.

43.

44. 45.

46.

47.

48.

49.

50.

51.


resonance imaging mapping of the motor cortex in patients with cerebral tumors. Neurosurgery 1996;39:515-20. Schudler M, Maldijian JA, Liu WC. Functional imagedguided surgery of intracranial tumors located in or near the sensorimotor cortex. J Neurosurg 1998;83:412-18. Boling W, Parsons M, Kraszpulski M, Cantrell C, Puce A. Whole-hand sensorimotor area: cortical stimulation localization and correlation with functional magnetic resonance imaging. J Neurosurg 2008;108:491-500. Brannen JH, Badie B, Moritz CH, Quigley M, Meyerand ME, Haughton VM. Reliability of functional MR imaging with word-generation tasks for mapping Broca’s area. Am J Neuroradiol. 2001;22:1711-18. FitzGerald DB, Cosgrove GR, Ronner S, Jiang H, Buchbinder BR, Belliveau JW, et al. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. Am J Neuroradiol 1997;18:1529-39. Roux FE, Boulanouar K, Lotterie J-A, Mejdoubi M, LeSage JP, Berry I. Language functional magnetic resonance imagine in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery 2003;52:1335-45. McKhann GM, II, Hirsch J. Comment on reference 43. Neurosurgery 2003;52:1346. Petrovich N, Holodny AI, Tabar V, Correa DD, Hirsch J, Gutin PH, et al. Discordance between functional magnetic resonance imaging during silent speech tasks and intraoperative speech arrest. J Neurosurg 2005;103:267-74. Kombos T, Suess O, Kern BC, Funk T, Hoell T, Kopetsch O, et al. Comparison between monopolar and bipolar electrical stimulation of the motor cortex. Acta Neurochir (Wien) 1999;141:1295-301. Yoo SS, Talos IF, Golby AJ, Black PM, Panych LP. Evaluating requirements for spatial resolution of fMRI for neurosurgical planning. Hum Brain Mapp 2004;21:34-43. Sabbah P, Foehrenbach H, Dutertre G, Nioche C, DeDreuille O, Bellegou N, et al. Multimodal anatomic, functional, and metabolic brain imaging for tumor resection. Clin Imaging 2002;26:6-12. Archip N, Clatz O, Whalen S, Kacher D, Fedorov A, Kot A, et al. Non-rigid alignment of pre-operative MRI, fMRI, and DT-MRI with intra-operative MRI for enhanced visualization and navigation in image-guided neurosurgery. Neuroimage 2007;35:609-24. Gati JS, Menon RS, Ugurbil K, Rutt BK. Experimental determination of the BOLD field strength dependence in vessels and tissue. Magn Reson Med 1997;38: 296-302. Rugg MD, Johnson JD, Park H, Uncapher MR. Encoding-retrieval overlap in human episodic memory: a functional neuroimaging perspective. Prog Brain Res 2008;169:339-52.

Imaging in Stereotactic Surgery

17 General Imaging Modalities: Basic Principles A. A. Gorgulho . W. Ishida . A. A. F. De Salles

Historical Landmarks and Principles Although not based on imaging, the basic principle of stereotactic surgery started with the work of Zernov in 1890 [1]. A Russian anatomist, he developed a map of the brain cortex depicted in a hemisphere that, when attached to the human head, would keep a constant relationship with corresponding functional areas of the cortex. This instrument allowed placement of the craniotomy guided by the patient’s symptoms [2]. Further studies of the function of the central nervous system and symptoms the diseases required a more precise approach than the one devised by Zernov. Precise placement of recording and stimulating electrodes in specific areas of the brain to unveil function of deep structures called for a mathematical approach, Horsley and Clark devised and reported it in 1906 [3]. The Cartesian coordinate system, X (lateral), Y (anterior-posterior), Z (cranial-caudal), was born and remains the basis of stereotactic surgery (> Figure 17-1). If one reads the original work of Horsley and Clark, a striking finding is seen to the modern eyes; no imaging is mentioned for targeting the deep structures of the brain of the experimental animal [4]. As imaging was not readily available at the time, the skull landmarks were used as stereotactic reference. X-rays had just been described in Germany by Roentgen in 1895, only ten years before the seminal description of the stereotactic technique through the collaboration of the two English scientists, a surgeon and a physicist [3,4]. As the information age was not #


as fast pace as it is today, years were necessary for the incorporation of scientific accomplishments to surgery. Although stereotactic surgery continued to be largely employed in the laboratory, using the Horsley and Clark’s interaural line and midline as reference [5], these skull-based reference points were too variable to allow a safe determination of a target in the depth of the human brain [6]. Moreover, little was known about the function of the deep structures of the brain to allow intervention in humans. The natural path of animal experimentation was necessary for confirmation of the effects of lesioning of brain structures before one would propose interventions in humans. It was approximately 20 years after the initial studies of Horsley and Clark on the functional anatomy of the deep brain structures that the theory of basal ganglia motor integration was put forward by Spatz [7]. This theory spearheaded the first attempts of surgery in the extrapyramidal system to control movement disorder [8].

Ventriculography and the Stereotactic Landmarks Parallel to these animal laboratory experimentations, imaging of the brain was being developed. Plain skull x-rays were followed by the description of ventriculography by Dandy in 1918 [9] and by angiography in 1927 by Egas Muniz [10]. These two monumental imaging modalities would dominate the landscape of stereotactic surgery for the following 50 years, firmly

252

17

General imaging modalities: basic principles

. Figure 17-1 The Horsley and Clark stereotactic apparatus. Although the X, Y, Z coordinates’ convention changed over the years, modern convention is as follows: the ‘‘X’’, ‘‘Y’’, and ‘‘Z’’ are right-left, antero-posterior and cranio-caudal displacement from the stereotactic space center respectively. (These pictures are a courtesy of the historical collection at UCLA Medical Library. This is the second Horsley and Clark apparatus assembled in history)

establishing the association of imaging and stereotactic localization. Since then stereotactic surgery has developed in parallel with imaging techniques. Early in their work, Spiegel and Wycis realized the importance of the stereotactic technique for morphological and functional neurosurgical interventions [6]. They described the need for improved imaging for visualization of deep brain structures, and actually developed methods of determination of stereotactic coordinates based on the calcification of the pineal gland, lately based on pneumoencephalography. The posterior commissure-pons line served as reference for their measurements. These measurements were used mainly for functional stereotactic surgery. While the development of functional stereotactic surgery was rapid with the perfection of their stereotactic frame (> Figure 17-2), the morphological applications evolved slowly because the visualization of lesions in the brain became available only with the incorporation of angiography to the stereotactic technique. Although few neurosurgeons still use ventriculography for functional neurosurgery, its use is practically a historical legacy.

. Figure 17-2 Spiegel and Wycis stereotactic device constructed in 1954, available at University of California in Los Angeles. As in Figure 19-1 notice the Cartesian coordinates, X, Y, and Z applied to human stereotactic surgery


The main legacy of ventriculography and pneumoencephalography in stereotoactic surgery is the anterior-commissure (AC) and posteriorcommissure (PC) line. AC is seen approximately 2 mm below the posterior border of the foramen of Monro, while PC is seen just cranial to the entrance of the aqueduct of Sylvius and just caudal to the pineal calcification. These two anatomical landmarks, now readily visualized on all plans of the MRI, specifically seen in the sagittal and axial plan [> Figure 17-3], supported the development of the main atlases of the human brain, Shaltelbrand and Wahren [11] and Talairach’s proportional atlas of the human brain [12]. Specific targets in the brain are described based on the distance of the midcommissural plane which is the intersection of the

17

Cartesian coordinates dividing the brain into eight quadrants. Talairach used the length of the AC-PC line to develop the proportional atlas of the human brain widely used in epilepsy surgery.

Angiography Angiography paralleled the developments in stereotactic surgery. Stereotactic angiography was introduced by Talairach’s group [13]. Those workers dedicated several years of research to developing safe methods of inserting recording electrodes and radioisotope-loaded catheters and mapping the cortical anatomy by means of angiography. Their pioneer work using orthogonal

. Figure 17-3 T1 MRI axial and sagittal sections passing through AC-PC planes. (a) and (b) are axial AC-PC planes, in (a) without correction for AC-PC plane angle, notice the arrows showing AC (upper arrow) and PC (lower arrow). (b) shows the MRI precisely at AC-PC axial plane as reconstructed by the stereotactic software. (c) and (d) are sagittal AC-PC planes, in (c) without correction to the horizontal plane, notice the arrows AC (left arrow) and PC (right arrow). (d) shows AC-PC aligned to the horizontal plane by the stereotactic software

253

254

17


approach to avoid the cerebral vasculature in functional and tumor stereotaxis established the grounds for several groups to use stereotactic angiography [14]. Specially trained neurosurgeons dedicated to epilepsy surgery followed Talairach’s work. Techniques of angiography mapping of the brain promptly brought to computerized stereotaxis, initially through the use of superimposition techniques [15,16], later by digitization or scanning of angiographic films [17], and more recently in DICOM format, even with three dimensional angiography [18]. Talairach’s group also concentrated on the understanding of the three-dimensional (3D) characteristics of the cerebral vasculature and its relationship with cerebral tumors aiming to develop diagnostic and therapeutic approaches, either with precisely placed craniotomy or by the use of stereotactic-guided placement of isotopes [14,19]. Because of the inherent two-dimensional nature of angiography, they relied on stereoscopic techniques to obtain the 3D information. The knowledge developed with stereotactic angiography led to the treatment of arteriovenous malformation (AVMs) with single dose radiation [20]. Angiography was not widely married with the stereotactic technique to approach intracranial lesions until the application of stereotactic radiosurgery for AVMs was described in 1972 [20]. Talairach described the implantation of isotopes for treatment of subcortical tumors using the blush of the angiogram [14], and Leksell described the use of external beam radiation directed stereotactically to obliterate intracranial targets and coined the term ‘‘Radiosurgery’’ [21]. Diagnostic procedures using the stereotactic technique were initiated only a decade later, despite the poor and only indirect visualization of structural brain lesions [22,23]. The number of morphological procedures surpassed the functional applications of stereotactic surgery with the advent of computed-imaging techniques which allowed direct visualization of

the target [24]. Angiography sponsored the fast development of radiosurgery and approaches for determination of the seizure foci in epilepsy surgery.

Computerized Era When computed tomography (CT) scan became available to stereotactic surgery, approximately 30 years after the first human stereotactic proedure, stereotactic surgery had a second revolution [24]. Now lesions could be visualized and the risk of approaching highly vascularized lesions became measurable. Biopsies of brain tumors, brachytherapy and especially radiosurgery dominated the time of stereotactic surgeons during the 1980s and 1990s [25]. CT scan also brought back the interest of neurosurgeons to lesioning the depth of the brain for symptom control in neurodegenerative disease such as Parkinson’s disease, since the precision and the safety of the stereotactic method improved [26]. The functional stereotactic landmarks well seen in Dandy’s ventriculography, which served the bases for all the electrophysiological studies of the specialty, were now well seen with the CT scan. The 1980s saw the resurrection of lesioning in the brain as a therapeutic option. The ventriculographic approach was compared with the computerized approach and the computerized era for functional neurosurgery was established [27]. However it was the morphological application of the stereotactic method that spearheaded this revolution and extension of the technique to common place in the regular operating room of the general neurosurgeon [28]. This came with the progressive abandonment of the stereotactic frame for image-guided surgery using triangulation methods and explosion of stereotactic radiosurgery as a minimally invasive technique for treatment of brain tumors and arteriovenous malformations [29,30].


Computer Tomography Stereotactic Principles Hounsfield rightly received the Nobel Prize for medicine in 1979 for his description of X-Ray computed tomography (CT) in 1973 [31]. The imaging modality revolutionized neuroscience and the knowledge of brain pathology, function, and the ability of the stereotactic neurosurgeon to approach the brain safely. The technique was developed for visualization and not for precise calculation of intracranial targets. Therefore, years of work of stereotactic surgeons was necessary to bring this image safely into stereotactic surgery. As CT provided axial images only, stereotactic surgeons could determine only two coordinates from the slice of interest, i.e., where the target was located, either the pathology or the functional site to be targeted. By convention, the ‘‘X’’ became the lateral coordinate and the ‘‘Y’’ the antero-posterior. The vertical coordinate was not seen in the chosen slice, and the stereotactic surgeon had to devise methods of ‘‘Z’’ determination. Initially stereotactic surgeons relied upon the movement of the CT scan table to calculate the ‘‘Z’’, however the manufacturers of the CT scans were not worried about the precision of movement of the table, since the scanners were built for diagnosis. Frames were developed to overcome this imprecise movement of the table. The best example of such strategies is the Laitinen’s device which had transverse bars calibrated to correct the imprecision of the table movements [32]. It was not until the clever oblique bar introduced to a localizing box attached to the stereotactic frame by a graduate student at the University of Utah that the problem of the ‘‘Z’’ coordinate could be solved (> Figure 17-4) [33]. The stereotactic frame with the localization box became standard for all stereotactic procedures, including functional and morphological, from fine lesioning of pathways in the brain, to

17

. Figure 17-4 Axial representation of the fiduciary system with explanation of the oblique fiducial of the stereotactic localizing box. The Brown-Roberts-Wells (BRW) localizing box allows for three-dimensional definition of a point in any imaging slice (insert). The 9 points fiduciary system became widely used because of the possibility to correct for frame misaligment. The X and Y can be directly extracted from the axial slices, while the Z is calculated using the distance between the oblique bar to the reference bar in each slice

implanting electrodes to biopsies and craniotomy [16,34]. All commercially available stereotactic frames were adapted for the use of the oblique fiducials for determination of the ‘‘Z’’ (vertical) coordinate. The accuracy of the method was compared to the most used frames and shown to be submillimetric [35]. CT is considered up to now the most precise method for determination of stereotactic coordinates. The nature of X-rays with its rectilinear path avoids the introduction of distortions in the calculations. Distortions are introduced when using magnetic resonance imaging (MRI).

MRI Principles As the MRI scan became available and its geometric distortions were controlled [36–38], this imaging technique was preferred by stereotactic

255

256

17


surgeons [39,40]. Because it presents exquisite visualization of the nuances of the brain pathways and nuclei and the consequences of the surgery [41–43], it has revolutionized the approach of functional stereotactic surgery, no longer depending so much on ventricular landmarks, but relying on direct targeting of the structure needing functional modification [44]. Brain-function visualization is the next frontier on the development of stereotactic targeting. The incorporation of the chronic electrical stimulation as a therapeutic approach, initially for treatment of behavioral disorders, then for refractory chronic pain and movement disorders and more recently again for psychiatric disorders, has decreased the serious complications of the approaches in the depth of the brain. Progressively lesions of nuclei and pathways are being replaced by the ability of electrical stimulation to modify function by focally modulating neurons and brain networks. Functional imaging becomes more important for modulation of functional diseases of the brain, such as the neurodegenerative disease, genetic pathologies and brain damage by ischemic or traumatic injuries. Functional MRI and the ability to operate inside the magnet using frame [39,41] or frameless techniques brought new opportunities for functional neurosurgery [45–47]. In the arena of morphologic stereotaxis the revolution was on imaging localization of tumors and malformations in the brain needing intervention. Initially stereotactic surgery was used for simple needle biopsy [48], then to aid resections and guidance [16]. Here also functional imaging and the exclusive visualization of fibers related to lesions are revolutionizing surgical resections and targeting functional stereotactic surgery (> Figure 17-5) [49,50]. These important imaging developments are readily applicable to stereotactic radiosurgery (SRS), currently representing substantial, if not the major application of the stereotactic technique [51]. Initially SRS

was dependent on the stereotactic frame [52] and now, similar to surgical resection, it is becoming independent of the frame approach [53].

Positron Emission Tomography and Stereotactic Procedures CT and MRI scans sometimes do not adequately demonstrate the regions of interest for the stereotactic procedures. Molecular imaging, capable of demonstrating pathologies not seen in morphologic imaging can complement the needs of stereotactic surgery. Positron emission tomography (PET) adds this important metabolic information, and when incorporated by fusion with CT and MRI, may allow more accurate targeting and treatment planning in stereotactic radiosurgery, tumor resection, and biopsy.

PET in Morphological Stereotactic Surgery Most PET scans use a radiotracer made up of a common metabolite, such as glucose or an amino acid, attached to a radioisotope such as 18F (Fluorine) or 11C (Carbon). The 18F-FDG (fluoro-deoxy-glucose) PET is most widely used, and when combined with CTor MRI, will demonstrate with exquisite anatomic precision regions of increased glucose metabolism. Since neoplasms and inflammatory lesions often have high glucose uptake matching that of the brain, differentiation of a lesion from normal surrounding brain may be limited. The amino acids, however, are selectively more utilized by neoplasms than normal brain. The 18F-DOPA (fluoro-phenylalanine) and 11C (Carbon) methionine PET scans utilize amino acid molecules, and have demonstrated increased radioactivity in neoplasms, when compared to normal brain [54,55]. Extent of surgical resection or radiosurgery targeting will sometimes


17

. Figure 17-5 3-D frame and fiber tracking of the pyramidal system used for subthalamic nucleus targeting (traced arrow). Notice the distortions that can happen in fiducial system of the stereotactic localizing box (full arrows), reconstruction with iPlan software (BrainLab, Germany). This is a Leksell frame with copper sulfide liquid in the fiducial system (Elekta, Sweden)

be modified significantly by incorporation of PET on CT and MRI imaging [56–58]. Fluorodopa PET, C-methionine PET, and other amino acid-based PET scans have proven to be more effective than FDG-PET for imaging of neoplasms [54,59]. Fluorodopa and C-methionine PETscans demonstrated sensitivity to low grade as well as high grade tumors, and may help to differentiate areas of radiation necrosis (> Figure 17-6) [59,60]. PET scans sometimes demonstrate evidence of tumor recurrence before CT or MRI. They have proven to be most helpful in the management of gliomas [56–58], but also useful in treatment of other malignancies. It can be of help with pituitary adenomas [61], meningiomas, and parasellar lesions, where proximity to the cavernous sinus makes differentiation of tumor

from normally enhancing structures difficult. PET has proven to be helpful with spinal cord tumors [62,63], particularly in the presence of instrumentation or in the patient not able to tolerate MRI (pacemaker or electrical stimulator) in preparation for radiosurgery.

PET Scans in Functional Neurosurgery The PET characteristics of brain anatomy under normal and abnormal conditions have provided clues to a better understanding of brain anatomy and physiology. Changes on PET scanning related to deep brain stimulation have added to the still rudimentary body of information relating

257

258

17


. Figure 17-6 Fluorodopa-PET to differentiate tumor recurrence from radiation necrosis. Nonsmall cell carcinoma brain metastasis treated with radiosurgery using 16 Gy prescribed to the 90% isodose line. (a) T1 MRI with gadolinium showing lesion growth with central hypo-intensity, possibly radiation necrosis. (b) Fluorodopa PET performed one year after the treatment and at the same time of the MRI in (a). (c) Coronal PET showing higher uptake of Fluorodopa in the lesion (full arrow) than in the basal ganglia (traced arrow), consistent with tumor recurrence. Histology of the resected specimen showed nonsmall cell carcinoma with focal necrosis

to pain, movement disorders, and behavioral problems. The noninvasiveness of PET makes it a valuable research modality. Striatal as well as extra-striatal dopaminergic activity in neurological and psychiatric disorders have been studied using PET biomarkers [64]. The PET tracers are very important in advanced research on problems of early and more specific diagnosis of major movement and psychiatric disorders, physiology of the dopaminergic system, evolution of disease processes, and response to medications and surgical interventions [65–68]. FDG-PET studies in patients with major depression have demonstrated increased glucose metabolism in the left amygdala and frontal limbic pathways, with evidence of decreased amygdala metabolism under antidepressant drug treatment [69] and/or vagus nerve stimulation [70]. Similar PET responses have been reported with deep brain stimulation of the anterior limbic system, such as the subcallosal cingulate gyrus [71]. These findings are consistent with the findings of PET blood flow studies in depressed patients by Mayberg et al. [72,73]. Mayberg et al. used the findings of increased blood flow in the sub-genual cingulate cortex, area AcG25, to realize a target

for deep brain stimulation to control the symptoms of medically refractory major depression. Mayberg et al’s early work demonstrated the integral role played by the subgenual cingulate cortex in both, normal, and pathological shifts in mood [73]. Increases in limbic and paralimbic blood flow (as measured using PET) occur in the subgenual cingulate cortex and anterior insula during sadness. There is a significant inverse correlation between blood flow in the subgenual cingulate cortex and right dorsolateral prefrontal cortex [74]. A clinical response to antidepressants is associated with limbic and striatal (subgenual cingulate cortex, hippocampus, insula, and pallidum) decreases in metabolism and dorsal cortical (prefrontal, parietal, anterior, and posterior cingulate cortex) increases in metabolism [69,72]. In 2005, DBS electrodes were bilaterally implanted in the subgenual cingulate cortex [75] of 6 patients with medically refractory major depression. When stimulation was on, patients reported positive emotional phenomena. In the acute postoperative period the patients experienced reproducible increases in activity and mood scores, changes that failed to occur during sham stimulation. Chronic stimulation at high


frequency, probably leading to suppression of function in the site, resulted in significant response and remission of depression in 4 of the 6 patients at 6 months. These well conducted studies showed the effectiveness of PET findings to enhance the knowledge of brain function leading to diagnosis and therapeutic measures. Anterior capsule deep brain stimulation has resulted in decreased (18) FDG-PET activity or decreased glucose uptake in the subgenual anterior cingulate gyrus and ventral striatum in a group of patients with refractory obsessive compulsive disorder [76]. OCD patients with hoarding behavior showed different patterns of cortical PET- FDG activity, compared to those with nonhoarding behavior [77]. Evidence is accumulating to support the use of PET in routine target determination and follow up of patients undergoing neurosurgical interventions for mental illness. Advances in radioligand technology [78] have provided radioisotope labeled molecules that enable study of neurotransmitters (serotonin, norepinepherine, dopamine and glutamate) and their receptors with PET [79–82]. Patterns and intensity of uptake have added to our understanding of movement and psychological disorders. As an example fluorodopa-PET uptake in the putamen would be decreased in idiopathic Parkinson’s Disease as well as in a Parkinson’s-Plus condition. However, greater loss of striatal D2 receptors in Parkinson’s-Plus on a (11)C-raclopride PET scan might help to identify the patient as being unsuitable for surgical treatment with stereotactic implant of DBS, pallidal or thalamic lesion. Radioligands and PET have advanced the study of neuroreceptors remaining a most valuable tool for ongoing studies and treatment of patients with motor and psychological disorders [83].

Image Fusion The ability to bring multimodality imaging to plan stereotactic procedures started with the work of Talairach. He obtaining information

17

from angiography to avoid vasculature for seizure placement of electroencephalographic electrodes for seizure focus determination [13]. The attempt to bring MRI and CT into the stereotactic space determined by plain X-ray was started for stereotactic radiosurgery with fusion of imaging by photographic magnification manipulation [15]. These efforts have culminated with computerized fusion with software algorithms developed based on contours, image intensity, and voxel matching. These precise approaches have allowed gathering detailed information prior to the procedure, facilitating the surgery planning. Before these techniques were available, the patient had imaging with the stereotactic frame in place and multimodality approaches were unyielding. Now a portable CT scan in the operating room has obviated the need of elaborate stereotactic operating rooms. The fusion techniques offer also the opportunity of atlas information integration to the patient’s image (> Figure 17-7). Moreover, real time information on brain shift and possible complication during the operation are obtained while operating inside the magnet [39,41]. Fusion of multimodality images is very important for correction of distortions of PET, digital angiography and MRI scans. The portable CT in the operating room can offer this correction [38,84].

Image-Guided Surgery The integration of multimodality imaging is possible without the stereotactic frame [85–87]. This capability has revolutionized not only stereotactic surgery but also general neurosurgery. It is now impacting in other specialties such as radiation oncology, orthopedic surgery, head and neck surgery and general surgery. Modern imaging technology brings presurgical information to the surgeon that obviates unknowns. Computer technology, using this information, provides that surgery can be performed virtually on a screen before the patient is even touched. In addition, surgery has advanced to a level

259

260

17


. Figure 17-7 (A) Shaltenbrand-Wahren coronal plate. (B) Superimposition of the anatomy in (A) onto the corresponding MRI coronal slice by iPlan stereotactic software (BrainLab, Germany). Notice the adjusting sliding scale under (A) capable of matching the atlas with the MRI

where minimal invasion and maximal effectiveness is routine. The term ‘guided surgery’ in the modern sense, should be viewed as ‘modern surgery’. Guided surgery’, however, is still seen by many as the use of computerized imaging, or traditional X-ray-based stereotactic techniques described above to bring the surgeon precisely to the pathology being operated on. The pressures of competition and multimillion dollar malpractice law suits have driven the modern medical centers to invest heavily in technology. This in turn has driven the price of medical procedures to almost unacceptable levels. The

hope is that applied technology can decrease the costs of each patient treatment. Image guided surgery is an area that may lead to substantial savings in medical dollars. The scope of the approaches and the realistic surgical undertaking may lead to shorter hospital stays due to fewer complications related to extensive surgeries, less need for long convalescent and rehabilitation periods, and, consequently, a faster return of the patient to the workforce. Ultimately, this results in decreasing the overall price of medical care. This concept has been exemplified with complex skull base disease.


These difficult tumors are treated now with transnasal procedures for skull base tumor resections [88] under real time imaging in the operating room, and followed by radiosurgery, reducing patient recovery time, decreasing morbidity, and offering the patient complete control of their disease [89]. The stereotactic developments throughout the 20th century as described in this chapter, spearheaded by the computerized imaging, provided remarkable noninvasive imaging techniques developed primarily for diagnostic studies. These techniques were adapted for surgery guidance with navigation using triangulation techniques [85–87]. Now infrared reflectors or magnetic field are used for real-time localization [90]. Fast computers and smart software packages permitted the introduction of these images to the operating field to guide the surgeon. Digital fluoros copy, ultrasound, computer tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET) are now brought to the operating room and combined with merging data set techniques, allowing the surgeon to take advantage of a wealth of information that was previously unavailable. The surgeons of the past relied upon the principle of ‘‘exposure, exposure, exposure’’, and their individual knowledge of gross anatomy to perform surgery. The surgeons of the present rely on their knowledge of anatomy, anatomical imaging, and functional anatomy to perform minimally invasive procedures and solve previously unapproachable problems [91].

Spine Stereotactic Surgery Invasive stereotactic fixation for radiosurgery of the spine was previously tried without acceptance from the stereotactic community [96]. The procedure proved to be too invasive and impractical to be largely applied. The development of imageguided surgery, as described above, provided the

17

base for the development of the spine stereotactic technique. Image fusion and computerized image are now applied to stereotactic radiosurgery of spine lesions [91,94,95] and for placement of hardware. Completely noninvasive, fiducial systems use infrared triangulation and online image fusion of oblique X-rays and CT reconstruction to provide real-time movement tracking and targeting of lesions in the spine and surrounding regions [97,98]. The technology has reached precision to treat intramedullary lesions (> Figure 17-8).

Future Directions We are on the verge of perfecting real-time imaging in surgery [39,46]. During the past decade, the information brought during surgery by plain X-rays, fluoroscopy, and ultrasound was maximized and their limitations were established. Surgeons have now turned their eyes to the wealth of possibilities brought by portable CT scans and operating rooms equipped with interventional MRIs. MRI offers the possibility of not only exquisite anatomical information during surgery, but also dynamic changes of this anatomy associated with real-time changes in function. It also carries the advantage of not being harmful to the medical personnel, as are techniques dependent on isotopes or X-rays. The operating room with multimodality imaging, also known as operating room of the future, is a focus of studies in major medical centers. The logistics and real advantages of bringing a complex technology such as MRI to the operating room, or bringing the operating room to the complex MRI environment, has become a subject of symposiums on modern surgery [41]. The evolving field of functional MRI has brought the possibility of deciding before surgery the location of a fine function in the brain in relation to pathology (> Figure 17-9). It has also allowed relating the complex wiring of the brain to the location of ‘brain pace makers’ (> Figure 17-5

261

262

17


. Figure 17-8 Medullary AVM in a 22-year-old woman who bled, developed tetraplegia and recovered after a C2–C7 laminectomy. (a) Sagittal contrasted T1 MRI before radiosurgery. (b) Sagittal CT with the radiosurgery plan (12 Gy, 90% isodose line, 1.60 cc lesion). (c) Sagittal contrasted T1 MRI 24-months post-radiosurgery. (d) Anteroposterior angiogram before treatment, notice the angiographic stereotactic fiducials (traced arrow). (e) Coronal CT showing the radiosurgery plan. (f) Anteroposterior angiogram post-treatment. Full arrows point the initial nidus (a), residual at 24 months (c) and residual at 26 months (f)

and > Figure 17-10). All this information is readily related to imaging during surgery. Products that integrate information from multiple imaging sources with diffusion of fluids through tissues, such as brain parenchyma for delivery of drugs after resection, have started to appeart on the market. This is achieved with stereotactic precision. Similar information is being generated by therapeutic thermal application, electrical current, and radiation. Laser or radiofrequency ablation, electrical stimulation with smart pacemakers capable of receiving and analyzing physiological clues, and radiation delivery with

modulating capabilities are all novel approaches being developed [92,93]. The operating room of the future for ‘guided surgery’ and modern stereotactic surgery, requires real-time anatomical imaging technology related to function of the tissue under therapy at the moment of resection, lesioning or modulation [42,93]. This allows maximization of resection, drug infusion, electrical tissue influence, biopsy and the optimal use of radiation strategies to manipulate biological systems [92,101]. The patient should be least invaded and most helped by the modern stereotactic techniques.


17

. Figure 17-9 (a) Functional MRI showing Broca’s area (traced arrows) just posterior to an Arteriovenous Malformation (full arrows). The same AVM is shown on anteroposterior (b) and oblique (c) angiogram. The AVM removed through a craniotomy guided by stereotactic triangulation with the patient awake for complete speech preservation, lateral angiogram (d)

. Figure 17-10 MRI fiber tracking from the subgeneal area (AcG25) recognized as the stereotactic target for implantation of deep brain stimulation electrodes for treatment of medically refractory depression, Notice the virtual electrode placement (open arrows). Notice the fibers going to prefrontal and orbitofrontal cortex and cingulate fasciculus (full arrows) [49,50]. (a) shows 3-D, sagittal and coronal reconstructions with axial MRI passing through posterior commissure, (b) same reconstruction passing through anterior commissure (traced arrows)

263

264

17


References 1. Zernov DN. L’ence´phalome´tre. Rev Gen Clin Ther 1890;19:302. 2. Kandel EI, Schavinsky YV. Stereotaxic apparatus and operations in Russia in the 19th century. J Neurosurg 1972;37:407-411. 3. Clark RH, Horsley V. On a method of investigating the deep ganglia and tracts of the central nervous system (cerebellum). Brit Med Journal 1906;2:1799-1800. 4. Horsley V, Clark RH. The structure and function of the cerebellum examined by a new method. Brain 1908;31:45-125. 5. Sun B, De Salles AAF, Medin P, Solberg T, Hoebel B, Felder-Allen M, Wie C-W, Ackerman R. Reduction of hippocampal-kindled seizure activity in rats by stereotactic radiosurgery.Experimental Neurology. 1998;154:691-695. 6. Spiegel EA, Wycis HT. Stereoencephalotomy, Part I. New York: Grune & Stratton; 1952. 7. Pia HW. In memory of Hugo Spatz 1888–1969. Acta Neurochir (Wien) 1970;23(2):183-6. 8. Gildenberg PL. Evolution of Basal Ganglia Surgery for Movement Disorders. Stereotact Funct Neurosurg 2006;84:131-135. 9. Dandy WE. Ventnculography following the injection of air into the cerebral ventricles. Ann Surg 1918;68:5-11. 10. Moniz E, de Carvalho L, Almeida Lima P. Angiopneumographie. Presse Med 1931;39:996-999. 11. Schaltenbrand G, Wahren AE. Stereotaxy of the human brain. Stuttgart: Thieme; 1982. 12. Talairach J, David M, Tournoux P, Corredor H, Kvasina T. Atlas d’anatomie stereotaxique. Paris: Masson; 1957. 13. Szikla G, Bouvier G, Hori T, et al. Angiography of the human brain cortex. Berlin: Springer-Verlag; 1977. 14. Talairach J, Aboulker P, Ruggiero G, et al. Utilization del methode radiostereotaxic pour le traitement radioactive in situ des tumeurs cerebrales. Rev Neurol 1953;90:656-658. 15. De Salles AAF, Asfora WT, Abe M, Kjellberg RN. Transposition of target information from the magnetic and computed tomography scan images to conventional X-ray stereotactic space. Appl Neurophyisol 1987;50:23-32. 16. Kelly PJ. Tumor Stereotaxis. Philadelphia: Sounders; 1991. p. 108-113. 17. Zhang J, Levesque MF, Wilson CL, et al. Multimodality imaging of brain structures for stereotactic radiosurgery. Radiology 1990;175:435-441. 18. Pedroso AG, De Salles AA, Tajik K, Golish R, Smith Z, Frighetto L, Solberg T, Cabatan-Awang C, Selch MT. Novalis Shaped Beam Radiosurgery of arteriovenous malformations. J Neurosurg 2004;101 Suppl 3:425-34. 19. Bancoud J, Morel P, Talairach J, et al. Interet de l’exploration functionelle stereotaxique dans la localization des lesions expansives. Rev Neurol (Paris) 1961;105: 219-220.

20. Steiner L, Leksell L, Greitz T, et al. Stereotactic radiosugery for cerebral arteriovenous malforations: Report of a case. Acta Chir Scand 1972;138:459-462. 21. Leksell L. The stereotactic method of radiosurgery of the brain. Acta Chir Scand. 1951;102:316-319. 22. Housepian EM, Pool JL. Application of stereotaxic methods to histochemical, electromicroscopic and electrophysiological studies of bran subcortical structures. Confin Neurol 1962;22:173-177. 23. Kalynaraman S, Gillingham FG. Stereotaxic biopsy. J Neurosurg 1964;21:854-858. 24. Gildenberg PL. The history of stereotactic neurosurgery. Neurosurg Clin N Am 1990;1(4):765-80. 25. De Salles AAF, Goetsch SJ (eds.). Stereotactic Surgery and Radiosurgery. Madison, Wisconsin: Medical Physics Publishing Corporation; 1993. p. 1-16. 26. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomhy in the treatment of Parkinson’s disease. J Neurosurg 1992;76(1):53-61. 27. Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computerized tomography-guided determinations of brain target in functional stereotaxis. J Neurosurg 1990;73:565-571. 28. Lunsford LD. The genesis of neurosurgery and the evolution of the neurosurgical operative environment: Part 1–Prehistory to 2003. Neurosurgery 2003;52(6): 1512. 29. De Salles AAF, Lufkin R. Minimally Invasive Therapy of the Brain. Thieme Medical Publishers; pp 1–6, 1997. 30. Lunsford LD. Radiosurgery as a future part of neurosurgery. Mayo Clin Proc. 1999;74(1):101-3. 31. Hounsfield GN. Computerized transverse axial scanning (tomography). I. Description of system. Br J Radiol 1973;46:1016-1022. 32. Laitinen LV, Hariz MI. Multi-purpose stereoadapter. Appl Neurophysiol 1987;50(1–6):68-76. 33. Heilbrun MP, Roberts TS, Apuzzo ML, Wells TH Jr, Sabshin JK. Preliminary experience with BrownRoberts-Wells (BRW) computerized tomography stereotaxic guidance system. J Neurosurg 1983;59(2):217-22. 34. Hariz MI. Correlation between clinical outcome and size and site of the lesion in CT-guided thalamotomy and Pallidotomy. Stereotactic and Funct Neurosurg 1990;54 (55):172-185. 35. Maciunas RJ, Galloway RL Jr, Latimer J, et al. An independent application accuracy evaluation of stereotactic frame systems. Stereotactic and Funct Neurosurg 1992;58:103-107. 36. Maciunas RJ, Fitzpatrick JM, Gadamsetty S, Maurer CR Jr. A universal method for geometric correction of magnetic resonance images for stereotactic neurosurgery. Stereotact Funct Neurosurg 1996;66(1–3):137-40. 37. Burchiel KJ, Nguyen TT, Coombs BD, Szumoski J. MRI distortion and stereotactic neurosurgery using the Cosman-Roberts-Wells and Leksell frames. Stereotact Funct Neurosurg 1996;66(1–3):123-36.


38. Alexander E 3rd, Kooy HM, van Herk M, Schwartz M, Barnes PD, Tarbell N, Mulkern RV, Holupka EJ, Loeffler JS. Magnetic resonance image-directed stereotactic neurosurgery: use of image fusion with computerized tomography to enhance spatial accuracy. J Neurosurg 1995;83(2):271-6. 39. De Salles AAF, Frighetto L, Behnke E, Sinha S, Bronstein J, Torres R, Subramanian I, Cabatan-Awang C, Frysinger R. Functional Neurosurgery in the MRI Environment. Minim Invasive Neurosurg. 2004;47 (5):284-9. 40. Starr PA, Christine CW, Theodosopoulos PV, Lindsey N, Byrd D, Mosley A, Marks WJ Jr. Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imaging-verified lead locations. J Neurosurg 2002;97(2):370-87. 41. Lee WYM, De Salles AAF, Frighetto L, Torres R, Behnke E, Bronstein J. Imaging techniques and electrode fixation methods for deep brain stimulation: Intraoperative MRI 0.2 T, 1.5 T and fluoroscopy. Minimally Invasive Neurosurgery 2005;48:1-6. 42. De Salles AAF, Brekhus SD, De Souza EC, Behnke EJ, Farahani K, Anzai Y, Lufkin R. Early postoperative appearance of radiofrequency lesions on magnetic resonance imaging. Neurosurgery 1995;36 (May):932-936. 43. De Salles AAF. Role of stereotaxis in the treatment of cerebral palsy. J of Child Neurology 1996;11:S43-S50. 44. De Salles AAF, Hariz M. MRI Guided Pallidotomy. In: Rengachary SS, editor. Neurosurgical Operative Atlas, Volume 7, Williams & Wilkins, AANS; 1998. pp 141-148. 45. Martin AJ, Hall WA, Roark C, Starr PA, Larson PS, Truwit CL. Minimally invasive precision brain access using prospective stereotaxy and a trajectory guide. J Magn Reson Imaging 2008;27(4):737-43. 46. Larson PS, Richardson RM, Starr PA, Martin AJ. Magnetic resonance imaging of implanted deep brain stimulators: experience in a large series. Stereotact Funct Neurosurg 2008;86(2):92-100. 47. Holloway KL, Gaede SE, Starr PA, Rosenow JM, Ramakrishnan V, Henderson JM. Frameless stereotaxy using bone fiducial markers for deep brain stimulation. J Neurosurg 2005;103(3):404-13. 48. Hariz MI, Bergenheim AT, De Salles AAF, Rabow L, Trojanowski T. Percutaneous stereotactic brain tumor biopsy and cyst aspiration with a non-invasive frame. British Journal of Neurosurgery 1990;4:397-406. 49. Sedrak M, Gorgulho A, De Salles AF, Frew A, Behnke E, Ishida W, Klochkov T, Malkasian D. The role of modern imaging modalities on deep brain stimulation targeting for mental illness. Acta Neurochir Suppl 2008;101:3-7. 50. Hauptman JS, DeSalles AA, Espinoza R, Sedrak M, Ishida W. Potential surgical targets for deep brain stimulation in treatment-resistant depression. Neurosurg Focus. 2008;25(1):E3. 51. De Salles AA, Gorgulho AA, Selch M, De Marco J, Agazaryan N. Radiosurgery from the brain to the

52. 53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65. 66.

67.

17

spine: 20 years experience. Acta Neurochir Suppl 2008;101:163-8. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta Chir Scand 1951;102:316-319. Agazaryan N, Tenn SE, DeSalles AA, Selch MT. Imageguided radiosurgery for spinal tumors: methods, accuracy and patient intrafraction motion. Phys Med Biol 2008;21;53(6):1715-27. Becherer A, Karanikas G, Szabo M, et al. Brain tumour imaging with PET: a comparison between [18F] fluorodopa and [11C] methionine. Eur J Nucl Med Mol Imaging 2003;30:1561-1567. Chen W, Cloughesy T, Kamdar N, et al. Imaging Proliferation in Brain Tumors with 18F-FLT PET: Comparison with 18F-FDG. J Nucl Med 2005;46:945-952. Grosu A, Weber W, Astner S, et al. 11C-Methionine PET Improves the Target Volume Delineation of Meningiomas Treated with Stereotactic Fractionated Radiotherapy. Int J Radiat Oncol Biol Phys 2006;66:339-344. Pirotte B, Goldman S, Dewitte O, et al. Integrated positron emission and magnetic resonance imagingguided resection of brain tumors: a report of 103 consecutive procedures. J Neurosurg 2006;104:238-253. Singhal T, Narayanan T, Jain V, et al. 11C-L-Methionine Positron Emission Tomography in the Clinical Management of Cerebral Gliomas. Mol Imaging Biol. 2008;10 (1):1-18. Chen W, Silverman D, Delaloye S, et al. 18 F-DOPA PET Imaging of Brain Tumors: Comparison Study with 18FFDG PET and Evaluation of Diagnostic Accuracy. J Nucl Med 2006;47:904-911. Terakawa Y, Tsuyuguchi N, Iwai Y, et al. Diagnostic accuracy of 11C-Methionine PET for differentiation of recurrent brain tumors from radiation necrosis after radiotherapy. J Nucl Med 2008;49(5):694-9. Tang B, Levivier M, Heureux M, et al. 11C-Methionine PET for the diagnosis and management of recurrent pituitary adenomas. Eur J Nucl Med Mol Imaging 2006;33:169-178. Shimizu T, Saito N, Aihara M, et al. Primary Spinal Oligoastrocytoma: A Case Report. Surg Neurol 2004;61:77-81. Wilmshurst JM, Barrington SF, Pritchard D, et al. Positron emission tomography in imaging spinal cord tumors. J Child Neurol 2000;15(7):465-72. Elsinga PH, Hatano K, Ishiwata K. PET tracers for imaging of the dopaminergic system. Curr Med Chem 2006;13(18):2139-53. Berg D. Biomarkers for the early detection of Parkinson’s and Alzheimer’sdisease.Neurodegener Dis 2008;5(3–4):133-6. Broussolle E, Dentresangle C, Landais P, et al. The relation of putamen and caudate nucleus 18F-Dopa uptake to motor and cognitive performances in Parkinson’s disease. J Neurol Sci 1999;166(2):141-51. Koerts J, Leenders KL, Koning M, et al. Striatal dopaminergic activity (FDOPA-PET) associated with cognitive

265

266

17 68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.


items of a depression scale (MADRS) in Parkinson’s disease. Eur J Neurosci 2007;25(10):3132-6. Moore RY, Whone AL, Brooks DJ. Extrastriatal monoamine neuron function in Parkinson’s disease: an 18Fdopa PET study. Neurobiol Dis 2008;29(3):381-90. Drevets WC, Bogers W, Raichle ME. Functional anatomical correlates of antidepressant drug treatment assessed using PET measures of regional glucose metabolism. Eur Neuropsychopharmacol 2002;12(6):527-44. Pardo JV, Sheikh SA, Schwindt GC, et al. Chronic vagus nerve stimulation for treatment-resistant depression decreases resting ventromedial prefrontal glucose metabolism. Neuroimage 2008;42(2):879-89. Lozano AM, Mayberg HS, Giacobbe P, et al. Subcallosal cingulate gyrus deep brain stimulation for treatment of resistant depression. Biol Psychiatry 2008;15:64(6): 461-7. Mayberg HS, Brannan SK, Tekell JL, Silva JA, Mahurin RK, McGinnis S, et al. Regional metabolic effects of fluoxetine in major depression: serial changes and relationship to clinical response. Biol Psychiatry 2000;48:830-843. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA, et al. Reciprocal limbic-cortical function and negative mood: converging PET findings in depression and normal sadness. Am J Psychiatry 1999;156:675-682. Liotti M, Mayberg HS, Brannan SK, McGinnis S, Jerabek P, Fox PT. Differential limbic—cortical correlates of sadness and anxiety in healthy subjects: implications for affective disorders. Biol Psychiatry 2000;48:30-42. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, et al. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45:651-660. Van Laere K, Nuttin B, Gabriels L, et al. Metabolic imaging of anterior capsular stimulation in refractory obsessive-compulsive disorder: a key role for the subgenual anterior cingulated and ventral striatum. J Nucl Med 2006;47(5):740-7. Saxena S, Brody A, Maidment K, et al. Cerebral Glucose Metabolism in Obsessive-Compulsive Hoarding. Am J Psychiatry 2004;161:1038-1048. Schou M, Pike V, Halldin C. Development of radioligands for imaging of brain norepinephrine transporters in vivo with positron emission tomography. Curr Top Med Chem 2007;7(18):1806-16. Cannon DM, Ichise M, Rollis D, et al. Elevated serotonin transporter binding in major depressive disorder assessed using positron emission tomography and (11C)DASB; comparision with bipolar disorder. Biol Psychiatry 2007;62(8):870-7. Meyer JH. Imaging the serotonin transporter during major depressive disorder and antidepressant treatment. J Psychiatry Neurosci 2007;32(2):86-102.

81. Rauch SL, Dougherty DD, Malone D, et al. A functional neuroimaging investigation of deep brain stimulation in patients with obsessive-compulsive disorder. J Neurosurg 2006;104(4):558-65. 82. Yu M. Recent developments of the PET imaging agents for metabotropic glutamate receptor subtype 5. Curr Top Med Chem 2007;7(18):1800-5. 83. Heiss W and Herlolz K. Brain Receptor Imaging. J Nucl Med 2006;47:302-312. 84. Bezrukiy NV, De Salles AAF, Dahlbom M, DeMarco J, Selch M, Smathers J. ‘‘Multimodality Image Fusion for Stereotactic Radiosurgery Planning and Follow-Up’’ Scientific paper exhibit - 87th RSNA annual meeting, Chicago, IL. Radiology 221(P):223, 2001. 85. Maciunas RJ, Galloway RL Jr, Fitzpatrick JM, et al. A universal system for interactive image-directed neurosurgery. Stereotactic Funct Neurosurg 1992;58:108-113. 86. Roberts DW, Strohbehn JW, Hatch JF, et al. A frameless stereotaxic integration of computerized tomographyic imaging and the operating microscope. J Neurosurg 1986;65:545-549. 87. Barnett GH, Kormos DW, Steiner CP, et al. Use of frameless, armless stereotactic wand for brain tumor localization with 20d and 3-D neuroimaging. Neurosurgery 1993;33(4):674-678. 88. Dusick JR, Esposito F, Kelly DF, et al. The extended direct endonasal transsphenoidal approach for nonadenomatous suprasellar tumors. J Neurosurg 2005;102 (5):832-841. 89. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for treatment of cavernous sinus meningioma. Int J Radiation Oncology, Biol Phys 2004;59:101-111. 90. Lionberger DR, Weise J, Ho DM, Haddad JL. How does electromagnetic navigation stack up against infrared navigation in minimally invasive total knee arthroplasties? J Arthroplasty 2008;23(4):573-80. 91. De Salles AAF, Pedroso AG, Medin P, Agazaryan N, Solberg T, Cabatan-Awang C, Espinosa DM, Ford J, Selch MT. Novalis Shaped Beam and Intensity Modulated Radiosurgery and Stereotactic Radiotherapy for Spine Lesions. J Neurosurg 2004;101Suppl 3:435-40. 92. De Salles AAF, Melega WP, Lacan G, et al. Radiosurgery with a 3 mm collimator in the subthalamic nucleus and substantia nigra of the vervet monkey. J Neurosurg 2001;95:990-997. 93. Anzai Y, Lufkin R, De Salles AAF, et al. Radiofrequency ablation of brain tumors using MR guidance. Min Invas Ther & Allied Technol 1996;5:232-242. 94. Kim CW, Lee YP, Taylor W, Oygar A, Kim WK. Use of navigation-assisted fluoroscopy to decrease radiation exposure during minimally invasive spine surgery. Spine J 2008;8(4):584-90. 95. Villavicencio AT, Burneikiene S, Bulsara KR, Thramann JJ. Intraoperative three-dimensional fluoroscopy-based computerized tomography guidance for


percutaneous kyphoplasty. Neurosurg Focus. 2005;15: 18(3):e3. 96. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady JR. Preliminary clinical experience with linear acceleratorbased spinal stereotactic radiosurgery. Neurosurgery 1995;36(2):311-9. 97. Adler JR Jr. Image-guided frameless stereotactic radiosurgery. In: Maciunas RJ editor. Interactive Image-guided neurosurgery. Park Ridge, IL: American Association of Neurological Surgeons; 1993. pp. 81-89. 98. Ryu S, Jin R, Jin JY, Chen Q, Rock J, Anderson J, Movsas B. Pain control by image-guided radiosurgery for solitary spinal metastasis. J Pain Symptom Manage. 2008;35(3):292-8.

17

99. Gorgulho A, De Salles AA, Frighetto L, Behnke E. Incidence of hemorrhage associated with electrophysiological studies performed using macroelectrodes and microelectrodes in functional neurosurgery. J Neurosurgery 2005;102(5):888-96. 100. Gorgulho A, Juillard C, Uslan DZ, Pegues D, Tajik K, Aurasteh P, Behnke E, De Salles AAF. Infection following deep brain stimulator implantation performed in the conventional versus in the MRI operating room. J Neurosurgery, in press. 101. De Salles AAF, Bittar G. Thalamic pain syndrome: anatomical and metabolic correlation. Surg Neurol 1994;41:147-51.

267

24 Image Reconstruction and Fusion B. A. Kall

A wide variety of digital radiological images may be used in stereotactic and functional surgery. These data are generally computed tomography (CT) and variations of magnetic resonance imaging (MRI, MRA, MRV, fMRI etc). Stereotactic data may also be incorporated from other imaging sources like positron emission tomography (PET), single photon emission computed tomography (SPECT) and ultrasound. The variety of applicable imaging sources will continue to grow. Images are created on an image scanning device and transferred to an image-guided system or other off-line computer system for image processing and procedure planning. These systems decode the images, offer a variety of image processing techniques including image reconstruction of two-dimensional slices into a threedimensional volumes, multimodality (CT and MR) or monomodality (T1 MR to T2 MR) image registration and image fusion. Single imaging volumes and/or fused imaging volumes may then registered to the coordinate system of an image-guided device for the surgical procedure or intervention or used postoperatively to confirm an outcome. Registration and Fusion are two image processing techniques that are often used interchangeably, but are not the same. Image registration is a prerequisite for image fusion. Registration is the process of determining a spatial transformation between two coordinate systems. Each imaging scanner produces images in its own three-dimensional coordinate system. Spatial transformations may be calculated between two or more imaging volumes (image registration), between imaging volumes and a stereotactic

#


device (stereotactic registration) or between two or more individual imaging volumes and then to a stereotactic device (image registration followed by stereotactic registration). Image fusion is the process of merging or overlaying two or more registered imaging volumes into a hybrid image. The main purpose of nearly all stereotactic planning and intraoperative image-guidance systems is to decode and reconstruct a set of two-dimensional slices into a three-dimensional volume, transform coordinate spaces between one (or more) imaging database(s) and a stereotactic device to precisely aid the stereotactician in the planning and performance of the surgical procedure or treatment. This chapter will review various image reconstruction, registration and fusion techniques utilized in stereotactic and functional surgery.

Introduction to Three-dimensional Radiological Imaging A typical image-guided radiological image dataset is a collection of (contiguous) two-dimensional slices. A medical image scanner defines the location, size and orientation of each slice in its own coordinate system (S). Each slice is typically 512 columns (X) by 512 rows (Y), denoted as 512 512. Some scanners produce slices with a smaller or larger number of rows and columns like 256 256, 128 128 or 1024 1024. Each dot in a slice at a particular row, column intersection is known as a pixel (picture element) and has a two-dimensional size in X and Y measured in millimeters (mm).

336

24

Image reconstruction and fusion

Pixels do not have to be square (same size in X and Y), but usually are collected in that manner for stereotactic procedures. Each pixel is represented internally as a number which represents a gray scale intensity or components of RGB (red, green blue) color so it may be displayed on a computer graphic display. The intensity or color value in some manner is a function of the type of scanner on which the image was produced (e.g., CT intensities are represented in Houndsfield units which are a measure of the attenuation of X-rays). Each image slice (and therefore each pixel) has a location (Z) and thickness in the coordinate system of the scanning device (both usually represented in millimeters). Each component in a medical image is a threedimensional cube known as a voxel (volume element). Each voxel has an X, Y, and Z coordinate and three-dimensional size. The anatomical orientation of a slice is also defined by the scanner (e.g., transverse, sagittal, coronal, oblique). Most image-guided systems and image processing systems for image registration and fusion prefer an imaging series to be scanned as a contiguous set of slices with a homogenous voxel size. In order to facilitate the transfer of images from a scanner to an external device like an image-guided system, the DICOM (Digital Imaging and Communications in Medicine) standard is utilized (http://medical.nema.org). This transfer may be accomplished with tapes, disks or over a network. DICOM adds a header (collection of tags) onto a computer file containing each image which, in addition to containing information like demographics also contains all the relevant information describing how the images were acquired and spatial information to reconstruct them into three-dimensional volumes. An image-guided or image processing system extracts information from the DICOM headers to detect the orientation, coordinate system and voxel dimensions of the imaging series. The individual slices may then be stacked into a three-dimensional volume matrix. In most

circumstances, the volume is reconstructed into homogeneous three-dimensional voxels by interpolation.

Registration Registration is the process of finding a spatial transformation that maps points from one coordinate system (C1) to those in another coordinate system (C2). Image registration is the process of finding a spatial transformation (T) that maps all points in one image/volume to homologous points in another image/volume (> Figure 24-1).

Methods of Registration One image volume is referred to as the reference (or fixed) volume and a second image volume to be registered to the reference volume is generally referred to as the moving (or working) volume when performing image registration. The second volume may be resampled volumetrically to match the reference volume and overlaid or fused with the reference volume and in some instances resliced to match the original slices in the reference volume following registration. . Figure 24-1 Image registration involves determining a spatial transformation (T) that maps points in the coordinate system from image volume 1 (C1) to a matching point in the coordinate system of image volume 2 (C2)


Image registration algorithms are categorized by the type of transformation method employed. Most registration techniques used in stereotactic surgery are rigid (linear) transformations. Rigid registration assumes the images are isotropic in that the original images are not warped, skewed or distorted. A series of rotations and translations are used to calculate the overall mathematical transformation to spatially register two image datasets. Deformable (or elastic) transformations apply additional warping components that may be applied either globally or locally to subsets of the image volumes. Deformable image registration techniques are used more often for intrasubject image registrations as well as to study disease progression (e.g., tumor volume growth) or to register anatomy that moves (heart). Deformable transformations are a useful method to correct geometric distortions in the original imaging data [1]. This chapter will focus primarily on rigid transformation methods while a deformable transformation component may be introduced if the image data is warped or distorted.

Stereotactic Frame Registration The earliest form of multimodality image registration utilized stereotactic frames [2]. The mathematical properties of registering medical images to a stereotactic frame are covered in much detail elsewhere in this textbook. Briefly, a head frame is rigidly attached to a patients head and a localization device with know geometric properties is attached to the head frame when a radiological scan is acquired. The localizer deposits marks in the imaging data. A computer can then transpose every voxel in stereotactically collected image data into the coordinate system of the particular frame using the geometric properties of the localizer. Multiple scans from multiple modalities may be stereotactically collected.

24

In some systems, the headframe can be removed and precisely reapplied to collect data on different days [3]. Once more than one image volume is registered to the coordinate system of the frame, points, volumes and entire scans may be related to each other using the coordinate system of the stereotactic frame [4–7].

Point Matching Registration Point matching registration, also known as iterative closest point registration, is the simplest method of registering two image volumes without a stereotactic frame [8,9]. Neurosurgeons will recognize point matching registration as the manner by which matching points are registered between locations selected on images (anatomical, stickon or screwed in fiducials) and corresponding locations touched on the patient with an imageguided probe in the operating room. A number of N corresponding anatomical points are identified by the user in each imaging set. The point matching algorithm iteratively alters a spatial transformation by applying a stepwise pattern of translations and rotations. The updated transformation is applied to transform coordinate points from the second set and the sum of squares distance between matching points in the two data sets is recalculated. The algorithm iterates (repeats) for new combinations of rotations and translations until the sum of squares measure is minimized usually measured as a root mean square difference (RMS).

Surface/Edge Matching Registration Surface matching is similar in nature to point matching except that the sets of points are extracted from the surfaces or edges of structures in the image data. These surfaces and edges are commonly determined by thresholding segmentation techniques. Surfaces or edges that are

337

338

24


commonly utilized are the skin surface, cortical surface or edges of the ventricular system for cranial procedures or the surface of vertebra for spinal imaging. This method does not require a set of homologous points (> Figure 24-2).

Surface/edge matching may be performed by variations of point matching as well as algorithms generally knows as the ‘‘hat and head’’ method originally developed for registration of CT, MR, and PET Images [10–14]. Surfaces and edges of

. Figure 24-2 Surface-based image registration: (a) determination of surface in Ictal and Interictal SPECT, (b) determination of MRI cortical surface, (c) SPECT hyperperfusion fused to MRI


matching structures are extracted from the imaging data. One set is represented as a cloud of threedimensional points (‘‘hat’’) and the other as a stack of slices (‘‘head’’). A spatial transformation is iteratively calculated by transforming the hat points onto the head surface until the closest fit of the hat on the head is determined. The square of the distances from points on the hat and its corresponding closest point on the head drawn toward the centroid of the head model is the metric that is minimized. This type of surface registration has also been used previously in registering imaging data to an image guided system [14].

Voxel-based Similarity and Mutual Information Registration One drawback to point and surface matching registration is that manual interaction is necessary to identify either matching anatomical points for point matching registration or surfaces/edges using segmentation methods for surface-based registration methods. These can be both time consuming and are often not precisely reproducible. One user may select slightly different matching anatomical points or the threshold parameters utilized may alter the surfaces or edges extracted using segmentation techniques. Voxel-based registration methods involve calculating a spatial transformation by optimizing some particular measure that may be determined directly from the voxel values rather than geometric relationship in point matching and surface matching registration. Woods et al. [15,16] first proposed voxel similarity registration measures based on the premise that the relative grey values of similar tissues in different images would correspond. Voxel similarity measures were further refined by Hill et al. [17] to define a joint histogram approach that shows changes as the alignment converges so that when anatomical structures overlap the histogram shows clusters around

24

the corresponding grey levels of those structures. Dispersions in the clustering decrease as the alignment approaches convergence. Unfortunately these voxel based-techniques assume that the intensities of corresponding anatomy in corresponding image volumes are linearly related, which is not generally true for multimodality registration. Measuresofdispersioninthetwo-dimensional histograms were then proposed. Hill et al. [18] proposed the third order moment and is a measure of the skewness of the distribution while Collignon [19] and Strudeholme [20] proposed entropy, a measure from information theory [21] as a metric of registration. Entropy, in general, is a measure of dispersion of a probability dispersion. Mutual information registration methods, or relative entropy, were then proposed by Viola and Wells [22–25] as well as Collignon et al. [26–28] at about the same time. Briefly, mutual information (MI) is measure of ‘‘the statistical dependence between two random variables or the amount of information that one variable contains about the other’’ [27]. In medical imaging, one random variable is a voxel intensity in one image volume and the second random variable is a voxel intensity in a second image volume. When two image volumes are geometrically aligned, the mutual information of the image intensity values of corresponding voxels is maximized. This image registration method can be performed without human manual interaction and it does not require any linear relationship of the grey level intensities of matching anatomical structures. Mutual information registration is also implemented as an iterative algorithm. An initial transformation is usually calculated to automatically align the centroids of the imaging volumes. Rotations and translations are applied during each iteration (for rigid transformation), the measure of mutual information is recalculated and the iterations continue in a stepwise optimized manner until the mutual information measure is maximized.

339

340

24


Discussion A number of monomodality or multimodality images may be acquired for a stereotactic or functional procedure. These imaging modalities primarily include CT and variations of MR imaging, but may include a variety of other modalities. The number of image sources considered when planning and performing a stereotactic procedure will continue to grow as will the volume of information from each imaging source. Registering mono or multimodality image volumes to correlate points, volumes or entire imaging volumes and in some instances fusing multiple imaging modalities into a hybrid image volume will continue to be a valuable tool for the stereotactician when planning or performing a stereotactic procedure or confirming the outcome. Rigid spatial registration methods are typically used to perform mono and multimodality imaging registration in stereotactic and functional surgery. Deformable transformation parameters may be introduced if geometric distortion or warping of the imaging data is suspected. Stereotactic frame image registration may be the most accurate and failsafe, but cannot be applied retrospectively. Manual registration methods include point matching and surface alignment registration. Point matching registration is the quickest way to register imaging volumes, but is highly dependent on the selection of matching landmarks. Some imaging systems prefer matching points around the area of interest and some work better when the matching points are dispersed around a larger area of the imaging volume. A small change in a threshold value used for extracting surfaces and edges for surface matching registration may produce slightly differing results. More recently developed voxel-based, and specifically mutual information registration techniques, offer an automated method of image volume registration that requires no assumptions about the underlying characteristics of

the imaging intensity information of a specific imaging acquisition. It is imperative that the result of any image registration is validated and verified. All systems that offer image registration and fusion options offer various mechanism for validation and verification and should be used carefully and comprehensively. Registration algorithms may provide a numeric root mean square (RMS) error estimation or graphically depicted ‘‘zones of confidence’’ to estimate how well the algorithm believes the registration aligns. These numeric and graphical confidence methods should be considered secondary to a comprehensive visual review of the registration accuracy over the entire image volume. The moving image volume is usually resampled to match the orientation of the fixed image volume for validation and verification. Typical visual methods for image registration verification and verification include synchronized comparison of internal and external anatomical locations (> Figure 24-3a) or graphical overlay/fusion techniques such as blending (> Figure 24-3b). The two registered volumes may also be validated and verified by displaying one volume in grayscale and one in color and fusing the registered volumes (> Figure 24-3c). Small angular misregistrations may only become apparent by verifying a large number of both internal and external locations throughout the entire image volume (> Figure 24-4). Errors in image registration and fusion may also occur because of underlying problems in the original imaging data. Image registration methods work best with contiguous, thin slice imaging acquisitions. Patient movement during the scan may result in ghost artifacts in the images. Even with high quality control, medical image scanners, and in particular MR scanners may produce imaging containing distortions because of gradient field nonlinearities [29] that may acceptable for diagnostic purposes, but not for quantitative purposes like image registration and


24

. Figure 24-3 Validation and verification techniques: (a) reviewing corresponding points, (b) blending of CT with stereotactic frame and MR frameless, (c) fusing one volume in grayscale (MRI) and one in color (SPECT)

stereotactic surgery. Image registration methods used in stereotactic and functional surgery generally assume that the images are not warped, skewed or distorted.

A routine diagnostic scan performed at a typical institution may not be a good candidate for an image registration because of the relative thickness of each image. Radiological technicians

341

342

24


. Figure 24-4 Small angular misregistration apparent by comparing matching points on CT and MRI

may also not follow appropriately prescribed imaging protocols and by the time the surgeon may want to do an image registration, it may be too late to repeat the imaging study over again. Automated methods of image registration may occasionally not converge on the correct result because the algorithm was not optimized correctly or because the orientation the two original image volumes are too far apart from one another (e.g., neck flexion significantly different between two clinical imaging volumes which is analogous to a straight CT gantry tilt in one image volume and 10–20 gantry tilt in another). Image reconstruction, registration and fusion techniques perform best with high quality images as inputs.

Summary A variety of images from multiple sources may be considered by the stereotactic and functional surgeon for planning a procedure or confirming results. Images from monomodality or multimodalities may be spatially registered and in some instances fused to provide useful information to improve the ability for a positive outcome. Comprehensive verification and validation of manually

or automatically registered or fused volumes is a mandatory step in performing these techniques. When used appropriately, image reconstruction, registration and fusion techniques may be a powerful tool for the stereotactic and functional surgeon.

References 1. Archip N, Clatz O, Whalen S, DiMaio SP, Black PM, Jolesz FA, Golby A, Warfield SK. Compensation of geometric distortion effects on intraoperative magnetic resonance imaging for enhanced visualization. Oper Neurosurg 2008;62(3):209-16. 2. Goerss S, Kelly PJ, Kall B, Alker GJ, Jr. A computed tomographic stereotactic adaptation system. Neurosurgery 1982;10:375-9. 3. KellyPJ, Earnest R, IV, Kall BA, Goerss SJ, Scheithauer B. Surgical options for patients with deep-seated brain tumors: computer-assisted stereotactic biopsy. Mayo Clin Proc. 1985;60:223-9. 4. Kall BA, Kelly PJ, Goerss SJ, Earnest, F, IV. Crossregistration of points and lesion volumes from MR and CT. In: Proceedings of the seventh annual meeting of frontiers of engineering and computing in health care, Chicago; 1985, p. 692–5. 5. Kelly PJ, Kall B, Goerss S, Alker GJ, Jr. A method of CT-based stereotactic biopsy with arteriographic control. Neurosurgery 1984;14:172-7. 6. Kelly PJ, Kall BA, Goerss S. Computer-assisted stereotactic biopsy utilizing CT and digitized arteriographic data. Acta Neurochirurgica 1984; Suppl 33:233–5.


7. Kall, BA, Kelly PJ, Goerss SJ. Comprehensive Computerassisted data collection, treatment planning and interactive surgery. Proc SPIE Med Imaging 1987;767:509-14. 8. Besl PL, McKay ND. A method for registration of 3D shapes. IEEE Trans Pattern Anal Mach Intell 1992;14 (2):239-56. 9. Zhang ZY. Iterative point matching for registration of free-form curves and surfaces. Int J Comput Vis 1994;13(2):119-52. 10. Pelizzari CA, Chen GTY. Registration of multiple diagnostic imaging scans using surface fitting. In: The use of computers in radiation therapy. North-Holland: Elsevier; 1987, p. 437-40. 11. Pelizzari CA, Chen GTY, Spelbring DR, Weichselbaum RR, Chen CT. Accurate three-dimensional registration of CT, PET and MRI images of the brain. J Comput Assist Tomogr 1989;13:1, 20-26. 12. Pelizzari CA, Levin DN, Chen GTY, Chen CT. Image registration based on anatomical surface matching. In: Interactive image-guided neurosurgery. American Association of Neurological Surgeons, New York: Theime; 1993. p. 47-62. 13. Levin DL, Pelizzari CA, Chen GTY, Chen CT, Cooper MD. Retrospective geometric correlation of MR, CT and Pet images. Radiology 1988;169:817-23. 14. Kall BA. Computerized visualization techniques in stereotactic neurosurgery. In: Textbook of stereotactic and functional neurosurgery. New York: McGraw Hill; 1998. p. 351-5. 15. Woods. RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992;16(4):620-33. 16. Woods RP, Mazziotta JC, Cherry SR. MRI-PET registration with automated algorithm, J Comput Assist Tomogr 1993;17(4):536-46. 17. Hill DLG, Hawkes NA, Harrison NA, Ruff CF. A strategy for automated multimodality image registration incorporating anatomical knowledge and imager characteristics. In: Proceedings of the 13th international conference on information processing in medical imaging. LNCS, vol. 687, Berlin: Springer; 1993. p. 182-96. 18. Hill DLG, Studholme C, Hawkes DJ. Voxel similarity measures for automated image registration, visualization in biomedical computing. Proc SPIE 1994;2359:205-16.

24

19. Collignon A, Vandermeulen D, Suetens P, Marchial G. 3D multimodality medical image registration using feature space clustering. In: Proceedings of the first international conference on computer vision, virtual reality and robotics in medicine. LNCS, vol. 905, Berlin: Springer; 1995. p. 195-204. 20. Studholme S, Hill DLG, Hawkes DJ. Multiresolution voxel similarity measures for MR-PET registration. In: Information processing in medical imaging. Dordrecht: Kluwer; 1995. p. 287-98. 21. Cover TM, Thomas JA. Elements of information theory. New York: Wiley; 1991. 22. Wells WM, III, Viola P, Kikinis R. Multi-modal volume registration by maximization of mutual information. In: Medical robotics and computer assisted surgery. New York: Wiley; 1995. p. 55-62. 23. Viola P, Wells WM, III. Alignment by maximization of mutual information. In: Proceedings of the fifth international conference on computer vision (ICCV 95), Boston, MA. IEEE Computer Society Press; 1995. p. 16-23. 24. Viola P. Alignment by maximization of mutual information. Ph.D. Thesis, Massachusetts Institute of Technology, Boston, MA; 1995. 25. Wells WM, Viola P, Atsumi H, Nakajima S, Kikinis R. Multi-modal volume registration by maximization of mutual information. Med Image Anal 1996;1(1):35-51. 26. Collignon A. Multi-modality medical image registration by maximization of mutual information. Ph.D. Thesis, Catholic University of Leuven, Leuven, Belgium; 1998. 27. Collignon A, Maes F, Delaere D, Vandermeulen D, Suetens P, Marchial D. Automated multimodality image registration based on information theory. In: Information processing in medical imaging. Dordecht: Kluwer; 1995. p. 263-74. 28. Maes F, Collignon A, Vandermuelen D, Marchal G, Suetens P. Multimodality image registration by maximization of mutual information, IEEE Trans Med Imaging 1997;16(2):187-98. 29. Wang D, Strungell W, Cowin G, Doddrell DM, Slaughter R. Geometric distortion in clinical MRI systems: part I: evaluation using a 3D phantom. Magn Reson Imaging 2004;22(9):1211-21.

343

23 Neurophysiologic Mapping for Glioma Surgery: Preservation of Functional Areas R. M. Richardson . M. S. Berger

The resection of tumors located within or adjacent to eloquent cortical regions is often necessary to alleviate focal neurological deficits secondary to mass effect and increased intracranial pressure. In addition, there is growing evidence that greater extent of tumor resection correlates with increased time to tumor progression and overall survival for glioma patients [1,2–4]. The surgical aim is to achieve maximal tumor removal without producing permanent morbidity. Neurophysiological mapping of functional areas is critical for minimizing the morbidity associated with removing abnormal tissue from eloquent cortex. Here we present techniques for intraoperative cortical and subcortical stimulation mapping to maximize safe removal of tumors located in motor and language cortex. Cortical stimulation techniques have been adapted from the pioneering methods of Penfield and Boldrey [5], while localization of subcortical motor and sensory tracts was first described by Berger et al. [6].

Asleep Craniotomy with Motor Function Mapping

pathways, it is important to perform both cortical and subcortical stimulation mapping. Regardless of the degree of tumor infiltration, swelling, apparent necrosis, and gross distortion by the tumor mass, functional cortex and subcortical white matter may be located within the tumor itself or the adjacent infiltrated brain [7]. When using stimulation mapping methods to identify subcortical pathways, the surgeon is able to achieve an acceptable risk of permanent motor deficits in patients with gliomas that are within or adjacent to motor tracts.

Preoperative Neurological Evaluation Although motor mapping will often not be useful in patients with severe hemiparesis, if antigravity movement is present preoperatively it is usually possible to stimulate both cortical and subcortical motor pathways intraoperatively. In children younger than 6 years of age, who may have cortical electrical inexcitability, somatosensory evoked potentials must be available and used to identify the central sulcus via phase reversal.

Indications Preoperative Functional Imaging Hemispheric tumors located within or adjacent to rolandic cortex, supplementary motor area (SMA), corona radiata, internal capsule and uncinate fasciculus constitute the major indications for intraoperative motor function mapping. Due to the risk of damaging descending motor #


A volumetric MRI scan is obtained preoperatively for use with an intraoperative navigational system. The relation of the tumor to primary motor cortex is assessed by identifying the central or Rolandic sulcus on the rostral cuts of the axial T2-weighted

326

23

Neurophysiologic mapping for glioma surgery: preservation of functional areas

MRI scan. This landmark is always present, regardless of mass effect, and is a reliable marker for the motor strip located within the gyrus directly anterior to the sulcus. The motor cortex can be found on midsagittal cuts by following the cingulate sulcus posteriorly and superiorly to its termination, at which point the motor cortex is directly anterior to this sulcus. On far lateral images, the inferior to mid portion of the motor cortex is localized to a region bisected by a perpendicular line drawn from the posterior corner of the insular triangle. Each of these MRI landmarks is useful for determining the proximity of the lesion to the motor cortex preoperatively. Preoperative functional imaging of primary motor cortex is achieved by magnetic source imaging (MSI), in which the source localization of functional cortical areas by magnetoencephalography (MEG) is coregistered with an anatomic MRI scan [8,9]. MEG is used to detect the magnetic field associated with neuronal activity itself, rather than relying on an indirect correlate such as the hemodynamic response upon which fMRI is based. MEG is generated by dipole currents associated with dendritic excitatory and inhibitory postsynaptic potentials, which produce frequency specific oscillations whose rhythms change upon brain activation. In this way, the somatosensory cortex and motor cortex are reliably localized preoperatively (> Figure 23-1). Resecting brain tumors involves the risk of damaging the descending subcortical motor pathways. Diffusion tensor image (DTI) fiber tracking is a noninvasive MRI technique that can delineate the subcortical course of the motor pathway by modeling three-dimensional local water diffusion along axonal membranes. DTI is used to visualize descending motor pathways starting from a functional cortical site and extending through the corona radiata, posterior limb of the internal capsule and cerebral peduncle [10] (> Figure 23-2). Fiber tracts delineated using DTI can be used to identify the motor tract in deep white matter and define a safety margin around

. Figure 23-1 Magnetic source imaging (MSI) shows the primary motor cortex involving the right second digit. The white tract mesial to the motor peak depicts the subcortical pathway subserving the motor cortex

. Figure 23-2 A diffusion tensor image (DTI) shows the subcortical motor pathway within the cerebral peduncle and upper pons (corticospinal tract) in relationship to a cystic pilocytic tumor involving the upper brain stem. The DTI tract is depicted in white

the tract during tumor resection, a method that has been validated using intraoperative subcortical stimulation mapping of the motor tract and magnetic source imaging [11]. DTI tractography, however, may be limited by tract termination or deviation in regions of peritumoral vasogenic edema and therefore should be used in combination with intraoperative stimulation mapping.


Surgical Technique After the dura is opened, and the contralateral arm, leg, and face are uncovered to observe for movement, stimulation mapping begins with identification of the motor cortex. A bipolar electrode (5 mm spacing, 60 Hz, 1ms phase duration) is placed on the surface of the brain for 2–3 sec with a current amplitude between 2 and 16 mA. The motor strip is stimulated in the asleep patient with a starting current of 4 mA and increased in 2 mA increments until a motor response is visually identified. A current above 16 mA has never been necessary to evoke sensory or motor responses and should be avoided [12]. The current is reduced to 2 mA when stimulating the awake patient and is raised in 1 mA increments for eliciting responses from both the motor and sensory cortex. At this point, cold Ringer’s lactate solution should be available for immediate irrigation of the stimulated cortex should a focal motor seizure develop. This will abruptly stop the seizure activity originating from the irritated cortex, without using short-acting barbiturates. Multichannel electromyography recording may be used for greater sensitivity in detecting muscle movements, allowing the use of a lower stimulation current and decreasing the risk of stimulation-induced seizure activity. Stimulated brain sites are marked with sterile numbered tickets (> Figures 23-3a and b). First, the inferior aspect of the rolandic cortex is identified by eliciting responses in the face and hand. For leg motor cortex, a strip electrode may be inserted along the falx and stimulated using the same current applied to the lateral cortical surface to evoke leg motor movements. This maneuver is safe because of a lack of bridging veins between the falx and the leg motor cortex. Similarly, a subdural strip electrode may be used to locate the motor strip when the craniotomy is near but not overlying this cortex. Once the motor cortex is defined, the resection proceeds with identification of the descending

23

tracts using similar stimulation parameters. Functional white matter axons are depolarized using the same current parameters applied to the cortex (> Figures 23-4a and b). When movements or paresthesias are evoked, the resection should cease because of the close proximity of intact functional pathways (current spread with bipolar stimulation is 2–3 mm from the electrode contacts). Following completion of tumor resection, a final stimulation of previously identified cortical motor sites confirms that underlying functional tracts have been preserved, which is equally valuable in cases where subcortical responses are not obtained. The presence of intact cortical and subcortical motor pathways implies that any surgery-related deficit would likely be transient, with resolution in days to weeks. In the senior author’s experience with surgery for hemispheric gliomas within or adjacent to the rolandic cortex, patients whose subcortical pathways were identified with stimulation mapping were more prone to develop an additional (temporary or permanent) motor deficit than those in whom subcortical pathways could not be identified (27.5% vs. 13.1%) [23]. Although motor deficits that lasted more than 3 months occurred in 7.4% of the patients whose subcortical pathways were identified, compared to 2.1% of those without subcortical responses, very few patients have been left with a dense paresis using this method.

Awake Craniotomy with Language Function Mapping Indications Due to the discrete localization of essential language areas in the individual patient and great variation in their location across the population, an awake craniotomy with cortical stimulation mapping is indicated for any lesion involving the dominant temporal, mid- to posterior frontal, and mid- to anterior parietal lobes. Stimulation

327

328

23


. Figure 23-3 Intraoperative mapping of the motor cortex involving the hand and shoulder is shown (a). The tumor is seen on the sagittal MRI scan to be located within the supplementary motor area (b)

mapping has shown that multiple discrete areas in perisylvian cortex of the dominant hemisphere as essential for language functions, with separation of areas for different aspects of language, including naming in two languages, different semantic classes, naming compared to reading, and language from verbal memory [12]. Sites where stimulation repeatedly interferes with naming have been localized to focal areas of approximately 1 cm2 in the dominant hemisphere cortex, frequently with one such site in perisylvian inferior frontal cortex and several others in temporoparietal cortex [13]. The exact location of these sites in the language dominant hemisphere varies substantially across the patient

population, particularly in temporoparietal cortex, with random distribution within the temporal lobe and in the inferior posterior frontal and anterior inferior parietal lobes. Additionally, no specific region of the temporal lobe cortex may be found to be essential for language function in many patients (55% in the study referenced above). With regard to bilingual patients, cortical stimulation studies have demonstrated the existence of both shared and distinct languagespecific cortical centers. In one study of 25 bilingual patients, primary and secondary language representations were similar in total cortical extent, but differed in anatomical distribution [14].


23

. Figure 23-4 Enhancing tumor is seen within the thalamus distorting the internal capsule laterally (a). The intraoperative map is shown following resection of the thalamic tumor and identification of the subcortical motor pathways within the internal capsule involving the upper extremity (b)

Secondary sites were located exclusively in the posterior temporal and parietal regions, while sites for the primary language were found throughout the mapped region. Other studies in bilingual patients have demonstrated both common and separate cortical anomia areas for both languages in temporoparietal and frontal areas [15], anomia sites for the second acquired language that were always colocalized with anomia sites in the first acquired language [16], and anomia sites for each language that were

always distinct and separate [17]. For these reasons, it is advisable to individually speech map each language in which the patient is fluent.

Preoperative Neurological Evaluation Patients with dominant hemisphere tumors in close proximity to language sites are ideal candidates for an awake craniotomy. Note that patients

329

330

23


with significant vasogenic edema and mass effect from their tumor may not be candidates for an awake craniotomy because of the potential for cerebral herniation out of the dural opening. Despite the use of osmotic diuretics, awake patients are at risk for developing alterations in arterial CO2 which may compromise the safety of the planned craniotomy and tumor resection. Swelling, herniation, and contusion may occur, resulting in termination of the procedure. The left hemisphere is dominant for language in 85% of the population, with rightsided dominance present in 6% and bilateral representation present in 9%. Ninety-eight to ninety-nine percent of right-handed individuals have left-sided language dominance. Thus, a Wada (intracarotid amytal) test is used almost entirely for verifying cerebral dominance in lefthanded patients. Those patients who undergo intraoperative mapping for language sites should be preoperatively tested for language errors by presenting the individual with a series of visual slides with common objects and words to be named and read, respectively. After confirming that the face motor cortex and Broca’s area are functional by asking the patient to protrude the tongue and count to ten, the slides of common objects and words are shown. Patients must be able to name common objects with a baseline error rate lower than 25%, with each slide presented at least three times. In patients who have moderate to severe dysphasia in either comprehension or expression, successful language mapping will not be possible. Therefore, these patients may either be asleep during surgery, without any attempt to do more than an internal decompression, or be challenged with steroids and diuretics for 7–10 days and reevaluated regarding their baseline naming error rate. An alternative approach is to biopsy the tumor, confirm histopathology, and radiate the lesion to reduce its size or stabilize its growth, in hopes of producing functional recovery sufficient to allow for intraoperative mapping.

Preoperative Functional Imaging Literature describing the correlation between cortical sites identified with language fMRI and those found by direct cortical stimulation at the time of surgery show variable agreement in localization [18], and therefore fMRI is not routinely used for preoperative planning. DTI of the superior longitudinal, inferior fronto-occipital and uncinatus fasciculi, reconstructed from anatomical landmarks, have depicted tracts occurring mostly at the periphery in high-grade gliomas, but frequently located inside the tumor mass in low grade gliomas [20]. As we move into an era of functional imaging to preoperatively map descending language pathways, it becomes important to keep in mind that the pathways identified are purely anatomical, and may not reflect the true functionality of the axonal bundles identified. In the above study, there was high correlation (97%) between DTI fiber tracking and intraoperative subcortical stimulation, the combined use of which may decrease the duration of surgery, patient fatigue, and intraoperative seizures.

Surgical Technique The initial primary goal during an awake craniotomy is to have a cooperative patient for speech mapping purposes. It is imperative that the patient be kept comfortably sedated when mapping is not being performed. A propofol and remifentanil infusion is titrated for patient sedation during the incision and craniotomy. Once the bone flap is removed, the dura is infiltrated with local anesthetic along the middle meningeal artery. The dura should remain closed until the patient is awake and alert; otherwise, coughing and straining during emergence from propofol may cause the brain to herniate outward, especially if tumor edema and mass effect are present. All sedatives are then discontinued to restore the


patient to an awake, cooperative state. During cortical mapping of language function, no sedatives are administered. If seizures occur during cortical mapping and are not controlled with cold Ringer’s lactate solution, propofol can be given for seizure suppression. After the motor pathways have been identified, the electrocorticography (ECoG) equipment is placed on the field and attached to the cranium. ECoG is used to monitor for after discharges induced by bipolar electrode stimulation of the cortex (> Figure 23-5). The presence of after discharge potentials indicates that the stimulation current is too high and must be decreased by 1–2 mA until no after discharge potential is present following stimulation. Using the ideal stimulation current, object-naming slides are presented and changed every 4 sec, and the patient is expected to correctly name the object during stimulation mapping. The answers are carefully recorded, and each cortical site is checked three times to ensure that there is no anomic or dysnomic stimulation-induced error. All cortical sites essential for naming are marked on the surface of the brain with sterile numbered tickets. Additionally, the patient is asked to count from 1 to 50 while the stimulation probe is placed near the inferior aspect of the motor strip . Figure 23-5 Electrocorticography equipment is shown in place, and electrodes are recording during stimulation of the cortex for after-discharge potentials

23

to identify Broca’s area. Interruption of counting (complete speech arrest), without oropharyngeal movement, localizes Broca’s area. Speech arrest is usually localized to the area directly anterior to the face motor cortex within a few centimeters. On occasion, however, stimulation-induced speech arrest can be found anteriorly in the pars opercularis or above the face motor cortex in the inferior frontal gyrus. Throughout language mapping, ECoG is continuously monitored for after discharge spikes to alleviate the possibility that naming errors are caused by the propagated effects of current spread or ongoing cortical depolarization (> Figure 23-6a–d). Subcortical stimulation may also be used for detection of eloquent white matter bundles which are essential for language function [19]. Routine use of subcortical language site identification has been reported to result in the identification of language related cortical sites in 59% of patients [20]. In the group of patients in whom a subcortical language site was identified during resection, the likelihood of developing a permanent deficit was 3.8% (7% in patients with a preexisting language deficit), independent of histology and location. When no subcortical sites were found at the time of surgery, no permanent deficits were noted, indicating that when a subcortical response is reliably detected, resection must stop. Cortical stimulation studies have shown that the distance of the resection margin from the nearest language site is likely the most important variable in predicting improvement in preoperative language deficits, duration of postoperative language deficits, and permanence of postoperative language deficits [21]. If the distance of the resection margin from the nearest language site is >1 cm, significantly fewer permanent language deficits occur (> Figure 23-7).

Summary Identification of functional cortical areas in patients with brain tumors provides the neurosurgeon with

331

332

23


. Figure 23-6 Preoperative MRI scan showing a diffuse infiltrative lesion involving the suprasylvian region of the dominant hemisphere (a). Intraoperative map demonstrating the face motor cortex (#1 and #2) (b). Stimulation-induced speech arrest, i.e., Broca’s area, was not identified anywhere within this region. The post-resection intraoperative map shows preservation of the face motor cortex and resection of the inferior aspect of the somatosensory cortex and the frontal operculum, with sites marked for stimulation on the underlying insular cortex (c). Stimulationinduced anomia was seen at site #30 in the superior temporal gyrus. Postoperative MRI scan showing the resection cavity with the face motor cortex isolated within the resection cavity

. Figure 23-7 1 cm rule: Resection within 1 cm of an essential language site results in permanent deficits in a small percentage of cases. Resecting adjacent brain greater than 1 cm from an essential site will result in temporary deficits; however, no new deficits are seen past 3 months [21]


the ability to achieve aggressive resections while preserving neurological function. The localization of intracerebral tumors via mapping of functional cortical and subcortical tracts has become an important tool in the preoperative assessment of patients with intrinsic cerebral tumors. The advantages of combining functional imaging information with a surgical navigation system are optimized when combined with intraoperative cortical and subcortical mapping. In the future, based on recent ECoG data showing task-specific changes in the spatial pattern of neuronal oscillations [22], it may be possible to localize function by correlating alterations in these rhythms with various behavioral tasks administered at the time of surgery, rather than by directly stimulating cortical tissue.

References 1. Berger MS, Deliganis AV, Dobbins J, Keles GE. The effect of extent of resection on recurrence in patients with low grade cerebral hemisphere gliomas. Cancer 1994;74:1784-91. 2. Keles GE, Anderson B, Berger MS. The effect of extent of resection on time to tumor progression and survival in patients with glioblastoma multiforme of the cerebral hemisphere. Surg Neurol 1999;52:371-9. 3. Keles GE, Lamborn KR, Berger MS. Low-grade hemispheric gliomas in adults: a critical review of extent of resection as a factor influencing outcome. J Neurosurg 2001;95:735-45. 4. Lacroix M, Abi-Said D, Fourney DR, Gokaslan ZL, Shi W, DeMonte F, Lang FF, McCutcheon IE, Hassenbusch SJ, Holland E, Hess K, Michael C, Miller D, Sawaya RA. multivariate analysis of 416 patients with glioblastoma multiforme: prognosis, extent of resection, and survival. J Neurosurg 2001;95:190-8. 5. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937;60:389-443. 6. Berger MS, Kincaid J, Ojemann GA, Lettich E. Brain mapping techniques to maximize resection, safety, and seizure control in children with brain tumors. Neurosurgery 1989;25:786-92. 7. Skirboll SS, Ojemann GA, Berger MS, Lettich E, Winn HR. Functional cortex and subcortical white matter located within gliomas. Neurosurgery 1996;38:678-84; discussion 675-84. 8. Benzel EC, Lewine JD, Bucholz RD, Orrison WW Jr. Magnetic source imaging: a review of the Magnes system

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

23

of biomagnetic technologies incorporated. Neurosurgery 1993;33:252-9. Gallen CC, Sobel DF, Waltz T, Aung M, Copeland B, Schwartz BJ, Hirschkoff EC, Bloom FE. Noninvasive presurgical neuromagnetic mapping of somatosensory cortex. Neurosurgery 1993;33:260-8; discussion 268. Berman JI, Berger MS, Mukherjee P, Henry RG. Diffusion-tensor imaging-guided tracking of fibers of the pyramidal tract combined with intraoperative cortical stimulation mapping in patients with gliomas. J Neurosurg 2004;101:66-72. Berman JI, Berger MS, Chung SW, Nagarajan SS, Henry RG. Accuracy of diffusion tensor magnetic resonance imaging tractography assessed using intraoperative subcortical stimulation mapping and magnetic source imaging. J Neurosurg 2007;107:488-94. Ojemann GA. The neurobiology of language and verbal memory: observations from awake neurosurgery. Int J Psychophysiol 2003;48:141-6. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in left, dominant hemisphere. An electrical stimulation mapping investigation in 117 patients. J Neurosurg 1989;71:316-26. Lucas TH II, McKhann GM II, Ojemann GA. Functional separation of languages in the bilingual brain: a comparison of electrical stimulation language mapping in 25 bilingual patients and 117 monolingual control patients. J Neurosurg 2004;101:449-57. Roux FE, Tremoulet M. Organization of language areas in bilingual patients: a cortical stimulation study. J Neurosurg 2002;97:857-64. Walker JA, Quinones-Hinojosa A, Berger MS: Intraoperative speech mapping in 17 bilingual patients undergoing resection of a mass lesion. Neurosurgery 2004;54:113-17; discussion 118. Bello L, Acerbi F, Giussani C, Baratta P, Taccone P, Songa V, Fava M, Stocchetti N, Papagno C, Gaini SM. Intraoperative language localization in multilingual patients with gliomas. Neurosurgery 2006;59:115-25; discussion 115-25. Roux FE, Boulanouar K, Lotterie JA, Mejdoubi M, LeSage JP, Berry I. Language functional magnetic resonance imaging in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery 2003;52:1335-45; discussion 1337-45. Duffau H, Capelle L, Sichez N, Denvil D, Lopes M, Sichez JP, Bitar A, Fohanno D. Intraoperative mapping of the subcortical language pathways using direct stimulations. An anatomo-functional study. Brain 2002; 125:199-214. Bello L, Gallucci M, Fava M, Carrabba G, Giussani C, Acerbi F, Baratta P, Songa V, Conte V, Branca V, Stocchetti N, Papagno C, Gaini SM. Intraoperative subcortical language tract mapping guides surgical removal of gliomas involving speech areas. Neurosurgery 2007;60:67-80; discussion 62-80.

333

334

23


21. Haglund MM, Berger MS, Shamseldin M, Lettich E, Ojemann GA. Cortical localization of temporal lobe language sites in patients with gliomas. Neurosurgery 1994;34:567-76; discussion 576. 22. Canolty RT, Edwards E, Dalal SS, Soltani M, Nagarajan SS, Kirsch HE, Berger MS, Barbaro NM, Knight RT. High gamma power is phase-locked to theta oscillations in human neocortex. Science 2006;313:1626-8.

23. Keles GE, Lundin DA, Lamborn KR, Chang EF, Ojemann G, Berger MS. Intraoperative subcortical stimulation mapping for hemispherical perirolandic gliomas located within or adjacent to the descending motor pathways: evaluation of morbidity and assessment of functional outcome in 294 patients. Journal of neurosurgery 2004;100(3):369-75.

28 Accuracy in Stereotactic and Image Guidance A. Hartov . D. W. Roberts

Stereotactic surgery has traditionally been defined by its unique mechanical instruments, by a subset of neurological and neurosurgical conditions, and by an expectation of accuracy and precision, but at its most fundamental level there is the underlying concept of coregistration. It is this ability to correlate an atlas or imaging study with the surgical field that enables the highly accurate, reliable, and safe procedures of the field today. This chapter will review the mathematical methodology underlying this process and the type of accuracy that can be achieved.

Methods of Stereotactic Registration A Cartesian coordinate system consists of a reference point, the origin, and orthogonal reference directions. One can express the location of any points in such a system in reference to the origin and along the reference directions by using the distances one travels in each to reach the point in question. The same point while fixed in absolute space can be expressed in any number of coordinate systems or frames of reference (frame for short). The coordinates of that point will be a function of the chosen frame. It is sometime useful to be able to express the coordinates of a point in a given frame F1 in those of a different frame F2. This is known as transforming the point from frame F1 to F2 and can be done by applying a translation and rotation to the given point: #


2

xF2

3

2

r11 r12 r13

32

xF1

3

2

Dx

3

7 6 76 7 6 7 6 4 yF2 5 ¼ 4 r21 r22 r23 54 yF1 5 þ 4 Dy 5 Dz zF2 r31 r32 r33 zF1 or PF 2 ¼ R PF 1 þ T

ð1Þ

For computational efficiency it is desirable to express points in a homogeneous coordinate system as 4-tuples with the last vector component 1 and simplify the combined rotation and translation operations into one matrix multiplication: 3 2 3 2 3 2 xF 1 r11 r12 r13 Dx xF 2 6 y 7 6 r r r Dy 7 6 y 7 7 6 F1 7 6 F 2 7 6 21 22 23 76 7¼6 7 6 4 zF 2 5 4 r31 r32 r33 Dz 5 4 zF 1 5 1

0

0

0

1

or PF 2 ¼ F 2 TF 1 PF1

1 ð2Þ

The two methods are exactly equivalent. Here we use a notational convention in which one can express the coordinates of a point P which is given in F1, hence PF1, in F2 using a transformation T which takes a point from F1 to F2, hence the notation F2 TF1 . This combined operation for rotation and translation is known as an affine transformation. A transformation consisting only of pure rotation and pure translation is said to be rigid because it maintains the relative position and orientation of points or objects transformed. The principal advantage of using affine transformations is that they can be combined by successive multiplications of the transformation matrices. There are 16 elements to a transformation matrix, but only 6independent parameters: ðDx; Dy; DzÞ and yx ; yy ; yz .

444

28

Accuracy in stereotactic and image guidance

Translation is expressed in the last the rotation terms have the form: 2 1 0 0 6 0 cosðy Þ sinðy Þ x x 6 Rx ¼ 6 4 0 sinðyx Þ cosðyx Þ 2

0 0 cosðyy Þ 0

6 0 6 Ry ¼ 6 4 sinðyy Þ 2

0 sinðyy Þ

1

0

0

cosðyy Þ

column and 0

3

07 7 7; 05

1 3 0 07 7 7; 05

0 0 0 cosðyz Þ sinðyz Þ

0

cosðyz Þ

0

0

1

1 3 0 07 7 7; 05

0

0

1

6 sinðy Þ z 6 Rz ¼ 6 4 0 0

ð3Þ

where yx ; yy ; yz are rotation angles about the respective axes. Any arbitrary rigid transformation can thus be expressed as a series of rotations about the z, x and y axes, followed by a translation: 3 2 1 0 0 tx 60 1 0 t 7 y7 6 tr TUS ¼6 7 4 0 0 1 tz 5 2

0 1

60 6 6 40 2

0 0 1 0 cosðyx Þ sinðyx Þ

0 0 cosðyy Þ 0

6 0 6 6 4 sinðyy Þ 2

1 0

0

0

3

07 7 7 cosðyx Þ 0 5 sinðyx Þ

0 1 3 sinðyy Þ 0 0 07 7 7 cosðyy Þ 0 5

0 0 0 cosðyz Þ sinðyz Þ

6 sinðy Þ z 6 6 4 0

0

0

1 3 0 0 0 07 7 7 1 05

0

0 1

cosðyz Þ

One can say of two frames of reference for which a transformation mapping points from one to the other is known that they are registered. In the following we present in greater detail methods used to compute the transformation between frames of reference, in the context of neuronavigation. We call such procedures co-registration. In the situations of interest to this discussion, we have a patient’s head attached rigidly to a frame which defines a coordinate system. A second coordinate system is defined by a preoperative imaging study, MRI or CT, which also represents the patient’s head. Although the two do not co-exist in a common real space, it is possible to co-register these two spaces using various techniques we will explore in the following.

Point Based Co-registration We assume that we have two sets of homologous points defined in two frames of references. In practice this can be done by using anatomical references, fiducial markers attached to the skin or bone implanted markers. The markers are identified in MRI/CT coordinates visually, and their position in operating room (OR) space can be recorded with a 3D tracker. This set of homologous points can be paired and the transformation matrix that best maps one set onto the other can be computed using several techniques. Given N point pairs, we can state the problem as follows: 2

xF2;1 xF 2;2 ::: xF 2;N

3

2

r11 r12 r13 Dx

3

7 7 6 6 6 yF2;1 yF 2;2 ::: yF 2;N 7 6 r21 r22 r23 Dy 7 7 7¼6 6 7 7 6 6z 4 F2;1 zF 2;2 ::: zF 2;N 5 4 r31 r32 r33 Dz 5 ð4Þ

Based on this definition, it is easy to see that to reverse a transformation one needs only take its inverse.

2

1

1 :::

1

3

0

0

0

1

xF1;1 xF 1;2 ::: xF 1;N 7 6 6 yF1;1 yF 1;2 ::: yF 1;N 7 7; or Pa ¼ F2 TF1 Pa 6 7 6z F2 F1 4 F1;1 zF 1;2 ::: zF 1;N 5 1

1 :::

1

ð5Þ


in which we use the a superscript to indicate the augmented matrices of points. There are 12 unknowns in this system of equations, the elements of the rotation sub matrix and the translation components. With 4 points, we could write 12 equations (the last row of this system of equations provides no information) and in principle have enough information to solve for all the unknowns. A simple method to solve this set of equations would be to solve the least squares problem using the pseudoinverse: 1 PaF2 ðPaF 1 ÞT PaF1 ðPaF 1 ÞT ¼ F2 TF 1 ð6Þ While this computation could work in well behaved cases, it often fails to produce a rigid transformation matrix, i.e., a matrix of pure rotation and translation, as described above. Formally, this solution does not enforce the orthonormality of the transformation matrix. The reason for this is that all the elements in the rotation sub matrix are treated as independent variables, while we have seen in equation (4) that they are related and that there are only 3 independent variables. A more reliable method of calculation is available based on the singular value decomposition of the de-meaned point coordinate matrices [1]. This approach minimizes the fiducial registration error (FRE) [2], the RMS distance between point pairs and it enforces orthonormality. The method can be summarized as follows, starting with N points expressed in 3x1 vectors: 3 2 xF1;1 .. . xF1;N 7 6 PF 1 ¼4 yF1;1 .. . yF1;N 5; zF1;1 .. . zF1;N 2 3 xF2;1 .. . xF2;N 6 7 PF 2 ¼4 yF2;1 .. . yF2;N 5; homologous points; zF2;1 .. . zF2;N

28

2

3 2 3 xF 1 xF2 mF 1 ¼ 4 yF 1 5; mF 2 ¼ 4 yF2 5; coordinate means zF 1 zF 2

Compute QF 1 ¼ PF 1 1N mTF1 ; QF 2 ¼ PF2 1N mTF 2 , the de-meaned coordinate vectors obtained with the use of the outer products of a Nx1 ‘‘1’’ vector and the means. Then compute H ¼ QF1 QTF 2 , the outer product of the de-meaned coordinate vectors, from which the 3x3 rotation matrix R is obtained: ½U S V ¼ svdðHÞ; R ¼ V U T using the singular values decomposition of the H matrix. At this stage, it is necessary to check whether the determinant of R is positive, if not the last column of V should be negated and R recomputed. The translation vector corresponds to the difference in the means, expressed in the same frame of reference: t ¼ mF 2 R mF1 , from this we can form the complete transformation matrix: 3 2 F2

6 TF1 ¼ 6 4

Rð33Þ 0

0

tð31Þ 7 7 5 0 1

This basic algorithm produces the best results and is the most reliable, in large part due to the robustness of the singular value decomposition algorithm. A different formulation of this algorithm, albeit exactly equivalent, and based on quaternions is given in [3]. A small but practical improvement can be made to this algorithm, which has to do with the need to correctly pair points in the two frames of reference. In practice, during surgery for example, it is often the case that a larger number of fiducial markers will be placed on the patient prior to preoperative imaging, knowing that not all may be reachable once in the operating room. This leads to the need for some labeling of the markers and some record keeping during the registration procedure, which can be a source of mistakes and result in a botched co-registration procedure.

445

446

28


We have devised a method based on the genetic algorithm which will compute the transformation based on randomly selected and randomly paired points, until a ‘‘good’’ result is obtained. This method relies to some extent on a non-symmetric arrangement of the markers, which otherwise could results in registrations seeming correct while rotated by a large angle [4]. Another approach that bypasses the need to match point pairs is to use the iterative closest point algorithm [3], but one must contend with the fact that it can settle on local minima in its optimization procedure. In order to discuss the accuracy of registration methods, it is useful to introduce specific measures of error. An extensive discussion of errors in co-registration can be found in [5]. We have alluded to one already, the fiducial registration error, or FRE. It is defined as the RMS distance between corresponding registration fiducials once the two sets have been merged into the same frame of reference. This is a useful measure in that it is easily obtainable anytime one computes the co-registration. It is also readily verifiable in laboratory tests. The fiducial location error (FLE) is a measure of the quality of the instrumentation, rather than that of coregistration. It is the error distance in the actual location of a fiducial and its location as reported by either the 3D tracking system or as it is obtained from an imaging study. This error will have an impact on the accuracy of the coregistration, although it is not possible to keep track of its effect outside of carefully conducted laboratory experiments. From the surgeon’s point of view the most useful error metric would be the target registration error (TRE), however it is not readily measured under practical OR conditions. The TRE is the error distance between the actual location of a point of interest, the target, which could be the tumor center for example, and its location following co-registration. This is again a difficult quantity to evaluate since a priori it is not possible to point to the tumor center prior to surgery and see how far off it is from its predicted location. One can compute an estimate for it,

based on one’s knowledge of the relative location of the fiducial marker in relation to the point of interest. It is in fact possible to produce an expected distribution of the TRE, given a setup and use this to gauge the quality of a coregistration. An extensive discussion of TRE and associated quantities is given in [5]. As we suggested in the preceding discussion, the error in point-based registration (FRE or TRE) will be a function of how accurately we can locate the points, in other words a function of FLE. The different methods that have been presented over time are known to produce more or less error, based on how well one can identify the registration points. The use of anatomical landmarks has proved to be the least accurate method because of the ambiguity inherent in identifying them on patients. The ambiguity translates into large FLEs in both frames of reference, that is on the patient when in the OR and when identifying the landmarks on the MRI or CT studies. Landmarks that have been used consisted of the medial and lateral canthi of the eyes, both tragii, the bridge of the nose and the tip of the ear projected on the scalp. From the description of these references it is apparent that these landmarks are not points but small regions on the anatomy; there may easily be a few mm of ambiguity in defining their location precisely. This is a large number compared to the 0.35 mm accuracy in locating a single point for a typical 3D tracking system [6]. A more accurate set of fiducial markers can be produced by using specially designed markers that are visible in MR or CT images and which are readily located in space. They are made of a toroid-shaped object soaked with contrast and encapsulated in an adhesive plastic capsule which can be attached anywhere on the skin. These markers define a precise location on the skin and their location can be identified with a ball-end stylus that fits exactly in the toroid center. This generally constitutes a significant improvement over the landmark-based registration but it still suffers from FLE due to the motion


of the skin in relation to the cranium. This is true during the MRI acquisition in which the patient is resting on his back and applying some tension of the skin, and in the OR where the head clamp attachment points may be near a marker and pull it slightly. It is also possible for an inexperienced user to apply some force with the stylus on the markers while conducting the co-registration measurements. In carefully conducted in vivo experiments in a CT suite, Lunn et al. were able to demonstrate a TRE of 1.5 mm [7]. Although early on this was probably a best case result, typical results in the operating room today are approaching this level. The method recognized as the most accurate consists of implanting specially designed markers in the bone so that they are rigidly coupled. The markers consist of a threaded stem which is screwed into the bone through a small incision in the skin and a concave spherical receptacle on the exposed end which is designed to receive a stylus tip. The center of the spherical shape defines uniquely and very precisely the registration point. The marker are designed to be visible in MR or CT and their appearance in imaging studies is large enough that an accurate estimate of the spherical center can be obtained by analyzing the curvature of the receptacle end, provided the images are of a sufficient voxel resolution. This method produced the best results, but it is not widely accepted in practice, due to the discomfort and risks associated with it. Using this technique, Maurer et al. have demonstrated clinical TREs of 0.74 0.44 mm with CT, and 1.25 0.45 mm with MRI [8], the difference in accuracy between the two being accounted for the higher spatial resolution CTs.

Surface Based Co-registration It is possible, rather than using discrete reference points (fiducials) to extract the surface of an object of interest such as a patient’s head in two frames of reference and to compute a

28

transformation matrix that best ‘‘matches’’ spatially these two surfaces. This approach implies means of representing and capturing surfaces. It also implies a measure of the spatial match between two surfaces. Furthermore the nature of the computation will be very different since we cannot base it on co-locating discrete points in two spaces and use well tested direct solutions to the least squares problem. The reason being that while we may have many points representing a surface, we do not have a correspondence between these points, nor does such a correspondence necessarily exist since surfaces can be interpolated differently between the points that represent them. Surfaces can be described parametrically as a smooth continuous function, or they can be discretized as a set of points. To produce a surface from a set of points one can view them as forming the vertices of triangular facets, which together form a surface. One sometimes refers to this construct as a surface mesh. In computer applications surface meshes are in widespread use, from computer graphics to engineering CAD systems and a large body of algorithms exists to process surfaces thus described. Surface digitization of a physical object can be achieved by several means. One can simply move the tip of a tracked stylus back and forth on the surface of interest and record points in an irregular pattern until enough points have been gathered [9]. This method lacks predictability and may result in unevenly and incompletely digitized surfaces if one is not attentive. A better method that has been used consists of using a separately tracked laser scanner. The surface is digitized in the coordinate system of the scanner, the scanner’s position in space is itself known and it is therefore possible to register the digitized surface points with the desired frame of reference. Such scanners come with very good resolution and produce very high point densities and thus describe the surface very well [10]. Another method that has been used to digitize surfaces is stereopsis. It is possible with two images of the

447

448

28


same scene taken from two separate points of view and with the same features identifiable in both to locate in relation to the cameras the 3D coordinates of these features in the scene. Most operating microscopes are binocular devices and allow one to attach video cameras on both optical paths. Using such an arrangement, it has been possible to digitize cortical surfaces, for example [11]. This approach has the distinct advantage that it is readily integrated physically into an operating microscope, which is normally used in neurosurgery. Extracting surface information from a preoperative imaging study, MRI or CT, is an easier problem, generally, in that segmenting the boundary between skin and air is a relatively simple programming task, usually solved by using thresholding, with some additional processing to smooth the resulting surface. One can follow this step with the marching cubes algorithm [12] which will produce the mesh describing the surface from the list of pixels corresponding to the skin/air interface. It is usually also desirable to decimate such a surface since it tends to have a very high point density. Given two surface descriptions, in the form of points, triangular patches or parametric formulation, the iterative closest point algorithm (ICP [3]) can produce a transformation matrix that minimizes a distance metric between the surfaces. Its operation is elegant and its convergence fast, compared to other optimization algorithms. We summarize the algorithm here. Given a set of points P ¼ fpi g; i ¼ 1:::N, measured from a physical object (e.g., a patient’s head or an object of interest digitized in a machine shop), and given a computer representation of the same object denoted X consisting of a 2nd set of points, it is possible to compute the shortest distance of any point pi in the set P from the surface in question by defining it as the smallest distance between that point and any point on the surface X denoted dmin ðpi ; X Þ. The specific details of the computation of this distance depends

on the internal representation of the surface X, it could be a parametric surface or a triangulated set of points, for example, nevertheless the distance can be computed. As a byproduct of computing the distance dmin ðpi ; X Þ, one can also record the set of points Q ¼ fqi g; i ¼ 1:::N corresponding to the closest points in X to the points pi in P. With these matched point pairs, one can compute the optimal transformation that will co-register points in P with their closest point matches in Q using the same algorithm we presented in the point based co-registration discussion. At this stage we can transform the initial set of points P0 to a new set P1 (here the index indicates the iteration number) and repeat the process until no improvements are obtained. This algorithm can be summarized as follows: 0) 1) 2) 3) 4)

Starting with K=0, given pi;k 2 Pk and X Compute dmin ðpi;k ; XÞ; i ¼ 1:::N and qi;k 2 X Compute Q TP which merges best pi;k qi;k Q Compute Pkþ1 ¼ TP Pk If RMS dmin ðpi;kþ1 ; X Þ > threshold go to 1)

One should note that this algorithm will work on two sets of points as well as surfaces. This makes it useful in point based registration as well as surface based registration. However, the biggest drawback of this algorithm is that it will deterministically drive the optimization procedure towards a local minimum corresponding to the starting point. Because of this it is necessary to start with different initial transformations Q TP which will reach different local minima, and increase the probability of finding the global optimum. While it is rare that the algorithm fails to converge properly, it can happen. The reported accuracy obtained in surgery for patient registration based on the digitization of the head surface is on the order of 2.4 1.7 mm typically [13].


Volume Based Co-registration Volumetric data is becoming increasingly available from many sources. MRI and CT are well known sources, but improved technology has made 3D ultrasound more widespread and it is becoming a valuable source of intraoperative data. Given the significantly lower cost of 3D ultrasound, compared to MRI or CT, and given its simple and fast use in the OR environment, its integration in neuronavigation systems is seen as a potential replacement for intraoperative MRI, for example. Given different sources of 3D imaging data of a same patient, it is sometime desirable to co-register these volumes, so as to overlap corresponding regions. To achieve this, one could use the methods we have outlined above, point based or surface based co-registration and probably with acceptable results when the methods are realizable. This is particularly likely to work well if the identified features are correctly recognizable in both modalities and, most importantly, if a rigid transformation is suitable to achieve the match. An example of such a situation would be trying to co-register two CT studies of a patient’s head at different times. There are some situations in which this is not possible, while the need for co-registering remains. A few examples of this would be that two MRI or CT studies of the same patient do not capture the same regions and only have a partial overlap. One can have, for example a MRI that covers a patient’s head completely and a series of a few high resolution CT slices that were selected to cover a tumor and omit the rest of the head. The CT in such a case represents a small slab of the volume encompassed by the MRI. Another situation that may occur is when trying to coregister 3D ultrasound data with MRI. Here the ultrasound acquired through a craniotomy will not have any of the surface features of the patient.

28

In practice, a physician can usually identify features in both imaging series, orient them properly and produce an overlap of the two volumes manually, if necessary. To produce this same result automatically represents some difficulties. The main reason for this difficulty is that generally, the numerical values of voxels corresponding to the same anatomical features are not the same in both imaging series. This is true to a small extent when working with the same type of images (e.g., CT with CT) in both studies, but it is much worse when dealing with different imaging modalities. The physical properties that are behind the various imaging technologies (e.g., density, proton density, acoustic impedance) result in different distributions of pixel values for different tissue types. When looking at images from the same patient in two modalities (e.g., MR and CT), even though corresponding features have different pixel values, identifiable regions (e.g., gray matter, white matter, bone, ventricles) tend to have uniform or at least consistent pixel values (mean variance) within modalities even if these differ between modalities. One can in fact establish statistically this correspondence between pixel values and use this knowledge to align image pairs or volumes. A statistical quantity called the mutual information between two pixel or voxel sets represents the degree of congruence between the two given a transformation that aligns them. For an understanding of mutual information based algorithms, one needs to start with the concept of information in an image. The concept is closely related to that of entropy, as defined by Shannon [14] in the context of communication theory: H¼

N X i¼1

pi log

N X 1 ¼ pi log pi pi i¼1

The probabilities pi in this equation represent the likelihood of an event (e.g., a pixel of a given value) occurring and H is the entropy or

449

450

28


information content of a given set of pixel (e.g., image). If an image is made of all pixels having the same value then its information content is 0 since the probability of every pixel value pi is N/N = 1 (using the frequency of occurrence as the measure of probability). The mutual information between two images is a measure of how much we can predict about image B given information from image A. Clearly this is maximized if images A and B are identical. One can predict the value of all the pixels in B if one knows A. One choice (there are several possible choices, see [15] for a detailed presentation) of a formal definition of mutual information is the Kullback-Leibler distance: IðA; BÞ ¼

X a2A b2B

pða; bÞ log

pða; bÞ pðaÞpðbÞ

where p(a,b) is the joint distribution of pixel values, the probability of pixel values a and b occurring simultaneously at the a same image location in images A and B, while p(a) and p(b) are the probabilities of a occurring in A and b occurring in B. Note that if A and B have independent probability distributions, e.g., knowing A has no predictive value regarding B, then p(a,b) = p(a) p(b). One can align two images or volumes by using mutual information as the objective function to maximize while trying different transformation matrices. Although conceptually this is a simple notion, there are many ways to achieve this and the methods that have been presented tend to be specific for certain types of images. A number of problems need to be worked out which will have an impact on the operation of the algorithm. These are: (1) the nature of interpolation when matching pixels or voxels, (2) the optimization method, (3) remaining within the range of convergence. When rotating slightly one image in relation to another, the pixel centers do not correspond exactly and one is left with a choice of how to establish the pixel value from the rotated

corresponding to the unrotated reference image. Another situation in which a direct correspondence between pixels is not readily defined is when the imaging modalities have different resolutions or pixel sizes. One can chose the nearest neighbor method, or one can interpolate the pixel values using any number of methods, with higher methods giving smoother results. The method selected will have an impact on the convergence of the optimization and more importantly on the type of optimization technique that can be used. Interpolation with higher order methods (splines) results in very smooth changes of the objective function given small changes in the transformation matrix, while the nearest neighbor method often results in a very ‘‘jumpy’’ objective function. The tradeoff here is the computational effort. The methods of optimization that have been used in conjunction with mutual information include most of the well known techniques. Powell’s method and derivative (gradient) based methods work in many implementations that have been presented. With more difficult cases such as aligning ultrasound data with MRI data, for example which have very dissimilar appearances, these techniques seem to readily find local minima from which they cannot escape. Other types of searches can be used in such situations, which are known to avoid this problem, such as the genetic algorithm (GA) and the method of simulated annealing (SA). A serious problem with the mutual information approach, which is present with all algorithms but to which the GA and SA are particularly prone is that the mutual information will be as large and sometime larger when overlapping regions are very small and thus tend to be much more uniform. This happens when shift and rotation are such that only a part of the two images overlaps (e.g., a corner outside of the region of interest, representing air) which results in a seemingly very good agreement between the two images. To avoid this, it is necessary to integrate some a priori information regarding the allowable range of translation and rotation to search during the optimization.


Other refinements have been presented which consist of preprocessing the two data sets to improve the performance. Preprocessing generally consists of applying some spatial filtering operation to the two images to smooth them. Wu et al. [16] have reported re-registration accuracy of 1.919 mm and 3.742 mm in two cases when aligning multiple tracked ultrasound images to a preoperative MRI. In one of the early paper to introduce the concept of mutual information for registration, Viola and Wells [17] conducted some experiments in which they obtained errors defined as the variance of registered points with respect to the reference in all three directions. Their results were 1.87 mm in X, 2.22 mm in Y, 14.19 mm in Z, and a combined angular variance 3.05 . These results were based on benchtop experiments and do not necessarily reflect what could be achieved in a clinical setting.

Non-rigid Transformations All the discussion that proceeded assumed that a rigid transformation was to be obtained. This is usually sufficient when aligning images of a patient’s head or co-registering a patient’s cranium with the corresponding preoperative MRI. When dealing with soft tissue data this may not be the case. In such situations, methods collectively classified as non-rigid registration may be required. We will not discuss them at length here, except to briefly outline the two broad categories of methods that are being developed for this purpose. In the first, one can alter the affine matrix so that it includes different scaling factors along the axes. The matrix 2

sx 60 T ¼6 40 0

0 sy 0 0

0 0 sz 0

3 0 07 7 05 1

will applying different scaling factors (sx, sy, sz) on each of the coordinates, resulting in a distorted

28

mapping. It is also possible to include shear parameters in the upper left 3x3 matrix so as to produce transformations in which the coordinates in the destination frame of reference are function of all three coordinates in the source frame: 2

1 4 syx 0

sxy 1 0

32 3 2 3 0 x x þ sxy y 0 54 y 5 ¼ 4 y þ syx x 5 1 1 1

In the example above, which we limited to the 2D case for simplification, we can see that the coordinate x is transformed to x + sxyy in the destination frame due to the off diagonal terms (sxy, syx) in the upper left 2x2 submatrix. Note that this is distinct from the translation operation in that the value of y in the source frame will affect the value of x in the destination frame. It is possible to combine all these terms and obtain a global non-rigid transformation incorporating rotation, translation, scaling and shear terms that will better match the deformation that exists between two sets of voxels. This is generally not a sufficient approach to obtain good matching between the source and destination frames. To better capture deformation, it is necessary to define ‘‘local’’ deformation parameters. One technique that has been presented [18] consists of defining a grid (u, v, w) which can have any desired resolution and which is used to parameterize the elements of the affine transformation matrix. In other words, the rotation, translation scaling and shear parameters can be adjusted locally using the (u, v, w) values and by interpolating them using a cubic spline, for example, as was done in [18].

Discussion and Conclusion The evolution of stereotactic neurosurgery from a frame-based methodology reliant upon frontal and lateral radiographs to the widespread utilization of its capabilities throughout neurosurgery

451

452

28


and beyond has been the result of computational resources becoming available in the operating room environment. A direct consequence of having direct access to sufficiently powerful and affordable computing ability has been the growth of transformational stereotaxy, utilizing quantitative methods to accomplish co-registration between high content three-dimensional data-sets. Clinical application today has largely employed rigid transformations based upon pairs of corresponding points identifiable in the datasets to be linked. Such methods are straightforward, reliable, and implementable using existing digitizers. Sufficient efficiency and accuracy have been achieved in this manner to make image-guided systems useful and practical in the operating room. As increasing computational resources become available and the algorithms are developed to support other registration methodologies, these more sophisticated strategies will bring only greater efficiency, lower costs, and improved capability. Different registration methods have their respective strengths and weaknesses, dependent upon the requirements of the digitizing technology employed, the nature of the preoperative datasets, the characteristics of the surgical field, the surgical procedure itself, and the needs of the surgeon.

References 1. Arun KS, Huang TS, Blostein SD. Least-squares fitting of two 3-D point sets. IEEE Trans Pattern Anal Mach Intell 1987;9:698-700. 2. Fitzpatrick JM, West J, Maurer CR, Jr. Predicting error in rigid body, point-based registration. IEEE Trans Med Imaging 1998;17:694-702. 3. Besl PJ, McKay ND. ‘‘A method for registration of 3D shapes.’’ IEEE Trans Pattern Anal Mach Intell 1992;14(2):239-56. 4. Hartov A, Roberts DW, Paulsen KD. ‘‘A comparative analysis of coregistered ultrasound and magnetic resonance imaging in neurosurgery.’’ Neurosurgery 2008;62 3 Suppl 1:91-9; discussion: 99–101. 5. West JB, Fitzpatrick JM. ‘‘The distribution of target registration error in rigi-body point-based registration.’’

6.

7.

8.

9.

10.

11.

12.

13.

14. 15.

16.

17.

18.

In: Kuba A editor. IPMI’99: information processing in medical imaging. Lecture notes in computer science, vol 1613. Berlin: Springer; 1999. p. 460-5. Polaris Optical Tracking System. ‘‘Application programmer’s interface guide.’’ Waterloo ON: Northern Digital Inc.; 1999. Lunn KE, Paulsen KD, Roberts DW, Kennedy FE, Hartov A, West JD. Displacement estimation with coregistered ultrasound for image guided neurosurgery: a quantitative in vivo porcine study. IEEE Trans Med Imaging 2003;22(11):1358-68. Maurer CR, Fitzpatrick JM, Galloway RL, Wang MY, Maciunas RJ, Allen GS. The accuracy of imageguided neurosurgery using implantable fiducial markers. In: Lemke HU, Inamura K, Jaffe CC, Vannier MW, editors. Computer assisted radiology. Berlin: Springer; 1995. p. 1197-202. (http://citeseer.ist.psu.edu/maurer95accu racy.html) Friets E. ‘‘The frameless stereotaxic operating microscope: system analysis, enhancements, and nonfiducial registration.’’ PhD Thesis, Dartmouth College, Hanover NH; 1993. Miga MI, Sinha TK, Cash DM, Galloway RL, Weil RJ. Cortical surface registration for image-guided neurosurgery using laser-range scanning. IEEE Trans Med Imaging 2003;22(8):973-85. Sun H, Lunn KE, Farid H, Wu Z, Roberts DW, Hartov A, Paulsen KD. Stereopsis-guided brain shift compensation. IEEE Trans Med Imaging 2005;24(8): 1039-52. Lorensen WE, Cline HE. Marching cubes: a high resolution 3D surface construction algorithm. Comput Graphics 1987;21:163-9. Raabe A, Krishnan R, Wolff R, Hermann E, Zimmermann M, Seifert V. Laser surface scanning for patient registration in intracranial image-guided surgery. Technique Assessments. Neurosurgery 2002;50(4): 797-803. Shannon CE. A mathematical theory of communication. Bell Syst Tech J 1948;27:379-423, 623-56. Pluim JPW, Maintz JBA, Viergever MA. Mutualinformation-based registration of medical images: a survey. IEEE Trans Med Imaging 2003;22(8):986-1004. Wu Z, Hartov A, Paulsen KD, Roberts DW. Multimodal image re-registration via mutual information to account for initial tissue motion during image-guided neurosurgery. Conf Proc IEEE Eng Med Biol Soc 2004;3:1675-8 (IEEE Cat. No.04CH37558). William PV, Wells M, III. Alignment by maximization of mutual information. Int J Comput Vis 1997;24(2): 137-54. Rueckert D, Sonoda LI, Hayes C, Hill DLG, Leach MO, Hawkes DJ. Nonrigid registration using free-form deformations: application to breast MR images. IEEE Trans Med Imaging 1999;18(8):712-21.

27 Anatomical and Probabilistic Functional Atlases in Stereotactic and Functional Neurosurgery W. L. Nowinski

Introduction Early stereotactic human brain atlases were constructed to support human stereotactic instruments [1]. The advent of computed tomography (CT) and magnetic resonance imaging (MRI) enabled imaging of patients’ brains. At present, direct visualization of stereotactic target structures is feasible [2–7] and their depiction, particularly on 3 Tesla systems, is of high quality [8,9]. MRI acquisitions, though superior to CT scans, result in unpredictable and non-reproducible deformations [10]. There are also certain controversies regarding a mismatch between imaging and electrophysiology as well as a target structure incompleteness in the scans [11–13]. Therefore, despite tremendous progress in diagnostic imaging, the stereotactic atlas, particularly in electronic (computerized) format, is still considered an important aid [8,14–17]. It took about two decades from the publication of the first print stereotactic atlas by Speigel and Wycis in 1952 [18] to have a computerized brain atlas available in a clinical setting in 1974 [19]. After the next two decades, at the end of the 1990s, the computerized brain atlases have become prevalent in neurosurgical workstations [20]. The content, role, and use of the atlas in stereotactic and functional neurosurgery (SFN) have been evolving over time in various aspects. This evolution spans: (1) from print atlases to electronic atlases to atlas databases with deformable atlases to population-based multiple complementary atlases with distributions of anatomy, #


function, connectivity, and vasculature as well as with clinical results, (2) from sparse anatomical plates to high resolution volumetric multi-modal atlases at various scales, (3) from atlas-assisted targeting to pre-, intra- and postoperative atlas support, (4) from atlas-assisted identification of a few deep brain targets to that in the whole brain, (5) from a single user, single site/resource tool to web-enabled applications to community-centric atlas building and sharing, and (6) from personal handwritten paper records to softcopy spreadsheets to globally shared databases. This chapter features anatomical and probabilistic functional human brain atlases in SFN. It contains three sections covering: (1) atlases including their comprehensive and up-to-date overview, construction, features, and limitations, (2) atlas-based applications along with atlas use and benefits, and (3) future directions. Through these sections we attempt to present a continuous evolution of the stereotactic human brain atlases and their growing potential, both present and future.

Brain Atlases Numerous human anatomical and functional brain atlases have been created for SFN. We review the print atlases as well as describe computerized atlases and their features. The construction of the Cerefy brain atlases, including the probabilistic functional atlas, is described in more detail along with their features, potential, and limitations.

396

27

Anatomical and probabilistic functional atlases in stereotactic and functional neurosurgery

Anatomical Brain Atlases Print Atlases A number of stereotactic human brain atlases in print format have been created since the first stereotactic atlas was published by Speigel and Wycis in 1952 [18]. They include: Talairach et al. atlas of deep gray nuclei in 1957 [21], Schaltenbrand and Bailey atlas of the brain in 1959 [22], Andrew and Watkins atlas of the thalamus and adjacent structures with a probability study in 1969 [23], Van Buren and Borke atlas of variations and connections of the thalamus in 1972 [24], Schaltenbrand and Wahren stereotactic atlas of the brain in 1977 [25], Szikla et al. atlas of vascular patterns and stereotactic localization in 1977 [26], Afshar et al. atlas of the brainstem and cerebellar nuclei in 1978 [27], Talairach and Tournoux co-planar stereotactic atlas of the brain in 1988 [28], Ono et al. atlas of the cerebral sulci in 1990 [29], Talairach and Tournoux atlas of stereotactic anatomical correlations for gray and white matter in 1993 [30], and Morel et al. multiarchitectonic and stereotactic atlas of the thalamus in 1997 [31]. The content of print atlases is static, non expandable, and non transferable. These atlases are typically not fully segmented and not completely labeled. A major limitation to their use in clinical practice is the difficulty in mapping the print plates into an individual brain.

Computerized Atlases and their Features Though the print stereotactic brain atlases have been available for almost six decades, these are the electronic atlases that have been integrated with neurosurgical workstations and adopted worldwide by the neurosurgical community. Computerized atlases overcome certain shortcomings of the print atlases, can be processed algorithmically,

and offer new features, such as fully segmented and labeled images, atlas to scan registration resulting in automatic segmentation and interactive labeling of patient’s images, cross-correlated presentation in two- (2D) and three-dimensions (3D), defining regions of interest for analysis, structure searching, and quantification. They also constitute reference frameworks enabling integration of information from multiple sources. Features of Computerized Atlases

The main features of computerized atlases are described and illustrated below. Atlas segmentation. Atlas segmentation refers to parcellation of atlas images or volume into individual structures. Each atlas structure is defined either by specifying its region by contouring it (> Figure 27-1d), color-coding its pixels/ voxels (> Figure 27-1b), or both (> Figure 27-1c). As a result, every location in the segmented atlas belongs uniquely to a certain structure (or to the background). Various parcellations are in use; e.g., the thalamus can be parcellated by Hassler’s [32] (> Figure 27-4b) or Walker’s parcellation [33] (> Figure 27-1b), or simultaneously by both (> Figure 27-15b). Atlas labeling. Atlas labeling refers to assignment of names (or generally some classes) to atlas structures (> Figures 27-1b and > 27-4b). Each structure is assigned a unique name (with synonyms, if necessary) which identifies it. All these names form an index. The existing stereotactic atlases, such as Schaltenbrand and Wahren [25] (> Figure 27-4) and Talairach and Tournoux [28] (> Figures 27-1– > 27-3) use their own nomenclatures. Other nomenclatures, such as Terminologia Anatomica [34], are also applicable for labeling. The labels may be full or abbreviated (to facilitate multiple labeling) placed in 2D or 3D (> Figures 27-1 and > 27-4). Atlas deformation (registration, warping). A deformable atlas can be warped (individualized) to match a patient’s scan. Once individualized, the segmentation and labeling content in the atlas is transferred to the patient-specific data


(> Figures 27-1f and > 27-4f). Atlas to data registration is discussed in Sections ‘‘Atlas to scan registration’’ and ‘‘Atlas to data registration.’’ Atlas representation. The segmented and labeled atlases can have various representations including bitmaps (images) with colored structures, bitmaps with delineated (contoured) structures, geometric contours of structures (e.g., in form of splines), 3D geometric surface (polygonal) models, and volumetric models (> Figures 27-1, > 27-3– > 27-5, and > 27-7d). These representations are useful in various atlas-assisted applications. For instance, an atlas in contour representation does not eclipse a scan when superimposed on it (> Figures 27-1f and > 27-4f ) in contrast to that in bitmap representation (> Figure 27-16a); geometric contours can be zoomed in without changing their line thickness which provides a more accurate separation of neighboring structures (> Figure 27-13); surface models are useful for 3D visualization and spatial exploration (> Figures 27-1e, > 27-3, > 27-4e, and > 27-5b); and volumetric models are suitable for non-rigid warping and reformatting in any plane (> Figures 27-7d and > 27-8c). Structure searching. Any structure in the index is searchable to localize it in the original or individualized atlas. The search can be performed in the current atlas image, across entire atlas or multiple atlases. Atlas-assisted quantification. The individualized stereotactic atlas placed in a coordinate system enables reading of stereotactic coordinates (of targets and other locations) and calculating distances. Two main coordinate systems in use are: the Schaltenbrand system [25] with the origin at the midcommissural point (> Figure 27-5a) and the Talairach system [28] with the origin at the anterior commissure (> Figure 27-1). Spatial cross-correlation. The atlas or, generally, a multiple atlas database facilitates spatial correlation across orientations, atlases, and dimensions. A location (e.g., a target point)

27

in a single atlas available in multiple orientations can be correlated across orthogonal planes (> Figure 27-14). Multiple atlases spatially coregistered and cross-correlated enable to examine the same structure or region in various atlases and explore multi-atlas complementarity (> Figures 27-5 and > 27-15a). Correlation can be across all three orientations (on the triplanar) (> Figures 27-2, > 27-5a, > 27-6a, and > 27-15a) or between 2D and 3D (enabling 3D surface or volumetric models to be correlated with the original or reformatted images) (> Figures 27-3a, > 27-15a, and > 27-16a). Computerized Atlases

To benefit from the electronic format, most of the print atlases have been converted into it, including: Schaltenbrand and Bailey atlas [35–37]; Schaltenbrand and Wahren atlas [35,38–40]; Afshar et al. atlas [41]; Van Buren and Borke atlas [35]; Talairach and Tournoux atlas [35]; and Morel atlas [42]. Moreover, electronic versions of the classic Thieme stereotactic print atlases [25,28–30] are included into the Cerefy brain atlas database (Section ‘‘Cerefy atlases’’). Computerized versions of the print atlases vary substantially ranging from a simple, direct digitization of the original printed material to a sophisticated, fully segmented, labeled, enhanced, and 3D extended deformable atlas (see also Section ‘‘Atlas-based applications’’). The existing stereotactic atlases are usually sparse with a variable inter-plate distance, varying from 0.5 to 4.0 mm for the Schaltenbrand and Wahren atlas, and from 2 to 5 mm for the Talairach and Tournoux atlas. However, to enable nonrigid warping or reformatting in an oblique plane, the atlas must be volumetric and of high resolution. Technically, image or contour interpolation of sparse atlases is feasible, enabling reconstruction of 3D models and generation of intermediate sections. For instance, the Schaltenbrand and Bailey atlas [22] was interpolated with 0.5 mm step in [37]. However, interpolation or 3D modeling does not compen-

397

398

27


sate for the intrinsic shortcomings of the original print material (as discussed in Section ‘‘Limitations of anatomical atlases’’). An analysis of the main target structures in the Schaltenbrand and Wahren atlas: the subthalamic nucleus (STN) [43], globus pallidus internus (GPi) [44] and ventrointermediate nucleus of the thalamus (VIM) [45], reconstructed in 3D shows that their shapes are not realistic (see e.g., > Figure 27-6c). To facilitate non-rigid warping and/or reformatting in any plane, some existing atlases have been extended and new stereotactic atlases in 3D geometric and high resolution volumetric representations have been developed [46–49]. A 3D atlas of the STN and its adjacent structures is described in [48]. A deformable atlas of the thalamus, called the Talairach 2000 atlas [47], is constructed by applying 3D geometric modeling to the Talairach and Tournoux atlas. Moreover, the Cerefy atlases contain 3D geometric and volumetric models of cerebral structures (Section ‘‘Cerefy atlases’’). A 3D deformable atlas of the basal ganglia is created from histological images [49]. The left hemisphere, after removal of the frontal and occipital lobes, was cut into 1.5 cm blocks. Each block was then cut into 800 coronal sections of a 70 mm thickness divided into two series. Every tenth section was stained. One series was Nissl-stained with cresyl violet and the other was immuno-stained for calbindin. Eighty structures, including basal ganglia nuclei, fiber bundles and ventricles, were traced on the histological sections. Another histology-based atlas of the thalamus and basal ganglia is built from 86 pairs of coronal sections cut with 0.7 mm intervals [46]. For each pair of sections, one section was stained with Luxol Blue for myelin and the other with a Nissl stain for cell bodies. The sections were manually segmented and labeled with 105 structures.

Cerefy Atlases The most prevalent computerized atlases in SFN are the Cerefy atlases (www.cerefy.com) [20,50– 53]. Their acceptance and adoption by the community resulted probably not only thanks to providing a high quality content, but also by proposing novel atlas-based solutions, and developing sophisticated yet user-friendly tools for planning, intraoperative support, and postoperative analysis (Section ‘‘Atlas-based applications’’). The Cerefy anatomical atlases have been derived from the classic print brain atlases edited by Thieme [25,28–30]. They are available in neurosurgical workstations [20] (as add-on libraries) and also distributed on CD-ROM [54–57] (by the publisher of the original print atlases), see Section ‘‘Atlasbased applications.’’ The Cerefy anatomical brain atlas database contains, among others, the electronic versions of: Atlas for Stereotaxy of the Human Brain by Schaltenbrand and Wahren (SW) [25] and CoPlanar Stereotactic Atlas of the Human Brain by Talairach and Tournoux (TT) [28]. To create these computerized atlases, the original print materials were intensely processed, enhanced, and extended. This processing involved: (1) scanning of the print images and compiling textual materials, (2) full segmentation (contouring or color coding) of all atlas structures, (3) complete labeling (naming) of all atlas structures, (4) arrangement of the atlas images into volumes, (5) atlas checking, correcting, enhancing, and extending, (6) constructing 3D versions, (7) developing various representations in 2D and 3D (bitmap, contour, polygonal, and volumetric), (8) mutual coregistration of all 2D and 3D atlases, and (9) studying atlas properties. Atlas construction required the development of several dedicated and sophisticated tools. To enable integration of the Cerefy atlases with the existing neurosurgical workstations, two add-on atlas libraries with the viewers are


developed: Cerefy Electronic Brain Atlas Library and Cerefy Brain Atlas Geometrical Models (www. cerefy.com) [20]. The Cerefy Electronic Brain Atlas Library contains the atlases in image representation and the Cerefy Brain Atlas Geometrical Models comprise the atlases in geometric contour and 3D polygonal representations. Cerefy Talairach and Tournoux Atlas

The TT atlas contains axial, coronal, and sagittal images of gross neuroanatomy derived from a single normal brain specimen. The brain had been sectioned and photographed sagittally, and the axial and coronal orientations were interpolated manually. To create the Cerefy TT (C-TT) atlas, the original print images were digitized with 0.2 mm resolution, and extensively processed, enhanced, and extended as follows: (1) the original grids, rulers, and annotations were removed, (2) each atlas structure was assigned a unique color-coded representation, in contrast to a mixture of contour, color-coded, and texture representations in the print atlas, (3) the left thalamic nuclei and the basal ganglia, not available in the print atlas on the axial and coronal plates, were outlined and color-coded, (4) the right hemisphere cortex on the axial plates was added by mirroring the left hemisphere cortex, and (5) Brodmann’s areas and gyri, which are labeled but not segmented in the print atlas, were segmented and color-coded on the axial plates [50]. The C-TT atlas has continuously been improving in terms of its quality, image [52] and textual content [56], representation, and spatial consistency [53]. Examples of the C-TT atlas in multiple orientations and representations are shown in > Figure 27-1. The GPi correlated across all three C-TT orthogonal planes is illustrated in > Figure 27-2. An example of 2D-3D cross-correlation in the C-TT atlas is presented in > Figure 27-3a. > Figure 27-3b shows an enhanced version of the 3D C-TT atlas suitable for interpolation [58]. Cerefy Schaltenbrand and Wahren Atlas

27

The SW atlas, based on 111 brains, contains photographic plates of macroscopic and microscopic sections through the hemispheres and brainstem. The macroscopic plates with gross anatomy provide the extent of variation of cerebral structures. The microscopic myelinstained sections show the deep brain structures in great detail. To create the Cerefy SW (C-SW) atlas, the microseries along with the overlays were digitized with 0.1 mm resolution. They were processed, enhanced, and extended as follows: (1) the digitized images were rotated (typically within 0.2–0.5 ) to ensure their proper alignment, (2) all structures on the overlays were contoured manually under a high magnification (of up to 20 times), (3) numerous contours open in the original SW atlas were closed, enabling interactive labeling and structure searching (the detailed list of the processed structures is enclosed in the User Guide of [56]), (4) geometric contour and 3D polygonal models were constructed, (5) each contour and polygonal model was assigned a label (or labels) consistently with the SW overlays, (6) the microseries and the contours with the corresponding labels were extended to cover both hemispheres: the axial (Brain LXXVIII, right hemisphere) and coronal (Brain LXVIII, right hemisphere) plates with the corresponding contours were mirrored along the midsagittal (interhemispheric) plane, and the sagittal plates (Brain LXXVIII, left hemisphere) with the contours were replicated for the right hemisphere; the 3D models were also replicated for the other hemisphere, and (7) corresponding sections were aligned and stacked to form a brain volume; the processed material was initially organized into 18 atlas volumes [50], but practically the axial, coronal, and sagittal microseries have been in use. The C-SW atlas has continuously been enhancing (see the User Guide of [56] and Section ‘‘PFAcentric combined atlas’’). Examples of the C-SW atlas in multiple orientations and representations are shown in > Figure 27-4.

399

400

27


Cross-correlation of and complementarity in multiple atlases are shown in > Figure 27-5.

Limitations of Anatomical Atlases Creation of stereotactic print atlases required monumental efforts, particularly that their creators

. Figure 27-1 C-TT atlas in multiple orientations and representations labeled in 2D and 3D: (a) digitized original axial image (note that the Talairach grid (Section ‘‘Talairach transformation’’) encompassing the brain is not exactly rectangular), (b) color-coded axial image with gyri, Brodmann’s areas, and subcortical structures labeled with full and abbreviated names, (c) coronal image in color-coded and contour representations along with the Talairach grid, (d) sagittal image in contour representation, (e) 3D C-TT atlas labeled in 3D, (f) sagittal scan segmented and labeled by the atlas in contour representation


27

. Figure 27-1 (Continued)

. Figure 27-2 The GPi correlated across all three orthogonal planes in the C-TT atlas. The label and coordinates at the intersection of the reference lines (the pointed location) are displayed. The description of the GPi is also shown

401

402

27


. Figure 27-3 3D C-TT atlas: (a) subcortical structures of the 3D C-TT atlas cross-correlated with the coronal images of the 2D C-TT atlas; the subthalamic nucleus and its surrounding structures are labeled in 3D, (b) enhanced version of the 3D C-TT atlas with subcortical structures and some white matter tracts (the thalamus is rendered semi-transparent to show its component nuclei)

did not have access to information technology. By today’s standards and requirements, however, some features of these atlases are regarded as limitations. The shortcomings of stereotactic atlases have been well studied [39,43–45,51,59–62]. These studies have been possible thanks to the advances in computers enabling a qualitative and quantitative atlas analysis. This analysis is facilitated by constructing electronic versions of print atlases, building 3D models of structures, developing tools for atlas quantification on the orthogonal planes and in 3D (> Figure 27-6), and formulating and computing measures characterizing the location, size, shape, and mutual relationships of the studied structures [43]. The SWaxial, coronal, and sagittal microseries derived from three different hemispheres are not spatially consistent, meaning that a given point in 3D may belong to more than one anatomical structure. The original printed material is not fully consistent in terms of plates (neuroanatomy),

grid, image-overlay correspondence (see e.g., > Figure 27-4d), and landmarks. Inaccuracies existing across all three SW microseries are clearly visible on the triplanar (> Figure 27-6a) or in 3D (> Figure 27-6c), and those within a given orientation in 3D (> Figure 27-6b). The SW axial plates were not acquired in the intercommissural (but in Reid’s) plane and are rotated 7 clockwise. The resulting inaccuracy for the P.m.i (GPi) 2.0 mm in front of the midcommissural point (i.e., on plate Fa 2.0) is +0.25 mm, where ‘‘+’’ means dorsal and ‘‘ ’’ ventral offset. The atlas inaccuracy for the V.im.e ranges from 0.49 mm (on plate Fp 4.0) to 0.86 mm on plate Fp 7.0; that for the Sth (STN) spans from +0.25 mm on plate Fa 2.0 to 0.86 mm on plate Fp 7.0. Some quantitative verification of the SW atlas was done in [39,59]. It demonstrates that the sagittally sectioned thalamus is 10% larger than the coronally sectioned thalamus and 40% larger than that sectioned axially. Another group


of studies (summarized below) quantifies the main stereotactic target structures in the C-SW atlas: STN [43], GPi [44], and VIM [45]. It partly concurs qualitatively though not quantitatively with [59].

27

The 3D models of the STN: 3D-A, 3D-C and 3D-S, reconstructed from the C-SW axial, coronal and sagittal microseries, respectively, are placed in the Schaltenbrand coordinate system, and compared quantitatively in terms of location

. Figure 27-4 C-SW atlas in multiple orientations and representations labeled in 2D and 3D: (a) digitized original axial image, (b) axial image in contour representation labeled with full and abbreviated names, (c) color-coded coronal image, (d) sagittal image in contour and image representations labeled (note a mismatch between the SW image and its corresponding overlay in the brachium colliculi superioris), (e) 3D C-SW atlas labeled in 3D, (f) axial scan segmented and labeled by the atlas in contour representation

403

404

27



(centroids), size (volumes), shape (normalized eigen values), orientation (eigen vectors), and mutual spatial relationships (overlaps and inclusions) [43]. These 3D STN models differ in location, size, shape, orientation, overlap size, and inclusion rate as follows. The 3D-S volume of 207.1 mm3 is 1.27 times larger than that of 3D-A and 1.38 times larger than that of 3D-C. The highest overlap size is between 3D-A

and 3D-S. The highest inclusion rates of 52.5 and 66.6% are for 3D-A and 3D-S. 3D-C has the lowest overlap size and lowest inclusion rates (about 20–30%), meaning that 3D-C is considerably displaced in comparison to 3D-A and 3D-S. The lateral centroid coordinate of 3D-C is 9.18 mm and that of 3D-S is 12.17 mm. Each model has some limitation: 3D-A in orientation, 3D-C in location, and 3D-S in shape realism.


27

. Figure 27-5 Cross-correlation of and complementarity in multiple atlases: (a) subthalamic nucleus in the C-SW and C-TT atlases mutually correlated (the axes of the Schaltenbrand reference system along with the coordinates are also shown), (b) complementarity of a 3D highly parcellated thalamus in the SW atlas with a gross neuroanatomy in the TT atlas

. Figure 27-6 Inconsistency in the C-SW atlas: (a) structure mismatch on the triplanar, (b) unrealistic shape of the 3D thalami reconstructed from the coronal contours, (c) unrealistic and inconsistent 3D VIM models reconstructed from axial (in blue), coronal (in red), and sagittal (in green) microseries (dorsal view)

405

406

27


The reconstructed 3D GPi models substantially vary in location, size, and inclusion rate [44]. The centroid of 3D-C is located more medially (15.6 mm) than these of 3D-A (17.5 mm) and 3D-S (18.2 mm), and that of 3D-A is more ventrally ( 2.3 mm) than those of 3D-C ( 0.1 mm) and 3D-S ( 0.4 mm). 3D-S has the smallest volume of 347.3 mm3, 3D-A is 1.18 and 3D-C 1.85 times larger. The highest inclusion rate is for 3D-S (54.3 and 56.3%) and the lowest for 3D-C (28.8 and 30.6%). The shape, orientation, and overlap size are less variable. The reconstructed 3D models of the VIM (> Figure 27-6c), VIM externum (VIMe), and VIM internum (VIMi) also differ significantly in location, size, shape, and inclusion rate [45]. The centroid of 3D-A/VIM differs considerably from those of 3D-C/VIM and 3D-S/VIM. The difference between the centroids of 3D-C/VIM and 3D-S/VIM is in laterality only: this of 3D-C/ VIM is located more medially (11.85 mm) than that of 3D-S/VIM (14.62 mm). 3D-A/VIM has the smallest volume (of 69.00 mm3), 3D-C/VIM is 3.71 and 3D-S/VIM 3.89 times larger. The overlap is also highly variable: 104.88 mm3 for 3D-C/ VIM with 3D-S/VIM, and very low (3.22 and 7.45 mm3) when 3D-A/VIM is involved. The highest inclusion rate is for 3D-C/VIM with 3D-S/VIM (39.10 and 40.97%) and the lowest for 3D-A/VIM with 3D-C/VIM (1.26 and 4.66%). The centroid of 3D-A/VIMe differs noticeably from those of 3D-C/VIMe and 3D-S/VIMe. The difference between the centroids of 3D-C/VIMe and 3D-S/VIMe is mainly in laterality: this of 3D-C/VIMe is located more medially (12.91 mm) than that of 3D-S/ VIMe (16.65 mm). 3D-A/VIMe has the smallest volume (of 49.87 mm3), 3D-S/VIMe is 3.24 and 3D-C/VIMe 3.36 times larger. The overlap sizes are low: 32.72 mm3 for 3D-C/VIMe with 3D-S/ VIMe, and very low (1.32 and 2.01 mm3) when 3D-A/VIMe is involved. The inclusion rates are also low: the highest is for 3D-C/VIMe with 3DS/VIMe (19.53 and 20.29%) and the lowest for 3D-A/VIMe with 3D-C/VIMe (1.19 and 4.01%).

There are substantial differences among the centroids of 3D-A/VIMi, 3D-C/VIMi and 3D-S/ VIMi. This of 3D-A/VIMi is located more anteriorly ( 1.92 mm) than that of 3D-C/VIMi ( 5.02 mm). The centroid of 3D-A/VIMi is located more ventrally (2.88 mm) than those of 3D-C/VIMi and 3D-S/VIMi (each at 5.34 mm). 3D-A/VIMi has the smallest volume (of 19.75 mm3), 3D-S/VIMi is 3.23 and 3D-C/VIMi 4.30 times larger. 3D-A/VIMi practically does not overlap with 3D-C/VIMi and 3D-S/VIMi. The inclusion rates for 3D-C/VIMi with 3D-S/VIMi are medium (32.63 and 43.43%). The overall conclusion from these three studies is that the SW atlas shows inter- and intraorientation spatial inaccuracies. Quantification of these inaccuracies may help enhancing the SW atlas and is useful in atlas registration (see Section ‘‘Combined anatomical-functional atlases’’). The original TT atlas, though constructed from a single brain specimen, is also not spatially consistent. This consistency was defined as uniformity of labeling across all three orientations at the common points and calculated for the entire C-TT atlas, majority of its structures, and cortical areas [62]. It was also analyzed in function of discrepancy measuring the spatial offset in labeling. The C-TT atlas has 27% consistency (three labels are common), 62% partial consistency (two labels are common), and 38% inconsistency. The thalamus with 86% consistency is the most consistent structure. The basal ganglia have a good consistency. The inconsistency of major subcortical gray matter structures is very low for 3 mm discrepancy. The inconsistency of all subcortical structures is higher (17%), caused mainly by a very high inconsistency of the white matter tracts. The entire atlas consistency increases by 20% for 1 mm discrepancy, then constantly grows by 10% for each of 2, 3, and 4 mm discrepancy, and finally slows down to 3% for each of 5 and 6 mm discrepancy. The consistency increase for the cortical areas is higher than that for the subcortical structures.


In addition to the above limitations, each of the SW and TT atlases exhibits shortcomings in terms of the landmarks defining its reference system. In the print SW atlas, the anterior commissure (AC) and posterior commissure (PC) landmarks are inconsistent. This atlas provides three sources of information about their location and distance: (1) guide attached to the atlas, (2) microseries plates, and (3) plates showing the unsectioned hemispheres. These sources give three different AC-PC distances [61]. Moreover, on the SW sagittal microseries (used most frequently for planning) the external outlines (semi-circles) of the AC and PC structures are marked only 2.5 mm away from the midsagittal plane (on plate Sl 1.5 the AC and PC are not marked at all, whereas the midsagittal plane, where the AC-PC distance should be measured, is missing). In addition, the intercommissural distance between the marked AC and PC outlines varies, e.g, 20.5 mm on plate Sl 2.5, 20.0 mm on plate Sl 3.5, and 20.5 mm on plate Sl 5.0. From a computerized atlas perspective, there are at least four problems with the original Talairach landmarks [63] (see Section ‘‘Talairach transformation’’ for their definitions). First, not all of them are available in the original atlas. Second, locations of some landmarks contradict their definitions. In particular, on the axial plates, the L and P landmarks are beyond the Talairach grid (which also is not exactly rectangular, see > Figure 27-1a), and the R landmark is not present at all. The atlas plates do not cover the entire Talairach space and, consequently, the AC, PC, S and I landmarks are not available on the axial plates, A and P landmarks are missing on the coronal plates, and L and R landmarks are not present on the sagittal plates. Third, the cortical landmarks are not defined constructively. Fourth, the intercommissural landmarks are located outside their own structures and, despite being defined precisely, their accurate constructions are not easily doable on a scanner console or neurosurgical workstation. To cope with these

27

problems, the modified landmarks were introduced [63] (Section ‘‘Atlas use’’). As a consequence of all these studies, an absolute and direct reliance on the original SW and TT atlases is unsafe and these atlases must be used with a great care and understanding of their features and limitations.

Probabilistic Functional Brain Atlases Despite their great usefulness in SFN, the current computerized anatomical atlases have two major limitations (besides the abovementioned shortcomings). First, they are constructed from a few brain specimens only. Second, these atlases are anatomical, while the actual stereotactic targets are functional. Anatomical variability studies were already included in several stereotactic print atlases and some of them were based on multiple brain specimens. The first stereotactic atlas by Speigel and Wycis provides data on anatomical and radiographic variability [18]. A statistical analysis of variations in skull-brain relationships, endocranical reference system, and intercommissural line is given in the Schaltenbrand and Bailey atlas [22]. The Andrew and Watkins atlas [23] includes a probabilistic study of variability of the thalamic nuclei and neighboring structures. The SW atlas, constructed from 111 brains, presents macroscopic variation of selected structures and provides a possible extent of variation. The outlines on the overlays demarcate both the least common region and overlapping parts in several structures of interest including the STN and GPi. The accompanying tables give measurements made on coronal sections of eight brains with the mean distances from the structure’s center to the midline at different levels. In addition, contours for several ventricles and skulls along with their average outlines are provided. Despite the wealth of material used to construct the

407

408

27


entire SW atlas, the microseries, which are the most often employed, are derived from two brain specimens only. There are several ongoing efforts aimed at developing electrophysiology databases and atlases. A database of subcortical electrophysiology [64] stores stimulation-induced responses and microelectrode recording data coded and plotted along stereotactic trajectories from 106 procedures (88 patients). This database contains only intraoperative data and does not include the best targets. The data points are displayed in 3D as clusters of color-coded spheres. The approach applies nonrigid warping to normalize brains [65] and does not rely on the AC and PC landmarks. It also does not take into account the geometry of electrode. An initial atlas of target points based on 18 patients [66] exploits nonrigid registration for brain normalization [67]. This effort also aims to create a database for storing all pre-, intra-, and postoperative information for the treated patients. The CASS system contains a well-established database of about 2,500 electrophysiological response points from microelectrode diencephalic stimulation and recording collected over a few decades [35]. To overcome both limitations of the current computerized anatomical atlases as well as to avoid a cumbersome dealing with a huge number of electrophysiology numerical values (including point coordinates and various responses), the probabilistic functional atlas (PFA) was introduced conceptually in [20] and algorithmically in [68] as a new addition to the Cerefy family of brain atlases. The PFA algorithm is able to convert thousands of numbers with the best contacts, their location, orientation and size as well as coordinates of the patient-specific landmarks into simple to use 2D maps and 3D volumetric models. The PFA opens new directions not only in planning but also in providing community-centric solutions in SFN.

PFA: Concept and Algorithm The PFA is calculated from intraoperative neuroelectrophysiology, pre- and intraoperative neuroimaging, and postoperative neurological assessment. The PFA algorithm converts the coordinates of the neurologically most effective contacts into probabilistic functional maps taking into account the geometry of a stimulating electrode and patient’s anatomy. This atlas provides the distribution of the best stereotactic targets in a normalized atlas space, enables determining the accuracy of targeting, and facilitates studying properties of functional structures [20,57,68–73] (see also Section ‘‘Atlas use and benefits’’). The PFA algorithm calculates the atlas from the selected best contacts in the following steps: (1) 3D reconstruction of contact coordinates, (2) contract normalization, (3) voxelization of the normalized contacts, (4) calculation of the atlas function, and (5) probability computing [68]. Each best contact is normalized by applying the corresponding patient-specific normalization parameters and placed in the common atlas space. The normalized best contacts are voxelized by a rapid and optimal algorithm for voxelization of a deformed cylinder [68]. The atlas function at a given point is defined as the number of the best contacts residing at this point. In a given voxel, the (discretized) atlas function is calculated by counting the number of the best contacts containing this voxel. The atlas probability is computed as a linear function of the atlas function and four probability definitions are given in [68]. The probability distribution is then presented as color-coded (> Figures 27-7 and > 27-8) or gray scale (> Figures 27-9, > 27-10, and > 27-19) maps. The PFA is dynamic and can rapidly be recalculated for arbitrary (userdefined) resolution and extended by adding new patient’s data. The atlas can easily be reformatted and warped to match patient-specific data. The detailed processing steps depend on a data acquisition process and stereotactic procedure. For


the acquired data, the PFAs were calculated for the STN (PFA-STN) and VIM (PFA-VIM) with a 0.25 mm3 spatial resolution and 0.25 mm accuracy.

Data Acquisition Multi-modal data were acquired pre-, intra-, and postoperatively during surgical treatment of Parkinson’s disease (PD) patients [74–76]. Preoperatively, lateral and antero-posterior X-ray ventriculography projections were acquired. The X-rays imaged the AC, PC, thalamus, and third ventricle, see > Figure 27-19b. The AC-PC distance, the height of the thalamus (HT), and the width of the third ventricle (V3) were measured on these X-rays for each patient, and subsequently used to normalize the best contact(s). Intraoperatively, the best contact was identified neuroelectrophysiologically and its position was imaged on two orthogonal X-rays. Its coordinates were measured on them and reconstructed in 3D. Two types of quadripolar electrodes were used: DBS 3387 with a 1.5 mm inter-contact gap and DBS 3389 with a 0.5 mm gap. For thalamic stimulation, the monopolar electrode was also employed. Postoperatively, each best contact was verified neurologically during a 3 month patient assessment follow-up and, if necessary, updated and re-measured on the X-rays. The selection of the best contacts aimed at improvements in akinesia, rigidity, and tremor [75,76].

PFA-STN Two versions of the PFA-STN were constructed. The first (main) version was built from all available 366 best contacts of 184 PD patients [70]. This version is available in [57] for clinical use (Section ‘‘Cerefy Clinical Brain Atlas: Enhanced Edition with Surgical Planning and Intraoperative Support (CCBA-Plan)’’). The second (bilateral) version was developed from 168 bilateral cases in two situations: with and without lateral compensation

27

against the V3 [71]. It is useful in studying STN properties and comparing the left and right functional STN [71] (Section ‘‘Atlas use and benefits’’). Contact normalization employs the landmarks identified during the data acquisition and used in neurosurgery planning: AC, PC, HT, and V3. In the atlas (normalized) space, the distance between AC-PC = 24 mm, HT = 16 mm, and V3average = 6 mm, where V3average is the average value of V3. Consequently for each case, the coordinates of the best contact(s) were scaled antero-posteriorly to match the atlas AC-PC distance and dorso-ventrally to match the atlas HT. Lateral compensation against the width of the third ventricle was performed by shifting its lateral coordinate by ((V3average V3)/2). Points of interest on the X-rays, measured with accuracy of about 0.25 mm, were reconstructed in 3D from two orthogonal projections. Under a simplified assumption that the patient’s midsagittal plane is ideally parallel to the lateral plane and perpendicular to the anteroposterior plane, the AC-PC distance and HT could be measured directly on the lateral projection, V3 on the antero-posterior projection, and the 3D electrode coordinates calculated easily from its two 2D ideally orthogonal projections. The 3D reconstruction problem, however, must be considered taking into account the stereotactic environment, data acquisition process, and potential mis-positioning of the patient’s head. Two 3D reconstruction methods taking into account these factors are formulated and their errors estimated in [70]. The validation of the PFA-STN in terms of the voxelization procedure by applying the Monte Carlo method and the correctness of the contact data is addressed in [70]. In general, the more contacts are used for PFA calculation, the better is the atlas from a statistical standpoint. However, the contact data may be inaccurate or even incorrect. Therefore, a suitable balance between the size of contact population and atlas quality has

409

410

27


to be maintained. Two quality criteria were applied in the process of contact selection and checking: (1) contact cluster formation (by visual inspection), and (2) reconstructed contact height preservation (by automatic calculation) [70]. The best contacts, normalized and put together in the atlas space, form typically a cluster. The outliers located outside this cluster required additional checking (these contacts sometimes were assigned a wrong electrode type resulting, due to a different inter-contact gap, in a wrong contact positioning). The second criterion is more reliable. The physical height of the contact is 1.5 mm, and its accurate reconstruction from two perpendicular X-ray projections (before spatial normalization) should result in the same height. Besides a finite accuracy of calculations, there are several other factors causing the reconstructed contact height differs from its physical one. First, the length of the electrode may not be measured accurately enough on the projections. Second, a wrong type may be assigned to the electrode resulting in a misplaced contact location. Third, 3D reconstructions from 2D projections may be distorted because of patient’s mispositioning, particularly when the patient’s midsagittal plane is not parallel to the plane of the lateral projection. Finally, 3D reconstructions may not be accurate because of incorrect input parameters of the stereotactic frame (frame size, size of angiographic localizers, or projections of angiographic localizers) and positions of the Xray sources and film plates. The height of each contact was reconstructed and checked against a given range of accuracy taken as 0.25 mm. If the reconstructed height was outside this range, the contact was re-examined by checking and correcting, if necessary, the electrode type, and re-measuring the electrode length on the X-rays. If after these operations the contact height was still outside the given accuracy range, this contact was rejected. Consequently, the contacts with the reconstructed height lower than 1.25 mm or higher than 1.75 mm were rejected in the process of PFA-STN construction. > Figure 27-7 shows

the PFA-STN including the normalized contacts, normalized best contacts, voxelized best contacts, and axial, coronal and sagittal color-coded maps.

PFA-VIM The same algorithm was applied to construct the PFA-VIM. It was built from 107 best contacts in two situations: with and without lateral compensation against the V3 [72]. This compensation slightly changes the laterality of the PFA-VIM mean value location from 13.99 to 13.83 mm for the left and from 14.13 to 13.84 mm for the right hemisphere. It also reduces the lateral coordinate of the standard deviation by 22% for the left and 15% for the right hemisphere. > Figure 27-8 illustrates the PFA-VIM. The algorithm for PFA calculation is fast. The complete PFA-VIM was calculated in 2 s for 0.5 mm resolution and in 14 s for 0.25 mm resolution on a standard PC [72]. An atlas update with new cases is feasible in a fraction of second without re-calculation of the entire atlas as the atlas function is linear [68].

Advantages of the PFA The PFA has a number of advantages. It is a truly probabilistic functional atlas calculated from intraoperative neuroelectrophysiology, preand intraoperative neuroimaging, and postoperative neurological assessment by converting the coordinates of the neurologically most effective contacts into probabilistic functional maps taking into account the patient-specific anatomy and geometry of a stimulating electrode. The PFA provides a quantitative spatial distribution of the best stereotactic targets in a normalized atlas space, enables determining the accuracy of targeting, and facilitates studying functional properties of structures (see Section ‘‘Atlas use and benefits’’). The PFA aggregates knowledge from the previously operated cases. This knowledge can


27

. Figure 27-7 PFA-STN (the origin of the coordinate system is at the PC, the AC-PC distance is divided into 12 units of 2 mm each, HT is divided into eight units of 2 mm each, and lateral units are in mm): (a) normalized contacts, (b) normalized best contacts, (c) voxelized best contacts, (d) axial (A), coronal (C), and sagittal (S) color-coded maps as well as the voxelized atlas (3D) along with the coordinate system and the locations of the orthogonal planes, (e) probability color bar

be aggregated individually by the neurosurgeon, within a group of users, or over the entire neurosurgical community by means of e.g., a PFAbased portal for SFN [69] (see Section ‘‘Research prototypes’’). This portal facilitates data sharing among functional neurosurgeons, calculates rapidly PFAs from local and/or global (shared) databases, facilitates comparison of data collected at various centers, and enables creation of the

PFA for various structures over the Internet by the neurosurgical community. This shifts a paradigm in atlas construction and extension from manufacturer-centric to community-centric. The PFA shows a distribution of the best contacts in image and volumetric representations calculated with a user-specified resolution. The PFAs were constructed for the STN and VIM with a high spatial resolution of 0.25 mm3 and

411

412

27


. Figure 27-8 PFA-VIM (the reference system and denotations as in > Figure 27-7): (a) normalized contacts, (b) normalized best contacts, (c) axial, coronal, sagittal color-coded maps and the voxelized atlas

accuracy of 0.25 mm. Consequently, the PFA is a volumetric atlas, consistent in 3D, and can easily be reformatted in any orientation. The PFA is dynamic and scalable. Any new cases can be added to the current PFA or various PFAs can easily be merged as the atlas function is linear (additive). Moreover, these calculations are fast. This enables a near real-time update of the PFA and its use in web-based applications. Finally, the PFA can easily be used for planning, see Section ‘‘Atlas use,’’ particularly in combination with the anatomical atlas (Sections ‘‘Combined anatomical-functional atlases’’ and ‘‘Atlas use’’).

Limitations of the PFA Although the PFA is a novel concept yielding a new type of atlas, its current implementation has several limitations in terms of data used, contact modeling, case normalization, and content.

The PFA does not contain the entire pre-, intra-, and postoperative data; neither does it include the complete electrophysiological findings from microrecording and stimulation. Therefore, attempts such as [66] aimed at storing all these data are of importance. However, to make this vast amount of information beneficial within neurosurgeon’s time constraints, it must be aggregated, knowledge extracted, and presented in a useful and easy way. The PFA, whose current version takes into account three symptoms only (akinesia, rigidity, and tremor), must generally be able to provide distribution of any individual symptom as well as any combinations of them. The atlas should be queried not only against a single value (probability at present) but also against any outcome, feature, and scoring system, such as UPDRS. In addition, it must be possible to mask the atlas with the regions that are positive (with improvement), negative (without improvement), and unexplored (not studied yet).


The current PFA algorithm assumes that the tissue activation region is limited to the shape of the normalized contact. Advances in imaging enable to visualize brain conductivity [77] and consequently predict the volume of tissue activated during deep brain stimulation (DBS) [78] on a patient-specific basis [79]. Subsequently this volume can be visualized in 3D along with the electrode, scan(s) and atlas(es), as e.g., in the Cicerone system [80] (Section ‘‘Research prototypes’’). Incorporation of any new geometric model of activated tissue into the PFA requires only substituting the algorithm for model voxelization, while the other calculations remain the same [68]. The brain normalization applied is limited to antero-posterior and dorso-ventral scalings, and lateral translation. It results from the anatomical features (AC, PC, HT, and V3) present in the acquired neuroimages and neurosurgery planning performed. Besides the speed in PFA normalization and fast mapping of the PFA onto patients’ scans, another advantage of this approach is that the optimal voxelization algorithm for the deformed (normalized) contact holds for this transformation [68], making the calculation of the entire PFA or its update fast. In principle, non-rigid normalization is feasible if a 3D scan is available, however, unpredictable and non-reproducible deformations in MRI [10] will limit the PFA accuracy (of 0.25 mm at present) and the current voxelization algorithm may not be valid any longer (whereas a direct voxelization of an arbitrarily deformed contact will take much more time). The PFA-GPi is not available yet. In addition, construction of PFAs for emerging targets, such as Brodmann’s area 25 [81] and pedunculopontine nucleus [82] has to be considered.

Combined Anatomical-functional Atlases Anatomical and functional atlases are complementary, so their combination is potentially useful. In

27

particular, the PFA and the SW atlas are highly complementary as the PFA, which is probabilistic, of a high spatial resolution (of 0.25 mm3), dynamic, composed from theoretically unlimited number of brain specimens, and consistent in 3D, complements a deterministic, sparse, static, based on two brains, and 3D inconsistent SW atlas. Conversely, a low parcellated PFA is enhanced by a highly parcellated SW atlas. Two approaches are feasible to combine the PFA with the SW atlas: (1) SW-centric by warping the PFA against the SW atlas [61], and (2) PFA-centric by warping the SW atlas against the PFA [83]. Both approaches were applied to the SW coronal and sagittal microseries; the SW axial plates were not considered as they are inclined by 7 (if they are needed, the corresponding PFA axial images suitably tilted must be generated first by interpolation). The TT atlas is (by construction) in spatial correspondence with the PFA created by normalizing its component cases to the AC-PC distance of 24 mm and HT of 16 mm (i.e., the same as in the TT atlas). Another effort towards building a combined anatomical-functional atlas is presented in [84]. The results for 28 patients with the implanted DBS electrodes in the STN were normalized nonrigidly [65] to form the postoperative maps of improvement in terms of the UPDRS increase. The maps were created for two groups of cases: less favorably (30% or higher improvement) and more favorably (50% or higher improvement).

SW-centric combined atlas The registration of the PFA to the SW atlas enhances the latter while not eliminating its limitations (Section ‘‘Limitations of anatomical atlases’’). The PFA was warped against the C-SW atlas by applying the same normalization transformation used for its construction to each brain separately (i.e., Brain LXVIII and Brain LXXVIII) for the normalization parameters determined in

413

414

27


[61]. The PFA was registered with the coronal microseries by scaling it dorso-ventrally by 19.4/ 16.0; the other two orientations remained unchanged. This is equivalent to stretching the size of the PFA voxel dorso-ventrally by 1.2125. The PFA was registered with the sagittal microseries by scaling its voxel dorso-ventrally by 19/16 and left-right by 23/24; the third orientation remained unchanged.

PFA-centric combined atlas The registration of the SW atlas to the PFA preserves spatial consistency of the PFA enabling planning in any orientation(s). It also reduces a SW inter-orientation spatial inconsistency. Moreover, its target structure dependence makes it more accurate. To register the C-SW atlas to the PFA, two criteria were applied: (1) normalization of the C-SW atlas, and (2) matching of the target structure’s centroid to the best target [83]. The C-SW atlas was treated as a component case of the PFA. Normalization of the C-SW atlas required its scaling in the antero-posterior and dorso-ventral directions to ensure the same AC-PC distance and HT in each atlas. The normalization parameters had been determined earlier [61]. The goal of the centroid-best target matching was to align the centroids of the target structures in the C-SW atlas with the locations of the best targets in the PFA. The scaling parameters were determined in the studies that quantified the main target structures STN [43], GPi [44], and VIM [45] in the C-SW atlas (see also Section ‘‘Limitations of anatomical atlases’’). The major difference between the C-SW coronal and sagittal microseries is in laterality, while the maximum difference between the posterior/anterior and ventral/dorsal centroid coordinates is small: 0.11 mm for the STN and 0.85 mm for the VIM. Hence, the centroid-best target matching was done laterally only and the other two

orientations were used for AC-PC and HT alignment. The lateral scaling factors differed considerably for the STN and VIM: the STN had to be stretched by 18% more than the VIM on the coronal microseries, and the VIM had to be compressed by 13% less than the STN on the sagittal microseries [83]. Therefore, the C-SW lateral scaling has to be target structure dependent (the other two scaling factors remain the same across target structures). The lateral scaling also matches (approximately due to the abovementioned small posterior/anterior and ventral/ dorsal mismatches) the target structure centroids for the C-SW coronal and sagittal microseries improving their mutual spatial consistency. The PFA-STN combined with the C-SW atlas is shown in > Figure 27-9. The PFA-VIM combined with the C-SW atlas is presented in > Figure 27-10.

Atlas-assisted Stereotactic and Functional Neurosurgery The usefulness of a ‘‘raw’’ atlas in SFN is limited. These are the atlas to scan registration and a battery of tools which make the atlas a valuable aid. Moreover, the deformable atlas along with these tools must be available to the neurosurgeon, preferable in a neurosurgical workstation.

Atlas to Scan Registration Overview Registration is an essential operation transferring the segmentation and labeling information from the atlas to the patient-specific data. Despite the existence of numerous techniques for brain warping, overviewed e.g., in [85,86], there is no nonrigid solution acceptable in clinical practice yet.


27

. Figure 27-9 PFA-STN combined with the C-SW atlas: (a) coronal view, (b) sagittal view. The probability is proportional to the gray scale and the locations of the images are shown in the top left corner

. Figure 27-10 PFA-VIM combined with the C-SW atlas: (a) coronal view, (b) sagittal view. The probability is proportional to the gray scale and the locations of the images are shown in the top left corner

The current practice in SFN it to use the Talairach transformation (Section ‘‘Talairach transformation’’) normalizing brains piecewise linearly [28] with a manual setting of the landmarks. This transformation is automated by developing the Fast Talairach Transformation which

maps the C-TT atlas onto a scan in 5 s [87] (Section ‘‘Fast Talairach transformation’’). Nonrigid brain warping methods are theoretically more accurate than piecewise linear warping and some of these methods are already employed in SFN [65,67,88], mainly for atlas construction.

415

416

27


Nonetheless, a landmark-based study [89] indicates that in the region between the AC and PC, where the main stereotactic targets are located, the difference between the linear and nonlinear (nonrigid) warping is practically negligible. The existing methods typically disregard the AC and PC, though the error due to their inaccurate localization is up to 0.5 cm [90]. The use of nonrigid methods is obviously advantageous in situations when the computational time is practically irrelevant, such as atlas construction, atlas-to-atlas pre-registration, or research applications. Their long computational time, however, is not acceptable in routine clinical procedures. Moreover, despite the increasing computing speed and development of more efficient brain warping methods, their ‘‘black box’’ nature limits their practical use in a clinical setting. The use of complex mathematical models and laws of physics with a little or no relevance to anatomy is barely to be understood and accepted. Unrealistically warped images do not help in trusting these methods. Finally, the need of setting multiple parameters prior to applying a method often restricts its use to its developers.

Talairach Transformation The Talairach transformation [28] normalizes brains piecewise linearly. It is based on the Talairach landmarks: two internal landmarks located on the midsagittal plane and six external landmarks lying on the smallest bounding box encompassing the cortex. The original Talairach point landmarks are: AC – the anterior commissure (point) is the intersection of the lines passing through the superior edge and the posterior edge of the anterior commissure (structure), PC – the posterior commissure (point) is the intersection of the lines passing through the inferior edge and the anterior edge of the posterior commissure (structure), L/R – most lateral point of the parietotemporal cortex for the

left/right hemisphere, A – most anterior point of the frontal cortex, P – most posterior point of the occipital cortex, S – most superior (dorsal) point of the parietal cortex, I – most inferior (ventral) point of the temporal cortex. The Talairach bounding box and the reference planes (i.e., the intercommissural plane, midsagittal plane, and coronal planes passing through the AC and PC) divide the brain into 12 regions (> Figure 27-12b). The Talairach transformation normalizes the brain by warping its scan within each region linearly to match the corresponding landmarks, resulting in an overall piecewise linear warping.

Fast Talairach Transformation To automate the Talairach transformation, the Fast Talairach Transformation (FTT), is developed which warps the C-TT atlas against a T1-weighted scan in 5 s [87]. The FTT exploits the modified Talairach landmarks [63] (see Section ‘‘Atlas use’’) and calculates them automatically in three steps: extraction of the midsagittal plane [91], identification of the AC and PC [92], and localization of the cortical landmarks [93]. Furthermore, the original Talairach transformation is extended by introducing two additional landmarks: the top of the corpus callosum and the most ventral point of the orbito-frontal cortex, and formulating an automated method for their calculation [94]. This extension, dividing the brain into 24 regions, enhances the quality of the FTT [94] and enables its use even when a complete brain axial scan is not available.

Atlas Use and Benefits The most often use of atlas is for targeting. The computerized atlases may provide additional benefits when employed properly and armed with powerful tools.


Atlas-derived best target points

27

PFA-determined accuracy of targeting

The computerized brain atlas enables determining the best stereotactic targets. An atlas-derived best target can be the mean value for a population of target points, the target structure’s centroid (or a centroid-related location), or the highest probability point or region for probabilistic atlases. The axial, coronal, and sagittal scatter plots of 54 DBS electrodes (of 29 patients) implanted into the STN are presented in [95]. The best contacts were those of being most effective during at least a 3 month postoperative follow-up. These plots allow for calculation of the populationbased (mean) best target which is (11.72, 1.62, 2.47), where the (right/left, posterior/anterior, and ventral/dorsal) coordinates (in mm) are in the Schaltenbrand reference system. The centroid coordinates of the 3D STN [43], 3D GPi [44], and 3D VIMe [45] in the CSW atlas are summarized in > Table 27-1. The PFAs versions constructed so far vary in terms of the number of contacts used and lateral compensation (applied or not). The mean best target points across these versions for the PFASTN and PFA-VIM determined in [83] are given in > Table 27-2.

As the atlas function is computable at any location, it is possible to calculate it in the neighborhood of the best target and study its behavior. The more level the function is (i.e., the wider ‘‘plateau’’ exists), the lower spatial accuracy of targeting is acceptable. A probability histogram (> Figure 27-11) shows that the decrease in probability around the best target voxel(s) is very high and, practically, the plateau does not exist [71]. Subsequently, a probability threshold was determined by comparing the size of the left and right PFA-STN [71]. Their ratio exhibits two behaviors: for low and medium probabilities it equals to one, and above probability of 0.77 it grows rapidly (up to 43 without lateral compensation against the V3 and up to 11 for lateral compensation), indicating a differentiation between the left and right STN in the regions, where the highest number of contacts were implanted (compare also > Figure 27-9a). This probability (of 0.77) divides the STN into the cold and hot STN, and the hot STN is taken as the target region. The size of the hot STN to that of the entire STN is between 1–2% indicating that targeting has to be done with a high spatial accuracy.

. Table 27-1 Centroid (right/left, posterior/anterior, ventral/dorsal) coordinates (in mm in the Schaltenbrand reference system) of the 3D STN, 3D GPi, and 3D VIMe in the C-SW atlas Orientation C-SW axial C-SW coronal C-SW sagittal

STN centroid (11.21, 0.64, 3.83) (9.18, 2.03, 3.95) (12.17, 1.92, 3.84)

GPi centroid (17.50, 5.44, 2.35) (15.58, 4.80, 0.11) (18.25, 4.06, 0.43)

VIMe centroid (13.77, 2.99, 2.35) (12.91, 5.66, 4.80) (16.65, 6.51, 4.91)

. Table 27-2 STN and VIM mean best target point coordinates (denotations as in > Table 27-1) PFA-STN mean best target Left

Right

(11.75,

( 11.75,

2.00,

2.75)

0.75,

2.75)

PFA-VIMe mean best target Left

Right

( 14.00,

( 14.00,

6.00, 1.75)

6.00, 1.00)

417

418

27


. Figure 27-11 Probability histogram of the PFA-STN calculated without lateral compensation against the V3. The length of each horizontal bar is proportional to the atlas volume for the probability determined by its color. Note that the high probability volumes (at the top) are very small in comparison to the low probability volumes (at the bottom). The probability color bar is on the right

Atlas use An early practical use of the stereotactic atlas for targeting and trajectory planning was to employ a single, usually sagittal, orientation scaled by means of a projector. Multiple complementary computerized brain atlases available in multiple orientations enable novel and beneficial ways of atlas use. Landmarks Stereotaxy requires a precise coordinate reference system. Therefore, the landmarks defining the coordinate system of a stereotactic atlas are critical. The error due to inaccurate localization of the AC and PC is up to 5 mm [90] and their lack may result even in a higher inaccuracy. In general, the landmarks must be defined uniquely, be easily identifiable in a scan, and their automatic identification should be computationally efficient. There are, however, several problems with the original landmarks in the SW and TT atlases as discussed in Section ‘‘Limitations of anatomical atlases.’’ To overcome these problems, a new, equivalent set of the Talairach landmarks was introduced (> Figure 27-12) and the resulting errors analyzed [63]. The new landmarks are defined in a more constructive way facilitating their efficient and automatic calculation. Two intercommis-

sural lines are defined on the midsagittal plane: central and tangential. The central intercommissural line is passing through the centers of the AC and PC structures, each approximated by a circle. The tangential intercommissural line is tangential dorso-posteriorly to the AC structure and ventro-anteriorily to the PC structure. In addition to the original Talairach AC and PC landmarks, three other AC and PC point landmarks (and the corresponding AC-PC distances) were introduced as follows: AC – is a point within the intersection of the anterior commissure (structure) with the midsagittal plane which can be: (1) central (the center of the anterior commissure structure), (2) internal (the most posterior point on the central intercommissural line), and (3) tangential (the tangent point of the tangential intercommissural line with the anterior commissure). PC – is a point within the intersection of the posterior commissure (structure) with the midsagittal plane which can be: (1) central (the center of the posterior commissure structure), (2) internal (the most anterior point on the central intercommissural


27

. Figure 27-12 Modified Talairach landmarks: (a) locations of the landmarks on three orthogonal planes: intercommissural, coronal passing thought the AC, and coronal passing through the PC, (b) reference planes and grid superimposed on a scan in 2D and 3D

line), and (3) tangential (the tangent point of the tangential intercommissural line with the posterior commissure). The locations of all modified Talairach landmarks are show in > Figure 27-12a. The modified landmarks can efficiently be calculated automatically (Section ‘‘Fast Talairach transformation’’). There are two factors speeding up their identification. First, two coordinates of the cortical landmarks are already fixed by definition so the third coordinate has to be calculated only. Second, in contrast to processing the entire brain, the identification of the modified landmarks is limited to three predefined planes only: intercommissural plane, and two coronal planes passing through the AC and PC landmarks, > Figure 27-12. This approximation is verified in [87]. The choice of the intercommissural line and distance is application-dependent. For SFN, the central intercommissural line and internal intercommissural distance are recommended [63].

The simultaneous use of axial, coronal, and sagittal orientations is advantageous and a computerized atlas is suitable for this purpose. The atlas structures can be presented on three cross-correlated images displayed individually (> Figure 27-2) or jointly (> Figure 27-6a) as a triplanar. This potentially enhances the accuracy of targeting by identifying and setting the landmarks on the orthogonal planes as well as increases neurosurgeon’s confidence. The detailed targeting steps for pallidotomy/ pallidal stimulation, thalamotomy/thalamic stimulation, and subthalamotomy/subthalamic stimulation performed simultaneously on all three orthogonal planes supported by global and local landmark-based registrations along with the selection of suitable local landmarks are given in [90]. The concept of global and local registrations is illustrated in > Figure 27-13. An example of a subthalamic stimulation targeted simultaneously on axial, coronal, and sagittal orientations by employing the C-SW Targeting on multiple orientations

419

420

27


. Figure 27-13 Global and local registrations of the C-SW coronal microseries in contour representation with an MRI scan. (a) Global (Talairach transformation-based) registration providing a correspondence between the atlas and the scan. The Talairach grid is attached to the contours and any corner of it is movable whose displacement causes a piecewise linear deformation of the contours in real-time. The globally registered head of the caudate nucleus and the optic tract in the atlas do not fit well the data. (b) Local registration enhancing the atlas-to-scan fit in the region of the GPi (highlighted) performed with the following landmarks: the head of the caudate nucleus to scale dorsally, and the optic tract to scale ventrally and laterally. This approach (illustrated on a single orientation) is applicable to all three planes simultaneously

atlas in contour representation is illustrated in > Figure 27-14. The target structure STN is delineated on all three orthogonal planes by the contours corresponding to atlas plates Hv 3.5, Fp 3.0, and S 12.0, and the target point is set manually to lie within all three contours. Typically a single stereotactic atlas is employed for planning. The use of multiple atlases with various complementary contents is beneficial by potentially improving the accuracy of targeting, increasing neurosurgeon’s confidence, and compensating for some atlas shortcomings [51,96]. Any orientation of a multi-atlas triplanar may come from any atlas. For instance, the triplanar may be formed from the SW axial and sagittal orientations and TT coronal orientation (as the SW coronal microseries differ significantly from the SW axial and sagittal microseries, Section ‘‘Limitations of anatomical atlases’’), > Figures 27-5a and > 27-15a. This approach also allows various parcellations and nomenclatures to be employed jointly for labeling, such as Hassler’s parcellation [32] in the SW Planning with Multiple Atlases

atlas and Walker’s parcellation [33] in the TT atlas, > Figure 27-15b. Atlas-assisted Preoperative Planning, Intraoperative Support, and Postoperative Assessment The most typical use of the stereo-

tactic atlas is for preoperative planning, mainly targeting. Intraoperatively, the atlas can also serve as a global positioning system and provide the neuroanatomy surrounding the target structure, list of structures along the stereotactic trajectory, and spatial relationships between the tip of electrode and critical structures, such as the optic tract. It also aids in data archival [57]. Postoperatively, the atlas assists in a spatial assessment of DBS/lesion placement [96]. The use of multiple complementary anatomical atlases in multiple orientations for preoperative planning, intraoperative support, and postoperative assessment is addressed in [51] and illustrated in > Figure 27-16. An integration of web technology with the PFA concept

Community-centric Atlas Construction


27

. Figure 27-14 Subthalamic stimulation targeted simultaneously in multiple orientations by means of the C-SW atlas in contour representation: (a) axial (plate Hv 3.5), (b) coronal (plate Fp 3.0), (c) sagittal (plate S 12.0). The target structure is highlighted, the reference axes are drawn, and the target point within the target structure is marked by a zoomable cursor

shifts a paradigm in atlas construction and extension from manufacturer-centric to communitycentric. This is enabled and illustrated by a community-centric PFA-based portal [69] providing an infrastructure for data collection and sharing as well as calculation of the PFA over the Internet, see also Section ‘‘Research prototypes.’’ The PFA is easily individualized to a patient’s scan by performing a transformation inverse to that applied for contact normalization as follows: (1) postero-anterior linear scaling to match the patient’s AC-PC distance,

PFA-based Planning

(2) ventro-dorsal linear scaling to match the patient’s HT, and (3) lateral translation matching the patient’s V3 (if a lateral compensation against the V3 was applied to create the atlas). The best target point is taken as the location with the highest probability (or, alternatively, for the STN as a point within the hot STN [71] (Section ‘‘PFAdetermined accuracy of targeting’’). The PFA may be superimposed on the patient’s scan such that the region of zero probability is rendered transparent and the non-zero probability region is displayed in gray scale proportionally to probability, see > Figure 27-19.

421

422

27


. Figure 27-15 Planning with multiple atlases. (a) Pallidotomy targeting with multiple atlases: C-SW atlas in axial (plate Hv 3.5) and sagittal (plate S 20.0) orientations, C-TT atlas in coronal orientation (plate TT 8.0), and 3D C-TT atlas with scan triplanar. The target point is marked by a zoomable cursor . (b) Multiple labeling in the thalamic region with two nomenclatures by Walker’s and Hassler’s

In the individualized anatomical-functional atlas, the SW atlas delineates the target structure and its surrounding neuroanatomy, while the PFA determines the target point within the target structure as the location with the highest probability (i.e., the highest density of the normalized best contacts for the previously operated cases). Consequently, the PFA enables precise targeting for the current and next tracks while the SWatlas serves as a global positioning system providing the neuroanatomy surrounding the target structure, list of structures traversed along the trajectory, and spatial relationships to critical structures. The anatomical-functional atlas created by the SW-centric approach (Section ‘‘SW-centric combined atlas’’) is probably easier acceptable as a well-known SW atlas is enhanced by the PFA. Its limitation is that planning on each orthogonal orientation has to be done separately. The PFA-centric approach (Section ‘‘PFA-centric combined atlas’’) produces a superior combined atlas allowing for planning with all orientations

Combined Anatomical-functional Atlas

simultaneously. In addition, it reduces spatial inconsistency among the SW microseries. The spatial correlation between the anatomical STN (from the SW atlas) and functional STN (PFA-STN) was studied quantitatively [73]. For probability p 0.3, more than 95% of the functional STN is inside the anatomical STN and for p 0.5 the complete functional STN is inside the anatomical STN. Therefore, the PFA-STN for p = 0.5 after registration to a scan can potentially be used for 3D identification of the STN in neuroimages substituting the SW atlas for this purpose.

Potential Benefits of Atlas Use The use of the stereotactic atlases as discussed above has several potential benefits, including reduced cost and time, reduced invasiveness, increased accuracy of targeting, increased neurosurgeon’s confidence, facilitated rapid planning, support for new procedures, facilitated


27

. Figure 27-16 Atlas-assisted preoperative planning, intraoperative support, and postoperative assessment: (a) subthalamic stimulation planning with multiple 2D and 3D atlases, (b) intraoperative support with the GPi (pallidum mediale internum) delineated and labeled, stereotactic trajectory (thin line) with the current position of the microelectrode (thick line) shown, and the structures traversed along this trajectory listed (left); and a 3D view of the microelectrode, target structure and atlas-scan triplanar (right), (c) postoperative assessment of a DBS placement in the GPi, (d) postoperative analysis of a thalamic lesion in all three planes

inter-clinician communication, and enabled building of community-centric solutions. The atlas notably lowers the surgical cost by reducing the duration of surgery by decreasing the number of stereotactic tracts. An initial evaluation suggested that the atlas could potentially reduce

the number of tracks from five to one per hemisphere [96] resulting in cost and time savings. This cost is further reduced by decreasing the number of microelectrodes inserted. This decrease also lowers the invasiveness of the surgical procedure, decreasing potential surgical complications by

423

424

27


reducing the risk of hemorrhage as well as surgically induced capsular and visual deficits. The atlas increases the accuracy of targeting by employing multiple, complementary atlases in multiple orientations, global and local registrations with any clearly visible landmarks set on three orthogonal planes, and sub-pixel accuracy measurement. This is further enhanced by the combined anatomical-functional atlas, such as the PFA co-registered with the SWatlas. Then, the PFA enables precise targeting (with a spatial resolution of 0.25 mm3) for the current and next tracks while the SW atlas serves as a global positioning system providing the delineated neuroanatomy surrounding the target structure, list of structures traversed along the track, and spatial relationships to critical structures. The individualized atlas increases neurosurgeon’s confidence pre-, intra-, and postoperatively by employing multiple atlases in multiple orientations, providing scan labeling, and allowing for an intraoperative measurement of distances to critical structures. This confidence is anatomical, functional, and spatial. Anatomically, the atlas delineates the structure of interest that, depending on acquisition, may be indiscernible or incompletely visible on the scan. The anatomical atlas also provides a detailed neuroanatomy labeled on all three orthogonal planes with a higher parcellation than that of the scan itself (> Figure 27-4f). Functionally, the atlas determines the target point within the target structure and provides the distribution of probabilities. Spatially, the atlas provides the triplanar, 2D-3D cross-correlation, and 3D relationships. The atlases and the approaches described above facilitate rapid planning, suitable for surgical procedures where the time between the scanning and the operation is short, so that the scanning, planning, and surgery can be done during the same session without removing the stereotactic frame. This also allows the neurosurgeon to plan more sophisticated trajectories by displaying the track on all three planes and in

3D, providing the list of structures traversed along it, and measuring distances to critical structures. The availability of the segmented and labeled Brodmann’s areas (BAs) in the C-TT atlas (> Figure 27-1b) facilitates localizing cortical areas, such as BA25 [81], or exploring them, for instance, BA4, BA6, BA24, BA32 [97]. While the understanding of an underlying neuroanatomy in the scan is easy for a neurosurgeon, communicating the individualized neuroanatomy to other clinicians may be tedious and time-consuming. A scan annotated by a deformable atlas is able to transfer this information to other clinicians including neurologists and neuroradiologists as well as to residents and medical students (> Figures 27-1f and > 27-4f). The concepts and initial solutions addressed here enable atlas construction in a communitycentric manner by enhancing and extending the PFAs for the existing and creating PFAs for new target structures (as discussed in Section ‘‘Limitations of the PFA’’). Finally, though the initial evaluation suggested that the atlas could potentially reduce the number of stereotactic tracks, further analysis and studies are required to quantify atlas benefits.

Atlas-based Applications Numerous atlas-based applications for SFN have been developed as research prototypes and commercial products. The use of atlas in these applications varies in numerous terms including:

Atlas(es) employed (2D SW, 2D TT, 3D SW, 3D TT, PFA, and/or others) Construction of the computerized atlas (from a directly digitized print plates to an enhanced, expanded, and 3D deformable atlas) Availability of all orthogonal orientations (available, not available)


Availability of multiple atlases (single atlas, multiple isolated atlases, multiple mutually co-registered atlases) Atlas representation (image, contour, polygonal, volumetric) Atlas-to-scan registration approach (interactive, automatic, mixed) Atlas-to-scan warping transformation (linear scaling, 3D piecewise linear scaling, nonrigid warping against an anatomically normal scan, nonrigid warping against a pathological scan) Structure labeling (not available, available at the border, available for the entire structure) Atlas display (atlas alone, atlas and scan sideby-side, atlas images overlaid onto a scan, atlas images overlaid onto a scan with usercontrolled blending, atlas contours overlaid onto a scan) Other supportive tools.

Research Prototypes Since the development of the first atlas-based tool in 1974 [19], numerous research atlas-assisted prototypes have been developed [69,80,96,98–103]. We briefly feature some of them. NeuroPlanner is an atlas-based software system for SFN that supports preoperative planning and training, intraoperative procedures, and postoperative analysis [96] (see > Figures 27-12b, > 27-13–> 27-16). NeuroPlanner contains multiple 2D and 3D Cerefy atlases mutually coregistered [50] with about 1,000 structures and 400 sulcal patterns. Numerous tools provide four groups of functions: data-related (data reading, interpolation, reformatting, image processing), atlas-related (real-time interactive atlas-to-data warping, multiple labeling in 2D and 3D), atlasdata exploration-related (interaction in three orthogonal and one 3D views, continuous dataatlas exploration), neurosurgery-related (targeting, path planning, measurements, simulating

27

electrode insertion, simulating therapeutic lesioning). NeuroPlanner along with the Cerefy atlases, trial licensed to several commercial, clinical, and research sites, has been playing an important educative role and influencing the design of commercial systems that incorporated the Cerefy brain atlases (Section ‘‘Neurosurgical workstations’’). BrainBench resulted from integration of the brain atlas with virtual reality [102]. It contains a suite of tools for SFN along with the 3D C-TT atlas. It employs a virtual workbench called the Dextroscope (see also Section ‘‘Future stereotactic environments’’) where the user reaches with both hands behind a mirror into a computergenerated 3D stereoscopic object, and moves and manipulates it in real time with natural hand movements. BrainBench facilitates preparing faster plans, provides a better and more accurate choice of target points, improves the avoidance of sensitive structures, has fewer sub-optimal frame attachments, and enables faster and more effective planning and training. The Cerefy Neuroradiology Atlas (CNA) [100] is a general purpose, public domain tool for rapid labeling and exploration of scans by means of the C-TT atlas warped by applying the Talairach transformation. The CNA provides interactive scan labeling, navigation on all three planes, zoomable triplanar, data-atlas blending, reading Talairach coordinates, mensuration, putting annotations, and drawing regions of interest. It saves an atlas-labeled and annotated scan in a Dicom or XML file for subsequent use by other clinicians or for presentations. A community-centric PFA-based portal for SFN [69] provides an infrastructure for data collection, sharing, and calculation of the PFA over the Internet. This portal links two (not necessarily exclusive) groups of neurosurgeons: these who are willing to share their data with those who would like to use data from others. A neurosurgeon is able to generate a customized PFA in three ways, as: (1) local PFA from the

425

426

27


neurosurgeon’s own data, (2) globally combined PFA from the data of others (shared in the global database), and (3) globally and locally combined PFA from the neurosurgeon’s own data combined with selected data of others. The portal provides tools for: (1) data input, transfer, and editing, (2) data selection for PFA creation, (3) PFA calculation, (4) PFA display and interactive manipulation, (5) target planning and probability histogram generation, and (6) parameter setting. This portal facilitates data sharing among functional neurosurgeons, calculates rapidly PFAs in three ways as stated above, facilitates comparison of data collected at various centers, and enables creation of the PFA for different structures over the Internet by the neurosurgical community (a server with this public domain portal was maintained for a few years; about 200 users registered, but no data was deposited into the global database for public use). An atlas-assisted software system described in [98] is dedicated to assist neurosurgical planning. It contains a 3D version of the TT atlas, is able to segment the cortex and ventricles in less than 5 min, and provides a nonrigid registration of the atlas to a patient’s MRI scan in a few minutes. This application is able to deform the atlas against brain tumors and is interfaced with a navigation system allowing for an intraoperative use of the atlas. CASMIL is a comprehensive, augmented reality software/toolkit with the Cerefy atlas for image-guided neurosurgeries [99]. It integrates a variety of modules and provides multiple tools for rigid and nonrigid registration (imageimage, image-atlas, and image-patient), automated 3D segmentation, brain shift prediction, knowledge-based querying, and intelligent planning. Brain shift is predicted by applying a patient-specific finite element model. CASMIL provides near real-time interaction with intraoperative MRI. It also has been securely webenabled and optimized for remote web and PDA access.

Cicerone is a software tool for stereotactic neuroelectrophysiological recording and DBS electrode placement [80]. It enables interactive 3D visualization of the co-registered MRI and CT scans, 3D brain atlases, neuroelectrophysiological microelectrode recordings, and DBS electrode(s) with the volume of tissue activated (VTA) as a function of the stimulation parameters. Preoperatively, for the intended anatomical target, Cicerone assists in selecting the optimal position on the skull for burr hole to maximize the likelihood of complete microelectrode and DBS coverage. Intraoperatively, it allows visualization of the electrode location in 3D relative to the surrounding neuroanatomy and neurophysiology. Moreover, Cicerone enables prediction of the VTA generated by DBS for a range of electrode trajectories and tip locations.

Commercial Products Two groups of atlas-assisted commercial products are featured below: neurosurgical workstations and CD-ROMs. Neurosurgical Workstations

Computerized brain atlases are commonly available in neurosurgical workstations. The Cerefy Electronic Brain Atlas Library and/or Cerefy Brain Atlas Geometrical Models are available in the StealthStation (Medtronic Surgical Navigation Technologies, Louisville, CO), Target and iPlan (BrainLAB AG, Feldkirchen, Germany), SurgiPlan (Elekta Instrument, Stockholm, Sweden), and SNN 3 Image-Guided Surgery System (Surgical Navigation Specialists, Mississauga, Ontario, Canada) as well as in the neurosurgical robot NeuroMate (Integrated Surgical Systems, Davis, CA). The Cerefy brain atlas libraries are also in a process of evaluation by Prosurgics (UK), Renishaw (UK), Cedara Software (Canada), and Z-KAT (USA). Other companies have developed their own digital versions of the SW and TT print atlases, including Tyco/Radionics


(Burlington, MA) and Stryker/Leibinger (Kalamazoo, MI). Electronic atlases are also available in the COMPASS System of Stereotactic Medical Systems [38] and in the CASS system of MIDCO [35]. Illustrations of the Cerefy atlases in the StealthStation and iPlan are shown in > Figure 27-17. CD-ROM Applications

Several atlas-assisted applications are available on CD-ROM. We feature some of them below. The ECBA [54] is a reference tool with the C-SW and C-TT atlases. It provides many features not available in the original atlases, including co-registered atlases; flexible display, manipulation, and printing of atlases in multi-atlas and triplanar modes; and about 17,000 structures pre-labeled on 1,500 atlas images. The foreword to this atlas was written by Dr. Jean Talairach (> Figure 27-18). The ECBA allows the individualized atlas to be generated without loading a scan and provides a simple targeting procedure [104]. The C-SW atlas is fit to the scan by means of 2D local deformation. First a rectangular region of interest (ROI), set among any clearly visible landmarks, is measured on the film or scanner console. The corresponding atlas image with the target structure is then deformed in real-time for the same landmarks such that the dimensions of the atlas and film ROIs are the same. The target point is then set on the individualized atlas image and its coordinates are read. The individualized image printed on a transparent foil can be overlaid on the film or, alternatively, this superimposition can be done electronically. The Electronic Clinical Brain Atlas (ECBA)

Functional imaging is an established technique in neurosurgery for studying the brain in health and disease. Identifying multiple activation loci on numerous functional images, determining their underlying cortical and subcortical anatomy, and reading their coordinates along with Brain Atlas for Functional Imaging (BAFI)

27

anatomical and functional values is a tedious, time-consuming, and error-prone task. The BAFI [55] was developed to facilitate this task by providing rapid atlas-assisted analysis of functional images [60] along with a locus-driven mechanism [105]. This CD-ROM allows the user to load anatomical and functional datasets, co-register them and place in the Talairach space, identify activation loci and label them automatically with Brodmann’s areas, gyri, and subcortical structures by using the C-TT atlas. Numerous tools are available for identification of activation loci, placing marks on them, editing and labeling of marks, and saving the results in electronic format. Cerefy Clinical Brain Atlas: Enhanced Edition with Surgical Planning and Intraoperative Support (CCBA-Plan) The CCBA-Plan [57] contains the

C-SW atlas, C-TT atlas, and PFA-STN. It provides standard atlas-related operations (> Figures 27-2 and > 27-5a) as well as is equipped with numerous tools for planning and intra-operative support by means of the combined anatomicalfunctional atlas. For a loaded patient-specific image (MRI, CT, or X-ray ventriculography), the CCBA-Plan allows the neurosurgeon to: warp the C-SW atlas in 1D, 2D, or 3D as well as translate, rotate and flip the image; plan the target point on an individualized C-SW atlas; plan the entry point on an individualized C-SW atlas by providing either its coordinates or the angles and distance to the target point; plan up to five electrode tracks; display trajectories in two orthogonal views; simulate electrode insertion with a label and PFA-STN probability readout at the electrode tip as well as provide an atlas display guided by tip movement; annotate the electrode track with recording and stimulation findings, label, and distances to the target and the intercommissural plane; and save the annotated electrode to an external file. These features allow the neurosurgeon to collect findings intraoperatively in electronic format and use the CCBA-Plan as his/her own local archive. Tra-

427

428

27


. Figure 27-17 Cerefy atlases in commercial neurosurgical workstations: (a) StealthStation: intraoperative use of the C-SW atlas in all three orientations (image courtesy of Dr. J. Henderson), (b) iPlan: trajectory verification based on side-by-side viewing of the scaled C-SW images superimposed on the corresponding reconstructions of a patient’s scan (image courtesy of T. Schwan)


27

. Figure 27-18 Demonstration of the ECBA (right bottom corner) by the author to Drs. Talairach (center) and Tournoux (right) at Saint Anne Hospital in Paris in 1996

jectory planning and simulation of electrode insertion are shown in > Figure 27-19.

new environments providing paradigm shifts in SFN.

Future Directions Tremendous technological advances impacting neurosurgery [106], particularly in information and communication technology, nanotechnology and molecular imaging, will drive the progress in SFN. This progress will also impact atlasassisted SFN. As the role of the atlas in SFN has been evolving, we predict that its importance and usefulness will be growing. We believe that the future efforts in atlas-assisted SFN should be carried out in three following directions: 1.

2.

3.

Construction of more accurate, more detailed, high resolution volumetric, multimodal probabilistic atlases at different scales Development of faster, more accurate, reliable, and automatic methods for atlas to data registration along with their validation Development of novel therapeutic procedures, more powerful applications, and

Future Stereotactic Atlases We believe that the future atlas for SFN must: be population based; contain anatomy, function (including neuroelectrophysiology), connectivity, and vasculature along with their variability; cover the entire brain and represent it at different scales (gross neuroanatomy, cytoarchitecture, myelination patterns, neurochemistry, neuroreceptors, gene expressions); provide distributions of the best targets in function of the symptoms treated, neurological findings, and clinical outcomes; differentiate regions into positive (with quantitative improvement), negative (without improvement), and unexplored (not studied yet); be self-updated dynamically and continuously with the new cases processed; be fully segmented and labeled, and highly parcellated; be consistent in 3D; be volumetric of high resolution; be correct and

429

430

27


. Figure 27-19 CCBA-Plan. (a) Trajectory planning by the combined anatomical-functional atlas. Three trajectories are set. The probability of the target point is displayed in the bottom left corner. (b) Simulation of electrode insertion on a sagittal ventriculogram. The distance to the target as well as the probability and structure name at the tip of electrode are displayed. The combined atlas is fit to the image by performing translation, rotation, and 3D scaling. The PFA images are rendered transparent for zero probability


validated; and provide powerful tools supporting its use and dynamic growth. The future atlas should be updated dynamically and continuously with the clinical cases, ideally globally without geographic, political, and/or vendor-imposed constraints (atlas without borders). The efforts on the PFA [68] and the community-centric PFA-based portal [69] partly illustrate this direction. The architecture of the future stereotactic multiatlas is presented in > Figure 27-20 (modified after [83]). To enable atlas potential in clinical practice, it must be equipped with numerous tools supporting a broad variety of operations including: fast and accurate registration (for atlas-to-data, datato-atlas, data-to-data, and atlas-to-atlas mappings), labeling with multiple features (reading and placing of meta labels); planar and curved reformatting (in the orthogonal planes, electrode planes (the ‘‘probe eye’s view’’), and arbitrary planes); rendering (triplanar display, and surface and volume rendering); exploration (of atlases, data, and other associated materials); incorporation and processing of clinical and imaging results and studies; readout (of stereotactic coordinates, probabilities, and scoring systems); quantification (of distances, angles, areas, and volumes); selection of any atlas subset (volume

. Figure 27-20 Architecture of the future stereotactic multi-atlas

27

of interests) according to given features/criteria; representation conversion (for image, contour, polygonal, and volumetric representations); and handling specific applications and/or devices (e.g., stereotactic frames, navigation devices). Towards constructing this type of atlas, a multi-atlas was proposed and its first version presented composed of the PFA, interpolated C-TT atlas, and enhanced C-SW atlas, mutually co-registered [83], see > Figures 27-3b, > 27-9, and > 27-10. Recently, new atlases are being constructed which eventually may find some future applications in SFN, including a probabilistic atlas of the basal ganglia and (yet unparcellated) thalamus [107], Cerefy atlas of cerebral vasculature with 3D arterial and venous systems and their variability [108], population-based atlas of white matter tracts [109], atlases of blood supply territories [110,111], and 3D probabilistic atlas of the cortical structures [112]. Though this review is limited to stereotactic human brain atlases, development of new, 3D, high resolution, stereotactic animal atlases for experimental surgical procedures is also important [113].

431

432

27


Atlas to Data Registration Registration is a key technique enabling the growth of atlas use. We believe that a clinically usefully registration must be rapid, reliable, accurate, automatic, and validated in a clinical setting. Warping techniques should handle anatomically normal and pathological cases, including brain tumors causing significant mass effect. Some initial solutions able to warp the atlas against a tumor are already proposed [98,99,114], providing the deformed neuroanatomy surrounding the tumor. Moreover, this tumor deformed neuroanatomy has to be correlated with tractography [115]. Validation of registration techniques in SFN is critical and efforts such as [116] are of importance.

Future Stereotactic Environments The future atlases with clinically accepted warping techniques will enable novel procedures and open new avenues provided that suitable applications and environments will be developed. This development should tackle data explosion, novel therapeutic procedures, and conceptually new stereotactic environments enabling paradigm shifts in SFN. To manage a rapidly growing amount of patient’s and atlas’ data, new ways of visualization of and interaction with the data are necessary. A potentially useful solution may be the Dextroscope originating from BrainBench (Section ‘‘Research prototypes’’) as a research prototype for SFN, extended subsequently to tumor stereotaxy and vascular malformations in [117], and becoming at present a product applied in various neurosurgical procedures [118,119]. Development of new procedures, particularly targeted therapy delivery (of drugs, antibodies, and biological agents) including gene/cell therapy (with a new wave of compounds under development), will require more detailed and parcellated atlases at different scales and very accurate warping techniques.

In a longer term, fundamental paradigm shifts must be proposed and new neurosurgical environments developed. We believe that these paradigm shifts should be along two directions: technology-related and patient-related. A technology-related paradigm shift is addressed in [106] by proposing a futuristic environment called DOTELL. The current surgical environment is both device- and information-centric. With a technological progress, the neurosurgeon will (sooner or later) be overwhelmed by data and equipment explosions, unless this burden is taken over by some intelligent assisting environment capable of acquiring, integrating, and processing the entire data as well as controlling and handling all instrumentation. This assistant should isolate the neurosurgeon from the instrumentation and information, and be able to perform two main groups of functions: TELL or show me and DO it (hence DOTELL). The architecture of DOTELL and its construction via technology integration are presented in [106]. DOTELL is an intelligent atlas-assisted device with robotic capabilities able to integrate hospital infrastructure, imaging systems, knowledgebased decision support, and therapeutic modalities. An example of DOTELL assisting a neurosurgeon in a bilateral subthalamic stimulation is given in [106]. It performs a broad spectrum of actions ranging from preoperative patient scheduling, scanning and surgery planning to intraoperative supervised robotic-based execution and visualization to postoperative ICU management and scanning. The second major change concerns the patient. We believe that it is the patient who should make all major decisions regarding his/her health, life, past, and future. The neurosurgeon armed with his/her knowledge, experience, and technology should assist the patient in his/her choices. This patient-centric concept of do-ityourself-neurosurgery is illustrated in the appendix as a futuristic neurosurgical procedure. Though it may still sound as science fiction (as


a head-mounted multi-modal scanner, knowledge and skill mapping, and memory preservation do not exist yet), nonetheless many other components are available today. In either technology-related or patient-related paradigm shift, the role of the atlas is tremendous: to equip a device with domain knowledge and ability to educate the patient on his/her specific demand. To continue the atlas growth and to keep exploiting its potential, the clinicians, researchers, and engineers have to work closely together.

Acknowledgments I am deeply grateful to Drs. J Talairach and P Tournoux for insightful discussions on their atlases. This work might not have been advanced without their initial enthusiasm about our atlases. I am truly indebted to Drs. AL Benabid, Grenoble, TT Yeo, Singapore, and AM Lozano, Toronto for the stimulating discussions and opportunity to observe their procedures. I was also inspired by the presentations and discussions at the 1997, 2001, and 2005 meetings of the World Society for Stereotactic and Functional Neurosurgery. The creation of the PFA and the PFA-portal for SFN was a joint effort with Dr. AL Benabid. The NeuroPlanner was developed within a joint project with Dr. TT Yeo of Tan Tock Seng Hospital, Singapore. The construction of the first electronic version of the TT atlas and development of the ECBA was a joint project with Dr RN Bryan of Johns Hopkins Hospital. The BAFI was developed in consultation with Dr. DN Kennedy of Massachusetts General Hospital. > Figure 27-17a of StealthStation is courtesy of Dr. J. Henderson of St. Louis University Health Sciences Center (now at Stanford University Medical Center). > Figure 27-17b of iPlan is

27

courtesy of Thomas Schwan of BrainLAB; iPlan is a registered trademark of BrainLAB AG in Germany and the US. Many ideas, solutions, atlases, and applications proposed would not materialize without their efficient implementation. Numerous persons from our Biomedical Imaging Lab, A*STAR, Singapore contributed to the development of tools for atlas construction and atlas-assisted applications, including A Thirunavuukarasuu (brain atlas CD-ROMs), D Belov (PFA, PFA-based portal for SFN, CNA), A Fang (atlas tools, 3D TT, initial version of NeuroPlanner), BT Nguyen (atlas tools), J Liu (interpolation, modeling, 3D TT), GL Yang (NeuroPlanner), L Serra (BrainBench), KN Bhanu Prakash (FTT), QM Hu (FTT), and GY Qian (FTT). I thank Aminah Bivi for her editorial assistance. This work has been funded by A*STAR, Singapore.

Appendix: Unlocked Brain: Do-it-yourself Neurosurgery Mr. Green, a successful venture capitalist in wireless communication, entered the famous Brainsterium. He had developed a malignant brain tumor as confirmed by a non-excisional optical biopsy. His feelings were mixed. He had expected to undergo a gene therapy, but he was offered a classic tumor resection. His colleagues in life sciences invested heavily in gene therapy, so he expected to benefit out of it. He realized that none of them invested in image-guided surgery. Thus, he felt that an obsolete technology would decide about his life. On the other hand, the Brainsterium had a brand name and an impeccable reputation to be the best. His first impression upon entering the Brainsterium was surprisingly positive. He rather expected a terrible hospital smell, miserable patients being moved around, and a noisy crowd of visitors, as he had experienced while

433

434

27


visiting a hospital the last time when his father had got a stroke. What he saw instead looked like a quiet high-tech lab. He headed towards the office of the chief neurosurgeon, Dr. Noki, who would explain to him the procedure and supervise the surgery. Supervise, not operate; this was the major difference that the Brainsterium offered to its patients. ‘‘It is you who decides what to keep in and what to remove from your brain. We just provide the right environment to do it’’ – started Dr. Noki. Mr. Green looked surprised. Do-it-yourself neurosurgery? He had no idea about this market segment. ‘‘Your brain will be unlocked and you will be able to view its content, such as knowledge, skills and memories, and examine how they are being invaded by the tumor’’ – continued Dr. Noki. ‘‘This surgery has two contradictory goals. One is to destroy all tumorous cells completely and the other is to maximally preserve the functions of your brain. I will provide you with two extreme brain resection regions: the conservative region with the core tumor only, and the aggressive region that contains the core tumor along with all tumorous cells that have migrated away from it. You have to balance between them to plan your postoperative life. When all tumorous cells are completely removed, the chances of physical survival are higher. This is the best for your body, but not necessarily for your mind and career. Preserving maximally your brain functions sounds more attractive but it puts your life at a higher risk.’’ Mr. Green suddenly visualized his brain as a financial asset and things became clear. As the exclusive owner of his brain and its content, he himself wanted to have a full control over this asset and decide about the associated risk. ‘‘Someone’s brain is much more valuable than his bank account’’ – he thought, ‘‘so why for decades has this been working differently?’’ He was more and more eager to understand this worth-investing technology. He once had

run an R&D department before becoming a venture capitalist. Dr. Noki provided more operational details. ‘‘You will be given access to our patients’ database and have permission to communicate with anyone who underwent this type of surgery. If you decide to proceed and accept our terms and conditions, you will be allowed to access all tutorials and simulators, and you can play back any previous surgery with, of course, no access to the brain contents of our patients.’’ ‘‘In the next step’’– continued Dr. Noki – ‘‘your brain will be unlocked by measuring its magnetic, electrical, chemical, and optical properties using a battery of techniques. They will produce the images of anatomy, vasculature, connectivity, function, pathology, and knowledge in your brain. If you are interested in technical details, refer to our tutorials. The extent of the tumor will be defined, and the conservative and aggressive resection regions prepared for you. You will be trained to understand the images employed to plan the resection. These images show tissue at the micron’s scale and at this resolution it is easy to distinguish normal from tumorous cells. The content of the resected brain region can be partly recovered. Your memories will be retrieved and saved on a disk. Play it later at home so that your brain will restore these memories in new locations. Remapping of the skills and knowledge is still at an experimental stage. At present, the content of the resected brain will be recorded and stored. We will acquire the knowledge distribution map of your brain later to find suitable locations for placing back the recovered content. There is a chance that some day, with the advancements in knowledge remapping and brain reconnection technologies, your skills and knowledge will be fully recovered. Finally, the tumor will be ablated with the collimated ultrasonic scalpel and removed without opening your skull. It will be dissolved and sucked out through the vasculature. All actions and operations will be


controlled by our revolutionary surgical environment DOTELL.’’ The introduction was over. Mr. Green was ready for this fascinating journey. He began it from the Resource Center ushered by a lion-robot. He was requested to put his fingers on the scanning plate and look into a camera to capture his biometrics. Mr. Green logged into the patients’ database and entered his personal wavelength. He had quite a broad band, which substantially accelerated operations. First, he registered with the system and displayed the list of patients who underwent a similar procedure. There were several thousands of them. No mortality, no technical failure during surgery; neurological deficits were quite variable, however. Whom to ask? He entered his year and place of birth. There was a familiar name, Jack Case. They had been schoolmates in grade six before his family moved to the West Coast. Mr. Green requested a videoconferencing session. He was lucky. Jack was in his garden and Mr. Green recognized his old friend. Jack had chosen the complete tumor removal 2 years back. He quit his job and was spending his days tendering his oceanside garden. Today Jack would opt differently. Mr. Green terminated the session and was led to the Brain Unlocking Center. A pretty nurse with an east European accent welcomed him. He was asked to provide a detailed list of his skills and related knowledge. The questionnaire was quite boring but Mr. Green realized that it was critical for an accurate planning of his brain stimulation and knowledge mapping. Next, the nurse put a bulky helmet on his head. ‘‘This must be that famous BCC, Brainsterium’s collector and collimator – one of the key unfair advantages of the Brainsterium. How did they manage to design this three-in-one gadget able to acquire multi-modal data and to collimate myriad of energy sources dynamically providing a non-invasive access to any location in the brain at micron’s accuracy for stimulation and excision?’’ – wondered Mr. Green. While

27

stimulated, he experienced unusual sensations. He saw some strange visual effects, heard funny voices, smelled oriental plants, and had an impression he was flying while the angels were singing. ‘‘No, not yet’’ – he said. ‘‘This is just a knowledge mapping procedure.’’ He looked at his body tightly attached to the chair. His excitement rocketed when he entered the Brain Exploration Center. It resembled a cyber cafe he had used to visit with his son a long time ago. Patients with helmets sitting in cubicles appeared playing games and navigating through some mazes. But everyone played seriously, as he could win or lose his past and future life. Mr. Green entered a cubicle and touched the start button on the screen. A welcome message with his name appeared and a colorful image of his brain showed up. ‘‘It recognizes my biometrics’’ – he thought. Three available functions were displayed, ‘‘explore your brain,’’ ‘‘plan your surgery,’’ and ‘‘preserve your memories.’’ Mr. Green started with the first one. He was astonished with the ease he could navigate his brain and how the Cerefy Atlas was able to give him the name of any tiny structure along with description of its function. Mr. Green began to appreciate its potential. It was time to start doing the job. He touched the second button. The nurse appeared and demonstrated how to distinguish on the images the normal from tumorous cells, and how to edit the resected region. It was quite easy with the Dextroscope stereoscopic display and a 3D reach-in, tactile user interface differentiating normal from pathological tissues. His hands reached into the brain space and worked as the tools reshaping the resected region. His future was really in his hands. He did not realize that he was the only person in the whole Brainsterium authorized to change the resection plan, as the system was monitoring the user’s biometrics. He started the inventory of his brain in the area of difference between the conservative and

435

436

27


aggressive resection regions. Some skills were there that he might lose, such as climbing, driving, and playing the piano. He had given up his dream to climb K2 a long time ago. His personal driver was doing well so he would keep him. Playing the piano – no compromise. He kept reshaping the region to be resected. The good-bye part of his brain was finally defined. Its knowledge and skills would be attempted to be recovered in a postoperative process. Now to memories and Mr. Green touched the last button. The button-called nurse appeared, put a BCC helmet on his head, and activated an array of transcranial magnetic stimulators. ‘‘I am lucky’’ – he thought – ‘‘she might have been a robot.’’ He projected his brain’s image, positioned the pointer within the region to be resected, and pressed the stimulation button. Nothing happened. He changed location and pressed the button again. Now he was watching the Titanic movie with his first love. He kept on pressing. It worked as a time machine moving him back to distant events and places. He could hardly believe that there were so many memories in such a small piece of tissue. Every memory he evoked was recorded. After the surgery, he would just play back any recorded piece to re-enter it into his brain. The surgery plan was completed. He touched the submit button and was asked to confirm the plan and accept the legal statement. The session was terminated and he was invited for tea. The surgery would start in half an hour’s time. He loved this stuff. ‘‘I have got to subscribe to the Brainsterium Club, so I can come here every weekend for some brain surfing and unlocking’’ – he thought. His new friend, the lion-robot ushered him to the Operating Rooms area. In this high-tech environment it looked so classic and trustworthy. Dr. Noki and the pretty nurse were already there. Mr. Green laid down on the operating table, a BCC helmet was put on his head, and some

monitoring probes attached to his body. ‘‘Are you ready?’’ – asked Dr. Noki and added – ‘‘Do not be afraid. Though the whole procedure is fully automatic, I will be controlling its every step.’’ Mr. Green pressed the start button initiating his own neurosurgery. The stereoscopic image of his brain was projected directly into his retinas and the resection plan prepared by him appeared. Initially, numerous sparkles surrounding the core tumor were visible. Later, he saw blood vessels feeding the core tumor being closed and the tumor separated from its surrounding tissues. Mr. Green felt an injection. He noticed a wire going towards the tumor through the biggest blood vessel which remained still open. ‘‘This has to be a catheter’’ – he recalled. A magnetic system guided its tip automatically towards the tumor. The tip reached the tumor. A small balloon was inflated closing the vessel. The dissolved tumor tissues started disappearing fast. The space previously occupied by the core tumor kept on shrinking. Finally, the last blood vessel was closed. The surgery was over. It was a long, eye-opening day for Mr. Green as a patient and investor. He had to stay overnight at the Brainsterium under monitoring. Scanning was being performed automatically on continuous basis. Everything was normal as expected. In the morning Mr. Green was discharged. He went into Dr. Noki’s office and looked at him in a way that only a few multi-billionaires and CEOs deserved so far. ‘‘So far, so good’’ – Dr. Noki welcomed him. ‘‘We still have some work to do to recover the content of the resected part of your brain. A disk was handed to Mr. Green. ‘‘My preserved memories’’ – he thought. Mr. Green was requested to keep on monitoring at home. He was given a wearable monitor capable of transmitting his scans from his home to the Brainsterium wirelessly. Finally he found some of his contribution.


Mr. Green’s driver brought him back home. He realized that despite many urgent messages, he made one call only asking his secretary to donate anonymously to the Brainsterium’s R&D Center. He sat at his old grand piano and started playing his favorite pieces. He was quite happy with his technical performance and got an impression that he played even with a greater passion than before. Now he knew what he was going to invest in.

References 1. Alexander E, III, Nashold BS, Jr. A history of neurosurgical navigation. In: Alexander E, III, Maciunas RJ, editors. Advanced neurosurgical navigation. New York: Thieme; 1999. p. 3-14. 2. Ashkan K, Blomstedt P, Zrinzo L, Tisch S, Yousry T, Limousin-Dowsey P, Hariz MI. Variability of the subthalamic nucleus: the case for direct MRI guided targeting. Br J Neurosurg 2007;21(2):197-200. 3. Kosta P, Argyropoulou MI, Markoula S, Konitsiotis S. MRI evaluation of the basal ganglia size and iron content in patients with Parkinson’s disease. J Neurol 2006;253 (1):26-32. 4. Reich CA, Hudgins PA, Sheppard SK, Starr PA, Bakay RA. A high-resolution fast spin-echo inversionrecovery sequence for preoperative localization of the internal globus pallidus. Am J Neuroradiol 2000;21(5): 928-31. 5. Sather MD, Patil AA. Direct anatomical localization of the subthalamic nucleus on CT with comparison to Schaltenbrand-Wahren atlas. Stereotact Funct Neurosurg 2007;85(1):1-5. 6. Starr PA, Vitek JL, DeLong M, Bakay RA. Magnetic resonance imaging-based stereotactic localization of the globus pallidus and subthalamic nucleus. Neurosurg 1999;44(2):303-13. 7. Zhu XL, Hamel W, Schrader B, Weinert D, Hedderich J, Herzog J, Volkmann J, Deuschl G, Mu¨ller D, Mehdorn HM. Magnetic resonance imaging-based morphometry and landmark correlation of basal ganglia nuclei. Acta Neurochir 2002;144(10):959-69; discussion 968-9. 8. Slavin, KV, Thulborn, KR, Wess C, Nersesyan, H. Direct visualization of the human subthalamic nucleus with 3T MR imaging. Am J Neuroradiol 2006;27(1):80-4. 9. Spiegelmann R, Nissim O, Daniels D, Ocherashvilli A, Mardor Y. Stereotactic targeting of the ventrointermediate nucleus of the thalamus by direct visualization with high-field MRI. Stereotact Funct Neurosurg 2006;84(1): 19-23.

27

10. Benabid AL, Koudsie A, Benazzouz A, Le Bas JF, Pollak P. Imaging of subthalamic nucleus and ventralis intermedius of the thalamus. Mov Disord 2002;17 Suppl 3:S123-9. 11. Cuny E, Guehl D, Burbaud P, Gross C, Dousset V, Rougier A. Lack of agreement between direct magnetic resonance imaging and statistical determination of a subthalamic target: the role of electrophysiological guidance. J Neurosurg 2002;97(3):591-7. 12. Dormont D, Ricciardi KG, Tande´ D, Parain K, Menuel C, Galanaud D, Navarro S, Cornu P, Agid Y, Yelnik J. Is the subthalamic nucleus hypointense on T2-weighted images? A correlation study using MR imaging and stereotactic atlas data. Am J Neuroradiol 2004;25(9):1516-23. 13. Richter EO, Hoque T, Halliday W, Lozano AM, SaintCyr JA. Determining the position and size of the subthalamic nucleus based on magnetic resonance imaging results in patients with advanced Parkinson disease. J Neurosurg 2004;100(3):541-6. 14. Coffey RJ. Stereotactic atlases in printed formats. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. p. 237-48. 15. Germano IM, Weisz DJ. Computer technology as an adjuvant for target location and validation: In: Israel Z, Burchiel KJ, editors. Microelectrode recording in movement disorder surgery. New York – Stuttgart: Thieme; 2004. p. 152-63. 16. Stancanello J, Romanelli P, Modugno N, Cerveri P, Ferrigno G, Uggeri F, Cantore G. Atlas-based identification of targets for functional radiosurgery. Med Phys 2006;33(6):1603-11. 17. Yelnik J, Damier P, Demeret S, Gervais D, Bardinet E, Bejjani BP, Francois C, Houeto JL, Arnule I, Dormont D, Galanaud D, Pidoux B, Cornu P, Agid Y. Localization of stimulating electrodes in patients with Parkinson disease by using a three-dimensional atlas-magnetic resonance imaging co-registration method. J Neurosurg 2003;99(1):89-99. 18. Speigel EA, Wycis HT. Stereoencephalotomy: part I. Methods and stereotactic atlas of the human brain. New York: Grune and Stratton; 1952. 19. Bertrand G, Olivier A, Thompson CJ. Computer display of stereotaxic brain maps and probe tracts. Acta Neurochir Suppl 1974;21:235-43. 20. Nowinski WL, Benabid AL. New directions in atlasassisted stereotactic functional neurosurgery. In: Germano IM, editor. Advanced techniques in image-guided brain and spine surgery. New York: Thieme; 2002. p. 162-74. 21. Talairach J, David M, Tournoux P. Atlas d’anatomie stereotaxique des noyaux gris centraux. Paris: Masson; 1957. 22. Schaltenbrand G, Bailey W. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Thieme; 1959.

437

438

27


23. Andrew J, Watkins ES. A stereotaxic atlas of the human thalamus and adjacent structures. A variability study. Baltimore: Williams and Wilkins; 1969. 24. Van Buren JM, Borke RC. Variations and connections of the human thalamus. Berlin: Springer; 1972. 25. Schaltenbrand G, Wahren W. Atlas for stereotaxy of the human brain. Stuttgart: Thieme; 1977. 26. Szikla G, Bouvier G, Hori T. Angiography of the human brain cortex: atlas of vascular patterns and stereotactic localization. Berlin: Springer; 1977. 27. Afshar E, Watkins ES, Yap JC. Stereotactic atlas of the human brainstem and cerebellar nuclei. New York: Raven Press; 1978. 28. Talairach J, Tournoux P. Co-planar stereotactic atlas of the human brain. Stuttgart – New York: Thieme; 1988. 29. Ono M, Kubik S, Abernathey CD. Atlas of the cerebral sulci. Stuttgart: Thieme; 1990. 30. Talairach J, Tournoux P. Referentially oriented cerebral MRI anatomy: atlas of stereotaxic anatomical correlations for gray and white matter. Stuttgart: Thieme; 1993. 31. Morel A, Magnin M, Jeanmonod D. Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387:588-630. 32. Hassler R. Anatomy of the thalamus In: Schaltenbrand G, Bailey W, editors. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Thieme; 1959. p. 230-90. 33. Walker AE. The primate thalamus. Chicago: University of Chicago Press; 1938. 34. Federative Committee on Anatomical Terminology (FCAT). Terminologia anatomica. Stuttgart – New York: Thieme; 1999. 35. Hardy TL, Deming LR, Harris-Collazo R. Computerized stereotactic atlases. In: Alexander E, III, Maciunas RJ, editors. Advanced neurosurgical navigation. New York: Thieme; 1999. p. 115-24. 36. Kazarnovskaya MI, Borodkin SM, Shabalov VA. 3-D computer model of subcortical structures of human brain. Comput Biol Med 1991;21:451-7. 37. Yoshida M. Three-dimensional maps by interpolation from the Schaltenbrand and Bailey atlas. In: Kelly PJ, Kall BA, editors. Computers in stereotactic neurosurgery. Boston: Blackwell; 1992. p. 143-52. 38. Kall BA. Computer-assisted stereotactic functional neurosurgery. In: Kelly PJ, Kall BA, editors. Computers in stereotactic neurosurgery. Boston: Blackwell; 1992. p. 134-42. 39. Niemann K, Naujokat C, Pohl G, Wollner C, von Keyserlingk D. Verification of the Schaltenbrand and Wahren stereotactic atlas. Acta Neurochir 1994;129 (1–2):72-81. 40. Sramka M, Ruzicky E, Novotny M. Computerized brain atlas in functional neurosurgery. Stereotact Funct Neurosurg 1997;69:93-8. 41. Niemann K, van den Boom R, Haeselbarth K, Afshar F. A brainstem stereotactic atlas in a three-dimensional

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

magnetic resonance imaging navigation system: first experiences with atlas-to-patient registration. J Neurosurg 1999;90(5):891-901. Niemann K, Mennicken VR, Jeanmonod D, Morel A. The Morel stereotactic atlas of the human thalamus: atlas-to-MR registration of internally consistent canonical model. NeuroImage 2000;12(6):601-16. Nowinski WL, Liu J, Thirunavuukarasuu A. Quantification of three-dimensional inconsistency of the subthalamic nucleus in the Schaltenbrand-Wahren brain atlas. Stereotact Funct Neurosurg 2006;84(1):46-55. Nowinski WL, Liu J, Thirunavuukarasuu A. Quantification and visualization of three-dimensional inconsistency of the globus pallidus internus in the SchaltenbrandWahren brain atlas. Stereotact Funct Neurosurg 2006;84(5–6):236-42. Nowinski WL, Liu J, Thirunavuukarasuu A. Quantification and visualization of three-dimensional inconsistency of the ventrointermediate nucleus of the thalamus in the Schaltenbrand-Wahren brain atlas. Acta Neurochir; 2008;150(7): 647-53. Chakravarty MM, Bertrand G, Hodge CP, Sadikot AF, Collins DL. The creation of a brain atlas for image guided neurosurgery using serial histological data. NeuroImage 2006;30(2):359-76. Henderson JM. Integration of stereotactic position with MER. In: Israel Z, Burchiel KJ, editors. Microelectrode recording in movement disorder surgery. New York – Stuttgart: Thieme; 2004. p. 130-7. Nakano N, Taneda M. Three-dimensional atlas of subthalamic nucleus and its adjacent structures. No Shinkei Geka 2005;33(7):683-92. Yelnik J, Bardinet E, Dormont D, Malandain G, Ourselin S, Tande´ D, Karachi C, Ayache N, Cornu P, Agid Y. A three-dimensional, histological and deformable atlas of the human basal ganglia. I. Atlas construction based on immunohistochemical and MRI data. NeuroImage 2007;34(2):618-38. Nowinski WL, Fang A, Nguyen BT, Raphel JK, Jagannathan L, Raghavan R, Bryan RN, Miller G. Multiple brain atlas database and atlas-based neuroimaging system. Comput Aided Surg 1997;2(1):42-66. Nowinski WL. Computerized brain atlases for surgery of movement disorders. Semin Neurosurg 2001;12 (2):183-94. Nowinski WL. Electronic brain atlases: features and applications. In: Caramella D, Bartolozzi C, editors. 3D image processing: techniques and clinical applications. Medical Radiology series, Heidelberg: Springer; 2002. p. 79-93. Nowinski WL. The Cerefy Brain Atlases: continuous enhancement of the electronic Talairach-Tournoux brain atlas. Neuroinformatics 2005;3(4):293-300. Nowinski WL, Bryan RN, Raghavan R. The electronic clinical brain atlas: multiplanar navigation of the human brain. New York: Thieme; 1997.


55. Nowinski WL, Thirunavuukarasuu A, Kennedy DN. Brain atlas for functional imaging: clinical and research applications. New York: Thieme; 2000. 56. Nowinski WL, Thirunavuukarasuu A. The cerefy clinical brain atlas on CD-ROM. New York: Thieme; 2004. 57. Nowinski WL, Thirunavuukarasuu A, Benabid AL. The cerefy clinical brain atlas: enhanced edition with surgical planning and intraoperative support. New York: Thieme; 2005. 58. Liu J, Nowinski WL. A hybrid approach to shape-based interpolation of stereotactic atlases of the human brain. Neuroinformatics 2006;4(2):177-98. 59. Niemann K, van Nieuwenhofen I. One atlas – three anatomies: relationships of the Schaltenbrand and Wahren microscopic data. Acta Neurochir 1999;141: 1025-103. 60. Nowinski WL, Thirunavuukarasuu A. Atlas-assisted localization analysis of functional images. Med Image Anal 2001;5(3):207-20. 61. Nowinski WL. Co-registration of the SchaltenbrandWahren microseries with the probabilistic functional atlas. Stereotact Funct Neurosurg 2004;82:142-6. 62. Nowinski WL, Thirunavuukarasuu A. Quantification of spatial consistency in the Talairach and Tournoux stereotactic atlas. Acta Neurochir; 2009 (in press), DOI: 10.1007/s00701-009-0364-8. 63. Nowinski WL. Modified Talairach landmarks. Acta Neurochir 2001;143(10):1045-57. 64. Finnis K, Starreveld Y, Parrent A, Sadikot A, Peters T. Three-dimensional database of subcortical electrophysiology for image-guided stereotactic functional neurosurgery. IEEE Trans Med Imag 2003;22(1):93-104. 65. Collins DL, Evans AC. ANIMAL: validation and application of non-linear registration-based segmentation. Int J Pattern Recogn Artif Intell 11(8):1271–94. 66. D’Haese P, Cetinkaya E, Kao C, Fitzpatrick J, Konrad P, Dawant B. Toward the creation of an electrophysiological atlas for the pre-operative planning and intra-operative guidance of deep brain stimulators (DBS) implantation. In: Proceedings of the international conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2004), Saint-Malo, France. LNCS, vol. 3216, Berlin: Springer; 2004. p. 729-36. 67. Rohde GK, Aldroubi A, Dawant BM. The adaptive bases algorithm for intensity based non-rigid image registration. IEEE Trans Med Imag 2003;22(11):1470-9. 68. Nowinski WL, Belov D, Benabid AL. An algorithm for rapid calculation of a probabilistic functional atlas of subcortical structures from electrophysiological data collected during functional neurosurgery procedures. NeuroImage 2003;18(1):143-55. 69. Nowinski WL, Belov D, Benabid AL. A communitycentric internet portal for stereotactic and functional neurosurgery with a probabilistic functional atlas. Stereotact Funct Neurosurg 2002;79:1-12.

27

70. Nowinski WL, Belov D, Pollak P, Benabid AL. A probabilistic functional atlas of the human subthalamic nucleus. Neuroinformatics 2004;2(4):381-98. 71. Nowinski WL, Belov D, Pollack P, Benabid AL. Statistical analysis of 168 bilateral subthalamic nucleus implantations by means of the probabilistic functional atlas. Neurosurg 2005;57 Suppl 4:319-30. 72. Nowinski WL, Belov D, Thirunavuukarasuu, A. Benabid AL. A probabilistic functional atlas of the VIM nucleus constructed from pre-, intra- and post-operative electrophysiological and neuroimaging data acquired during the surgical treatment of Parkinson’s disease patients. Stereotact Funct Neurosurg 2006;83(5–6):190-6. 73. Nowinski WL, Thirunavuukarasuu A, Liu J, Benabid AL. Correlation between the anatomical and functional human subthalamic nucleus. Stereotact Funct Neurosurg 2006;85:88-93. 74. Benabid AL, Pollak P, Gross C, Hoffmann D, Benazzouz A, Gao DM, Laurent A, Gentil M, Perret J. Acute and long-term effects of subthalamic nucleus stimulation in Parkinson’s disease. Stereotact Funct Neurosurg 1994;62(1–4):76-84. 75. Benabid AL, Koudsie A, Benazzouz A, Fraix V, Ashraf A, Le Bas JF, Chabardes S, Pollak P. Subthalamic stimulation for Parkinson’s disease. Arch Med Res 2000;31 (3):282-9. 76. Krack P, Fraix V, Mendes A, Benabid AL, Pollak P. Postoperative management of subthalamic nucleus stimulation for Parkinson’s disease. Mov Disord 2002;17 Suppl 3:S188-97. 77. Tuch DS, Wedeen VJ, Dale AM, George JS, Belliveau JW. Conductivity tensor mapping of the human brain using diffusion tensor MRI. Proc Natl Acad Sci USA 2001;98 (20):11697-701. 78. McIntyre CC, Mori S, Sherman DL, Thakor NV, Vitek JL. Electric field and stimulating influence generated by deep brain stimulation of the subthalamic nucleus. Clin Neurophysiol 2004;115(3):589-95. 79. Butson CR, Cooper SE, Henderson JM, McIntyre CC. Patient-specific analysis of the volume of tissue activated during deep brain stimulation. NeuroImage 2007;34 (2):661-70. 80. Miocinovic S, Noecker AM, Maks CB, Butson CR, McIntyre CC. Cicerone: stereotactic neurophysiological recording and deep brain stimulation electrode placement software system. Acta Neurochir Suppl 2007;972: 561-7. 81. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, Schwalb JM, Kennedy SH. Deep brain stimulation for treatment-resistant depression. Neuron 3 2005;45(5):651-60. 82. Mazzone P, Lozano A, Stanzione P, Galati S, Scarnati E, Peppe A, Stefani A. Implantation of human pedunculopontine nucleus: a safe and clinically relevant target in Parkinson’s disease. Neuroreport 2005;16(17):1877-81. 83. Nowinski, WL. Towards construction of an ideal stereotactic brain atlas. Acta Neurochir 2008;150(1):1-14.

439

440

27


84. Chakravarty MM, Sadikot AF, Mongia S, Collin DL. Towards a multi-modal atlas for neurosurgical planning. In: Proceedings of the ninth international conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2006), Copenhagen, Denmark. LNCS, vol. 4191, Berlin: Springer; 2006. p. 389-96. 85. Toga AW, editor. Brain warping. San Diego: Academic Press; 1998. 86. Warfield SK, Guimond A, Roche A, et al. Advanced nonrigid registration algorithms for image fusion. In: Toga AW, Mazziotta JC, editors. Brain mapping. The methods. 2nd ed. San Diego: Academic Press; 2002. p. 661-90. 87. Nowinski WL, Qian G, Bhanu Prakash KN, Hu Q, Aziz A. Fast talairach transformation for magnetic resonance neuroimages. J Comput Assist Tomogr 2006;30 (4):629-41. 88. Christensen GE, Joshi SC, Miller MI. Volumetric transformation of brain anatomy. IEEE Trans Med Imaging 1997;16(6):864-77. 89. Grachev D, Berdichevsky D, Rauch SL, Heckers S, Kennedy DN, Caviness VS, Alpert NM. A method for assessing the accuracy of intersubject registration of the human brain using anatomic landmarks. NeuroImage 1999;9(2):250-68. 90. Nowinski WL. Anatomical targeting in functional neurosurgery by the simultaneous use of multiple Schaltenbrand-Wahren brain atlas microseries. Stereotact Funct Neurosurg 1998;71:103-16. 91. Hu Q, Nowinski WL. A rapid algorithm for robust and automatic extraction of the midsagittal plane of the human cerebrum from neuroimages based on local symmetry and outlier removal. NeuroImage 2003;20(4): 2154-66. 92. Bhanu Prakash KN, Hu Q, Aziz A, Nowinski WL. Rapid and automatic localization of the anterior and posterior commissure point landmarks in MR volumetric neuroimages. Acad Radiol 2006;13(1):36-54. 93. Hu Q, Qian G, Nowinski WL. Fast, accurate and automatic extraction of the modified Talairach cortical landmarks from MR images. Magn Reson Med 2005;53: 970-6. 94. Nowinski WL, Bhanuprakash KN. Dorso-ventral extension of the Talairach transformation and its automatic calculation for MR neuroimages. J Comp Assist Tomogr 2005;29(6):863-79. 95. Saint-Cyr JA, Hoque T, Pereira LC, Dostrovsky JO, Hutchison WD, Mikulis DJ, Abosch A, Sime E, Lang AE, Lozano AM. Localization of clinically effective stimulating electrodes in the human subthalamic nucleus on magnetic resonance imaging. J Neurosurg 2002;97(5): 1152-66. 96. Nowinski WL, Yang GL, Yeo TT. Computer-aided stereotactic functional neurosurgery enhanced by the use of the multiple brain atlas database. IEEE Trans Med Imaging 2000;19(1):62-9.

97. Fukuda M, Mentis M, Ghilardi MF, Dhawan V, Antonini A, Hammerstad J, Lozano AM, Lang A, Lyons K, Koller W, Ghez C, Eidelberg D. Functional correlates of pallidal stimulation for Parkinson’s disease. Ann Neurol 2001;49(2):155-64. 98. Ganser KA, Dickhaus H, Metzner R, Wirtz CR. A deformable digital brain atlas system according to Talairach and Tournoux. Med Image Anal 2004;8(1):3-22. 99. Kaur G, Tan J, Alam M, Chaudhary V, Chen D, Dong M, Eltahawy H, Fotouhi F, Gammage C, Gong J, Grosky W, Guthikonda M, Hu J, Jeyaraj D, Jin X, King A, Landman J, Lee J, Li QH, Lufei H, Morse M, Patel J, Sethi I, Shi W, Yang K, Zhang Z. CASMIL: a comprehensive software/toolkit for image-guided neurosurgeries. Int J Med Robot 2006;2(2):123-38. 100. Nowinski WL, Belov D. The cerefy neuroradiology atlas: a Talairach-Tournoux atlas-based tool for analysis of neuroimages available over the internet. NeuroImage 2003;20(1):50-7. 101. Shabalov VA, Kazarnovskaya MI, Borodkin SM, Kadin AL, Krivosheina VY, Golanov AV. Functional neurosurgery using 3-D computer stereotactic atlas. Acta Neurochir Suppl (Wien) 1993;8:65-7. 102. Serra L, Nowinski WL, Poston T, Ng H, Lee CM, Chua GG, Pillay PK. The brain bench: virtual tools for stereotactic frame neurosurgery. Med Image Anal 1997;1(4): 317-29. 103. St-Jean P, Sadikot AF, Collins L, Clonda D, Kasrai R, Evans AC, Peters TM. Automated atlas integration and interactive three-dimensional visualization tools for planning and guidance in functional neurosurgery. IEEE Trans Med Imaging 1998;17(5):673-80. 104. Nowinski WL, Yeo TT, Thirunavuukarasuu A. Microelectrode-guided functional neurosurgery assisted by electronic clinical brain atlas CD-ROM. Comput Aided Surg 1998;3(3):115-22. 105. Nowinski WL, Thirunavuukarasuu A. A locus-driven mechanism for rapid and automated atlas-assisted analysis of functional images by using the brain atlas for functional imaging. Neurosurg Focus 2003;15(1): Article 3. 106. Benabid AL, Nowinski WL. Intraoperative robotics for the practice of neurosurgery: a surgeon’s perspective. In: Apuzzo ML, editor. The operating room for the 21st century. Rolling Meadows: American Association of Neurological Surgeons; 2003. p. 103-18. 107. Ahsan RL, Allom R, Gousias IS, Habib H, Turkheimer FE, Free S, Lemieux L, Myers R, Duncan JS, Brooks DJ, Koepp MJ, Hammers A. Volumes, spatial extents and a probabilistic atlas of the human basal ganglia and thalamus. NeuroImage 2007;38(2):261-70. 108. Nowinski WL, Thirunavuukarasuu A, Volkau, Marchenko Y, Runge VM. The cerefy atlas of cerebral vasculature. New York: Thieme; 2009. 109. Lawes IN, Barrick TR, Murugam V, Spierings N, Evans DR, Song M, Clark CA. Atlas-based segmentation of


110.

111.

112.

113.

114.

115.

white matter tracts of the human brain using diffusion tensor tractography and comparison with classical dissection. NeuroImage 2008;39(1):62-79. Kim SJ, Kim IJ, Kim YK, Lee TH, Lee JS, Jun S, Nam HY, Lee JS, Kim YK, Lee DS. Probabilistic anatomic mapping of cerebral blood flow distribution of the middle cerebral artery. J Nucl Med 2008;49(1):39-43. Nowinski WL, Qian G, Bhanu Prakash KN, Thirunavuukarasuu A, Hu QM, Ivanov N, Parimal AS, Runge VL, Beauchamp NJ. Analysis of ischemic stroke MR images by means of brain atlases of anatomy and blood supply territories. Acad Radiol 2006;13(8):1025-34. Shattuck DW, Mirza M, Adisetiyo V, Hojatkashani C, Salamon G, Narr KL, Poldrack RA, Bilder RM, Toga AW. Construction of a 3D probabilistic atlas of human cortical structures. NeuroImage 2007;39(3): 1064-80. Chan E, Kovacevıć N, Ho SK, Henkelman RM, Henderson JT. Development of a high resolution threedimensional surgical atlas of the murine head for strains 129S1/SvImJ and C57Bl/6J using magnetic resonance imaging and micro-computed tomography. Neuroscience 2007;144(2):604-15. Nowinski WL, Belov D. Towards atlas-assisted automatic interpretation of MRI morphological brain scans in the presence of tumor. Acad Radiol 12:1049-57. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen AG, Fahlbusch R. Preoperative and intraoperative diffusion tensor imaging-based fiber

116.

117.

118.

119.

27

tracking in glioma surgery. Neurosurg 2005;56(1): 130-7; discussion 138. Castro FJ, Pollo C, Meuli R, Maeder P, Cuisenaire O, Cuadra MB, Villemure JG, Thiran JP. A cross validation study of deep brain stimulation targeting: from experts to atlas-based, segmentation-based and automatic registration algorithms. IEEE Trans Med Imaging 2006;25(11):1440-50. Kockro RA, Serra L, Yeo TT, Chan C, Sitoh YY, Chua GG, Ng H, Lee E, Lee YH, Nowinski WL. Planning and simulation of neurosurgery in a virtual reality environment. Neurosurg 2000;46(1): 118-37. Anil SM, Kato Y, Hayakawa M, Yoshida K, Nagahisha S, Kanno T. Virtual 3-dimensional preoperative planning with the dextroscope for excision of a 4th ventricular ependymoma. Minim Invasive Neurosurg 2007;50(2): 65-70. Wong GK, Zhu CX, Ahuja AT, Poon WS. Craniotomy and clipping of intracranial aneurysm in a stereoscopic virtual reality environment. Neurosurg 2007;61(3): 564-8; discussion 568-9.

441

29 Development of a Classic: The Todd-Wells Apparatus, the BRW, and the CRW Stereotactic Frames J. Arle

In many ways, the history of refined stereotactic surgery, particularly in the US, is intimately related to the concurrent progression in development of three devices, from the Todd-Wells apparatus, through the BRW frame, to ultimately the CRW frame system. It would involve the important use and feedback of many pioneers in stereotactic and function neurosurgery including Apuzzo (who used both BRW frame #1 and CRW frame #1) (Cosman E, 2008, personal communication), Gildenberg, Heilbrun, Nashold, and others. A catalogue of devices was created over the years for human stereotaxy, from the Zernov encephalometer in 1889 [1] to the multitude of instruments that arose from the publication in 1947 by Spiegel et al. [2], extending the ability of progressive and innovative neurosurgeons to attempt reliable intracranial targeting throughout the 1950s and 1960s, performing tens of thousands of stereotactic procedures [3]. However, these attempts were inevitably hampered by rudimentary imaging of intracranial anatomy. The best technique remained ventriculography, developed by Walter Dandy between 1916 and 1919 [4], though these were relatively uncomfortable procedures by most accounts, and invasive, and limited in their reliability by patient positioning at the time of the stereotactic procedure relative to when imaging was performed, and potentially by the anatomical consistency of locating intracranial structures with reference to #


ventricular widths and margins [5]. Despite these limitations, and combined with sophistication in human brain atlas detail, an emergence and, in breadth and sheer numbers, dominance of stereotactic procedures occurred over two decades in the 1950s and 1960s, pushing the field beyond the open and unpredictable procedures attempted earlier (e.g., anterior choroidal artery ligation by Cooper [6] or open ansotomy by Meyers [7]).

The Todd-Wells Apparatus Trent Wells, Jr. (> Figure 29-1a) had begun an engineering firm in that same seminal year of 1947 and he became involved in developing instruments for stereotaxy in animals at UCLA with Horace Magoun and Jack French [9]. Fate would eventually lead him to be involved in the design solutions and manufacturing of many important stereotactic devices, in addition to developing Gardner-Wells tongs for spinal traction. Early on, he found himself collaborating with a young neurosurgeon named Edwin Todd (> Figure 29-1b), who had become interested in surgery for movement disorders during his neurosurgical training at the Cleveland Clinic. As expressed by Wells and Todd themselves [9], ‘‘it was at this time, in the early 1960s, that the fortuitous meeting of a prospective stereotactic surgeon and a highly motivated stereotactic

454

29

Development of a classic: the Todd-Wells apparatus, the BRW, and the CRW stereotactic frames

. Figure 29-1 (a) Trent Wells, Jr, holding the BRW frame. Few were ever as involved in so many critical design innovations within the history of stereotaxy. He provided creative and durable engineering and machining for the development of at least five different stereotactic systems with three different neurosurgeons (Rand, Todd, and Roberts) [8] (b) Edwin M. Todd, neurosurgeon and instrumental motivation in the design and refinement of the Todd-Wells stereotactic apparatus at UCLA. Developed at the height of the era using ventriculography, this frame became one of the most widely used in the world [8]

apparatus inventor occurred; the former knew what he wanted, but not how to get it and the latter had the experience and the engineering ingenuity to turn seminal idea into reality.’’ All stereotaxy must develop a reliable coordinate system, whether Cartesian or not, that allows one to link such a system to particular anatomical landmarks, creating invariance that survives the vagaries of human anatomy and asymmetries of device placement. Ultimately, it also became clear that such invariance needed to achieve approximately 1 mm of spatial accuracy with every use. Left with only X-ray technology at the time as an imaging modality, Wells realized nonetheless that accuracy using biplanar X-rays depended not only on the refinement of the instrument per se, but on the ability to eliminate variation in X-ray magnification by placing the film cassettes at the same distance from the source in both planes every time. This could

be achieved by fixing the X-ray sources within the operating room, and placing the target at the precise focal point of both film planes each time by moving the head with the head frame. Wells had already begun part of his intracranial targeting education in the 1950s when he worked with neurosurgeon Robert Rand to develop and manufacture a transverse arc system for performing pallidectomies and then a second device for hypophysectomies [10]. Modifying this from a bur hole mounted system with an arc, to using a head mounted frame, and using Lucite reticules similar to gunsights on fighter aircraft (Wells had been an accomplished fighter pilot in WWII) in both lateral and AP directions, the desired target could be positioned along the arc axis and then accessed by any number of approaches on that side of the hemisphere. This device was the first modular human stereotactic instrument, using


29

. Figure 29-2 Conceptual schematic of the Todd-Wells apparatus, showing the focal point, defined by the intersection of precalibrated AP and lateral imaging with ventriculography. The head is then moved in the base ring to the make the desired target meet the focal point. Reliability can be ensured by fixing the X-ray sources and the frame mount in the OR [9]

two different arcs and allowing access to a target from virtually any direction. Ideally, however, fixed X-ray sources in the OR could be set up, adjusted appropriately by collimators, and with the multiaxial movement provided by the head holder, create a very versatile and reliable stereotactic device (see > Figures 29-2–29-4). It was technically similar, it turns out, to the device developed by Schaltenbrand in the Wurzburg operating suite in 1959, whereby X-ray sources were fixed 4m away and the target was moved to the focal point by moving the base ring [11]. The Todd-Wells device was subsequently used by several others in developing stereotactic foreign body removal, placement of depth electrodes,

electrothrombosis and metallothrombosis of intracranial aneurysms, radiofrequency lesioning of multiple targets, and even percutaneous cordotomy [9]. Given the imaging constraints of that time, this device was likely one of the most reliably accurate means of achieving accurate targeting for standard biopsies as well as the other procedures developed using the device, functional and otherwise. Although reliance on X-rays was widely practiced for many years, particularly in conjunction with ventriculography, a consistent and reliable means of ensuring accurate intracranial access was brought to bear in the Todd-Wells device, first presented to the neurosurgical community at large during the

455

456

29


. Figure 29-3 The Todd-Wells apparatus. Rigid, and accurate when X-ray sources are appropriately calibrated, and reliable [9]

International Neurological Surgery Meeting in Copenhagen in 1965 [9]. Commercialization of the device was initially procured by Codman [12] but eventually transferred to Radionics, helping broaden use of the device by multiple centers. Freidman and Coffey referred to it as one of the most widely used traditional stereotactic devices in the world [13].

The BRW But the nature of imaging changed radically in the 1970s with the development of the CT scanner, allowing for the first time, the potential of obtaining digital information for all three axes. The key was redesigning stereotactic devices to take advantage of this new informational largess, and Wells was perhaps the best – positioned engineer in the world for this task. Many were considering how to exploit this new development, realizing that one obvious advantage was that if one could extract relative coordinates from one axial scan slice to another, then the necessarily fixed referentiality between frame,

target and imaging source could be decoupled, and the inaccuracies and asymmetries of frame placement, human anatomy, and imaging calibrations could impact accuracy less overall, and make the process of obtaining stereotactic spatial coordinates easier in general. Through the collaborative work of Dr. Ted Roberts, chairman of neurosurgery at the University of Utah and Russell Brown, then a 3rd year medical student at the university, several key new concepts were developed that culminated in a new CT-compatible stereotactic frame, presented initially in 1979. Roberts and Brown introduced their conceptualizations to Trent Wells in 1978, and given his expertise in design translation, machining, and manufacturing of stereotactic devices for animal research over 30 years, Wells was able to distill their key innovations into a real device, capable of exceptional accuracy, consistency, sterilizability, and sturdiness. This became known as the BRW (BrownRoberts-Wells) frame and CT localizer system. Roberts was motivated by the need for more reliable and improved accuracy in stereotactic procedures, particularly in tumor biopsies [14], and the promise that CT brought was the ability better than ever to see lesions, with the additional potential to provide a coordinate structure to access them. Initially, Brown worked on using a 3-D multiplanar imaging system developed by Evans and Sutherland Computer Corporation in Salt Lake City, Utah, to explore methods of improving biopsy accuracy and yield [14]. He created software to use outlines of CT images of the brain and lesions, graphically stacking them together for manipulation [14]. When the Varian V-360-CT body scanner became available in Utah, he instead shifted to developing a prototype frame that could use the CT planar information as a potential coordinate reference frame [14]. This initial prototype was made by Brown in Lucite (> Figure 29-5) and incorporated versions of several innovations described below. The first


29

. Figure 29-4 Drawings showing the versatility of the Todd-Wells system. The wide array of procedures enabled by this frame helped create its widespread use. Note, however, the difficulty involved in taking a low lateral trajectory to a target. This limitation would prove to be one of the motivations in development of both the BRW and CRW later [9]

innovation was in finding a means to relate each axial planar slice uniquely to the frame. This was accomplished by Brown’s idea to use two rings suspended apart by three sets of rods arranged in ‘‘N’’ patterns. The ‘‘localizer ring’’ as it was called attached to the frame base in only one way with unique ball clamps, minimizing human error. As a separate unit, the localizer continued the modularity of the system, the sturdiness, and the relative ease of sterilization of system components. The initial Lucite prototype used four vertical rods with three diagonal rods between

them, running contiguously along the ring margin. The eventual metal commercial version would use six vertical epoxy-graphite rods with the three pairs of them separated by a diagonal rod and spread in these 3-rod sets equidistantly around the localizer ring. Such a configuration, widely adapted in modified ways by others later, allowed each axial slice to also have a unique set of nine nearly-circular cross-sections of the rods around the edge of the scan image (> Figure 29-6). The distances between the outer two rods and the center of the diagonal

457

458

29


rod, for each of the three sets, allowed for precise determination of the axial plane orientation to the frame, thus allowing any point in space within the localizer ring to have a unique x, y, and z . Figure 29-5 Prototype of the BRW frame, created by Russell Brown. Made of Lucite, it was composed of interlocking arcs and incorporated the innovative ‘‘N’’ shaped fiducial rods along the edge. Note that in this prototype the configuration of fiducials uses four contiguous rods with three diagonals interspersed. This was ultimately changed in the final localizer ring design [14]

set of coordinates. The transformation matrix from the two-dimensional axial image to full 3-D coordinates is given in > Figure 29-7. Importantly, the ‘‘z’’ or vertical height from the frame base was precisely related to the relative x and y inter-rod distances, freeing users from both the requirement that the patient be fixed to the scanner table and from, as Roberts put it, ‘‘the unreliable scanner table movements required by earlier CT equipment’’ [14]. Brown ultimately went on to patent this aspect of the design and other aspects as well [16]. The second main innovation derived from the motivation to allow targeting to all locations in the head from virtually any access point and trajectory through the hemispheres, a limited characteristic in most frames until then. This was accomplished primarily by allowing the . Figure 29-7 Transform matrix to convert from 2-D to 3-D coordinates. The vertical ‘‘Z’’ value is intrinsically related to the relative distances between the center diagonal fiducial and the two outer fiducials [15]

. Figure 29-6 Drawing showing the BRW and localizer ring in place, denoting to the right axial images at three different vertical reference lines with fiducial configurations. The interfiducial distances allow unique coordinates to be computed for any point within that localizer volume [14]

. Figure 29-8 Schematic showing the four angles that need to be used in computing the trajectory with the BRW frame. The fifth degree of freedom is the probe length. Despite the cumbersome nature of this solution, it allowed a much wider variation of target access [15]


rotation of the probe holder along a rotational base ring and including several other adjustable angulations and a probe holder pivot, in order to reach the desired target. Calculations were required involving four unique angles (alpha, beta, gamma, and delta), but once learned, the method was trustworthy and consistent. This was not a target-centered device and calculations of each of the necessary four angles to finalize the trajectory to the target was at first cumbersome, and required running the computations for several minutes on an adjunct DEC

29

machine, the equivalent of one of today’s desktop computers (Cosman E, 2008, personal communication). The device uses a polar coordinate system with 5 degrees of freedom (> Figure 29-8). Alpha refers to the base-ring angle of rotation. Beta refers to an up to 30 pivot in the same plane as Alpha. Gamma refers to a measure of angle perpendicular to the base ring, up to 180 , and Delta refers to an up to 90 pivot of the probe holder in the same plane as Gamma. The fifth degree of freedom is the probe length. Values can be read off the engraved vernier scales

. Figure 29-9 (a) BRW frame mounted on the base ring [14] (b) BRW base ring with the BRW localizer ring attached [14] (c) BRW phantom base allowed users to check trajectories and target alignment before the procedure as an internal safety check and calibration [14]

459

460

29


. Figure 29-10 (a) Original letter written by Dr. Eric Cosman to Trent Wells Jr in 1980 before the Houston stereotactic and functional meeting explaining that he believed he had formulated the solutions to transform the BRW to an arc-radius type of device and giving the derivation for a simplified determination of the x, y, and z coordinates of the target (courtesy of Dr. Eric Cosman) (b-d) Original handwritten calculations from ‘‘27 Sept. 1980’’ within which Dr. Cosman derives the CT-to-frame coordinate transformations, which ultimately showed the feasibility of the CRW design (courtesy of Dr. Eric Cosman)


on the arc system itself. Several efficiencies were made in short order to ease calculations, however, with first the introduction of the laptop-sized Epson Hx-20, followed by code written for the HP-41cv handheld calculator by Dr. Eric Cosman (Cosman E, 2008, personal communication), securing relative ease of use by many more surgeons within the field of stereotaxy. Wells had superbly transformed Brown’s prototype in Lucite into a robust reliable instrument, relatively free of X-ray artifact. Between 1981 (when the BRW was first commercially offered) and 1988,

29

the system became the most popular stereotactic frame system in the world.

The CRW As the liberal use of brain imaging progressed in quality and clinical importance, market forces also began to play a role in improvements from one manufacturer to another. While the BRW enjoyed relative success worldwide during the 1980s, two other frame systems also made a

. Figure 29-11 (a) Original hand drawn schematics of the CRW arc system, emphasizing several of its key features – the ability to translate the arc itself, (b and d) the ability of the trunion and arc composite to be relocated to a lateral aspect of the base thereby allowing direct lateral or otherwise difficult trajectories (c) and the ability to move the probe holder along the arc maintaining the same target (courtesy of Radionics, Inc)

461

462

29


. Figure 29-12 Photograph showing (left to right) Cosman, Roberts, and Wells behind the CRW system at the Radionics booth at the Toronto AANS meeting in 1988, the year the CRW was introduced commercially (courtesy of Dr. Eric Cosman)

relatively successful transition to CT, and then MRI – the Leksell frame and the ReichertMundinger frame. Both of these systems contributed important innovations to frame technology that were subsequently incorporated into modifications to the BRW frame. Leksell had already created what came to be called a ‘‘target centered’’ arc-radius system in 1949 [17], whereby an arc is mounted on the moveable frame which has a semicircular configuration. The probe, or other device, is mounted on the arc and can be moved along the arc into almost any desired location – the innovation being that the probe is made to be equal in length to the radius of the arc, thereby making every location for its trajectory terminate at the desired target. This greatly simplifies the preoperative calculations necessary for planning the setup of the system. Several others successfully incorporated this concept into frame designs, and then modified them later as well for use with CT [18,19], but they had limitations in certain trajectories, setup, or were simply never commercialized to any significant degree. The innovation of Mundinger, working with Reichert and Wolff, was the use of a phantom

. Figure 29-13 (a, b) Photographs of a similar target trajectory using the CRW (left) and BRW (right) frames, but highlighting the difference between the arc-radius type of system (CRW) and the original polar coordinate-based BRW [20]


29

. Figure 29-14 Four different examples showing the versatility of the CRW system, allowing a wide variety of trajectories to target: (a) arc translation, (b) straight lateral, (c) posterior translation, and (d) below the base ring transnasal [20]

base. This was also made available in the BRW system and served as a convenience for evaluating the trajectory and arc setup beforehand but also, more importantly perhaps, created an internal check making sure the surgeon had set up coordinates correctly (> Figure 29-9). Cosman understood, perhaps better than anyone, that neurosurgeons in general may have been willing to use the more complicated angular calculations involved with the BRW, but if faced with an easier to use and equivalent device, they would inevitably switch allegiances. Moreover, some trajectories, such as straight lateral or a rising transnasal approach, were not possible with

earlier versions or other frames. He tried to convince Wells to consider a target-centered arc system. Wells was reluctant at first (Cosman E, 2008, personal communication), but Cosman calculated out the transformation from the BRW semi-polar coordinate system to the target-centered semicircular arc-radius device he envisioned, and sent Wells a putative drawing to consider. > Figure 29-10 shows the handwritten calculations and the original letter Cosman wrote to Wells after solving the coordinate transforms. > Figure 29-11a–d shows his original drawings of the CRW and how it solved these remaining access problems that had eluded other frames

463

464

29


and even the BRW to some degree. Wells came around, and agreed he could build such a device (Cosman E, 2008, personal communication), producing what then replaced the BRW system in 1988 commercially – the Cosman-Roberts-Wells (CRW) stereotactic frame (> Figure 29-12). While the nature of the arc system and initial planning calculations changed, the localizer ring remained essentially the same (though with only two sets of vertical rods and three diagonals), as this innovation was proven to be a streamlined means of extracting the CT, and MRI, axial fiducial information. The CRW localizer ring was a modest simplification of the BRW localizer ring, allowing one (if the frame base is aligned nearly parallel with scanner gantry angle) simply to read off the vertical value by noting the distance between the diagonal middle rod and the closest vertical rod. The CRW may still be used with the BRW localizer ring, however, without any change in technique. The base of the CRW system, and this may be one of its most significant innovations, is a square with slots to hold the trunion apparatus equivalently in either the AP direction or the lateral direction, and the ring slides were reproduced on both sides (a benefit by averaging out potential error that might occur if only one were used). Even the BRW could not go to a horizontal trajectory or below. The arc can slide from one laterality to the other, and the probe-holder may be positioned from one end of the arc to the other, a combination that allows the widest possible number of trajectories of any frame, all 160 mm from the target with the standard probe holder. This distance was chosen by Cosman based on using anthropomorphic model heads he had in his lab (Cosman E, 2008, personal communication), adjusting the arc size to allow enough distance to mount certain accessories if needed, but not so much distance that the arc becomes unwieldy, especially in an OR environment where an efficient design pays dividends. > Figure 29-13a,b show the essential difference in

the geometrical solutions to solving the stereotactic targeting and trajectory problem between the BRW and the simpler, more versatile arcradius CRW [20]. Since its inception in the late 1980s, the CRW system has been updated in materials and accessories several times. The versatility of the system for accessing intracranial targets is unparalleled. > Figure 29-14 shows only four such variations possible with the standard system [20]. The current version of the CRW frame (shown in > Figure 29-15), has become lighter, but retains the same degree of broad neurosurgical applicability. Typically, the current system is used with a software interface provided by Radionics (Stereocalc and OmniSight Excel, copyright, Radionics Corporation) that not only calculates the appropriate coordinate transform from the fiducials on the scan with a fiducial automatic search and algorithm, but also graphically shows the frame itself in an interactive window, allows . Figure 29-15 Photograph of the current CRW system on the phantom base (courtesy of Radionics, Inc)


29

. Figure 29-16 Software included now with the CRW system, example screens shown in (a) and (b), allow the surgeon to automatically locate the centroid of the fiducials and localize the scan in the system, graphically pick targets, trajectories, fuse imaging modalities, visualize in 3-D graphics, overlay anatomical atlases, and plan many aspects of surgery with a user interface that underscores the enormous leaps technology has made in the last 30 years (courtesy of Radionics, Inc)

465

466

29


for a virtually error-free image fusion with other imaging modalities (fusing a CT in the frame, for example, with a prior MRI series done previously without the frame applied), and gives the user access to a digitized brain atlas, measurement techniques and 3-D brain surface renderings for sophisticated operative planning (> Figure 29-16).

Presently, the CRW system, with its clean design, robust manufacturing and quality control, wide array of adjunctive instruments (from angiographic localizers, repeat fixation kits, laser holders, interstitial radiotherapy holders, brain retractor attachments, depth electrode placement kits, and wide array of biopsy needles, microdrives, and offset probe holders), accuracy, and

. Figure 29-17 (a) Photographs showing the typical setup of the frame, (b) frame application, (c) alignment to intended target using the phantom base in the OR, and (in only one of many uses for the CRW frame) (d) mounting of a microdrive for microelectrode recording for placement of a DBS lead in the STN of a Parkinsons Disease patient – a modern incarnation of what has been taking place in one form or another for more than 60 years


sterilizability, has risen to pre-eminence as a stereotactic device. For example, a recent study of centers performing deep brain stimulation (DBS) for Parkinsons Disease found that almost half of the centers responding to questionnaires regarding their stereotactic technique for DBS used the CRW frame (> Figure 29-17). These 36 centers had performed over 4,500 DBS implantations [21]. The system continues to sell well on a worldwide basis. There will be further refinements and advancements but the story of progress, from the Todd-Wells apparatus to the CRW system, remains a fantastic example of brilliant collaboration between neurosurgery, engineering, technology, and commercialization.

References 1. Speigel EA, Wycis HT, Marks M, Lee A. Stereotaxis apparatus for operations on the human brain. Science 1947;106:349-50. 2. Gildenberg PL. Stereotactic surgery: present and past. In: Heilbrun MP, editor. Stereotactic neurosurgery. Baltimore, MD: Williams & Wilkins; 1988. p. 1-15. 3. Dandy WE. Ventriculography following the injection of air into the cerebral ventricles. Ann Surg 1918;70:378-84. 4. Harris MI, Bergenheim AT. A comparative study on ventricular graphic and computerized tomography-guided determination of brain targets and functional stereotaxis. J Neurosurg 1990;73:565-71. 5. Cooper IS. Ligation of the anterior choroidal artery for involuntary movements of parkinsonism. Psychiatr Q 1953;27:317-19. 6. Meyers R. Surgical experiments in the therapy of certain ‘extrapyramidal diseases’. Acta Psychiat Neurol 1951;26:1-42.

29

7. Apuzzo MLJ. A fantastic voyage: a personal perspective on involvement in the development of modern stereotactic and functional neurosurgery (1974–2004). Neurosurgery 2005;56(5):1115-33. 8. Wells TH, Todd EM. The Todd-Wells apparatus. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional surgery. New York: McGraw-Hill; 1998. p. 95-9. 9. Rand RW. A stereotactic instrument for pallidothalamectomy. J Neurosurg 1961;18:258-60. 10. Schaltenbrand G. Personal observations on the development of stereotaxy. Conf Neurol 1975;37:410-16. 11. Freidman WA, Coffey RJ. Stereotactic surgical instrumentation. In: Heilbrun MP, editor. Stereotactic neurosurgery. Baltimore, MD: Williams & Wilkins; 1988. p. 55-72. 12. Kandel EI. Stereotaxic apparatus and operations in Russia in the 19th century. J Neurosurg 1972;37:407-11. 13. Roberts TS. The brw/crw stereotactic apparatus. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional surgery. New York: McGraw-Hill, 1998. p. 65-71. 14. Brown RA, Roberts TS, Osborn AG. Stereotaxic frame and computer software for CT-directed neurosurgical localization. Ivest Radiol 1980;15:308-12. 15. Brown RA. US patent #4608977, issued 1986. 16. Leksell LA. Stereotactic apparatus for intracerebral surgery. Acta Chir Scand 1949;99:229-33. 17. Van Buren JM. A stereotaxic instrument for man. Electroencephalogr Clin Neurophysiol 1965;19:398-403. 18. Gouda K, Gibson RM. New frame for stereotaxic surgery: technical note. J Neurosurg 1980;3:256-9. 19. Cosman ER. Development and technical features of the Cosman-Roberts-Wells stereotactic system. In: Pell MF, Thomas DGT, editors. Handbook of stereotaxy using the CRW apparatus. Baltimore, MD: Williams & Wilkins; 1994. p. 1-52. 20. Ondo W, the DBS Study Group. The North American survey of placement and adjustment strategies for deep brain stimulation. Stereotact Funct Neurosurg 2005;83:142-7.

467

26 Electronic Stereotactic Atlases J. Yelnik . E. Bardinet . D. Dormont

Localizing structures and functions in the brain is a quest which has concerned human being since its early history. Although it is known that trepanations have been performed in primitive societies such as in the Neolithic period (7000 year BC), in Egypt (3000 years BC) or in Peru (2000 year BC), these practices were motivated by religious reasons rather than by medical reasons, as far as we know. The first rational approaches of brain anatomy and physiology were those of Hippocrates (400 years BC) although the knowledge of brain anatomy was still rudimentary at this period. Surgical interventions in the brain have remained very empirical during the middle-ages with the notable exception of Ambroise Pare´, who was able to treat fractures of the vertebral column and to make trepanations of the skull. Neurosurgery in fact began with Victor Horsley (1857–1924), while in an overlapping period Claude Bernard (1813–1878) developed his fundamental approach of experimental physiology, which opened an access to an understanding of brain functions. Localization of brain functions had for a while lost its way in the concepts of ‘‘Phreńologie’’ or ‘‘organology’’ of Franz Joseph Gall (1758–1828) and Johann Spurzheim (1776–1832) who proposed that the most complex human functions such as compassion, moral sense, vanity, feeling of property, kindness, and benevolence could be localized in specific parts of the brain, and even identified as a prominent development of the external shape of the skull in individual subjects with specific development of such functions. Brilliant neurologists, such as Paul Broca (1824–1880), Jean-Martin Charcot (1825–1893), John Hughling Jackson (1835–1911), and Joseph Babinski

#


(1857–1932) have finally provided the bases of a really scientific exploration of the localization of brain functions. This scientific approach has developed rapidly in the domain of neurology based on the anatomo-clinical method which consists of a systematic comparison between the neurological symptoms observed in a living patient and the anatomic lesions discovered in this brain after death. In the domain of neurosurgery, brain localization is a central issue. The technique by which a given region is localized in the brain of a living patient is called stereotaxy, which comes form the Greek stereo for three-dimensional and taxis for disposition, hence ‘‘localize in space.’’ The first stereotactic operations were performed on the basis of bony landmarks as proposed with the system of Horsley-Clarke [1]. This bony system is still in use for all experimental studies in the rat with the most commonly used atlas of Paxinos [2]. In the human, the bony landmarks have been shown in the 1950s by Spiegel and Wycis to be insufficiently accurate for human stereotactic surgery. They developed a method based on the Horsley-Clarke system [3], which relied on their own atlas of the human brain (> Figure 26-1) and on intracerebral landmarks visible on ventriculography. This pioneer neuro-imaging modality, introduced in 1918 by the American neurosurgeon Walter Dandy [4] and still in use in some neurosurgical centers, consists in taking X-ray images of the ventricular system after injection of filtered air directly into one or both lateral ventricles of the brain via one or more small trephine holes drilled in the skull under local anesthesia. The landmarks they selected were

374

26

Electronic stereotactic atlases

. Figure 26-1 A frontal unstained section of the first stereotactic atlas of the human brain, that of Spiegel and Wycis (Reference 5). The three-dimensional localization of cortical areas or deep brain structures is given by the medio-lateral and dorso-ventral millimetric grids and by the position of the section in the series

the epiphysis and the foramen of Monro [5]. Afterwards Jean Talairach at the Sainte-Anne Hospital in Paris, France, developed a system based on two different and more reliable ventricular landmarks, the anterior and posterior commissural points (AC-PC) [6–9], a coordinate system which has become the gold standard in the field of human stereotactic neurosurgery, even when ventriculography was later replaced by magnetic resonance imaging (MRI). The ACPC system of coordinates has been widely used since the mid-twentieth century for neurosurgical interventions for movement or psychiatric disorders [10–13].

In the 1980s, the development of MRI completely transformed the concept of in vivo brain structure localization. MRI uses magnetic fields and radio frequencies to produce high quality two- or three-dimensional images of brain structures. When the main magnetic field is imposed, each point in space has a unique radio frequency at which the signal is received and transmitted. Sensors read the frequencies, which are then used to build an image. With MRI, it has become possible to image both surface and deep brain structures with a high degree of detail. This technical progress has completely transformed the concept of brain anatomy, targeting, or localization, and consequently, the notion of brain atlases itself. Another important development in the domain of neurosurgery had a major influence on the issue of brain localization when AlimLouis Benabid in Grenoble, France, observed in 1987 that brain lesioning (e.g., thalamotomy) could be replaced by chronic high frequency electrical stimulation of the same region for the treatment of tremor [14]. Chronic stimulation had already been used for treating pain in the early 1970s [15], but Benabid first proposed to combine the implantable pacemaker technology (developed for cardiac pacemaker by 1960) with chronically implanted deep brain electrodes [16]. The advantages of stimulation over lesion were that the method did not produce a definitive lesion, the electrode could be removed or displaced if adverse events occurred, the electrical parameters could be adjusted to obtain the best possible clinical result. In stereotactic functional deep brain stimulation (DBS) neurosurgery, three successive stages that are linked to each other work toward a precise identification of deep brain structures. The preoperative step is that of targeting, which consists of identifying a given deep brain structure, previously chosen on the basis of theoretical and physiopathological arguments, in the brain of a given patient. The peroperative step is that of exploration during which the particular characteristics of the target in the concerned patient


are investigated using electrophysiological and clinical testing. The postoperative step is that of localization which determines the exact position of each implanted electrode within the deep brain structures of the operated patient. The first DBS target was the thalamic nucleus Vim for tremor, either essential tremor or parkinsonian tremor [17]. It was proposed by Alim-Louis Benabid following the peroperative observation that stimulation of the Vim nucleus at high frequency (130 Hz) resulted in the immediate suppression of tremor, thus mimicking the classical thalamotomy of the same nucleus in a reversible way. He then had the great idea of a permanent stimulation of the Vim through a stimulator implanted in the subclavicular region. Then, on the basis of previous results of pallidotomy for movement disorders [10], the internal globus pallidus was proposed for the treatment of the other parkinsonian symptoms, rigidity, akinesia, and Ldopa-induced dyskinesias [18]. Finally, following experimental researches in non human primates [19,20], the subthalamic nucleus (STN) came forward as the best target for the treatment of the whole symptomatology of Parkinson’s disease [21]. These successive targets rapidly raised problems of localization. The Vim thalamic target was first identified on the basis of the AC-PC system of coordinates, the underlying assumption being that all brains, at least the deep brain structures, can be considered similar after spatial normalization (i.e., deformation) based on the proportional Talairach system or other similar consideration. The pallidal target raised particular problem due to its lateral localization, far from the midline. Indeed, the AC-PC distance did not provide any information about a precise mediolateral localization of the target. Finally, the STN raised difficult problems because of its small size (12 5 3 mm3) [22], its complex oblique orientation, and the numerous axonal bundles by which it is surrounded. At the present time, DBS has greatly developed. The number of patients operated has

26

increased dramatically over the past 20 years, with DBS centers being now present in a huge number of different countries. The large series of patients operated have provided an invaluable source of anatomo-clinical data from which the relationship between electrode localization and the clinical outcome could be studied in great details. The varying effects that were obtained, which depend on both the precise localization of the electrodes and the electrical settings that were applied, have been studied thoroughly, which has increased at the same time the fine knowledge of DBS mechanisms and the need for more refined electrode localization. Today, electrodes are implanted not only within a given target (a nucleus as the STN or the globus pallidus) but within a given anatomofunctional subdivision of the nucleus. In addition, the volume of tissue activated (VTA) can now be analyzed by taking into account the electrical parameters that are applied and the fine anatomical structure of the region stimulated [23]. Neuro-imaging during the DBS procedure is required at two stages: preoperatively to determine the position of the target in a given patient, which is done using ventriculography and/or MRI, and postoperatively to determine the precise localization of the electrodes and their four contacts in the patient’s brain, which is done using MRI or CT scan which has been proposed as an alternative following a 2002 FDA alert about DBS and postoperative MRI. This latter possibility has become necessary in the particular context of DBS studies, due to accidents that occurred for patients submitted to MRI acquisition with the electrodes and the stimulator activated. In addition, peroperative local electrophysiological exploration (in the neighborhood of the planned target) and estimation of the VTA [23] provide a better volumetric vision of the neighborhood of the targets, including gray matter nuclei and white matter bundles. These different data contribute to improvement of current knowledge of the mechanisms of DBS and of anatomo-clinical correlations.

375

376

26


At the same time, these data raise a problem of scale. Indeed, the resolution of MRI or scanner are roughly of the same magnitude (about 1 mm) but are all insufficient with regard to the required definition of the DBS targets whose definition becomes progressively more and more accurate and restricted to subportions of deep brain nuclei. Peroperative micro-electrode recordings are close to this level of resolution, while the accuracy of definitive stimulating macro-electrodes is lower. Taking into consideration these various scales of resolution is a challenging issue for any system intending to localize precise targets for stereotactic functional neurosurgery, and electronic stereotactic atlases are in a central position in this process.

method stains both the cell bodies (pink to violet) and the myelin sheaths (blue to green). Immunohistochemical methods stain specific compounds of the membrane or cytoplasm, as for example, the enzyme tyrosine hydroxylase in dopaminergic neurons, the calcium-binding proteins parvalbumin, calretinin, and calbindin D28K particularly present in the basal ganglia. Finally, histology is the best method because it reveals the real cellular structure of the brain. However, it cannot be applied to living brains and therefore require other methods.

Histology: An Accurate Method to Identify Deep Brain Structures

The resolution and contrast provided by a MRI machine depends largely on the strength of its magnetic field. In this domain, clinical practice and advanced MRI research must be distinguished. The most common MRI machines encountered in neuroradiology departments are 1.5 T MRI machines. Today, 3 T machines have begun to replace 1.5 T machines in most research imaging centers, and they will progressively be installed almost everywhere for use in clinical practice. Some advanced research imaging centers are already equipped with 7 T machines. MR images are in fact created by using different sequences that provide different contrasts. T1 and T2 being relaxation times of proton spins after radio-frequency excitation, different combinations of T1 and T2 weights can be applied to obtain different contrasts. The T1-weighted MRI is commonly used to reveal the 3D anatomy of the entire brain. At 1.5 T, it clearly reveals the caudate nucleus and putamen. Optimized T2-weighted sequences at 1.5 T have been proposed that reveal the STN and substantia nigra as zones of hyposignal, whereas they are invisible in T1-weigthed sequences. This property is used to target the STN at the preoperative step of neurosurgery for Parkinson’s disease [37]. Other

Histology is the most accurate method to identify deep brain structures because it makes it possible to identify the regions in which cell bodies of neurons group together to form nuclei. This is the cytoarchitectonic method, that has been first used for establishing a parcellation of the cerebral cortex [24] and then of thalamic nuclei [25–27] or basal ganglia structures [28]. Histology can also reveal axon fascicles by staining the myelin sheaths, which has been used to construct brain atlases [29,30]. In addition, functional subdivisions can be revealed by using immunohistochemical methods such as the revelation of calcium-binding proteins [31–36]. Histology is applied to thin slices of brain tissue obtained from a dead specimen and most often submitted to a formalin fixation. There are a vast number of methods that can reveal different components of the nervous tissue. The Nissl method, with cresyl violet, thionin, or toluidin blue stains the acid components, DNA and RNA, and therefore of the cell bodies in which they are localized. The Weigert or Weil methods stain the myelin sheaths hence the axon bundles, while the Kluver–Barrera

MRI: Can Deep Brain Structures be Identified in MRI?


T2-weighted sequences have been proposed to target the globus pallidus [38]. Deep brain structures can be identified on such brain images using a number of different methods proposed by the medical image processing community, mostly on 1.5 T MR images [39–45]. These methods are referred to as direct methods as they allow one to identify structures directly in the patient’s brain MRI. On the contrary, indirect methods, referred to as atlas-based methods, require an atlas previously constructed and its adaptation to patient’s MRI. Obviously, direct methods are intrinsically limited as only the structures that are visible or partially visible in these clinical practice MR images (e.g., lateral ventricles, caudate nucleus, putamen, thalamus, optic tract) can be actually identified with some certainty. Diffusion tensor imaging (DTI) has recently appeared as a highly promising imaging technique. Based upon the diffusion properties of water molecules submitted to a magnetic field, it measures the anisotropy coefficient that depends upon the more or less privileged direction of the nervous tissue. In grey matter, nervous tissue has no privileged direction whereas in white matter, axon bundles impose a strong orientation. It is therefore possible to detect this direction as a fraction of anisotropy (FA) and to reconstruct the entire trajectory of individual axon fascicles by determining a privileged direction common to adjacent voxels in a DTI acquisition. This property has been used to identify subnuclei in the thalamus [46]. Whatever the sequence used, MRI has two limitations. First, only some deep brain structures are visible on MRI. Many brain structures remain invisible. This is particularly obvious for 1.5 T acquisitions. Development of more powerful magnetic fields provides more MR signal and allows increasing the resolution. 3 T machines, which represent the future of clinical practice, clearly provide better anatomical (T1-weighted) images, but differentiation of deep brain structures such as external and internal parts of the globus pallidus or thalamic nuclei still remains difficult. 7 T

26

research machines provide really fascinating images but many technological as well as safety issues need to be resolved before 7 T machines can be used in clinical practice. A second limitation of MRI is that the exact relationship between the nature of the MRI signal and the real nature of the cytological architecture of the nervous tissue has never been demonstrated. It must be underlined that a 1 cubic millimeter of MRI is characterized by a single parameter, namely the grey level corresponding to a combination of relaxation times of all the protons present in this cubic volume. In the real nervous tissue, the same cubic millimeter consists of hundreds of neurons with their cell bodies, local axonal arborizations and dendritic arborizations, of glial cells and of afferent axonal arborizations, which are all inter-connected in highly specific networks of synaptic contacts (> Figure 26-2). Although it has not yet fully demonstrated that MRI contrasts actually reflect the exact histological structure of the nervous tissue, it is widely accepted in the neurological and neurosurgical communities that MRI provides a reliable image of the brain of individual living subjects or patients. Researches are in progress to better understand this MRI/histology relationships in particular with the T2-weighted sequences [47,48]. Nowadays, MRI provides an irreplaceable tool to visualize the anatomy of the deep brain structures of individual patients, but MRI has nevertheless intrinsic limitations in terms of resolution. This is the main reason for which brain atlases are still being developed, together with methods to adapt the atlas to the particular configuration of the brain of a given patient. Different types of brain atlases have been developed that are presented here, and different adaptation (or deformation) methods will be described and discussed later.

2D Printed Atlases The first published atlas was that of Spiegel and Wycis [5], which was to be used in coordination

377

378

26


. Figure 26-2 Upper left box: one cubic millimeter of the globus pallidus consists of hundreds of neurons with their cells body (here revealed in blue with the Nissl method) and dendritic arborization (redrawn from Yelnik J, Percheron G, François C, A Golgi analysis of the primate globus pallidus. II. Quantitative morphology and spatial orientation of dendritic arborizations, J Comp Neurol., 227:200-213, 1984). Lower left box: the localization of this one cubic millimeter is shown in a frontal section of a MRI acquisition. Lower right box: the region of the globus pallidus is enlarged to show the one millimeter resolution of the MRI. Upper right box: the same one cubic millimeter of pallidal tissue is represented by a uniform grey level in the MRI

with the stereotactic apparatus they developed from that of Horsley and Clarke [3]. The atlas consisted of photographs of regularly-spaced sections in the sagittal, frontal, and horizontal planes. The slices were unstained and allowed identification of only gross subdivisions of grey and white matter (> Figure 26-1)Their stereotactic system was based on two intracerebral landmarks: the calcified pineal gland and the foramen of Monro by which the third and the lateral ventricles communicate. This atlas was soon followed by the atlas of Talairach [7] who first proposed the AC-PC system as the most reliable ventricular landmarks coordinate system. The first version of the atlas of Schaltenbrand appeared in 1959 [49]. It was the first atlas with

three series of histological sections with myelin sheath staining in the frontal, sagittal, and horizontal planes. The atlas provided photographs of the sections and transparent pages with a delineation of the nuclei and fiber fascicles, which made it possible to compare histology and tracings. This atlas has rapidly become a gold standard in the field of stereotactic neurosurgery, although its three-dimensional coherence was very approximate [50–53]. A second version of the same atlas was published later [29]. The atlases of Guiot of the globus pallidus and thalamus in 1961 [54], Van Buren of the basal ganglia in 1962 [55], Andrew and Watkins of the thalamus in 1969 [56], Afshar of the brain stem and cerebellum in 1978 [57], and Hassler for surgery of Parkinson’s disease in


1979 [58] followed. More recently (in 1997), an atlas based on myelin sheath staining of the human brain has been published by Paxinos and coworkers [30]. In these atlases, the number of sections and the interval between sections varied greatly (> Table 26-1), which makes the anatomical accuracy and the sampling of the 3D structure of the brain highly variable from one atlas to another.

3D Electronic Atlases The major limitation of printed atlases is that they consist of 2D information, whereas the brain is a 3D structure. Four solutions have been proposed to overcome this problem.

Construction of 3D Atlases from 2D Printed Histological Atlases In 1987, Yoshida at Kurume, Japan, was probably the first who proposed the creation of a 3D atlas [59] by interpolation of the Schaltenbrand and Bailey’s atlas [49]. In 1997, Nowinski started to work on an ideal digital stereotactic atlas, which fuses the Talairach and Schaltenbrand atlases into a common navigation software tool (> Figure 26-3) . Table 26-1 Inter section interval in different histological atlases Existing printed atlases

Inter section interval

Spiegel and Wycis [5] Talairach et al. [7] Schaltenbrand and Bailey [49] Guiot et al. [54] Van Buren and MacCubbin [55] Andrew and Watkins [56] Schaltenbrand and Wahren [29] Afshar et al. [57] Hassler et al. [58] Talairach and Tournoux [9] Mai et al. [30] Morel et al. [31]

5 mm 3–6 mm 1–4 mm 0.5–1 mm 5 mm 1 mm 1–4 mm 1 mm 2 mm 4 mm 0.7–2.5 mm 0.9–1.8 mm

26

[60,61]. In 1998 the group of Collins and Peters at the Montreal Neurological Institute (MNI) of McGill University in Montreal, Canada [62], constructed an atlas from the Schaltenbrand atlas [29] by first aligning the successive sections on thebasis ofthegridpresentin theatlas,thencreating a 3D volume and submitting it to an interpolation operation using spline parameterization. The group of Ganser in Heidelberg, Germany, in 2003, constructeda3DdigitalversionoftheTalairachatlas [9]. They scanned the series of coronal sections, segmented manually the different structures, interpolated additional cross-sections, and created shell surfaces of each structure. The problem with such printed atlases is that their 3D accuracy and coherency is primarily dependant on the quality of the 2D individual sections of the original atlas but above all on the carefulness with which successive sections have been aligned one onto the other. Unfortunately, the two most widely used atlases, those of Schaltenbrand and Talairach, do not fulfill these requirements. The atlas of Schaltenbrand is well-known for its 3D inconsistency [50,51], and the atlas of Talairach is based on photographed sagittal sections of the brain of a 60year-old woman from which coronal and axial sections were obtained by manual interpolation.

Construction of 3D Atlases from MRI In order to build fully 3D brain atlases, several groups have proposed MRI-based atlases. These types of atlases are useful because they allow easy three-dimensional navigation in the brain. Also, as MRI is nowadays the reference in vivo brain imaging, it is very useful to be able to study correspondence between the MR signal and subcortical structure delineation. Started in 1990, the project of the Surgical Planning Laboratory (MGH, Harvard Medical School, Mass., USA) aimed at proposing a detailed

379

380

26


. Figure 26-3 The electronic atlas developed by Wieslaw Nowinski (References 60,61) consists of a digitized version of the atlas of Schaltenbrand and Wahren (Reference 29). a shows a frontal section the original atlas, b shows the electronic contours in both hemispheres, c shows color-coded images and d shows a 3D view of the thalamic nuclei

morphological brain atlas built from a T1-weighted (i.e., anatomical) MRI. The first goal of this project was to develop a tool for education. It is today referred to as the SPL anatomy browser and it is freely available on the Internet (www.spl. harvard.edu). It consists of a browser that permits navigation in a T1-weighted MRI volume in which numerous cortical and subcortical structures are included (> Figure 26-4). This atlas has also been used for presurgical planning and segmentation tasks. In 2002, the Neurofonctional Imaging Group (GIN, UMR6095, CYCERON, Caen, France)

proposed a macroscopic anatomical parcellation of the MNI MRI single subject brain used by the functional brain imaging community. This atlas, referred to as the AAL (Anatomical Automatic Labelling), has been included in the Statistical Parametric Mapping (SPM, Wellcome Department of Imaging Neuroscience, UCL, London, UK) software as a toolbox. Although this atlas is devoted to fMRI studies and cortical mapping, it includes some subcortical structures (caudate nucleus, putamen, pallidum, thalamus). More generally, many medical image processing groups have developed homemade


26

. Figure 26-4 The Surgical Planning Laboratory (SPL) browser developed at the Harvard Medical School (www.spl.harvard.edu) consists of a T1-weighted MRI volume with cortical and subcortical structures digitized and labeled

MRI-based 3D brain atlases that they have included in brain segmentation methods. These methods can be grouped under the naming ‘‘pseudo-direct methods,’’ as they all work directly in the living patient’s brain MRI, but use a priori information provided by these homemade atlases. Among these methods, Pitiot et al. proposed in 2004 an expert knowledge-guided system that allowed to robustly identify the corpus callosum, lateral ventricles, hippocampus, and caudate nucleus on in vivo MRI [39]. In 2005, Zhou et al. proposed a feature-based method using fuzzy templates built from a training set, which allowed to segment five subcortical structures, thalamus, putamen, caudate, hippocampus, and amygdala [63]. Fischl et al. in 2002

have presented a very nice and powerful whole brain segmentation algorithm [43]. The method is defined in a probabilistic framework and assigns 1 of 37 labels to each voxel of the brain volume. Subcortical structures include the caudate, putamen, pallidum, thalamus, hippocampus, and amygdala (> Figure 26-5). This algorithm is available as it has been included in the FreeSurfer software (MGH, Harvard, Mass., USA). In a radiotherapy context, Bondiau et al. [64] have also designed a whole brain atlas that comprises organs at risk, including thalamus, caudate, putamen, and pallidum, and is embedded in a segmentation pipeline [64]. In 2005, Lemaire and coworkers of Clermont Ferrand, France, proposed another solution based

381

382

26


. Figure 26-5 An automated algorithm for the probabilistic segmentation of a MRI volume has been developed by Fischer and coworkers (reference 43, http://surfer.nmr.mgh.harvard.edu/). Each voxel is assigned to one of the 37 structures available

on a MRI obtained from a post mortem specimen submitted to both a prolonged high resolution 4.7 T acquisition and a standard 1.5 T acquisition. The MRIs were studied by comparison with available printed histological atlases and imaging criteria to identify deep brain structures directly in the 1.5 T MRIs of individual patients were described [65]. However, in the context of DBS clinical practice, MRI is limited in terms of spatial resolution and contrast, which makes these approaches less accurate. As mentioned earlier, histology thus remains the best tool to reveal the anatomy of the basal ganglia, but histology can provide only bidimensional images of the brain. The challenge is therefore to obtain a ‘‘three-dimensional histology’’ by any possible tool.

Construction of 3D Atlases from Histology The elaboration of a 3D atlas from 2D histological sections requires first a good histology (choice of staining techniques, careful processing of individual sections, and reliable tracing by experts in brain anatomy), and second a method that would allow to align the whole series of individual sections into an anatomically

consistent 3D block. Indeed, sectioning of an anatomical specimen for histology provides a series of disconnected 2D slices, whose original 3D shape is lost. Several methods have been proposed to solve this problem, one of which being the use of an accurate system of landmarks, another one being coregistration of successive 2D slices [66]. Histological 3D atlases have been built by three different teams. An atlas of the human thalamus was constructed in 1997 in Zurich, Switzerland, on the basis of cyto- and myeloarchitectonic criteria and on the use of calcium-binding proteins (parvalbumin, calretinin, calbindin D-28K) as functional markers (> Figure 26-6). [25]. Three series of successive sections were traced (sagittal, horizontal, and frontal) from three anatomical specimens and were subsequently normalized with each other (frontal and horizontal to sagittal) to obtain an internally consistent Canonical model of the atlas [67]. Two atlases of the basal ganglia and thalamus have been published successively; one by the McGill Hospital in Montreal, Canada in 2006 [68], and the other one by the authors of this review in the Salpeˆtrie`re Hospital, Paris, France in 2007 [28]. The Canadian atlas was developed by Louis Collins as a continuation of his previous work on the Schaltenbrand atlas [62]. It consisted


of sections stained alternatively with Luxol Blue for myelin and with a Nissl stain for cell bodies (n = 86 pairs of sections) and were aligned through a slice-to-slice nonlinear registration, . Figure 26-6 A 3D histological atlas of the thalamus constructed by Morel and co-workers (reference 25) and transformed into a digitized 3D stereotactic atlas (Reference 67) is used to define a trajectory that pierces the central lateral nucleus of the thalamus (indicated by the arrow)

26

which was optimized by minimizing the mean distance between the segmented contours in adjacent pairs of slices. Then, 3D geometric objects were created by tessellation to represent different anatomic regions (> Figure 26-7). An additional feature of the French atlas was a MRI acquisition (T1- and T2-weigthed sequences) performed previous to brain extraction, which provided a reliable anatomical reference for both the construction of the atlas from stained histological sections and the adaptation of the atlas to the MRIs of individual patients. Histological sections were stained alternatively with Nissl stain and the calcium binding protein calbindin D-28K (n = 80 pairs of sections), and were aligned by piecewise linear coregistration with the MRI and a cryo block constructed from photographs taken during cryosectioning. Contours of basal ganglia and thalamus were traced from histological sections

. Figure 26-7 The atlas developed by Louis Collins and co-workers (Reference 68) consists of anatomical regions digitized from histological data and that can be transferred to the MRI volumes of different patients

383

384

26


and digitized. A specific procedure combining a multimodal optimization (MRI, cryoblock, nissl, calbindin) and a 3D optimization was implemented to assure an optimal 3D coherency [28]. Anatomically and geometrically consistent 3D

surfaces of each traced region were constructed by shape-based interpolation (> Figure 26-8). The specific qualities of the French atlas are a histological level of resolution, the inclusion of functional information based on calbindin

. Figure 26-8 The atlas developed by the authors of this review (Reference 28) consists of anatomical regions digitized from histological sections of a post-mortem specimen. Registration between atlas and patients is calculated from the MRI acquisitions of the same specimen obtained before histological sectioning. The structures transferred to a patient’s MRI are shown in 3D (upper line), in a T1- (middle line) and a T2-weighted acquisition (bottom line)


staining, a large number of sections (160 sections with 0.35 mm interval), and MRI acquisitions of the same specimen that allowed the construction of truly continuous 3D surfaces. This differs from the Talairach atlas in which the contours of cerebral structures were traced from one specimen sectioned in the sagittal plane (section interval section 4 mm) and were extrapolated in the coronal and axial planes by point-to-point projection (interval section 5 mm). In the Schaltenbrand and Wahren atlas, three series of sections are available (sagittal, coronal, and axial) but the number of sections is low (18, 20, 20), the section interval is high and variable (1–4 mm), and the 3D coherency is very low. The atlas of Collins and coworkers was derived from a set of serial histological data (0.7 mm section interval, 86 pairs of slices). It closely resembles the French atlas except that it lacks a MRI of the brain specimen. To overcome this problem, a pseudoMRI was created from the reconstructed voxellabeled atlas volume and used for adapting the atlas to patients through nonrigid registration.

Functional Population-based Atlases Another strategy consists in building an atlas from functional data collected in a population of subjects or patients. The data can be peroperative electrophysiological recordings or clinical exploration data (e.g., points that provoke arm dyskinesias or somesthetic perceptions on the hand) or postoperative tuning data (e.g., a contact that provokes heat sensation or diplopia). Once these data have been collected, they are placed in a common reference space that defines the atlas. Terry Peters and colleagues, who originally worked at the MNI and participated to the histology-based atlas [62], moved to London, Ontario, and started to develop a functional population-based strategy [69,70]. Electrophysiological and clinical peroperative data obtained from a

26

series of 88 patients operated for DBS at the London Health Science Centre were used to define this atlas which is, in turn, integrated into a computeraided system for DBS targeting (> Figure 26-9). A somewhat similar system has been developed by the group of Benoit Dawant at Vanderbilt University, Nashville, Tennessee. Pierre-François D’Haese has developed during his PhD thesis an atlas-based method for the automatic determination of DBS targets [71,72]. Their atlas is based on micro electrode recordings, stimulation parameters and final implant positions. In addition, a population-based electrophysiological map has been created by analyzing automatically the peroperative micro electrode recordings with signal processing techniques (wavelets-based de-noising, spike detection). Nowinski and Benabid have developed an atlas that they have called the PFA (probabilistic and functional atlas), which combines pre-, per-, and postoperative neuro-imaging data with peroperative electrophysiological data from 274 parkinsonian patients operated at the Joseph Fourier University School of Medicine in Grenoble, France [73–75]. As mentioned earlier, the strategy for building such population-based atlases is to place data coming from different patients in a common reference space that is to say to ‘‘normalize.’’ This is done by applying spatial normalization algorithms on the patients’ images (most often MR images), which allows one to compensate for inter-individual brain shape variability. Spatial normalization consists in computing a deformation between the patient’s image and a reference image chosen arbitrarily. More generally, computing a deformation between a reference and the patient is the necessary remaining step, whatever the atlas retained. Indeed, by definition an atlas is a template, thus a unique set of data which must be adapted, or deformed, to fit the particular brain geometry of each subject or patient. The crucial issue is to make the best possible choice of a reliable deformation procedure to adapt one of the available atlases to the

385

386

26


. Figure 26-9 The atlas developed by Tony Peters and co-workers (Reference 70) consists of electrophysiological and clinical preoperative data obtained in a series of 88 patients, here indicated as small spheres localized within a 3D MRI volume

individual brain anatomy of each patient, and to validate this choice.

From Atlases to the Brain of Living Patients An atlas can consist of series of anatomical 2D maps, 3D regions segmented in a reference MRI, 3D surfaces or functional data localized in a reference space. The brain geometry of subjects or patients is generally available in the form of a 3D MRI or a scanner. Deformation of an atlas can be performed using different strategies but all strategies are not appropriate for all types of atlases. There is also a ‘‘cultural’’ aspect in choosing among the variety of deformation strategies available nowadays. Indeed, adapting a brain atlas to the brain of a patient in the DBS context appears to be at the crossroad of two, rather different scientific communities: on the one side, the neurosurgical and neuroanatomical communities who traditionally have used landmark-based deformation methods, and on the other side the medical image processing community who is particularly expert in the development of automatic deformation, or registration algorithms.

Visual Deformation This is the most ancient way of using the information contained in an atlas. The user has a printed version of the atlas in one hand, generally a series of 2D anatomical plates with labeled structures, and a series of anatomical sections of the brain subject to be labeled in the form of MRI sections (or histological sections for a dead specimen) in the other hand. He must decide visually what atlas section best corresponds to a given subject section and then places mentally or using a tracing paper the atlas section on the subject section. Then he can attribute to a given point of the subject section an anatomical label coming from the atlas section. Such a procedure is in fact an atlas adaptation, but a totally manual one with mental/visual identification of the required landmarks. A somewhat original alternative relies on the direct analysis of MRI acquisitions. Deep brain structures are directly identified in the MRI acquisition of the subject by comparison with a previous study of a dead brain specimen MRI [41,65]. This procedure requires a skilled expertise of the user and above all, raises the question of the actual anatomic significance and reliability of the MRI signal.


Manual or Semiautomatic Linear Deformations The most widely used system in the DBS community to adapt a brain atlas to the individual anatomy of a living subject is the proportional system of Talairach [9]. It relies primarily on the AC-PC distance, i.e., the length between the anterior and posterior commissural points, well-identifiable on a ventriculography or a mid-sagittal section of a MRI acquisition. As the Talairach atlas has been constructed in reference to the AC-PC line, the user has just to measure the AC-PC distance in the living brain and to adapt the antero-posterior length of the atlas to that of the brain. In addition, the proportional system comprises an adaptation along the medio-lateral and infero-superior dimensions, which is based on the overall size of the brain, including the entire cerebral cortex, along these two axes. The three adaptations are independent of one another, which takes into account the size and the particular shape of each studied brain (e.g., dolichocephalic or brachycephalic brains). A grid system is proposed that makes the practical application of the proportional system easily applicable even without any computer. The system has been largely applied all over the world as for example in the context of lesional surgery for the determination of the targets for thalamotomy or pallidotomy [54]. The proportional system of Talairach is therefore a reliable system, although it is inhomogeneous since the adaptation along the antero-posterior dimension is based upon two deep brain ventricular landmarks, whereas adaptation along the medio-lateral and inferosuperior dimensions depends on the overall size of the cerebral cortex. This is due to the fact that with ventriculography, internal landmarks are less clear along these dimensions (the height of the thalamus and width of the third ventricle are the best possible landmarks). Another difficulty of the AC-PC system is the definition of the two commissural points, which can vary slightly but

26

significantly from one atlas to another and from one user to another. AC-PC length and angle can be defined from the center of each commissural point or from the most anterior and posterior points, i.e., minimizing the AC-PC length. In the Schaltenbrand [29] atlas, for example, the ACPC line of the horizontal series passes by the upper border of AC and the lower border of PC, whereas it passes through the center of the two points in the two other series (frontal and sagittal). This makes a 7 angle difference and a significantly different topological and metrical aspect of cerebral regions in the 2D atlas sections. Other linear semiautomated methods have been proposed. In our studies of the localization of the DBS electrodes in the GPi [76] and STN [77], we proposed a method of tri-linear deformation of the Schaltenbrand atlas [29] based on internal brain landmarks, namely the AC-PC landmarks along the antero-posterior dimension, the superior limit of the putamen and the individualization of the cerebral peduncles along the supero-inferior dimension, the lateral limit of the putamen, the anterior column of the fornix, and the mamillo-thalamic tract along the medio-lateral dimension [77]. These landmarks, also used by others [67], are more appropriate than the outer limits of the cerebral cortex which in fact has not strong relationships with the deep brain nuclei. In the functional atlas of Nowinski and Benabid [73,74], the landmarks are the ACPC distance, the height of the thalamus, and the width of the third ventricle and the deformation of atlas data also follows a tri-linear procedure. Besides the aforementioned atlas-based localization studies, it is worth noting that several groups have conducted localization studies that aimed at defining the optimal DBS target from a population of patients, but without any reference to an atlas. These studies included spatial normalization transformations, because for comparing the optimal contacts in a group of patients, it is necessary to place all the data available in a common or template space (just the same as for

387

388

26


the construction of population-based atlases). These works can therefore be viewed as population-based studies. Lanotte et al. [78] studied a series of 14 consecutive patients and expressed mean positions and distances (e.g., position of the central point of the most effective contact) with respect to the midpoint AC-PC line. No spatial normalization was applied to the individual data before their transfer in the reference space. Saint-Cyr et al. [79] studied a series of 29 patients (54 postoperative macro-electrodes). Spatial normalization of electrode contact positions was done along the antero-posterior axis using the AC-PC length. In Hamel et al. [80], 25 patients were included, spatial coordinates expressed with respect to the mid-commissural point, and spatial normalization was performed along the antero-posterior axis.

Automatic Methods of Deformation Several groups of the medical image processing community have developed brain atlases, some of which are dedicated to DBS [62,71,81–83]. These groups have a high level of expertise in the development of automatic deformation (or registration) algorithms. Therefore, it was almost natural for them to apply these algorithms to the atlas-to-patient adaptation problem. An automatic registration algorithm is based on the comparison of features (grey-level values, points, lines, graphs, etc.) present in the two images to be registered. The algorithm is defined by three main characteristics: the similarity measure, the space of allowed deformations (the number of degrees-of-freedom (DoF) of the deformation, e.g., 6 DoF for a 3D rigid transform), and the optimization method that is used. Deformations can be very constrained (limited number of DoF), e.g., linear scaling (7 DoF) or not, like elastic, fluid, or even free-form deformations. These last types of deformations are

often referred to as morphing or warping transforms. The choice of the most adequate deformation type is important, as it directly influences the quality of the atlas-to-patient result. Kikinis, Dengler and coworkers have proposed to deform their MRI-based brain atlas (the core of the SPL anatomy browser) by an elastic registration procedure consisting of a warping of the atlas image onto the patient image [84]. Ganser et al. have developed a nonrigid registration method to deform their 3D digital version of the Talairach atlas, consisting of the automated establishment of point correspondences between atlas and patient (these points being defined on the skull and the ventricles) and then interpolation of the corresponding displacement vectors using radial basis functions (> Figure 26-10) [85]. Louis Collins, at the McGill University Montreál, Canada, developed in 1994 the ANIMAL (Automated Nonlinear Image Matching and Anatomical Labelling) algorithm [86]. The algorithm is twofold: first a constrained affine transformation (translations, rotations, scales), followed by the calculation of a 3D nonlinear deformation field . Figure 26-10 The atlas developed by Ganser and coworkers (Reference 85) proposes a nonrigid registration method with which the atlas of Talairach (Reference 9) is deformed to fit the brain MRI geometry of a given patient


in a piece-wise fashion, fitting cubical neighborhoods in sequence. This algorithm has been applied to register the patient and atlas volumes of St-Jean [62]. It has also been used by Peters and colleagues for deforming their populationbased atlas on patients [70]. The same task was solved by a nonrigid registration algorithm based on radial basis functions in Dawant et al. [71]. These methods have in common a characteristic which is to register the images of the entire brains. On the contrary, Bardinet et al. propose a method that, after a first constrained alignment computed on the whole head, allows registering only the region of interest centered on the atlas of the basal ganglia [83].

Validation of the Atlas-to-patient Deformation Image processing algorithms, whether they are segmentation or registration, must be validated, which is always a difficult task. People of the medical image processing community have developed validation procedures adapted to a given type of algorithm. Image registration, which is crucial in atlas-to-patient deformation, can be validated by geometrical criteria (see for example, Hellier et al. [87]). But the key issue in validation is to clearly identify the context, the application and the question which is addressed, and to adapt the validation procedure to the problem to be solved. In the context of functional neurosurgery, the application is DBS. We have described different types of atlases and different tools to adapt a given atlas to the brain of a patient. The deformed atlas is supposed to give a precise and detailed description of the anatomy of the patient’s brain. The question that must be answered is therefore how to insure that the deformed atlas gives a reliable and anatomically plausible representation of the individual anatomy of the patient? This obviously depends on the type of the atlas used. If 2D printed atlases are used to identify brain regions in the MRI of a patient, no validation can

26

be performed and the expert’s opinion is the only reference available. If 3D electronic atlases are used, quantitative validation becomes possible as the atlas can be superimposed on the patient’s brain MR image. As the deformed atlas provides a segmentation of the brain, validation can be performed separately for each segmented structure. Segmentation of brain structures that are at least partially visible on MRI (such as the caudate or putamen) can be validated by comparing manually outlined brain structures with the same structures automatically segmented by the deformed atlas. Overlap indices, e.g. the Dice similarity coefficient, can be computed. For structures that are not visible on MRI, given by histological or functional population-based atlases, validation has to be done differently. How to insure that histological contours have been mapped accurately to an MRI with a mean voxel size of 1 mm? One way is to validate the MRI-based registration method that has been proposed [71,81]. But the ideal validation would be based on data that are comparable in scale to the level of definition of the histology. In DBS, these data are available and consist in the peroperative micro-electrode recordings. Indeed, these recordings, consisting of stereotactic positions and anatomical labels noted by the electrophysiologist provide a local representation of the brain structural organization that can be confronted to the information given by a histological atlas. Chakravarty et al. [82] have used intraoperative recordings of the sensory thalamus (during thalamotomy for Parkinson’s disease) to locally validate the deformation of their atlas. Bardinet et al. [83] confront their atlas with reconstructed electrophysiological recordings of the STN area collected during STN DBS of Parkinson’s disease patients.

Conclusion The perspective in the domain of brain localization remains completely open, with the goal of filling

389

390

26


in the gap that still exists between histology and imagery. (1) To date, histology remains the sole method that can reveal not only the precise anatomical structure of the nervous tissue from the cellular to the regional level, but also the functional aspect of the basal ganglia, for example, its sensorimotor, associative and limbic subdivisions that immunohistochemical methods can reveal. Progress in this domain should concern individual variations of deep brain nuclei size and shape and particularly at different ages of life and in different pathological conditions. (2) Imagery has become the irreplaceable method for visualizing the brain anatomy of a given living patient and particularly its characteristic three-dimensional architecture. Progress in this domain is brought into play before our eyes and concerns both the level of resolution of MRI, which will probably increase tremendously in the next few years, and the nature of imaging with the spectacular development of DTI-based tractography. (3) The decisive progress is in the hands of the image processing community, which should now develop more and more powerful techniques of image deformation to adapt a more and more sophisticated histological and immunohistological knowledge to individual imagery at the finest level of resolution. A sort of 3D histology could thus be obtained for individual living patients, from which a probabilistic population-based brain anatomy could emerge.

References 1. Horsley V, Clarke RH. The structure and function of the cerebellum examined by a new method. Brain 1908;31:54-124. 2. Paxinos G, Watson C. The rat brain in stereotaxic coordinates. 4th ed. London: Academic press; 1998. 3. Spiegel EA, Wycis HT, Marks M, et al. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 4. Dandy WE. Ventriculography following the injection of air into the cerebral ventricles. Ann Surg 1918;68(1):5-11. 5. Spiegel EA, Wycis HT. Stereoencephalotomy (thalamotomy and related procedures). New York: Grune and Stratton; 1952.

6. Talairach J, de Ajuriaguerra J, Hecaen H. Etude topographique des images ventriculaires en fonction des noyaux gris centraux. Annales de M,decine 1950;51: 83-100. 7. Talairach J, David M, Tournoux P, Corredor H, Kvasina T. Atlas d’anatomie ste´reótaxique. Paris: Masson; 1957. 8. Talairach J, Szikla G. Atlas d’anatomie ste´reótaxique du te´lence´phale. Paris: Masson; 1967. 9. Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain. New York: Thieme; 1998. 10. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger JA. Physiological observations in stereotaxic operations in extrapyramidal motor disturbances. Brain 1960;83:337-50. 11. Hassler R, Dieckmann G. Stereotaxic treatment of compulsive and obsessive symptoms. Confin Neurol 1967;29 (2):153-8. 12. Hassler R, Mundinger F, Riechert T, Gillingham FJ, Donaldson IML. Target reliability of the stereotaxic method as compared with autopsie findings. In the Third Symposium on Parkinson’s Disease. Livingstone; 1969:206–10. 13. Hassler R, Dieckmann G. Stereotactic treatment of different kinds of spasmodic torticollis. Confin Neurol 1970;32:135-43. 14. Benabid AL, Pollak P, Louveau A, Henry S, De Rougemont J. Combined (thalamotomy and stimulation) stereotactic surgery of the VIM thalamic nucleus for bilateral Parkinson disease. Appl Neurophysiol 1987;50:344-6. 15. Hosobuchi Y, Adams JE, Rutkin B. Chronic thalamic stimulation for the control of facial anesthesia dolorosa. Arch Neurol 1973;29(3):158-61. 16. Perlmutter JS, Mink JW. Deep brain stimulation. Annu Rev Neurosci 2006;29:229-57. 17. Benabid AL, Pollak P, Gervason C, et al. Long-term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991;337 (8738):403-6. 18. Kumar R, Lozano AM, Montgomery E, Lang AE. Pallidotomy and deep brain stimulation of the pallidum and subthalamic nucleus in advanced Parkinson’s disease. Mov Disord 1998;13 Suppl 1:73-82. 19. Bergman H, Wichmann T, DeLong MR. Reversal of experimental parkinsonism by lesions of the subthalamic nucleus. Science 1990;249:1436-8. 20. Benazzouz A, Gross C. F.ger J, Boraud T, Bioulac B. Reversal of rigidity and improvement of motor performance by subthalamic high-frequency stimulation in MPTP-treated monkeys. Eur J Neurosci 1993;5:382-9. 21. Limousin P, Pollak P, Benazzouz A, et al. Effect of parkinsonian signs and symptoms of bilateral subthalamic nucleus stimulation. Lancet 1995;345(8942):91-5. 22. Yelnik J. Functional anatomy of the basal ganglia. Mov Disord 2002;17 Suppl 3:S15-S21.


23. Butson CR, Cooper SE, Henderson JM, McIntyre CC. Patient-specific analysis of the volume of tissue activated during deep brain stimulation. NeuroImage 2007;34 (2):661-70. 24. Brodmann K. Vergleichende Lokalisationslehre der Grosshirnrinde: in ihren Principien dargestellt auf Grund des Zellenbaues. Leipzig: Barth; 1909. 25. Morel A, Magnin M, Jeanmonod D. Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387(4):588-630. 26. Jones EG, Burton H. Cytoarchitecture and somatic sensory connectivity of thalamic nuclei other than the ventrobasal complex in the cat. J Comp Neurol 1974;154:395-432. 27. Percheron G. The motor thalamus. J Neurosurg 1997;87 (6):981-2. 28. Yelnik J, Bardinet E, Dormont D, et al. A three-dimensional, histological and deformable atlas of the human basal ganglia. I. Atlas construction based on immunohistochemical and MRI data. NeuroImage 2007;34 (2):618-38. 29. Schaltenbrand G, Wahren W. Atlas for stereotaxy of the human brain. Stuttgart: Georg thieme Verlag, 1977. 30. Mai JK, Assheuer J, Paxinos G. Atlas of the human brain. San Diego: Academic Press; 1997. 31. Morel A, Loup F, Magnin M, Jeanmonod D. Neurochemical organization of the human basal ganglia: anatomofunctional territories defined by the distributions of calcium-binding proteins and SMI-32. J Comp Neurol 2002;443(1):86-103. 32. François C, Yelnik J, Percheron G, Tande´, D. Calbindin D-28k as a marker for the associative cortical territory of the striatum in macaque. Brain Res 1994;633:331-6. 33. Prensa L, Richard S, Parent A. Chemical anatomy of the human ventral striatum and adjacent basal forebrain structures. J Comp Neurol 2003;460(3):345-67. 34. Karachi C, Francois C, Parain K, et al. Threedimensional cartography of functional territories in the human striatopallidal complex by using calbindin immunoreactivity. J Comp Neurol 2002;450(2):122-34. 35. Augood SJ, Waldvogel HJ, Mnkle MC, Faull RLM, Emson PC. Localization of calcium-binding proteins and GABA transporter (GAT-1) messenger RNA in the human subthalamic nucleus. Neuroscience 1999;88 (2):521-34. 36. Hontanilla B, Parent A, Gim,nez-Amaya JM. Parvalbumin and calbindin D-28k in the entopeduncular nucleus, subthalamic nucleus, and substantia nigra of the rat as revealed by double-immunohistochemical methods. Synapse 1997;25(4):359-67. 37. Bejjani BP, Dormont D, Pidoux B, et al. Bilateral subthalamic stimulation for Parkinson’s disease by using three-dimensional stereotactic magnetic resonance imaging and electrophysiological guidance. J Neurosurg 2000;92(4):615-25.

26

38. Hariz MI, Krack P, Melvill R, et al. A quick and universal method for stereotactic visualization of the subthalamic nucleus before and after implantation of deep brain stimulation electrodes. Stereotact Funct Neurosurg 2003;80 (1–4):96-101. 39. Pitiot A, Delingette H, Thompson PM, Ayache N. Expert knowledge-guided segmentation system for brain MRI. NeuroImage 2004;23 Suppl 1:S85-S96. 40. Shen D, Herskovits EH, Davatzikos C. An adaptive-focus statistical shape model for segmentation and shape modeling of 3-D brain structures. IEEE Trans Med Imaging 2001;20(4):257-70. 41. Barra V, Boire JY. Automatic segmentation of subcortical brain structures in MR images using information fusion. IEEE Trans Med Imaging 2001;20(7):549-58. 42. Bueno G, Musse O, Heitz F, Armspach JP. Threedimensional segmentation of anatomical structures in MR images on large data bases. Magn Reson Imaging 2001;19(1):73-88. 43. Fischl B, Salat DH, Busa E, et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 2002;33(3):341-55. 44. Gonzalez Ballester MA, Zisserman A, Brady M. Segmentation and measurement of brain structures in MRI including confidence bounds. Med Image Anal 2000;4 (3):189-200. 45. Mangin JF, Poupon F, Duchesnay E, et al. Brain morphometry using 3D moment invariants. Med Image Anal 2004;8(3):187-96. 46. Behrens TE, Johansen-Berg H, Woolrich MW, et al. Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nat Neurosci 2003;6(7):750-7. 47. Dormont D, Ricciardi KG, Tande D, et al. Is the subthalamic nucleus hypointense on T2-weighted images? A correlation study using MR imaging and stereotactic atlas data. AJNR Am J Neuroradiol 2004;25(9):1516-23. 48. Meadowcroft MD, Zhang S, Liu W, et al. Direct magnetic resonance imaging of histological tissue samples at 3.0 T. Magn Reson Med 2007;57(5):835-41. 49. Schaltenbrand G, Bailey P. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Georg Thieme Verlag; 1959. 50. Nowinski WL, Liu J, Arumugam T. Quantification and visualization of three-dimensional inconsistency of the globus pallidus internus in the Schaltenbrand-Wahren brain atlas. Stereotact Funct Neurosurg 2006;84(5– 6):236-42. 51. Nowinski WL, Liu J, Thirunavuukarasuu A. Quantification and visualization of the three-dimensional inconsistency of the subthalamic nucleus in the Schaltenbrand-Wahren brain atlas. Stereotact Funct Neurosurg 2006;84 (1):46-55. 52. Niemann K, van Nieuwenhofen I. One atlas – three anatomies: relationships of the Schaltenbrand and

391

392

26 53.

54.

55.

56.

57.

58. 59.

60.

61.

62.

63.

64.

65.

66.

67.

68.


Wahren microscopic data. Acta Neurochir (Wien) 1999;141(10):1025-38. Niemann K, Naujokat C, Pohl G, Wollner C, von Keyserlingk D. Verification of the Schaltenbrand and Wahren stereotactic atlas. Acta Neurochir (Wien) 1994;129(1–2):72-81. Guiot G, Brion S, Akerman M. Anatomie ste´reótaxique du pallidum interne du thalamus et de la capsule interne. Etude des variations individuelles. Ann Chir 1961;15:557-86. Van Buren JM, MacCubbin DA. An outline atlas of the human basal ganglia with estimation of anatomical variants. J Neurosurg 1962;19:811-39. Andrew J, Watkins ES. A stereotaxic atlas of the human thalamus and adjacent structures. Baltimore: Williams and Wilkins; 1969. Afshar F, Watkins ES, Yap JC. Stereotaxic atlas of the human brain-stem and cerebellar nuclei. A variability study. New York: Raven Press; 1978. Hassler R, Mundinger F, Riechert T. Stereotaxis in Parkinson syndrome. Berlin: Springer; 1979. Yoshida M. Creation of a three-dimensional atlas by interpolation from Schaltenbrand-Bailey’s atlas. Appl Neurophysiol 1987;50(1–6):45-8. Nowinski WL. Towards construction of an ideal stereotactic brain atlas. Acta Neurochir (Wien) 2008;150 (1):1-13. Nowinski WL, Fang A, Nguyen BT, et al. Multiple brain atlas database and atlas-based neuroimaging system. Comput Aided Surg 1997;2(1):42-66. St-Jean P, Sadikot AF, Collins L, et al. Automated atlas integration and interactive three-dimensional visualization tools for planning and guidance in functional neurosurgery. IEEE Trans Med Imaging 1998;17 (5):672-80. Zhou J, Rajapakse JC. Segmentation of subcortical brain structures using fuzzy templates. NeuroImage 2005;28 (4):915-24. Bondiau PY, Malandain G, Chanalet S, et al. Atlas-based automatic segmentation of MR images: validation study on the brainstem in radiotherapy context. Int J Radiat Oncol Biol Phys 2005;61(1):289-98. Villeger A, Lemaire JJ, Boire JY. Localization of target structures through data fusion applied to neurostimulation. Conf Proc IEEE Eng Med Biol Soc 2005;2:1508-11. Malandain G, Bardinet E, Nelissen K, Vanduffel W. Fusion of autoradiographs with an MR volume using 2-D and 3-D linear transformations. NeuroImage 2004;23(1):111-27. Niemann K, Mennicken VR, Jeanmonod D, Morel A. The Morel stereotactic atlas of the human thalamus: atlasto-MR registration of internally consistent canonical model. NeuroImage 2000;12(6):601-16. Chakravarty MM, Bertrand G, Hodge CP, Sadikot AF, Collins DL. The creation of a brain atlas for image guided

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

neurosurgery using serial histological data. NeuroImage 2006;30(2):359-76. Guo T, Finnis KW, Parrent AG, Peters TM. Development and application of functional databases for planning deep-brain neurosurgical procedures. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8(Pt 1):835-42. Finnis KW, Starreveld YP, Parrent AG, Sadikot AF, Peters TM. Three-dimensional database of subcortical electrophysiology for image-guided stereotactic functional neurosurgery. IEEE Trans Med Imaging 2003;22 (1):93-104. D’Haese PF, Cetinkaya E, Konrad PE, Kao C, Dawant BM. Computer-aided placement of deep brain stimulators: from planning to intraoperative guidance. IEEE Trans Med Imaging 2005;24(11):1469-78. D’Haese PF, Pallavaram S, Niermann K, et al. Automatic selection of DBS target points using multiple electrophysiological atlases. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8(Pt 2):427-34. Nowinski WL, Belov D, Thirunavuukarasuu A, Benabid AL. A probabilistic functional atlas of the VIM nucleus constructed from pre-, intra- and postoperative electrophysiological and neuroimaging data acquired during the surgical treatment of Parkinson’s disease patients. Stereotact Funct Neurosurg 2005;83(5–6):190-6. Nowinski WL, Belov D, Pollak P, Benabid AL. A probabilistic functional atlas of the human subthalamic nucleus. Neuroinformatics 2004;2(4):381-98. Nowinski WL, Belov D, Benabid AL. An algorithm for rapid calculation of a probabilistic functional atlas of subcortical structures from electrophysiological data collected during functional neurosurgery procedures. NeuroImage 2003;18(1):143-55. Yelnik J, Damier P, Bejjani BP, et al. Functional mapping of the globus pallidus. Contrasting effect of stimulation in the internal and external pallidal nuclei in Parkinson’s disease. Neuroscience 2000;101(1):77-87. Yelnik J, Damier P, Demeret S, et al. Localization of stimulating electrodes in patients with Parkinson disease by using a three-dimensional atlas-magnetic resonance imaging coregistration method. J Neurosurg 2003;99 (1):89-99. Lanotte MM, Rizzone M, Bergamasco B, Faccani G, Melcarne A, Lopiano L. Deep brain stimulation of the subthalamic nucleus: anatomical, neurophysiological, and outcome correlations with the effects of stimulation. J Neurol Neurosurg Psychiatry 2002;72(1):53-8. Saint-Cyr JA, Hoque T, Pereira LC, et al. Localization of clinically effective stimulating electrodes in the human subthalamic nucleus on magnetic resonance imaging. J Neurosurg 2002;97(5):1152-66. Hamel W, Fietzek U, Morsnowski A, et al. Deep brain stimulation of the subthalamic nucleus in Parkinson’s


disease: evaluation of active electrode contacts. J Neurol Neurosurg Psychiatry 2003;74(8):1036-46. 81. Guo T, Finnis KW, Parrent AG, Peters TM. Visualization and navigation system development and application for stereotactic deep-brain neurosurgeries. Comput Aided Surg 2006;11(5):231-9. 82. Chakravarty MM, Sadikot AF, Germann J, Bertrand G, Collins DL. Anatomical and electrophysiological validation of an atlas for neurosurgical planning. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8(Pt 2):394-401. 83. Bardinet E, Bhattacharjee M, Dormont D, et al. A threedimensional, histological and deformable atlas of the human basal ganglia. II. Atlas deformation strategy and evaluation on retrospective series of parkinsonian

84.

85.

86.

87.

26

patients treated by deep brain stimulation. J Neurosurg, In press. Iosifescu DV, Shenton ME, Warfield SK, et al. An automated registration algorithm for measuring MRI subcortical brain structures. NeuroImage 1997;6(1):13-25. Ganser KA, Dickhaus H, Metzner R, Wirtz CR. A deformable digital brain atlas system according to Talairach and Tournoux. Med Image Anal 2004;8(1):3-22. Collins DL, Neelin P, Peters TM, Evans AC. Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space. J Comput Assist Tomogr 1994;18(2):192-205. Hellier P, Barillot C, Corouge I, et al. Retrospective evaluation of intersubject brain registration. IEEE Trans Med Imaging 2003;22(9):1120-30.

393

33 Laitinen Stereotactic Apparatus M. I. Hariz . L. V. Laitinen

The Laitinen stereotactic system consists of the Stereoadapter and Stereoguide with auxiliaries for various radiological and surgical uses. The Stereoadapter is mainly used for stereotactic imaging with computed tomography (CT), magnetic resonance imaging (MRI), and stereotactic angiography. Together with the biopsy kit, the Stereoadapter is used for localization, biopsy, or stereotactic resection of brain lesions. It has also been used with the linear accelerator for fractionated stereotactic irradiation of brain tumors and arteriovenous malformations. The Stereoguide is used together with the Stereoadapter in functional neurosurgery such as pallidotomy, thalamotomy, and deep brain stimulation for movement disorders and pain as well as anterior capsulotomy and hypothalamotomy for psychiatric disorders.

Noninvasive Multipurpose Stereoadapter The stereoadapter was developed in 1982–1983 [1]. The original aim was to design a noninvasive, imaging-compatible, relocatable instrument for biopsy of brain tumors. It soon became evident that the Stereoadapter was accurate enough for functional neurosurgery, and since 1987 it has been used in all functional neurosurgery without ventriculography or plain radiology. The key for the high localizing accuracy is that the three reference points of the skull – i.e., the external auditory meatus and the bridge of the nose – give a good relocating stability. The second reason for the high accuracy is that the

#


imaging fiducials lie immediately on the scalp and not at a far distance from it, as is the case with other stereotactic frames. The Stereoadapter is mounted to the patient’s head by means of a nasion support, two earplugs, and a strapping band at the occiput. Neither general nor local anesthesia is needed. Repeated mountings of the Stereoadapter have shown a high degree of reproducibility and tolerability [2–4]. The Stereoadapter is made of an aluminum alloy and reinforced plastic. It consists of two lateral triangular components with four transverse bars each, a connector plate, a nasion support component, and frontal laterality indicator pins (> Figure 33-1). The transverse bars are 2 mm thick in a dorsoventral direction and lie 25 mm apart from each other. They connect the anterior and posterior ear arms and are perpendicular to the posterior ear arm. Cogwheel cases join the lateral triangle components to the nasion support arms, which have millimeter scales. By winding the cogwheel screws, the nasion support is pressed against the bridge of the nose. The earplugs lie at the posterior ear arms and are pressed against the external auditory meati by means of a threaded lever in front of the nasion support. The connector plate at the vertex joins the lateral triangle components together and serves to press the triangle components against the scalp. A frontal pin mounted between the vertex connector plate and the nasion support serves as the laterality reference structure (> Figure 33-1). When the Stereoadapter has been mounted on the head, its position is recorded by the

512

33

Laitinen stereotactic apparatus

. Figure 33-1 The Stereoadaptermounted on the head. Left: the patient’s head is lying on a plastic cushion, and the Stereoadapter is immobilized by a multijoint mechanism prior to the imaging study. The star indicates the posterior ear arm of the right triangular component. TB = transverse bar. Right: frontal view. CP = connector plate, holding the lateral triangular components pressed to the scalp

symmetrical millimeter scales on both sides of the nasion support arms and on the connector plate of the Stereoadapter. The Cartesian reference structures of the Stereoadapter are the sagittal midplane passing through the frontal pin for the laterality x coordinate, the frontal plane between the anterior borders of the right and the left posterior ear arms for the antero-posterior (AP) y coordinate, and the transverse plane between a pair of transverse bars for the dorsoventral z coordinate. With the Stereoadapter mounted to the head, the dorsal most transverse bar level corresponds to the average height level of the cingulum, the second dorsal most bars to the average height level of the intercommissural line, and the third bar pair to the average level of the amygdala. For CT, MRI, or angiography study as well as for stereotactic irradiation, the Stereoadapter is mounted to the head and then immobilized to a plastic plate with a multijoint mechanism (> Figure 33-1). For MRI studies, thin tubes containing 2 mmol copper sulfate or olive oil are attached to the reference structures of the Stereoadapter. These fiducials give an artifactfree sharp image. For single photon emission

computed tomography (SPECT) or positron emission tomography (PET) studies, the tubes mentioned above can be filled with an appropriate solution of isotope, thus providing visible reference marks on the respective pictures. For angiography, either a conventional or digital angiography, the ordinary earplugs of the Stereoadapter are replaced by similar earplugs containing a 1-cm lead pin, and, a 1-cm lead pin is placed on the forehead of the patient. These lead pins provide the magnification factor on the AP view. On the side view, the magnification is given by the already known distance of 25 mm between two sets of the transverse bars [5]. For stereotactic irradiation, plastic plates are attached to the triangular components of the Stereopadapter (> Figure 33-2). These plates have lines indicating the position of the transverse bars and posterior ear arms, respectively, and serve to align the patient’s head according to the lateral laser beams of the linear accelerator. Instead of the ordinary frontal pin, a ruler with sliding millimeter scales and a cone are used (> Figure 33-2). This device serves to align the brain target in a lateral direction according to the frontal laser beam. The laser beam must coincide


. Figure 33-2 The stereoadapter mounted to the head prior to stereotactic irradiation. Plastic plates to indicate the brain target’s y and z coordinates are attached to the lateral triangular components. The laser cross lines are aligned with the reference markings of the AP and height coordinates on the plate. The cone on the forehead is attached to a laterality ruler. Both can slide according to the laterality of the target. The couch is moved until the frontal laser line hits the tip of the cone

33

cannulas. The probe carrier is rigidly attached to the connector plate of the Stereoadapter for either an anterior (frontal) or posterior (parietooccipital) approach (> Figure 33-3). The ordinary rod of the probe carrier can be replaced by a long curved one should a lateral or posterior fossa approach be needed. The Sedan biopsy probe consists of an outer cannula 250 mm long and 2 mm thick, with a 5-mm side opening at its distal end. An inner cannula with a sharpened end functions as a guillotine when it is advanced through the outer cannula during aspiration of a brain tumor specimen. After setting the phantom base according to the CT/MRI coordinates of the target, the tip of the biopsy cannula is directed against the phantom target, after which the hinge clamps are tightened (> Figure 33-3).

Stereoguide

with the tip of the cone. Thus, the cone allows for a visual verification of the proper alignment of the target when the couch is rotated for multiplanar irradiation.

Tumor Biopsy Kit The biopsy kit consists of a phantom base, probe carrier, twist drill, diathermy probe, and biopsy cannula. The phantom base is mounted between the right and left transverse bars of the Stereoadapter at a desired height level (> Figure 33-3). It has two slide components with millimeter scales for the x and the y coordinates, respectively. The slide component for the y coordinate has a millimeter-scaled vertical rod for the z coordinate. The probe carrier has two steel rods, two hinge clamps, and two concentric probe-guiding

The Laitinen Stereoguide is a stereotactic frame that functions according to the arc-radius principle [6]. It consists of an oval base ring fixed to the head with four steel pins (> Figure 33-4). Cylindrical components with millimeter scales are mounted on the lateral sides of the base ring. Cogwheel mechanisms permit the cylinder components to slide in such a way that their common axis coincides with the y and z positions of the intracranial target. A vernier scale ensures an accuracy of 0.25 mm. A semicircular arc carrying the electrode, endoscope, or other probe is mounted to the cylindrical components. It can slide in a lateral direction to bring the probe tip to the lateral x position of the target (> Figure 33-4). Thus the target lies at the center of the spherical system of the Stereoguide and can be reached from any suitable direction. If needed, the base ring of the Stereoguide, mounted on the patient, may be attached rigidly to a floor stand for surgery. Then, if intraoperative

513

514

33


. Figure 33-3 The biopsy procedure. Left: the biopsy kit mounted to the Stereoadapter. The phantom base is mounted at the level of the second dorsal most transverse bars and shows the x, y, and z positions of the brain target. The probe carrier is mounted to the connector plate of the Stereoadapter for a parietal approach. The biopsy needle points to the phantom target and the hinge clamps are tightened. Right: the Stereoadapter mounted to a dummy. The probe carrier is attached and the guiding cannula is directed toward the intracranial target

radiology is needed, the central beams of suitably adjusted lateral and AP x-ray tubes will pass through the y, z, and x origins of the frame, respectively [7].

Stereotactic Radiological and Surgical Applications Nonfunctional Stereotaxis The stereotactic management of a brain tumor, abscess, cyst, deep brain hematoma, etc., begins by mounting the Stereoadapter to the head and performing a stereotactic CT (or MRI) study [8–10]. The scanning is performed throughout the tumor area with 2- or 3-mmthick contiguous slices parallel to the transverse bars of the Stereoadapter (> Figure 33-5). The Stereoadapter is then detached from the head. The calculation of target coordinates may be done manually or using the software of the CT

or MRI machine. The y coordinate of the target is its distance from the interaural plane, the x coordinate is the distance from the medial border of the right posterior ear arm, and the z is the distance from the CT scan containing the target point and to that showing the nearest pair of transverse bars (> Figure 33-5) [8]. Surgery may take place at any suitable time after the CT/MRI study. The phantom base and probe carrier are set according to the CT/MRI target coordinates, after which the phantom base is detached (> Figure 33-3). The Stereoadapter with the probe carrier locked to it is remounted to the head. The inner probe-guiding cannula is replaced by a similar sterile one. Using local anesthesia, the skull is trephined with a 2.15mm-thick twist drill introduced through the guiding cannula. Diathermy coagulation is applied to the dura and cortex. The biopsy needle is then introduced to the target. Tumor specimens are obtained from various depths along the track of the needle. The procedure usually


. Figure 33-4 The Stereoadapter and Stereoguide mounted to a patient during a functional stereotactic procedure. A. The base ring of the Stereoguide is fixed to the head with screws. A steel pin introduced through the cylindrical component points at the y and z origins of the Stereoadapter (arrow) during setting of the surgical y and z coordinates. B. The probe carrier arc is attached to the cylindrical components. The electrode points to the frontal pin of the Stereoadapter (arrow) during setting of the surgical x coordinate (see text for details)

33

. Figure 33-5 A 2-mm-thick stereotactic CT scan at the level of the second dorsal most transverse bars of the Stereoadapter. The dotted line xb indicates the laterality x of the biopsy target, measured from the right transverse bar (see text). The AP coordinate of the target is indicated by y. The target’s laterality measured from the sagittal midplane of the Stereoadapter is indicated by x

the Stereoadapter can be drawn on the scalp using the same technique. The Stereoadapter is then removed and a small centered craniotomy is done [9]. takes 25–35 min [8]. The same instrumentation and technique can be used for the stereotactic placement of a drainage catheter in cases of cyst or abscess. Similarly, for a deep-seated tumor scheduled for stereotactic craniotomy and resection, a guiding catheter can be placed stereotactically at the edge of the tumor, after which the Stereoadapter is detached. The catheter is cut along the surface of the skin, a small centered flap is created, and a craniotomy is done. The catheter is then followed toward the tumor using small spatulas and routine microsurgical techniques. For resection of small superficial brain tumors, the location of the tumor in relation to

Fractionated Stereotactic Irradiation When a brain lesion is scheduled for stereotactic irradiation, the stereotactic CT scanning ought to include not only the area of the brain pathology but also the whole calvarium. With an arteriovenous malformation (AVM), stereotactic angiography with the Stereoadapter is done. The target coordinates are indicated on special side plates attached to the triangular components of the Stereoadapter. The lateral x coordinate is measured on the CT scan in relation to the midsagittal plane of the Stereoadapter (> Figure 33-5).

515

516

33


The Stereoadapter is remounted to the head with the patient lying on the couch of the linear accelerator (> Figure 33-2). The couch is moved so that the two laser cross lines from each side and the vertical laser beam from the ceiling indicate that the isocenter of the accelerator coincides with the Stereoadapter markings of the brain target – that is, the markings on the side plates and the frontal cone (> Figure 33-2). The Stereoadapter is then locked to the couch with a multijoint mechanism similar to that used for CT and MRI. During irradiation, the patient is monitored using a video camera. Each irradiation session, comprising five or six multiplanar fixed beams with variable collimation, lasts for about 30 min. The procedure may be repeated according to the schedule of fractionation [5,11,12].

Basically, MRI and CT scanning with the Stereoadapter are very similar. The positioning of the patient into the coil is, however, not as crucial as positioning into the CT gantry as far as the right-left alignment is concerned. The sagittal survey image of the triangular components should be obtained first. On this image, the tubes filled with copper sulfate or olive oil, indicating the references for the AP and height coordinates, are visualized (> Figure 33-6). The axial MRI scanning should also be done parallel to the transverse bars, and coronal scanning should be parallel to the posterior ear arms. In our experience, a stereotactic MRI study with axial thin slices takes about 15 min.

Functional Stereotaxis

Enlarged film copies of the CT/MRI scans are obtained from the area between the foramina of Monro (FM) and the proximal aqueduct, including the second pair of transverse bars. On the CT scan, the anterior commissure (AC) is localized according to the method of Laitinen and coworkers [1] on a slice lying 4 mm ventral to the ventral most margin of the foramina of Monro. If the scanning plane is parallel to the intercommissural line (ICL) – i.e., the transverse bars of the Stereoadapter are parallel to the ICL – the posterior commissure (PC) is seen on the same slice. If the scanning plane is not parallel to the ICL, adjacent CT/MRI slices of the area are studied in order to visualize the beginning of the aqueduct; the last slice before the appearance of the aqueduct is chosen to represent the level of the PC. This film copy is superimposed on that where the AC had been marked, and the position of the PC is transferred to the latter. Thus, the level of the ICL is determined in relation to the scanning level [3]. The mean angulation between the transverse bars and intercommissural line is 0.75 with a range of 7 [7].

The CT and MRI Studies After mounting the Stereoadapter, the head, resting on a plastic cushion, is aligned so that the connector plate of the Stereoadapter is parallel to the transverse laser beam of the gantry, after which the Stereoadapter is locked to a plastic plate on the CT table and a lateral Scoutview of the head is obtained. The cursor line is brought to the level of and parallel to the dorsal most transverse bars of the Stereoadapter, so that the scanning plane is parallel to the transverse bars and perpendicular to the posterior ear arms of the Stereoadapter. The visualization of the transverse bars is checked on the first CT slice to assure parallel alignment. Beginning from the level of the dorsal most transverse bars, 1.5- or 2-mm thick slices are scanned in 2-mm steps until the second transverse bars and the proximal part of the aqueduct are visualized, following which the Stereoadapter is detached. The CT study lasts for 10–15 min.

Determination of the Ventricular Landmarks and Target Coordinates


33

. Figure 33-6 CT and MRI scans of one patient performed on consecutive days. Both scans are 2 mm thick at a dorsoventral level 4 mm ventral to that of the anterior commissure-posterior commissure line and at the level of the second dorsal most transverse bars of the Stereoadapter. The left posteroventral pallidal target is indicated by an encircled dot

The anatomic position of the brain target – be it the ventrolateral thalamus, the posteroventral pallidum, or the anterior internal capsule, etc. – can now be plotted on the appropriate CT/MRI slice in relation to the AC and PC. Then, the coordinates of the target point are measured in relation to the reference structures of the Stereoadapter (> Figure 33-6). The y coordinate of the target is its distance from the interaural plane of the Stereo-adapter. The x coordinate is measured in relation to the sagittal midplane of the Stereoadapter, formed by projecting the frontal laterality indicator pin perpendicularly onto the interaural line. The z coordinate is the distance between the target level and the level of the second pair of transverse bars. In routine procedures, calculation of the functional brain target coordinates on either CT or MRI scans lasts for 5–10 min.

Surgery The surgery may be performed at convenience after the CT/MRI study. The Stereoadapter is

remounted on the head. The base ring of Laitinen’s Stereoguide is mounted around the Stereoadapter fairly parallel to its transverse bars. By means of two adjustable lateral support components, the base ring is so positioned that it lies as symmetrically as possible in relation to the Stereoadapter. Under local anesthesia, the base ring is rigidly fixed to the skull by means of four percutaneous pins. The patient is placed on the surgical table. The cylinder components of the Stereoguide are mounted on the left and the right sides of the base ring. Plastic cylinder blocks with axial steel pins are introduced through the cylinders, which are moved into such a position that the steel pins point to the y and z origins of the Stereoadapter – i.e., to the intersection of the second transverse bar and the anterior margin of the posterior ear arm. In this way, the y and z origins of the Stereoadapter are transferred to the cylinder components, and recorded on the corresponding millimeter scales of the Stereoguide (> Figure 33-4). The CT coordinates y and z of the surgical target are then added to the y and z readings of the Stereoadapter’s origin, after which the cylinder components of the

517

518

33


Stereoguide are positioned according to the sum. Thus the steel pins of the cylinder blocks point to the brain target, the y and the z coordinates of which are read on the millimeter-scales of the Stereoguide. The semicircular arc of the Stereoguide is mounted to the cylinder components. The electrode carrier on the arc is moved into a 90 position. The surgical probe is directed toward the frontal pin (> Figure 33-4). The position of the pin is recorded on the lateral millimeter scale of the cylinder components. Then the CT x value of the target is added to the recorded position of the frontal pin; the sum is the final stereotactic x coordinate. The arc is moved into this position. The Stereoadapter is then detached, and the surgery may proceed as usual [3].

Remote Postoperative Imaging Studies The noninvasive Stereoadapter permits a stereotactic imaging study to be performed months

after surgery for control of the final size and site of the stereotactic lesion [13]. This unique feature is of outmost importance, since a remote postoperative imaging study may show the final shape of the lesion after the complete resolution of postoperative edema (> Figure 33-7). The stereotactic imaging study is performed in a manner similar to the preoperative study. In this way the lesion can be accurately assessed in relation to the preoperative target point and in relation to the reference structures of the third ventricle (> Figure 33-7). Since the Stereoadapter’s position on the head will be the same postoperatively as it was pre- and intraoperatively, an exact radiological correlation between pre- and postoperative scanning can be made [13]. In cases where a permanent electrode for chronic electrical stimulation had been implanted in the brain, a plain x-ray performed in a stereotactic manner with the Stereoadapter remounted to the head provides a stereotactic control of the exact position and coordinates of the electrode tip at any time after the surgery.

. Figure 33-7 Preoperative and remote postoperative stereotactic CT scans of one patient. On the left scan, which is 2 mm thick, the preoperative pallidal target is indicated by a dot (arrow). The right scan, performed 4 months after surgery, represents the superposition of two contiguous, 2-mm-thick CT slices of the same area. The pallidal radiofrequency lesion is thus enhanced


Conclusions The Laitinen system is based on stereotactic imaging using the noninvasive Stereoadapter. The Stereoadapter is not individualized and fits most heads [2]. The reference structures of the Stereoadapter lie extremely close to the head and therefore to the target, which is a unique feature of this system. Calculation of target coordinates can be done easily and quickly by using the inherent software of the CT/MRI machine or manually with a ruler, an ink pen, and a mini calculator. The noninvasive design permits flexible and rational planning of different diagnostic and therapeutic stereotactic procedures. The technique obviates the need for surgery inside the CT machine. Patients can be operated on or irradiated when suitable for them, the surgeon, the radiotherapists, and the involved staffs. For brain biopsy, there is no need for an additional frame, since the Stereoadapter as such also functions as a probe carrier. This simplifies the procedure and markedly reduces the duration and costs of surgery. For stereotactic open resection of small brain tumors, the Stereoadapter does not interfere with the craniotomy, since it is removed from the head once the position of the tumor has been indicated on the scalp or by a catheter. The reproducibility of results from the noninvasive Stereoadapter permits an accurate repositioning of the brain target into the isocenter of the linear accelerator for fractionation of stereotactic irradiation. Furthermore, the use of the Stereoadapter for target localization can be combined with a neurosurgical navigation system by providing relocatable Cartesian references [14]. A great advantage to the patient in functional stereotaxy is the avoidance of ventriculography [15,16]. The accuracy of the method permits a functional stereotactic procedure to be carried on with minimal side effects and short hospital stays [16,17]. The noninvasive Stereoadapter makes possible a remote post-operative stereotactic

33

CT study for checking the site and the size of the final radiofrequency lesions and also for assessing the accuracy of the whole stereotactic procedure [2,3,13]. It is important to keep in mind that the noninvasive fixation of the Stereoadapter to the head requires good cooperation of the patient unless sedation is used [2]. The pressure exerted by the earplugs on the external auditory meatus and by the nasion support on the bridge of the nose, although generally well tolerated, may be uncomfortable to some patients. Besides, the noninvasivity calls for great care on the part of the surgeon in mounting the frame, positioning the patient, and closely supervising the scanning or irradiation procedure. Through 10 years of intensive use on more than 1,000 patients for all stereotactic imaging, surgical, and radiotherapeutic procedures, it is felt that the Laitinen system has been versatile, reliable, easy to use, time-saving, and inexpensive. The lack of sophistication and simplicity of the system have been experienced as an advantage rather than an inconvenience. However, as in any system, the surgeon should be well acquainted with it and learn to profit from its advantages while avoiding its pitfalls.

References 1. Laitinen LV, Liliequist B, Fagerlund M, Eriksson AT. An adapter for computed tomography-guided stereotaxis. Surg Neurol 1985;23:559-66. 2. Hariz MI. A non-invasive adaptation system for computed tomography-guided stereotactic neurosurgery. Thesis, Umea˚ University Medical Dissertations, New series no 269, ISSN 0346–6612. Umea˚, Sweden: Umea˚ University Printing Office; 1990. 3. Hariz MI. Clinical study on the accuracy of the Laitinen’s non-invasive CT-guidance system in functional stereotaxis. Stereotact Funct Neurosurg 1991;56:109-28. 4. Hariz MI, Eriksson AT. Reproducibility of repeated mountings of a noninvasive CT/MRI stereoadapter. Appl Neurophysiol 1986;49:336-47. 5. Bergenheim AT, Hariz MI, Henriksson R, Lo¨froth P-O. Fractionated stereotactic irradiation of brain tumors and

519

520

33 6.

7.

8.

9.

10.

11.


arteriovenous malformations using the linear accelerator and a non-invasive frame. In: Lunsford LD, editor. Stereotactic radiosurgery update. Elsevier; New York: p. 73-5. Laitinen LV. The Laitinen system. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1998. p. 99-116. Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computed tomography-guided determinations of brain targets in functional stereotaxis. J Neurosurg 1990;73:565-71. Hariz MI, Bergenheim AT, DeSalles AAF, et al. Percutaneous stereotactic brain tumor biopsy and cyst aspiration using a non-invasive frame. Br J Neurosurg 1990;4:397-406. Hariz MI, Fodstad H. Stereotactic localization of small subcortical brain tumors for open surgery. Surg Neurol 1987;25:345-50. Nguyen J-P, Decq P, Brugie`res P, et al. A technique for stereotactic aspiration of deep intracerebral hematomas under computed tomographic control using a new device. Neurosurgery 1992;31:330-5. Delannes M, Daly NJ, Bonnet J, et al. Fractionated radiotherapy of small inoperable lesions of the brain

12.

13.

14. 15.

16.

17.

using a non-invasive stereotactic frame. Int J Radiat Oncol Biol Phys 1991;21:749-55. Hariz MI, Henriksson R, Lo¨froth P-O, et al. A noninvasive method for fractionated stereotactic irradiation of brain tumors with linear accelerator. Radiother Oncol 1990;17:57-72. Hariz MI. Correlation between clinical outcome and size and site of the lesion in CT-guided thalamotomy and pallidotomy. Stereotact Funct Neurosurg 1990;54:172-85. Takizawa T. Neurosurgical navigation using a noninvasive stereoadapter. Surg Neurol 1993;40:299-305. Hariz MI, Bergenheim AT. Clinical evaluation of CT-guided versus ventriculography-guided thalamotomy for movement disorders. Acta Neurochir Suppl 1993;58:53-5. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. Hariz MI, Bergenheim AT, Fodstad H. Airventriculography provokes an anterior displacement of the third ventricle during functional stereotactic surgery. Acta Neurochir 1993;123:147-52.

30 Leksell Stereotactic Apparatus L. D. Lunsford . D. Kondziolka . D. Leksell

Background Any description of the Leksell stereotactic system must begin with a historical vignette that describes the genesis of its creation. Lars Leksell was a brilliant, innovative, and persistent pioneer in the emerging field of neurological surgery. His creative genius eventually led him to become professor of neurological surgery at the Karolinska Institute in Stockholm, Sweden. His career, however, began under the direction of Ragnar Granit, a Nobel Prize-winning neurophysiologist with whom Leksell collaborated. Leksell’s doctoral thesis presented the first description of the spinal cord gamma motor neuron system. Following this work, Leksell’s neurosurgical training began in the department of Herbert Olivecrona. At this time, in the early 1940s, the alarming mortality rate for routine neurosurgical procedures approximated 40%. Poor outcomes, difficulty with anesthesia (neurosurgical patients were allowed to ventilate spontaneously) and severe blood loss during surgery were features that made a lasting impression on Leksell. He was struck by the paradox of crude instrumentation that was poorly designed for the delicate central nervous system. His subsequent career was committed to the development of less invasive surgical techniques that facilitated management of a wide variety of intracranial problems. During the subsequent four decades, he became one of the most original contributors to the field of neurological surgery. In 1947 Leksell studied in Philadelphia with Spiegel and Wycis, who developed the first human stereotactic apparatus [1]. When he returned to Stockholm a year later, he developed his

#


first-generation stereotactic guiding device, which was ‘‘easy to handle and practical in routine clinical work’’ (> Figure 30-1). His first device was reported in 1949 [2]. Ease of use and practicality were concepts that remained preeminent principles of all subsequent Leksell systems as they evolved. The designs have focused on utility, accuracy, and versatility. The continued development of stereotactic instruments responded to the challenge of new surgical needs presented by deep brain surgery, diagnostic biopsy, radiosurgery, and functional neurosurgery. In addition, constantly improving neurodiagnostic imaging tools such as ultrasound, encephalography, angiography, computed tomography (CT), positron emission tomography (PET), magnetic resonance imaging (MRI), magnetic resonance angiography (MRA) [3], and magnetic source imaging required repeated revisions in the system design.

The Arc-Centered Principle Lars Leksell’s original description is itself an eloquent and simple analysis of the arc centered concept (> Figure 30-2): ‘‘Essentially, it consists of a semicircular arc with a movable electrode (probe) carrier. The arc is fixed to the patient’s head in such a manner that its center corresponds with a selected cerebral target. The electrodes (probes) are always directed towards the center and hence to the target. Rotation of the arc around the axis rods in association with lateral adjustment of the electrode (probe) carrier enables any convenient point of entrance of the

470

30

Leksell stereotactic apparatus

. Figure 30-1 The original Leksell stereotactic coordinate frame first reported by Lars Leksell in ‘‘The principle of the semicircular arc is identified in this first model stereotactic instrument’’

. Figure 30-2 The Leksell arc-centered design comprises a movable instrument carrier that slides on a semicircular arc. The carrier can be swung along the left-right ‘‘arc’’ angle and in an anterior-posterior ‘‘ring’’ angle

electrodes (probes) to be chosen independent of the site of the target [4].’’ First, an imaging-compatible base ring is attached by pins to the patient’s head. Next, the base ring is coupled with a fiducial system to enable target localization using a wide variety of radiological techniques. The frame has been modified

steadily over the past 50 years in order to be compatible with each advance in neurodiagnostic imaging. Versions of the frame over the years (currently model G) have included improvements in instrument carriers and coordinate fixation attachments as well as the use of non-ferromagnetic materials enabling use with MRI, PET, and CT.


As of early 2008 the Leksell stereotactic system was used at more than 1,300 neurosurgical centers in more than 50 countries worldwide. A semicircular arc is attached to the base ring of the frame such that the arc can be swung around its rings in the anteroposterior direction, or left and right if mounted on the anterior and posterior parts of the frame base. The instrument carrier slides along the arc. The radius of the arc is 190 mm. In essence, an intracranial target can be approached from any entry point. The most important principle is to select a safe trajectory to reach the defined target. Within the Leksell stereotactic family of instruments, the only exception to this is the Leksell Gamma Knife, first used by Leksell in 1967 [4]. Because the locations of the multiple cobalt-60 sources are fixed, the patient is moved into the point of beam intersection at the center of the collimation system. This position is determined by the stereotactic X, Y, and Z coordinates of the target. The same base ring is used for open image-guided surgery, functional neurosurgical procedures and stereotactic radiosurgical operations.

The Model G Instrument The currently available Leksell stereotactic instrument can be used interchangeably with all currently available imaging techniques. The frame consists of a rectangular base ring 190 210 mm (> Figure 30-3). The X (left-right) coordinate is set on the semicircular arc, which is attached to the frame at the chosen Y (anteroposterior) and Z (superoinferior) coordinates. Either a straight front piece or a curved front piece that accommodates the nose is used (> Figure 30-4). The X, Y, and Z axes of the coordinate system conforms to the X, Y, and Z geometry of CT, MRI, and PET scans. An imaginary frame origin (X, Y, and Z = 0) is in the upper posterior right side of the coordinate frame (> Figure 30-5). A fiducial system is attached to the base ring during the

30

. Figure 30-3 The current base ring for the model G Leksell stereotactic instrument

. Figure 30-4 The base ring has an front plate (upper) to which the semicircular arc can be attached in an antero-posterior position. Usually the curved front piece is used and the semicircular arc attached to the right and left sides

471

472

30


. Figure 30-5 The coordinate system of the Leksell model G stereotactic base ring. At the center of the frame the X, Y, and Z coordinates are 100. An imaginary ‘‘zero’’ exists superior, posterior and on the right side of the fiducial system

. Figure 30-6 The patient is positioned for CT localization. The fiducial system can be seen attached to the base ring

. Figure 30-7 The MRI fiducial system with frontal, posterior, lateral, and superior plates for determination of the X, Y, and X coordinates in multiplanar MRI imaging. The fiducial consists of a plastic tube filled with diluted copper sulfate solution

neurodiagnostic imaging component of the procedure (> Figure 30-6). A variety of fiducial systems are available for different imaging modalities, including systems for multiplanar MRI studies (> Figure 30-7). A radio opaque scale is available for conventional radiographic examination during the now rarely performed contrast encephalography for functional procedures (> Figure 30-8). A fiducial system compatible with computer-derived coordinate techniques is also available for digital subtraction angiography, which is often used during Gamma Knife radiosurgical procedures for arteriovenous malformations (> Figure 30-9). A variety of surgical instruments can be attached to the semicircular arc (> Figure 30-10). The instrument carrier can be moved to any arc angle and the arc rotates to any ring angle, both of which can be determined using a computer simulation technique. The working length of all probes is 190 mm, which corresponds to the radius of the semicircular arc. When the probe stop on


the instrument carrier of the arc is set at zero, the operative end of the probe will reach the target with an accuracy of 0.7 mm. The accuracy of reproducibility and precision of stereotactic . Figure 30-8 Angiographic localizers for computer based determination of the coordinates (left) or conventional coordinate determination (right) with a radio opaque system that is attached to the base ring

30

targeting using the Leksell system meets the guidelines established by the American Society for Testing and Manufacturing. Ultimately, accuracy is dependent on the imaging resolution, which is 0.5 mm for 512 512 matrices used during digital imaging acquisitions such as CT.

Application of the Coordinate Frame The stereotactic coordinate base ring is attached to the patient’s head using four pins (either aluminum screws with hard metal tips, titanium screws or disposable aluminum screws of various lengths (> Figure 30-11). The location of the pins is adjusted by changing the position of the posts on the base ring. (> Figure 30-12). The position of the base ring on the patient’s head is adjusted using ear bars that are temporarily placed in the external auditory canal. Although symmetrical placement is not critical,

. Figure 30-9 Antero-posterior and lateral x-rays visualize the coordinate x-ray indicators on the x-ray film. The X, Y, and Z coordinates are determined by simple calculation or a graphic method

473

474

30


. Figure 30-10 The Leksell model G expanded stereotactic arc can be attached to the base ring in the lateral or anteroposterior direction

. Figure 30-11 Application of the stereotactic frame requires that pins be inserted through the skull and into the outer table of the skull under local anesthesia. After application, gently pulling on the frame ensures that the frame and head move together

these ear bars allow for appropriate symmetrical placement of the frame on the patient’s head. Cushioning the ear bars with a small amount of foam in the external auditory canal helps to reduce discomfort from frame torsion during pin placement. Surgeons should view the preoperative imaging in order to make decisions relative to frame placement. Although appropriate shifting of the frame relative to the patient’s head is of little consequence for most open stereotactic or functional neurosurgical procedures, it is very

. Figure 30-12 The position of the pins is adjusted by changing the position of the posts on the base ring

important for procedures done with earlier Gamma Knife models. During such procedures, the surgeon attempts to place the lesion as close as possible to the center of the frame (X, Y, and Z = 100). Such frame application issues are not as relevant for procedures performed with the latest generation Gamma Knife Perfexion, since there is a greatly expanded working space with little or no chance of frame or pin interference with the collimating system. In the vast majority of patients, the stereotactic head frame is applied within approximately 5 min. Most patients benefit from mild sedation (either oral or intravenous) in order to reduce mild discomfort and anxiety. Some surgeons mix the local anesthetic with sodium bicarbonate to reduce the initial burning sensation associated with injection of the mixture of local anesthetic agents. For those patients who express significant anxiety (especially young men who seem to have a propensity for vasovagal symptoms), both additional sedation and an intravenous anticholinergic medication may be beneficial. We routinely prep the entire head only with isopropyl alcohol. No hair shave is performed.


The base ring and all pins are sterilized in advance of placement. Screw-type pins reduce frame application time to a matter of minutes. The pins can be used satisfactorily in children over the age of two without difficulty but should be used carefully in patients under that age, since the skull may still be compressible. We have used blood Vacutainer stoppers against the scalp (to diffuse pressure from the pin) in children as young as 5 months. In addition, the pins must be inserted in such a way that excessive distraction of the support bars does not occur. We tighten the pins diagonally, two and two at the same time, alternating between the pairs. No torque instrument is necessary. Once the pins are firmly secured, subsequent movement of the head will not result in frame dislodgment. Proper application can be assured by gently lifting the frame base ring and making sure that the patient’s head and the frame move together. In general, we apply the frame with the patient on a stretcher in the semi-sitting position with the legs slightly flexed at the hips, which enhances comfort and allows immediate transfer to the imaging site.

Conventional Radiographic Target Localization The x-ray-compatible indicator box is attached to the base ring. Anterior-posterior, and lateral images during planar radiography, encephalography, or conventional, single, and bi-plane digital subtraction angiography will result in recognition of proximal and distal X, Y and Z coordinates. Leksell’s original ingenious construction of a geometric spiral diagram to determine the target coordinates was never met with great acceptance. It may even have helped to perpetuate the myth that stereotactic surgery was excessively complex. The planar film localization technique has been replaced by scaled graphs, tabletop stereotactic calculators or computer based planning systems

30

(Leksell SurgiPlan for open Stereotactic procedures and Leksell GammaPlan for Gamma Knife surgery). It is critical to understand the position of the x-ray source referable to the frame so that the surgeon can distinguish between proximal and distal frame values. Magnification is irrelevant, and any distance between x-ray source, frame, and film can be chosen, provided that all fiducials are visualized on the resultant images. Selection of the central beam at an X, Y, and Z of 100 helps to facilitate easy visualization of the frame fiducials. Rotation is not a problem as long as all fiducials are recognized in the films. Film subtraction techniques are especially valuable during angiography.

Computed Tomography and Magnetic Resonance Imaging The Leksell system was one of the first instruments to be compatible with the development of computed tomography. In 1977, Lars Leksell together with Bengt Jernberg and Hans Sundquist, his engineering colleagues, began to redesign the frame to make it CT-compatible [5]. This required reduction in the amount of metal within the frame then in use (Model B). They developed imaging-compatible pins (plastic or carbon fiber) and an appropriate fiducial system that would allow conversion of CT scanner pixel information into accurate twodimensional stereotactic imaging. The model D was the first MRI-compatible frame [6,7]. In the current model G instrumentation, a CT or MRI-compatible fiducial system is anchored to the base ring using snap-on clips. Four slots in the base of the fiducial system prevent incorrect positioning of the fiducial system. Lateral side plates are critical and an anterior or posterior plate may be used to provide stability during the initial CT or MRI. During the imaging examination, the responsible surgeon takes time to confirm that the frame is not significantly rotated

475

476

30


referable to the scanner plane. All scans are performed parallel (or rarely perpendicular) to the base ring, as can be seen from the initial CT scout film or the MRI localizer scan. Because of the rapid acquisition time of current CT 64 slice scanners (entire head with 2 mm slices in 4 s), the scan time is very brief. Similarly for MRI, a volume acquisition of the head at 1.5–2 mm slices can be performed in 8–10 min depending on the number of excitations. Most imaging can be performed with a 25-cm field of view in order to increase target resolution. The frame-center (X, Y = 100) central pixel is defined and the target coordinates are determined referable to the frame center. Most brain lesions undergoing stereotactic surgery can be visualized with 2 mm-thick images. Narrower images (e.g., 1 mm) can be used but, while they improve spatial resolution, they do so at the expense of contrast resolution. Enhancement of the brain or the target by intravenous contrast sometimes helps gray-white differentiation and facilitates targeting of contrast enhancing lesions. In essence, CT is very satisfactory for lesions that can be seen well on preoperative CTs, while MRI is the preferable imaging tool for lesions that are best identified by MRI scan and for imaging during functional neurosurgery. Determination of the superoinferior stereotactic plane (Z coordinate) is done by the measuring of the fiducials and requires simple distance measurement functions available on the computer. The Leksell system is very stable in the field of 1.5 Tesla imaging, so that in most cases additional CT imaging is not necessary. If both MR and CT imaging are used, fusion of the two data sets is easy in the current planning systems. Direct calculation of the target can be performed on routinely available imaging software at the imaging site, using measure distance functions that allow determination of the target coordinates referable to the frame center. Alternatively, the images can be transferred by CD or Ethernet to the planning system [8].

SurgiPlan permits rapid calculation of stereotactic targets; however, in general, the scanner software itself can be used to accurately and rapidly make target coordinate determinations during the actual image acquisition interval. As noted modern-generation CT scanners (spiral scanners and high-resolution head techniques) often allow CT scan data acquisition in a matter of seconds. When using CT, effort should be made to place the pins outside of the imaging area of interest in order to reduce artifacts on the images. Magnetic resonance imaging stereotactic localization requires a consistent commitment to quality assurance, daily evaluation of image accuracy, and recognition of potential sources for distortion [3,9–11]. This modality has proven to be an excellent and reliable neurodiagnostic imaging tool for stereotactic surgery with the Leksell system. MRI-guided stereotactic surgery has several clear advantages, including significant reduction in image artifacts (since the pins and frame give no MR signal). In addition, the ability to do multiplanar imaging, to verify target coordinates between both coronal and axial locations, and to visualize certain targets by variations in imaging techniques related to T1 and T2 signal or contrast agents, clearly establish MRI as superior to CT in most instances. During CT imaging, the head frame is fixed in a stereotactic adapter compatible with the various commercially available scanners [12]. The patient must be leveled in advance and must be screened to make sure that there are no significant imaging artifacts, which would reduce image quality. One of the advantages of the Leksell MRI compatible stereotactic system is that the fiducial system for coordinate calculation is close to the patient’s head and to the center of the magnetic field. As both magnetic susceptibility and other distortions appear greater at the periphery of the field, other systems with fiducials located far from the center of the field may be associated with a greater risk of distortion. Even with axial


imaging, warping of the fiducials can be seen when the target is very close to the base ring. Keeping the base ring low is generally advisable. The magnet itself must be properly maintained and checked. Residual distortions of the field left by hairpins, surgical clips, and so on can result in image degradation. Both scientific investigations and accumulated clinical experience with hundreds of thousands of patients worldwide have validated the role of MRI stereotactic surgery. Multiplanar direct and reformatted imaging facilitates simulation of probe pathways in advance of the procedure itself. These virtual reality techniques have been incorporated into advanced surgical planning systems.

Versatility of the System Since the semicircular arc in general describes a sphere where the surgical target is at the center, any probe approach, regardless of entry point, will reach the target. The main principle is that the trajectory should be safe and should minimize the number of pial, ependymal, and critical brain structures that the probe traverses. Convexity, vertex, full lateral, suboccipital, and transsphenoidal routes are readily possible using the system. With the model G system, the expanded Leksell stereotactic arc can be placed in both the conventional left-right and in the antero-posterior positions. The arc can be reversed for supraorbital approaches so that the coordinates as well as the arc and ring angles can be read easily. Patients undergoing posterior fossa biopsy procedures usually do so in the semi-sitting position. We normally recommend a transcerebellar approach to intra-axial lesions at the level of the middle cerebellar peduncle or transcortical coronal approaches to midline brainstem targets [13]. Twist drill craniotomy is used almost exclusively for most functional, diagnostic, and therapeutic stereotactic surgery. Such an approach embodies the concept of minimal invasiveness [14–24].

30

Accessories A tremendously diverse variety of instruments are available for diagnostic, therapeutic, and functional brain surgery. In addition to the twist-drill craniotomy set (> Figure 30-13), electrodes of various sizes and shapes [25], microelectrodes [26], and a wide variety of lesion-generating devices are available [27]. For therapeutic and diagnostic surgery, the Backlund biopsy set (> Figure 30-14), a forceps system and a Sedan-type needle aspiration biopsy system are available. It is important that the user review and understand each instrument in advance. The distance from the stop to the operative end of the probe, i.e., the working length, is always 190 mm. However, most biopsy probes extend some distance beyond the probe operative end and this must be taken in consideration when deciding on probe placement. Percutaneous small needles are available for craniopharyngioma aspiration or puncture [15– 17,28–30]. The Backlund hematoma evacuation kit can be used for percutaneous evacuation of deep intracerebral hematomas (> Figure 30-15). Special probes for aspiration of colloid cysts and brain abscesses are available [20,31]. Imaging . Figure 30-13 The percutaneous twist-drill craniotomy set for the majority of conventional open stereotactic procedures. Burr holes are rarely used

477

478

30


. Figure 30-14 The Backlund biopsy set includes both cyst puncture needles and the ‘‘spiral’’ biopsy system

. Figure 30-15 The Backlund hematoma evacuator includes an ‘‘Archimedes screw’’ to remove deep seated intracerebral hematomas

compatible probe holders are available and permit intraoperative imaging with a probe at the target position [32]. Accessories to the operating room table lock the frame to the standard Mayfield adapter.

Stereotactic Microsurgery The availability of new diagnostic imaging tools has promoted less invasive strategies to

reach both deep and superficial targets of the brain. The issues relative to localization of the target within the brain have largely been resolved by preoperative planning. The development of refined tools with reduced risk to patients has been exemplified by the combination of stereotactic technology and microsurgery. Deep targets in the brain can be reached through limited-exposure craniotomies. Microsurgical or endoscopic evacuation of colloid cysts, excision of cavernous malformations, and removal of brain tumors are among the examples of the usage of current stereotactic techniques. At the University of Pittsburgh, more than 2,500 brain operations have been done using intraoperative CT guidance. Currently we use a General Electric 64 slice CT scanner that is located in the operating room suite itself. Imaging is done immediately prior to the surgical intervention, occasionally during surgery itself, and always afterward in order to detect potential complications [33–36]. Since the risk for complications in most stereotactic procedures is less than 1%, the concept of microsurgical intervention coupled with highresolution imaging has opened the doors to a wide variety of successful surgical techniques. > Table 30-1 illustrates the experience using the Leksell Stereotactic system during a 28 year interval at the University of Pittsburgh Medical Center. > Table 30-2 shows the observed complications after frame based stereotactic procedures during the same time interval. During the past 15 years we have not had a patient with intraoperative bleeding requiring evacuation. We have also not seen a scalp or bone infection in the 15 year interval since we switched to using twist drill craniotomies. > Table 30-3 is a comparison between frame based and frameless image-guided biopsy techniques. Various other accessories are available for the Leksell stereotactic system. These include retractors, small trephine craniotomy systems, a laser guide (> Figure 30-16), endoscopic adapters

30


and probe holders for catheter arrays. These types of instruments have largely supported the concept that present and future neurosurgery will be almost exclusively image-guided. . Table 30-1 Image-Guided Frame-Based Stereotactic Procedures (UPMC-Presbyterian, 1979–2007) Diagnostic

No. of procedures

Brain tumor biopsy Infection Degenerative disease Hematoma Stroke Subtotal Therapeutic Cyst aspiration Abscess drainage Isotope implantation Movement disorder Lesion DBS* Pain DBS Epilepsy depth electrodes Miscellaneous Subtotal Total

1,584 58 6 12 4 1,664 197 97 145 219 148 24 95 62 987 2,651

The Leksell Gamma Knife The Leksell Gamma Knife was developed through the fertile collaboration between Borje Larson, a physicist and radiation biologist, and Lars Leksell in the 1960s. First working with cross-firing of proton beams, Larson and Leksell eventually decided to create a dedicated radiosurgical instrument which could be used with ease in the standard hospital environment. The initial 179 cobalt-60 source prototype Gamma Knife was used primarily to create radio necrotic lesions within the deep gray and white matter tracts of the brain. Primary indications were intractable obsessive compulsive and anxiety neuroses, chronic pain from cancer, and movement disorders such as Parkinson’s disease. The second-generation device had 179 beams but a spherical dose profile (Leksell Gamma Unit, model U). A dramatic increase in the development of radiosurgical technology resulted in the 1980s and 1990s, when the redesigned 201 source Models B and C Gamma Knives became available (> Figure 30-17). In 2007 the completely redesigned and completely

. Table 30-2 Complications Related to Frame Based Stereotactic procedures: UPMC-Presbyterian; 1979–2007 Procedure type

Total

Hemorrhagea

Diagnostic biopsy Cyst aspiration Radiation implant Brain abscess Catheter and cyst reservoir insertion Hematoma aspiration Frame based craniotomy Pallidotomy Thalamotomy Deep brain stimulation (movement disorders) Depth electrodes for seizures Deep brain stimulation(chronic pain) Cell transplantation Mesencephalotomy/Capsulotomy Total

1,664 197 145 97 19 9 10 147 72 148

43 (2.58%) 5 (2.53%) 2 (1.37%) 1 (1.03%) 2 (1.36%) 1 (1.38%) 1 (0.67%)

95 24 20 4 2,651

55 (2.07%)

a

Six patients (0.36%) required Craniotomy and hematoma evacuation Includes cerebritis, meningitis

b

Seizure

Infectionb

6 (0.36%) 3 (1.52%) 1 (0.68%) 1 (1.38%) 11 (0.41%)

2 (0.12%) 2 (1.01%) 1 (0.68%) 4 (4.12%)

Total complications

1 (0.67%)

51 (3.06%) 10 (5.07%) 4 (2.75%) 5 (5.15%) 0 2 (1.36%) 2 (2.76%) 2 (1.34%)

1 (1.05%) 11 (0.41%)

1 (1.05%) 77 (2.90%)

-

479

480

30


. Table 30-3 Comparison of Image-Guided Brain Biopsy Techniques

Image compatibility Image integration Fixation of head Precise trajectory pre-plotting Target calculation method Software calculation Cranial target access Anesthesia Accuracy Ease of performing craniotomy Hemorrhage risk requiring craniotomy Accessible targets Lobar Deep Neuroimaging/neuropathy correlation . Figure 30-16 The stereotactic hemiarc attached to the base ring can be used to guide a laser beam or an endoscope to a target

Frame based

Frameless

MRI, CT, PET, MEG Intraoperative 4 point (frame) Yes Frame fiducials Yes Twist drill Local 0.7–1 mm Low Low ( Figure 30-18). As of early 2008, close to 300 Leksell Gamma Knives were in use in more than 50 countries

SurgiPlan is an image-integrated surgical planning computer program that converts stereotactically acquired images (e.g., MRI, CT) to actual Leksell stereotactic frame space. After imaging is complete, the data sets are transferred by CD or Ethernet to the surgical workstation. Conventional angiographic films can be digitized into the planning system as well. SurgiPlan allows preoperative viewing of images either in original orientation or in any reformatted plane. Three-dimensional representations of the whole brain can be rotated and the surgical target reviewed from a variety of


30

. Figure 30-17 Left: the 201 cobalt-60 source Leksell model B Gamma Knife. Right: the outer collimator helmet slides into the inner collimator containing the cobalt sources. The redesign of the unit facilitate reloading of the cobalt sources between 5 and 10 years after initial source placement

directions. The surgical image interaction is virtually in real time, since the image handling by the computers is extremely rapid. SurgiPlan increases the accuracy of target recognition and allows preplotting of simulated probe trajectories (> Figure 30-19). Using SurgiPlan, an optimal trajectory can be pre-plotted prior to actual placement of the probe. These ‘‘virtual reality’’ probe trajectories can be varied in order to minimize the risks. SurgiPlan is also valuable during functional neurosurgical procedures, especially

with MRI-acquired data sets. With such information, the anterior and posterior commissures, inter-commissural plane and functional targets can be determined and visualized in multiple or even probe’s-eye-view ‘‘planes’’ (> Figure 30-20).

Summary In a prior publication, we posed eight questions that should be answered referable to any stereotactic

481

482

30


. Figure 30-18 Left: the completely robotic 192 cobalt-60 source Leksell Gamma Knife Perfexion. Right: the sources are fixed in eight independently movable sectors mounted on the collimator body, containing collimators of 4, 8, and 16 mm

system [37]. Although time has passed, these eight questions remain pertinent: (1) Is it a complete system?, (2) Is it simple?, (3) Is it dependable?, (4) Is it versatile?, (5) Is it accurate?, (6) Is it compatible with multiple current imaging modalities?, (7) Is it both computer-compatible and independent?, and (8) Is its development keeping pace with developments in other related technologies?

The evolutionary Leksell stereotactic system provides affirmative answers to all these questions.

Acknowledgment The authors are indebted to Elekta Instrument AB for help in preparation of the figures.


30

. Figure 30-19 SurgiPlan is a versatile surgical planning system that determines the X, Y, Z coordinates and preplots probe trajectories

. Figure 30-20 Functional targets can be localized with Leksell SurgiPlan using AC-PC based formulas which generate the corresponding stereotactic coordinates

483

484

30


References 1. Spiegel EA, Wycis HT, Marks M, Lee ASJ. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 2. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta Chir Scand 1949;99:229-33. 3. Kondziolka D, Lunsford LD, Kanal E, et al. Stereotactic magnetic resonance angiography for targeting in AVM radiosurgery. Neurosurgery 1994;35:585-91. 4. Leksell L. Stereotaxis and radiosurgery: an operative system. Springfield, IL: Charles C. Thomas; 1971. 5. Leksell L, Jernberg B. Stereotaxis and tomography: a technical note. Acta Neurochir (Wien) 1980;52:1-7. 6. Leksell L, Leksell D, Schwebel J. Stereotaxis and nuclear magnetic resonance. J Neurol Neurosurg Psychiatry 1985;48:14-18. 7. Leksell L, Lindquist C, Adler JR, et al. A new fixation device for the Leksell stereotactic system. J Neurosurg 1987;66:626-9. 8. Peters TM, Clark JA, Pike GB, et al. Stereotactic neurosurgery planning on a personal-computer-based workstation. J Digital Imaging 1989;2:75-81. 9. Kondziolka D, Dempsey PK, Lunsford LD, et al. A comparison between magnetic resonance imaging and computed tomography for stereotactic coordinate determination. Neurosurgery 1992;30:402-7. 10. Lunsford LD. MRI stereotactic thalamotomy: report of a case with comparison to CT. Neurosurgery 1988;23:363-7. 11. Lunsford LD, Martinez AJ, Latchaw RE. Stereotaxic surgery with a magnetic resonance and computerized tomography compatible system. J Neurosurg 1986; 64:872-8. 12. Latchaw RE, Lunsford LD, Kennedy WH. Reformatted imaging to define the intercommissural line for CT-guided stereotactic functional neurosurgery. Am J Neuroradiol 1985;6:429-33. 13. Kondziolka D, Lunsford LD. Stereotactic biopsy for intrinsic lesions of the medulla through the long axis of the brainstem. Acta Neurochir 1994;129:89-91. 14. Backlund EO. A new instrument for stereotaxic brain tumor biopsy. Acta Chir Scand 1971;137:825-7. 15. Backlund EO. Stereotactic treatment of craniopharyngiomas: a 15 year material. In: Proceedings of the 32nd annual meeting of the scandinavian neurosurgical society, Linkoping, Sweden, 1979. 16. Backlund EO. Stereotaxic evacuation of hematomas (letter). J Neurosurg 1985;62:460-1. 17. Backlund EO. Studies on craniopharyngiomas: III. Stereotactic treatment with intracysticyttrium-90. Acta Chir Scand 1973;139:237-47. 18. Duma CM, Kondziolka D, Lunsford LD. Image-guided stereotactic management of non-AIDS related cerebral infection. Neurosurg Clin N Am 1992;3:291-302.

19. Engle D, Lunsford LD. Brain tumor resection guided by intraoperative computed tomography. J Oncol 1987; l4:361-70. 20. Hall WA, Lunsford LD. Changing concepts in the treatment of colloid cysts. An 11-year experience in the CT era. J Neurosurg 1987;66:186-91. 21. Lunsford LD, Coffey RJ, Cojocaru T, Leksell D. Image guided stereotactic surgery: a ten year evolutionary experience. Stereotact Funct Neurosurg 1990;54–55: 375-86. 22. Lunsford LD, Deutsch M, Yoder V. Stereotactic interstitial brachytherapy – current concepts and concerns in twenty patients. Appl Neurophysiol 1985;48:117-20. 23. Lunsford LD, Gumerman LW, Levine G. Stereotactic intracavitary irradiation of cystic neoplasms of the brain. Appl Neurophysiol 1985;48:146-50. 24. Lunsford LD, Somaza S, Kondziolka D, Flickinger JC. Brain astrocytomas: biopsy then irradiate. Clin Neurosurg 1995;42:464-79. 25. Roberts DW. Epilepsy: Deep brain electrodes. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1988. p. 413-23. 26. Ohye C. Selective thalamotomy for movement disorders. Microrecording stimulation techniques and results. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1988. p. 315-40. 27. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotaxic thermolesions in the pallidal region: a clinical evaluation of 81 cases. Acta Psychiatry Neurol Scand 1960;35:358-77. 28. Lunsford LD. Stereotactic treatment of craniopharyngioma: intracavitary irradiation and radiosurgery. Contemp Neurosurg 1989;11:1-6. 29. Lunsford LD, Pollock BE, Kondziolka DS, et al. Stereotactic options in the management of craniopharyngioma. Pediatr Neurosurg 1994;21:90-7. 30. Pollack IF, Lunsford LD, Slamovitz T, et al. Stereotaxic intracavitary irradiation for cystic cranipharyngiomas. J Neurosurg 1988;68:227-33. 31. Lunsford LD, Nelson PB, Rosenbaum AB. Stereotactic aspiration of a brain abscess using the therapeutic CT scanner: case report. Acta Neuro Chir (Wien) 1982;62:25-9. 32. Lunsford LD, Leksell L, Jernberg B. Probe holder for stereotactic surgery in the CT scanner: a technical note. Acta Neurochir 1983;69:297-304. 33. Lunsford LD. A dedicated CT system for the stereotactic operating room. Appl Neurophysiol 1982;45:374-8. 34. Lunsford LD. Advanced intraoperative imaging for stereotaxis: the surgical CT scanner. Acta Neurochir (Wien) 1984;33:573-5. 35. Lunsford LD, Parrish R, Albright L. Intraoperative imaging with a therapeutic CT scanner: technical note. Neurosurgery 1984;15:559-61.


36. Lunsford LD, Rosenbaum AE, Perry JP. Stereotactic surgery using the ‘‘therapeutic’’ CT scanner. Surg Neurol 1982;18:116-22. 37. Lunsford LD, Leksell D. The Leksell stereotactic system In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1988. p. 27-46. 38. Kondziolka D, Lunsford LD. Guided neurosurgery using the ISG viewing wand. Contemp Neurosurg 1995;17:1-6.

30

485

34 Miniframe Stereotactic Apparatus M. A. Madera* . W. D. Tobler*

Introduction As stereotactic localization is increasingly used in cranial neurosurgery, its applications now include biopsy of mass lesions, aspiration of hematomas/cystic lesions, catheter placement, and planning small, strategically placed craniotomy flaps. Another important, but less often used, stereotactic application includes placement of electrodes for stimulation of deep structures. A more rarely used application is lesion creation in the treatment of movement disorders. Options for stereotactic targeting have increased in recent years with the introduction of image-guided systems, commonly known as frameless stereotaxis. There has been a rapid evolution from framebased, mechanical stereotactic devices to imageguided stereotactic systems. Although large, frame-based devices may still be preferred in some centers, image guidance has overall displaced mechanically-based systems in popularity. Our review of the evolution of miniframe stereotactic devices begins with the earliest freehand stereotactic-guided procedures in which small devices attached to the skull served as probe holders. During the frame-based stereotactic era, miniframe stereotactic devices were largely overlooked as large frame-based devices were promoted as the gold standard. With the introduction of image-guided techniques, the smallest miniframe devices have become important tools for stereotactic surgery.

*The authors have no financial relationships to any of the companies mentioned. #


Becoming increasingly popular, miniframe devices not only generate a stereotactic trajectory but also provide the rigidity requirements of stereotactic procedures. The concept of a small, ball-and-socket device fixed to the skull through a burr hole for stereotactic targeting was one of the earliest stereotactic devices. The trajectory was generated by rotating the ball and its axis through the lesion, which was chosen on standard X-rays, and then by passing the probe to the calculated target depth. With the wide proliferation and availability of image-guided systems, the use of the ball-andsocket device seemed to provide a simple solution for firm fixation and trajectory generation for stereotactic surgery. This type of miniframe device is perhaps the most intuitive, compact, yet simple stereotactic device available for even the most complex stereotactic procedures. The history of technological evolution and increasing supportive literature points to an emerging era of image-guided surgery using miniframe-like devices for frameless stereotaxis.

Burr-Hole Mounted Ball-andSocket Probe Holders In 1956, Austin and Lee presented their three-piece, plastic, burr-hole-mounted, stereotactic device (> Figure 34-1) that was threaded into a trephine craniotomy [1]. Adjustments were made using plain X-rays to create a desired trajectory to a chosen entry point. Once the trajectory was determined, the probe was passed to the target. There are no additional publications of experience with

522

34

Miniframe stereotactic apparatus

. Figure 34-1 Schematic drawing of the Austin and Lee three-piece ball-and-socket stereotaxic instrument. The device is threaded and tapped into a trephined hole in the skull. The plastic ball-and-socket joint is mounted within the shell and a needle is passed to the desired depth [1]

this technique, which leads to the conclusion that this device never became widely used. However, there has been a series of reports, published from 1955 to 1975, of similar devices for stereotactic trajectory generation fixed to the skull [2–5]. Most of theses authors used ventriculography for target localization; applications included lesion creation for Parkinson’s disease and cryosurgery for lesion creation and tumor treatment. In a 1982 study on a series of 26 CT-guided biopsies, Greenblatt et al. reported the use of a skull-mounted, ball-and-socket holding device for one biopsy [6]. This device was useful for multidirectional aiming of a biopsy needle, especially for superficial lesions, and thus prevented needle wobbling or dislodging during the procedure. In 1983, a skull-mounted ball-and-socket device was used as a needle holder for CT-guided biopsy in ten cases [7]. When compared with the contemporary frames, it was described as a less

cumbersome stereotactic device but criticized for its lack of true stereotactic functionality. The trajectory was chosen based on a CTscan obtained before needle insertion into the brain while the biopsy probe remained in the ball-and-socket holder. The trajectory, aimed at the lesion, was adjusted on subsequent scans until it was advanced into the target. In contrast with stereotactic frames in which the procedure was simulated in a threedimensional (3D) phantom-coordinate system derived from the CT dataset, the ball-and-socket device served as a holder for the probe, which was then advanced by hand. Although the ball-andsocket device was deemed better than using a freehand biopsy technique, it fell short of recognition as a stereotactic device. In a series of reports in the early 1980s, Patil described a succession of devices that were smaller and less complex than traditional frames. Patil placed the Z coordinate of the lesion in line with the head clamp, or in later versions, with the larger Patil frame itself. The earliest version was a head clamp that was fixed to the patient in line with the laser beam of the CT scanner; since the head clamp and probe holder were in the same plane, one CT image showing these and the lesion allowed planning of the trajectory without lengthy calculations [8]. Although it was deemed useful for large lesions, its accuracy was criticized [9]. Subsequent versions of the device improved upon the degree of freedom of the probe holder at the expense of somewhat more cumbersome mechanics. With this device and thinner CT slices, Patil achieved greater accuracy and reported advantages over other frames including less artifact and simpler calculations for trajectory planning [10,11].

First Burr-Hole Mounted (miniframe) Mechanical Stereotactic Device The Pelorus system (Schaerer Mayfield USA, Cincinnati, Ohio) is a skull-mounted, true


stereotactic device conceived as an early alternative to traditional frame-based devices. Like traditional frame-based stereotactic devices, translation of the target coordinates from CT scan to a biopsy frame space was achieved mechanically on a special phantom frame and did not require any proprietary computer software (> Figure 34-2). The Pelorus system was based in part on the ball-and-socket device introduced by Austin and Lee in 1956 [1]. In the phantom frame, the trajectory through the balland-socket device was determined by pointing and advancing the probe to the target, which was set in the frame by a simple arithmetic calculation. The depth to the target was measured directly in the phantom. Unlike the large framebased devices, the small ball-and-socket was attached with a special fitting to a ring, measuring less than 4 cm in diameter; this ring was fixed to

. Figure 34-2 The Pelorus stereotactic system components: double transfer plate (a), reference ball and socket (b), phantom frame with x, y and z coordinates (c), target point (d), arc carrier post (e), arc (f) and adjustable biopsy ring (g) (printed with permission from Mayfield Clinic)

34

the skull with self-securing, bi-cortical screws that were placed with the patient under local anesthesia (> Figure 34-3a). When first reported on by Carol et al. in 1985, advantages of the Pelorus stereotactic system were unobstructed trajectories to the temporal lobe or posterior fossa, greater ease of application, and better patient tolerance [12]. In addition to the ball-and-socket burr-hole mounted device, an arc adaptor could be attached to the skull mount to permit center-ofarc targeting. Such targeting allowed for multiple entry points that could be changed at the time of surgery and was especially useful for stereotactic craniotomies. In our experience with the Pelorus stereotactic system at the Mayfield Clinic/University of Cincinnati Department of Neurosurgery between 1989 and 1995, the device was used in more than 400 stereotactic procedures for stereotactic biopsy, hematoma aspiration, placement of depth electrodes, and numerous stereotacticguided craniotomies (> Figure 34-3b). Our experience confirmed the Pelorus stereotactic system offered advantages over the frame-based systems, such as simplicity of application and ease of accessibility to cranial targets, especially in the temporal lobe and posterior fossa since no frame impeded the trajectory. As an alternative to frame-based systems that have limited working space and trajectory options, we reported our experience with the Pelorus apparatus with the arc adapter for implantation of depth electrodes for seizure monitoring [13]. In confirming target accuracy during surgery or postoperative CT evaluation for 96 stereotactic biopsy and/or aspiration procedures, we obtained nondiagnostic tissue (indicating a possible targeting error) in three patients; that is, diagnostic rate was 96.9% [14]. While the Pelorus stereotactic system never gained the popularity or mainstream acceptance of the large frame-based stereotactic devices, it served our stereotactic requirements in a most satisfactory manner.

523

524

34


. Figure 34-3 Illustration of the Pelorus stereotactic system. (a) Single-transfer plate is attached to the patient’s head with bicortical screws. (b) Ball-and-socket is mounted to the transfer plate with the arc carrier post locked at the exact coordinates. The surgeon can swivel the arc adapter along the post to determine the optimal trajectory and entry point to the targeted lesion (printed with permission from Mayfield Clinic)

Miniframe Devices in the Image-Guided Era Image-Guide Surgery In the mid-1990s, image-guided surgery was introduced into the operating room. Computing power enabled the creation of 3D image sets of the brain by using MRI or CT that could be viewed and manipulated in all three orthogonal planes, then allowing surgeons to navigate on these images. In contrast with the mechanical phantoms of the frame-based systems, these image-guided computers created a virtual phantom; the surgical workstation used a 3D phantom for simulation of the stereotactic procedure using the patient’s own anatomy. Using digitizing

instruments, surgeons could match the fiducial markers placed on the patient with the same markers identified on the 3D image dataset (CT or MRI). Specifically, the surgeon could move a pointer around the outside of the registered target on the patient’s skull and compare that with the position of the pointer tip on the workstation image of the patient’s brain. This new surgical technique, quickly termed frameless stereotaxis, enabled the surgeon to navigate in the operating room during the selection of a precise entry point for a small stereotactic craniotomy flap. The first digitizers were attached to mechanical arms that had spatial sensors embedded in their joints. Although these devices were effective, they were cumbersome and were soon replaced by active optical digitizers, which use infra-red


light to track LEDs placed on the surgical probes. Later replacement of the LEDs with reflective spheres permitted passive optical tracking, thereby eliminating the power cord to the probe so as to fashion a wireless navigational tool. Electromagnetic digitizers were developed to help eliminate the optical interference that can be problematic with optical digitizers, but they are not yet widely used [15]. These frameless stereotactic systems were first used to navigate outside the patient’s skull. Soon after their introduction, true stereotactic procedures using image-guided techniques were developed. The digitizing probe was easily positioned on the surface of the skull, an entry point was selected, and the computer then created a virtual trajectory to an intracranial target, including the depth calculated to that target. Using a fixed ball-and-socket device (miniframe apparatus) to generate a trajectory and to firmly hold the probe was an intuitive solution for live navigation on a 3D image of the brain for true stereotactic procedures.

Miniframe Frameless Stereotaxis At the University of Cincinnati, we modified the Pelorus ball-and-socket device so that it could be used with an optical tracking system. For a frameless stereotactic procedure, the Pelorus ring would be positioned over the entry point chosen during the actual navigation. Then the ring would be fixed to the skull at this location with self-securing screws. Using a simple computer program and an adaptor to hold the optical probe, we could place the probe in the balland-socket device. While navigating live and moving the probe in the ball-and-socket device, we then aligned the virtual trajectory from the tip of the probe to intersect the target. Fine adjustments could be made and the ball would be locked in place. The computer then would calculate the depth from the probe tip to the target. After opening the scalp and drilling a

34

small hole, the biopsy probe was passed to the target and the procedure performed. Modification of the Pelorus system to create a noninvasive stereotactic option resulted in a more versatile design called the AccuPoint (Schaerer Mayfield USA, Cincinnati, Ohio) ball-and-socket device. It was fitted with an adaptor that could be attached to the Mayfield Headrest System or the Budde Halo retractor system (Integra LifeSciences Corporation, Plainsboro, New Jersey). The AccuPoint device could be moved around the outside of the skull and positioned over the entry point, immediately against or close to the scalp, without placing screws into the scalp [16] (> Figure 34-4). Our cumulative experience with the Mayfield ACCISS (Schaerer Mayfield USA, Cincinnati, Ohio) image guided-system was reported for 300 cranial stereotactic procedures [16].

Stereotactic Biopsy with Frameless Stereotaxis Adaptation of the AccuPoint sphere for stereotactic procedures using image-guided surgery raised the question, ‘‘Could image-guided, frameless stereotactic techniques replace frame-based stereotactic procedures with the safety and accuracy profiles established for frame-based stereotaxis?’’ In our 2000 report of 79 patients who underwent frameless stereotactic biopsy using these techniques, three (4.4%) patients had biopsies that were nondiagnostic or a 95.6% diagnostic rate [17]. Our findings demonstrated nondiagnostic rates equivalent to those reported for frame-based systems and were confirmed in the ensuing years from other centers. Multiple other series validating the efficacy of stereotactic surgery have appeared recently in the literature. In 2001, Paleologos et al. reported a 97.6% diagnostic rate in 125 cases using a similar frameless technique [18]. Woodworth et al. reported similar a 89% diagnostic yield that was comparable for both frameless (110 cases) and frame-based (160 cases) techniques [19]; the SNN-Olivier FreeGuide (Philips Medical Systems,

525

526

34


. Figure 34-4 Mayfield ACCISS stereotactic system. (a) AccuPoint Targeting Sphere attached to the Budde Halo Retractor System by a mechanical arm. (b) With rotation of the mechanical arm, the surgeon can view the trajectory and target displayed on the computer workstation. (c) Once the optimal trajectory is determined and locked, the mechanical arm is removed and a biopsy needle can be passed to the target through the sphere (printed with permission from Mayfield Clinic)

Best, The Netherlands), used in the frameless cases, is essentially a probe holder (not a ball-and-socket device) that attaches to the Mayfield head holder and secures to the calvarium with a metal pin. In a retrospective review of their 10-year experience with stereotactic biopsies in 391 patients, Dammers et al. reported a combined diagnostic rate of 89.4%, with no differences between 227 frame-based and 164 frameless procedures [19]; their experience mirrored the findings of Woodworth et al. [20]. Interestingly, in a 9-year retrospective review of

213 consecutive stereotactic biopsies, Smith et al. reported equivalent diagnostic rates and complications between 139 frame-based and 74 frameless procedures [21]. At their institution, hospital stays were longer for the patients who underwent frameless procedures, which were performed under general anesthesia, than for patients who underwent frame-based procedures, which were mostly under local anesthesia. Noting the cost differential between the two groups, the authors recommended that frame-based techniques should be the first-line


consideration in stereotactic biopsy. In contrast, in their comparative study of 76 frameless and 79 frame-based biopsy procedures, Dorward et al. reported shorter operative times and significantly fewer complications in the frameless group [22]. They opined that superior imaging, better target visualization, and improved flexibility of frameless techniques translate into tangible benefits. Frameless stereotaxic biopsy seems to have come of age. The combination of the computerbased, image-guided, frameless technique with mainframe devices is now a more intuitive, user-friendly method than mechanical stereotactic frames. Our experience and mounting clinical evidence reported by others supports the safety and efficacy of frameless stereotaxic biopsy.

34

. Figure 34-5 Schematic drawing of a MRI-compatible SnapperStereoguide device. The base of the device is inserted into a 15-mm burr hole and secured with the swivel (a). Fixation wheel (b) secures the instrument guide, with a moveable angle of 35 degrees being provided by a spherical joint (c). Optical tracking system (d) is coupled to the instrument guide. Biopsy cannula (e), composed of carbon fiber-reinforced composite, is inserted into the instrument guide (with permission from [23])

Miniframe Devices with Intraoperative Imaging Intraoperative imaging is a solution for stereotactic updating in real time. Image-guided systems rely on historical data, that is, images acquired before entering the operating room. Innovation in MRI technologies now provide a sophisticated way to acquire updated information during the course of the surgical procedure. This updated anatomical information can then be downloaded into the image-guided system. To accomplish this, Bernays et al. devised the Snapper-Stereoguide, a multi-component assembly composed of an MRI-compatible plastic [23]. After its insertion into a burr hole, LEDs for image-guidance are attached to the assembly for intraoperative scans (> Figures 34-5 and > 34-6). Their 2000 report noted no adverse outcomes in 20 patients. In the authors’ experience, this system combined the advantages of intraoperative imaging with a small, MRI-compatible mounted system to improve operative time and maintain accuracy. A less expensive, less complex alternative to intraoperative MRI is CT whose advantages are no need for room shielding, lower cost, and less

. Figure 34-6 Photograph of the Snapper-Stereoguide (b) coupled to the optical tracking system (a) (with permission from [23])

527

528

34


logistical difficulty. At Mayfield Clinic, we have intraoperatively used the MobileSCAN CT (distributed by Schaerer Mayfield USA, Cincinnati, Ohio). Since the CT scanner is in the operating room, the patient, surgeon, anesthesiologist, staff, and hospital benefit from the efficiency of being able to immediately obtain images for verification of biopsy site, hematoma

aspiration, or tumor resection (> Figure 34-7). In our experience, accuracy has been excellent, and information gained from intraoperative scanning has helped guide the progress of the operation (‘‘Use of Mobile Intraoperative Computed Tomography Scanner for Intracranial and Spinal Procedures,’’ poster presentation, American Association of Neurological Surgeons, 2003).

. Figure 34-7 Intraoperative CT (iCT) in the operating room. (a) MobileSCAN iCT and image-guided system. (b) Position of the MobileScan CT for a biopsy. (c) Preoperative MRI of a 12-mm thalamic glioma. Six contiguous slices from an intraoperative scan (d) from the MobileSCAN CT shows the biopsy cannula in position. Contrast was not given for this biopsy (printed with permission from Mayfield Clinic)


Other Miniframe Devices The Navigus trajectory guide (Medtronic, Minneapolis, MN) is currently widely used for frameless stereotactic applications. Initially designed for use in stereotactic biopsy, the guide rigidly attaches directly to the skull and has attachments for different image-guided systems (> Figure 34-8). The Navigus is also conceptually identical to the Austin-Lee device. There are internal (fixed inside a 14-mm burr hole) and external (fixed to the surface of the skull) versions of the skullmounted part of the trajectory guide. Guides may be used for biopsy, shunt placement, endoscope insertion, or functional stimulation procedures. Quinones-Hinjosa et al. validated

34

the accuracy of this device for stereotoactic targeting [24]. The Nexframe (Medtronic, Minneapolis, MN) is a skull-mounted miniframe apparatus for accurate stereotactic targeting for functional stereotactic procedures (> Figure 34-9). In an important validation study, Henderson et al. demonstrated equivalent accuracy with the Nexframe device to the published results of accuracy for stereotactic frames in the laboratory setting [25]. Holloway et al. further showed equivalent accuracy to frame-based stereotaxis for functional procedures using the Nexframe device in a multicenter evaluation of 38 patients who underwent deep brain stimulation [26]. These devices are commercially available.

. Figure 34-8 Coronal (a) and along-the-probe (b) views during a biopsy procedure are shown using the Navigus trajectory guide (c) with inserted stereotactic probe (d) (reprinted with permission of Medtronic, Inc. ß 2008)

529

530

34


. Figure 34-9 Photo of the Nexframe stereotactic device (a), which is attached to the skull with the rigidly attached reference device (b) and microelectrode driver (c) (Nexdrive, Medtronic, Minneapolis, MN) (reprinted with permission of Medtronic, Inc. ß 2008)

Summary The history of stereotactic surgery is rich with technology and innovation and continues to evolve at a rapid pace. Beginning with those early, large, and often awkward looking devices for stereotactic localization, frames have always occupied center stage. From the outset, however, the ball-and-socket trajectory device, the original precursor miniframe apparatus, has been an enduring, though not mainstream, concept. These various, small devices that attach to the skull are now finding increasing popularity in the area of image-guided surgery and are likely to permanently displace stereotactic frames.

References

Other interesting and custom solutions for the challenges of functional neurosurgery have been reported. Eljamel et al. described the use of a small polyethylene cube containing five small parallel channels implanted into a burr hole for micro-electrode placement using frameless stereotaxis [27]. Winkler et al. reported their experience with this polyethylene cube in one patient who underwent deep brain stimulation [28]. Fiducial-like anchors were implanted into the skull before image acquisition. A custommade Micro Targeting platform is created for the patient to establish an individualized trajectory based on image-fusion data from CT, MRI, and target selection. In surgery, the platform is attached to the patient, and the stereotactic procedure is performed using the preplanned trajectory built into this custom made, one-time use stereotactic device.

1. Austin GM, Lee AS, Grant FC. A new type of locally applied stereotaxic instrument. J Am Med Assoc 1956;161:147-8. 2. Cooper IS. Chemopallidectomy: an investigative technique in geriatric parkinsonians. Science 1955;121: 217-18. 3. McCaul IR. A method for the localization and production of discreet destructive lesions in brain. J Neurol Neurosurg Psychiatr 1959;22:109-12. 4. Rand RW. A stereotaxic instrument for pallidothalamectomy in Parkinson’s disease. J Neurosurg 1961;18:258-60. 5. Kandel EI. New stereotactic apparatus and cryogenic device for stereotactic surgery. Confin Neurol 1975; 37:128-32. 6. Greenblatt SH, Rayport M, Savolaine ER, et al. Computed tomography-guided intracranial biopsy and cyst aspiration. Neurosurgery 1982;11:589-98. 7. Levy WJ. Simple plastic stereotactic unit for use in the computed tomographic scanner. Neurosurgery 1983; 13:182-5. 8. Patil AA. Computed tomography stereotactic head clamp. Acta Neurochirurgica 1982;60:125-9. 9. Patil AA. Computed tomography-oriented stereotactic system. Neurosurgery 1982;10:370-4 (comment). 10. Patil AA. Computed tomography (CT) oriented rotary stereotactic system. Acta Neurochirurgica 1983;68:19-26. 11. Patil AA. Computed tomography plane of the target approach in computed tomographic stereotaxis. Neurosurgery 1984;15:410-4. 12. Carol M. A true ‘‘advanced imaging assisted’’ skullmounted stereotactic system. Appl Neurophysiol 1985;48:69-72.


13. Yeh HS, Taha JM, Tobler WD. Implantation of intracerebral depth electrodes for monitoring seizures using the Pelorus stereotactic system guided by magnetic resonance imaging. Technical note. J Neurosurg 1993;78:138-41. 14. Tobler WD. The Pelorus apparatus. In: Gildenberg, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill, Health Professions; 1998. p. 119–26. 15. Mascott CR. Comparison of a magnetic tracking and optical tracking by simultaneouos use of two independent framelesss stereotactic systems. Neurosurgery 2005;57 (4 suppl):295-301. 16. Tobler WD. Surgical Navigation with the OMI System. In: Schulder M, Gandhi CD, editors. Handbook of stereotactic and functional neurosurgery. New York: M. Dekker; 2003. p. 91-101. 17. Tobler WD. To demonstrate that image-guided frameless stereotactic biopsy can be performed with precision equivalent to traditional frame-based stereotaxy. Presented at the XIV Congress: European society for stereotactic and functional neurosurgery, London; 25–27 October 2000. 18. Paleologos TS, Dorward NL, Wadley JP, et al. Clinical validation of true frameless stereotactic biopsy: analysis of the first 125 cases. Neurosurgery 2001;49:830-8. 19. Dammers R, Haitsma IK, Schouten JW, et al. Safety and efficacy of frameless and frame-based intracranial biopsy techniques. Acta Neurochir (Wien) 2005;150:23-9. 20. Woodworth GF, McGirt MJ, Samdani A, et al. Frameless image-guided stereotactic brain biopsy procedure: diagnostic yield, surgical morbidity, and comparison

21.

22.

23.

24.

25.

26.

27.

28.

34

with the frame-based technique. J Neurosurg 2006;104: 233-7. Smith JS, Quinones-Hinojosa A, Barbaro NM, et al. Frame-based stereotactic biopsy remains an important diagnostic tool with distinct advantages over frameless stereotactic biopsy. J Neurooncol 2005;73:173-9. Dorward NL, Paleologos TS, Alberti O, et al. The advantages of frameless stereotactic biopsy over frame-based biopsy. Br J Neurosurg 2002;16:110-18. Bernays RL, Kollias SS, Khan N, Romanowski B, Yonekawa Y. A new artifact-free device for frameless magnetic-resonance imaging-guided stereotactic procedures. Neurosurgery 2006;46(1):112-7. Quinones-Hinojosa A, Ware ML, Sanai N, et al. Assessment of image guided accuracy in a skull model: comparison of frameless stereotaxy techniques vs. frame-based localization. J Neurooncol 2006;76:65-70. Henderson JM, Holloway KL, Gaede SE, et al. The application accuracy of a skull-mounted trajectory guide system for image-guided functional neurosurgery. Comput Aided Surg 2004;9:155-60. Holloway KL, Gaede SE, Starr PA, et al. Frameless stereotaxy using bone fiducial markers for deep brain stimulation. J Neurosurg 2005;103:404-13. Eljamel MS, Tulley M, Spillane K. A simple stereotactic method for frameless deep brain stimulation. Stereotact Funct Neurosurg 2007;85:6-10. Winkler D, Strauss G, Helm J, et al. MicroTargeting1 platform: An individual stereotaxic device in functional neurosurgery. Int J Comput Assist Radiol Surg 2007;1: 295-9.

531

Stereotactic Targeting

25 Printed Stereotactic Atlases, Review R. J. Coffey

Introduction There is something so satisfying and practical about having a book that one can carry from place to place, open to and flip back and forth between whatever pages one pleases, that it is hard to imagine a world without stereotactic atlases in print format. The continued use of printed stereotactic atlases to guide functional neurosurgical operations still represents the legacy of nineteenth century neurological localizationalists and their phrenolological forebears – after more than 100 years of advances in neuroimaging and electrophysiological technology. Even apparently simple neurological processes involve the interaction of integrated functional systems that have nuclei in different geographical parts of the central nervous system, and that are connected by complex pathways. Although the concept of discrete anatomic ‘‘centers’’ for specific functions or behaviors is now outmoded, several well-defined intracerebral targets have retained their therapeutic utility since the advent of modern human stereotaxis in the mid-twentieth century. Unlike the pioneers of stereotactic surgery, who depended on the capricious appearance of normally calcified midline landmarks (pineal gland or habenular commissure) to navigate the brain, and unlike their immediate successors, who depended on the positive or negative shadows cast by air- or contrast-filled ventricles on x-ray film, contemporary neurosurgeons work directly from computed tomographic (CT) scans and magnetic resonance images (MRI) of the brain itself. Some – although the list is diminishing as imaging technology improves – important functional stereotactic #


targets are still indistinguishable from surrounding structures on CTor MRI; they remain invisible, or at least well camouflaged. Functional neurosurgeons solve this dilemma by referring to one of the excellent contemporary or historic stereotactic atlases or to a computerized atlas – the subject of Chapters 26 and 27. Experience with stereotactic atlases conjures up images of ponderous and expensive folio-sized volumes of high-quality photographic plates of unstained whole brain sections or, more often, magnified and stained sections of the thalamus, basal ganglia, and upper brain stem. Useful stereotactic atlas sections are cut at regular intervals in two or more planes. Of necessity, more than one brain is required if sagittal, frontal, and horizontal photographic sections are presented; over 100 brains were used to produce some atlases. Each section or photographic plate has a twodimensional scale in the margins, on a clear overlay, or on an accompanying line drawing. The numerical coordinate of the section plane and the two-dimensional coordinates of the target structure within the plane determine the set of three-dimensional coordinates required to reach the target with a stereotactic instrument. The fact that dimensions of individual brains vary from each other and from the idealized or standard brains depicted in atlas sections raises a contradiction for functional stereotactic atlases: For all the effort and expense required to produce a brain atlas, the results are not perfectly accurate (within less than a few millimeters) for any particular patient. In most instances, a purely atlasguided operation is not sufficiently accurate to justify the placement of a permanent therapeutic

348

25

Printed stereotactic atlases, review

lesion or stimulating electrode. Modern MR imaging of the individual patient comes to the rescue in most contemporary cases. However, the neurosurgeon still must undertake the sometimes tedious task of obtaining physiological corroboration of the probe’s position. This usually involves the use of intraoperative electrical stimulation, evoked potential recording, or other electroanatomical techniques on an awake, cooperative patient. The correlation of intraoperative electroanatomic phenomena with a stereotactic atlas, in light of the surgeon’s knowledge and the published results of previous surgeons’ experiences, is the essence of functional stereotaxis. Based on intraoperative observations during each probe trajectory, the surgeon must decide in which direction and by what distance to adjust the target for each subsequent trajectory. In this manner, the stereotactic atlas and a compendium of prior electroanatomic observations (memorized by the surgeon) are complementary tools that guide the successful performance of the surgical procedure. With few exceptions, judging from their intraoperative responses, most patients appear to have studied a stereotactic atlas as well.

A Chronology of Selected Stereotactic Atlases in Print Format Historic and Out-of-print Stereotactic Atlases 1952, 1962: Spiegel and Wycis Stereoencephalotomy (thalamotomy and related procedures), Part I: Methods and stereotaxic atlas of the human brain (Spiegel and Wycis, 1952)

In the spring of 1947 E.A. Spiegel, a clinical and research neurologist who had considerable experience in experimental animal stereotaxis, and H.T. Wycis, a neurosurgeon who had worked in Spiegel’s laboratory at Temple University in Philadelphia, performed the first modern stereotactic operation on a human patient; they had operated more

than 100 additional patients by 1952. By that time Spiegel and Wycis realized that functional stereotactic surgery ‘‘requires an exact preoperative calculation of the electrode position, and such a calculation depends on two conditions: (1) determination of a reference point by means of an X-ray picture taken under definite standard conditions, and (2) an exact knowledge of the position of the area to be destroyed in relation to the reference point. Thus . . . a stereotaxic atlas of the human brain is presented’’ [1]. Their landmark publication of the first human stereotactic atlas became the foundation for their own future work and for all who followed. Brains destined for inclusion in an atlas had to be fixed in situ as soon as possible after death. Before opening the cranium, the authors applied a stereotactic frame in order to pass metal rods completely through the skull and cerebrum at known distances from each other and at known stereotactic reference points in one or more planes. In this manner, shrinkage after fixation, freezing, or other processing could be quantified precisely and corrected by photographic enlargement or manipulation during preparation of the photographic atlas plates. Another major contribution was their demonstration of considerable variability in the contours and dimensions of the thalamus and other brain regions independent of variations in skull morphology. Even the cerebral midline deviated unpredictably from the midline of the skull. Brain atlases based on external cranial landmarks were suitable for small and medium-sized laboratory animals; that was not the case for humans. Thus, Spiegel and Wycis began a systematic search for reliable, radiographically demonstrable reference points on which to base stereotactic atlas planes of section and surgical procedures. Their initial reference point – the center of the pineal gland calcification on plain x-ray films – varied by 12 mm or more in the anteroposterior dimension and by as much as 16 mm relative to the interaural plane. They also


utilized the habenular calcification in some operations. In others, by means of lumbar pneumography, they visualized the posterior commissure (CP or PC), the foramen of Monro (FM), and rarely, the anterior commissure (AC). Spiegel and Wycis employed a line connecting the center of the PC with the pontomedullary sulcus at the posterior border of the pons (PO), to define the CP-PO line – an imaginary baseline – to construct their first atlas (> Figure 25-1). They cut their 5-mm-thick frontal (coronal) unstained sections and their 2–4-mm-thick myelin-stained frontal sections parallel to their so-called average cerebral directional line (inclined 4 degrees behind the CP-PO line). Oblique unstained sections 5 mm thick were cut through the brainstem at an angle of 30 degrees anterior to the CP-PO line and centered at the PC. A series of myelin-stained oblique sections 0.5 mm thick were cut parallel to the same . Figure 25-1 Landmarks for the intracerebral coordinate system used in Spiegel and Wycis’s first stereotactic atlas. Ch, commissura habenularum; Cp, posterior commissure; Cp-Po, posterior commissure – pons line; cran. 1, cran. 2, cranial direction lines; h1, horizontal line perpendicular cran. 1; ho, horizontal line perpendicular to Cp-Po; i + i, angles of inclination; Ob, medulla oblongata; Po, pons; Th, taenia habenulae (from [1], reproduced with permission)

25

plane. Unstained 5-mm-thick sagittal sections were cut parallel to the median plane. Unstained horizontal sections were cut perpendicular to both the median plane and the CP-PO line. This was as complicated as it sounds – and within 10 years Spiegel and Wycis and others eventually settled upon a simpler and more useful coordinate system. The final two chapters of Stereoencephalotomy (Part I) consist of anatomic and radiographic variability studies in 30 normal brains, plus postmortem examples of accurately and incorrectly placed lesions. In the variability studies, the authors catalogued the range of coordinates at which the borders of various nuclei, tracts, and other selected structures could be found in relation to both the center of the pineal gland and the posterior commissure. The 15 anatomic structures studied in this manner included the head of the caudate nucleus; the putamen; the globus pallidus; the anterior, dorsomedial, and ventrolateral thalamic nuclei; the pulvinar, tuber cinereum, mammillary bodies, corpus Luysii (subthalamic nucleus), substantia nigra, red nucleus, and medial and lateral geniculate bodies; and the mesencephalic spinothalamic tract. Interest in some of these targets would wane (dorsomedial nucleus of the thalamus, for example) and other more refined targets would emerge (the thalamic nucleus ventralis intermedius and the periaqueductal-periventricular gray matter, to cite a few). However, Spiegel and Wycis established a valuable precedent. Variability studies by other investigators – a process that continues into the present – have provided valuable insights into how stereotactic operations can go awry. The postmortem studies, especially the offtarget ‘‘misses,’’ illustrated the shortcomings of radiographic localization based on a single point such as the pineal gland, habenular calcification, or CP-PO line. By 1962, when Spiegel and Wycis published their second volume, Stereoencephalotomy (Part II) [2], general acceptance of the anterior commissure-posterior commissure line (AC-PC line (intercommissural line, IC line)) as the standard stereotactic reference system

349

350

25


had overcome the limitations inherent in the earlier method. Nevertheless, Stereoencephalotomy (Part I) probably did more to stimulate widespread interest and advancement in the field of stereotaxis than any other single publication. Early successes of Spiegel and Wycis were all the more remarkable given the nearly complete lack of previous experience in human stereotaxis between 1947 and 1952. Stereoencephalotomy, Part II: Clinical and physiological applications (Spiegel and Wycis, 1962)

Stereoencephalotomy (Part II), which was published 10 years after Part I, was primarily a textbook and only secondarily a revised and updated brain atlas [2]. The series of myelin-stained sections had appeared two years earlier in Confinia Neurologica [3,4]. In the interval, Spiegel and Wycis refined their stereotactic instrument (Stereoencephalotome Model V), improved and simplified their radiographic localization technique, and most importantly, embraced the Franco-German (Talairach and Schaltenbrand and Bailey) use of the intercommissural line as a stereotactic reference [5–10]. Most of the book was devoted to stereotactic techniques and clinical results, including honest morbidity and mortality figures for commonly performed stereotactic operations. Indications included psychosurgery, pain, involuntary movement disorders, epilepsy, and subcortical tumors. Perhaps the most valuable feature was presentation of post-mortem findings correlated with a patient’s radiographic, clinical, and surgical findings.

1957: Talairach and Colleagues Atlas D’Anatomie Stereotaxique: Reperage Radiologique Indirect des Noyaux Gris Centraux des Re´gions Mesencephalo-sou-Optique et Hypothalamique de L’Homme (Talairach, David, Tournoux, Corredor, and Kvasina, 1957)

Talairach’s first stereotactic atlas in book format appeared in 1957 (a magnificently produced and bound folio-sized volume), however, his pub-

lished scientific work on the subject dated back at least to 1949 [8,9]. Among the most important of Talairach’s contributions to stereotaxis were the introduction and popularization of combined positive-contrast and air ventriculography to demonstrate the AC and PC reliably, the invention of an accurately relocatable stereotactic instrument that utilized teleradiographic techniques and a ‘‘double grid’’ localization system, and the integration of angiography and ventriculography to create the most advanced stereotactic system in the pre-CT era. His technical developments shaped Talairach’s first and subsequent stereotactic atlases and radiographic-anatomic research over more than four decades [10]. One cannot overemphasize the importance of Talairach’s elegant demonstration that the deep structures of interest to stereotactic neurosurgeons bear a generally consistent relationship to the intercommissural line and its derivative planes [midsagittal plane, horizontal intercommissural plane, and the two vertical planes passing through the AC (VCA) and PC (VCP), respectively]. Later investigators would abandon Talairach’s two vertical planes in favor of a single intercommissural plane. Still, because of variation in length of the IC line between individuals (range, 23–28 mm from the center of AC to the center of PC; mean, 25.5 mm in Talairach’s work), one usually finds stereotactic coordinates listed as a distance anterior or posterior to PC (less often AC) as well as in relation to the mid-IC plane. Thus, Talairach’s system has exerted a lasting influence – even on workers who believe they have abandoned it for a more modern one. > Figure 25-2 reproduces Talairach’s illustrations of how his intracerebral reference system (the IC line and the VCA and VCP planes) and the locations of deep cerebral structures could bear a firm anatomic relationship to each other, yet vary considerably from the antiquated Horsley-Clarke reference system based on external landmarks. Even in a small sample of human ventriculograms, the axis of


25

. Figure 25-2 (a) Talairach’s illustration of the three early human stereotactic reference planes based on the same osseous cranial landmarks as used in animal stereotaxis. These were the infraorbitomeatal plane (Reid’s base line or the Frankfort line), the interaural plane (zero plane of Spiegel and Wycis), and the midline plane of the cranium. (b) Talairach’s demonstration of the variability in the major intracerebral axis (intercommissural line) from the osseous (Reid’s, or Frankfort) base line (from [9], reproduced with permission)

the IC line varied between 11.5 and 18.5 degrees from the infraorbitomeatal line (Frankfort line, Reid’s base line) (> Figure 25-2b). Talairach used his double-grid stereotactic instrument to create perforations in craniocerebral specimens at known locations and distances in the frontal and lateral planes. After this procedure, accurate coordinate measurements and profiles were derived for deep cerebral nuclei, subnuclei, and tracts, and the stereotactically marked brains were cut in either parasagittal or frontal sections along Talairach’s standard planes. He then mapped the three-dimensional profiles of the thalamic nuclei and other structures on millimeter-ruled diagrams. By presenting each sectional drawing next to the corresponding unstained or myelin-stained photographic plate at the same magnification, Talairach set the standard for all future stereotactic atlases. In this sense, even the apparently novel utilization of transparent overlays by Schaltenbrand and Bailey, and later by Schaltenbrand and Wahren [5,6], is derivative of Talairach. Talairach also invented a method to localize the ventral tier of thalamic nuclei in any patient. Using Talairach’s rules to proportionately subdivide the simple geometric

forms outlined by the IC line and the roof of the thalamus seen on lateral ventriculogram films, a neurosurgeon could draw or scratch Talairach’s diagram directly on the film, thereby recreating a properly scaled atlas template from which to derive stereotactic coordinates. Those who do not read French may be intimidated by Talairach’s atlas. However, the illustrations, captions, and labels are so clear that most of the essential data require no translation.

1959: Schaltenbrand and Bailey Introduction to stereotaxis with an atlas of the human brain (3 volumes) (edited by Schaltenbrand and Bailey, 1959)

The Schaltenbrand and Bailey atlas probably was the world’s most widely used compendium of brain maps during the nostalgic epoch of stereotactic surgery for involuntary movement disorders in the pre-L-dopa era (early 1960s to mid-1970s) [5]. Two oversized loose-leaf folio volumes contained the highest-quality myelinstained and unstained photographic atlas plates available at the time. The accompanying text, in both German and English, contained scholarly

351

352

25


treatises on neuroanatomy, physiology, and stereotactic techniques by the editors and 24 other contributors from the United States, Germany, and Italy. Despite having one editor and a plurality of contributors from the United States, the Schaltenbrand and Bailey atlas is decidedly Teutonic in style and content. Schaltenbrand’s stereotactic suite in Wurzburg contained an elaborate bidirectional optical projection system to superimpose atlas plates, anatomic outlines, adjustable magnification scales, and the patient’s ventriculogram images on the same translucent screen. In the decades before computer graphics, the atlas sections and scales could be optically modeled to match an individual patient’s intercommissural distance and other anatomic features. Volume II, the major portion of the atlas, will never become obsolete. Although the first three series of maps in this volume (in each of three orthogonal planes) contain unstained sections that are of limited usefulness to stereotactic surgeons, the next three series of maps contain the most magnificent myelin-stained atlas plates of

their day. The frontal series, cut orthogonal to both the mid-sagittal plane and the intercommissural line (and parallel to the midcommissural plane) (> Figure 25-3), begins with plate 36. The sections are presented four per page at 4 magnification, with a scaled and labeled transparent overlay attached to each page. The 16 sections, each 1–4 mm thick and all cut from the same brain, span the region from 16.5 mm anterior to 16.5 mm posterior to the midcommissural plane. The sagittal series, beginning with plate 42, is presented in the same manner, except that one or two sections appear on each page. The 18 sections are cut at 0.5–2.5-mm intervals, spanning the region between 2.0 and 27.5 mm lateral to the midline. Schaltenbrand and Bailey’s myelin-stained sagittal series has been a stereotactic bible of sorts for the past 50 years because the majority of functional stereotactic operations involve a transfrontal (precoronal) approach to the thalamus or upper midbrain through a parasagittal entry point. Among the 18 sections in this series,

. Figure 25-3 Schaltenbrand and Bailey’s three cardinal reference planes and their relation to the anterior and posterior commissures – a system that appears derivative of Talairach’s. With few exceptions (e.g., Afshar et al. [11], Andrew et al. [12]) every English language stereotactic atlas published since 1959 employed this same reference system. The illustration also shows the size of the central block of the brain (in millimeters) used to prepare the myelin-stained photographic atlas plates (from [5], reproduced with permission)


there is something almost magical about plate 47 (brain LXXVIII), which depicts the 13.5-mm and 15.0-mm planes. In most individuals, the hand area (median nerve territory) of the thalamic somatosensory relay nucleus resides in one of these planes (usually at 13.5 mm) and corresponds to the laterality at which the therapeutic lesion most often should be inflicted to relive parkinsonian tremor or essential tremor of the upper extremities (> Figure 25-4). Given the widespread use of the Schaltenbrand and Bailey atlas in both the printed format and as the basis for computerized software, the original owner of brain LXXVIII made an immense contribution to functional stereotactic surgery.

25

The myelin-stained horizontal series begins with plate 52 and, like the frontal series, is presented at four planes per page at 4 magnification. The 20 sections, all cut from a single brain, span the region from 16 mm above to 9.5 mm below the midcommissural point. Unfortunately, although the sections are all parallel to each other, they deviate from the axis of the IC line by approximately +7 anteriorly. Therefore, the +0.5-mm plane crosses the intercommissural plane within 0.5 mm of the midpoint of the IC line, but crosses the anterior commissure approximately 2.0 mm above its midpoint. Volume III contains 10-cresyl-violet-stained frontal sections and eight sagittal sections prepared

. Figure 25-4 Plate 47, brain LXXVIII, myelin-stained sagittal section 13.5 mm from the midline, with ruled and labeled acetate overlay. In the pre-CT stereotactic era (between 1959-late 1970s) this single atlas section probably guided more stereotactic operations – specifically, thalamotomies for tremor – than all of the other brain maps in the world combined (from [5], reproduced with permission)

353

354

25


in the same manner, each magnified 20 times. Aside from occasional use by the odd neurophysiologist or comparative anatomist, most copies of volume III spend a lifetime undisturbed on a library shelf.

1969, 1978: Andrew, Afshar, and Watkins A stereotaxic atlas of the human thalamus and adjacent structures; a variability study (by Andrew and Watkins, with the collaboration and histological assistance of Tomlinson, 1969)

A. Earl Walker wrote in his Foreword to this book that ‘‘In spite of thousands of stereotactic operations, the maps of target areas have been based upon very few anatomical preparations so that variations in size, shape, and position of the subcortical nuclei are inadequately established. It has been noted that the variations are so great that, using stereotactic coordinates alone, the chances of placing an electrode in a given small nucleus are indeed slight’’ [12]. Likewise, the authors noted that as of 1961 – when work on this book commenced as a project to delineate borders of the centromedian nucleus of the thalamus for surgery to relieve pain – even the recently published Schaltenbrand and Bailey atlas [5] contained relatively few thick-section variability diagrams based on only seven human specimens. The authors’ initially modest aims expanded into a more comprehensive textbook-sized study. The Methods section details how the authors – in a manner similar to the work of previous investigators – prepared, preserved insitu, indexed, and cut their specimens to permit calculation of- and compensation for shrinkage after formalin fixation. They ended up with 19 brains (38 hemispheres) after discarding specimens that were imperfectly cut, or were otherwise unsuitable for study. Because of their focus on thalamic nuclear variability, and because measurements based on dimensions of the thalamus were associated with less variability than measurements based upon the intercommissural (AC – PC) line,

the authors recorded AC-PC in all of their specimens, but preferred to base their variability measurements differently from classical Talairach space. In addition to AC – PC they measured: 1) the distance between the postero-inferior margin of the foramen of Monro to the midpoint of the ventricular (anterior) surface of the posterior commissure – the so-called FM – PC line or distance; and 2) the distance between FM and the tip of the pulvinar to determine the total thalamic length, or T.ThL. Readers should note that the postero-inferior margin of FM corresponds to the rostral limit of the thalamus (anterior nucleus). Most of the book consists of tabular and statistical analyses of the size, shape, location, and borders of thalamic nuclei and nearby structures (e.g., basal ganglia, optic tract). Probability tables and planar representations of each structure estimate the likelihood of finding a structure of interest at reasonable neighborhood coordinates relative to ventriculographically demonstrable landmarks – namely, FM, PC, the midcommissural plane, and the authors’ mid-thalamic plane (> Figure 25-5). From our perspective in the present high field strength MRI era, when one can see exquisite images of each patient’s deep brain structures immediately before (or even during) stereotactic operations, it is difficult to imagine what an eye opener the Andrew and Watkins atlas was for neurosurgeons of the 1960s. The actual atlas portion appears in the final two chapters – approximately 112 of the more than 250 pages – which contain full page 2.5 magnified line drawings and on the facing page, a corresponding stained photographic section (Nissl and or myelin-stained, from a representative specimen). This compendium of 21 coronal line drawings and macro photographs at 1.0 mm intervals from 1 to 21 mm behind FM, the anterior tip of the thalamus, and 17 sagittal line drawings and macro photographs at 1.0 mm intervals from 3–20 mm from the midline spans the entire thalamus, adjacent basal ganglia, and medial temporal lobes (> Figure 25-6). Only the principal nuclei and fiber tracts that were the


25

. Figure 25-5 Coronal and sagittal variability profiles of the subthalamic nucleus in the coronal plane 12 mm posterior to the foramen of Monro (FM, upper illustration) and in the parasagittal plane (lower illustration). White stippled shading represents 1 standard deviation (S.D.) and gray stippled shading represents 2-S.D., a range into which 66% of the population will fall relative to the FM-PC line and the mid-thalamic plane. (from [12] (Andrew, Tomlinson and Watkins), reproduced with permission)

subject of statistical analysis in the first portion of the book are illustrated. The outlines of the various structures are solid (based upon data from 70% or more of the specimens), light dashes (based on 40–70% of specimens), heavy dashes (based on 90 to 0%) of finding major brainstem nuclei and tracts at particular coordinates in any sectional plane. The atlas proper contains 54 myelin-stained brain stem sections photographed at 5 magnification and presented one per page. All sections are cut parallel to the fastigium-floor line (FFL) and orthogonal to the ventricular floor plane. An outline diagram with stereotactic coordinates and anatomic labels accompanies each photographic plate (in the style of Talairach, but organized more clearly). The 1-mm-thick sections extend from 23 mm rostral to the FFL (corresponding to the level of the red nucleus) to 30 mm caudal to the FFL (corresponding to the spinal tract of the trigeminal nerve, caudal to the gracile and cuneate nuclei) (> Figure 25-8). A similar treatment of the four


deep cerebellar nuclei appears immediately after the brain stem variability study. The variability studies and statistical profiles are illustrated in parasagittal graphic illustrations as well as in the transverse planes parallel to the FFL. The 12 myelin-stained cerebellar atlas plates are presented in exactly the same manner and at the same magnification as the brain stem photographs. They cover the range of transverse coordinates from 1 mm rostral to 10 mm caudal to the FFL. The anatomic detail presented in this atlas makes it the most valuable resource for brain stem stereotactic coordinates available in book format.

1972: Van Buren and Borke Variations and connections of the human thalamus, volumes 1 and 2 (Van Buren and Borke, 1972)

Van Buren and Borke produced this fine text and atlas while working at the National Institutes of Health in Bethesda, Maryland, and the folio-sized volumes were printed in Germany by SpringerVerlag [13]. Their work joined that of Schaltenbrand and Bailey [5], Schaltenbrand and Wahren [6], and Talairach and coworkers [9] to round out the ‘‘big four’’ stereotactic atlases of the pre-CT era, all printed in Europe, three in Germany.

25

Volume I includes a comprehensive textbook and a cytoarchitectonic study of the thalamus. Hundreds of high-quality photomicrographs of cresyl violet, myelin, and Golgi preparations are presented. In addition, extensive postmortem material from 54 patients shows both the original (therapeutic) cerebral lesions and the site and extent of secondary thalamic degeneration. Volume 2, Variations of the Human Diencephalon, continues the stereotactic atlas in the three orthogonal planes relative to the intercommissural line: sagittal, horizontal, and transverse (‘‘frontal’’ in the terminology of Schaltenbrand and Bailey). The atlas catalogues thalamic nuclei and their stereotactic coordinates relative to the IC line and the sagittal plane. Each series of cresyl-violet-stained plates is reproduced, one per page, at relatively high (8–10) magnification. The sagittal series consists of ten slices at 10 magnification, and spans the region between 2 and 25 mm lateral to the midline at 0.5–4 mm intervals. Outlines of nuclear groups and tracts are printed directly on the photographic plates, along with anatomic labels, coordinate index marks, and a magnification scale. Although the cell-stained sections at first appear unusual to surgeons accustomed to working from myelin-stained atlas plates, the stereotactic coordinates derived from this atlas correspond

. Figure 25-8 (a) Reference planes for measuring structural borders rostral to the 15-mm level (15 mm caudal to FB in > Figure 25-7). (b) Reference planes for structures at 15 mm or more caudal to the fastigial level are measured from the midline and the posterior medullary or spinal cord surface. a, anterior; l, lateral; m, medial; p, posterior (from [11], reproduced with permission)

357

358

25


closely to those obtained from other references. The cresyl-violet-stained horizontal series, presented at 8 magnification, consists of eight sections cut precisely parallel to the intercommissural plane. Sections approximately 3.5 mm thick span the region from 17 mm above to 8.1 mm below the IC line. The ten transverse (frontal) sections are photographed at 10 magnification, the same as the sagittal series. The plates cover a distance of 28.1 mm, from 23.4 mm anterior to PC to 4.7 mm posterior to PC. Individual nuclear profiles from five or six brains (depending on the plane of section) are mapped in the next chapters on minified (0.7) grids in the three orthogonal planes at 5-mm intervals on one or two pages each. The final chapter presents similar data for the gross anatomic structural outlines of 25 hemispheres normalized to either the AC or the PC. Simplified diagrams showing the region of densest overlap (median values) and extreme ranges also are presented. This feature allows a surgeon to visualize potential variations in the dimensions and locations of many structures conveniently (> Figure 25-9). But caution is in order; differences among individual brains tend to correct themselves when recalculated as fractional proportions of the IC distance (and thalamic height). For example, a point 6 mm anterior to PC in a brain with a 24-mm IC distance could be mapped in exactly the same relative (proportional) position if it were located 6.5 mm anterior to PC in another brain with a 26-mm IC distance. Both points are exactly one-fourth of the intercommissural distance anterior to PC. Thus, if one bears in mind the lessons of Talairach and others, the Van Buren and Borke atlas is a valuable tool.

1975, 1982: Tasker and Colleagues The human somesthetic thalamus, with maps for physiological target localization during stereotactic neurosurgery (Emmers and Tasker, 1975)

Owing to difficult accessibility during preparation of this chapter for the first edition, the present author omitted a discussion of Emmer’s and Tasker’s masterpiece of photographic, microscopic, physiological, and stereotactic spatial correlation of anatomy and function within the human somesthetic thalamus. This lavish folio edition presents the ten most useful (circa mid-to-late 1960s) somesthetic maps in the context of preCT, pre-computer – and of course, pre MRI – stereotactic surgery for Parkinson’s disease, other involuntary movement disorders, and intractable pain [14]. The book opens with a detailed description of Tasker’s stereotactic localization techniques using Leksell’s instrument, plus his own innovative attachments and accessories. Physiological localization is then explained as an iterative process during which each awake patient contributed new data to the aggregate responses obtained from electrical stimulation along similar (and nearby) trajectories in previous individuals. Tasker and others of his generation achieved extraordinary accuracy, confirmed by autopsy in cases with lesions inflicted for cancer pain, in adjusting the placement of thalamotomy lesions to account for variations in the dimensions and shape of the thalamus and adjacent structures. The particular planes selected for detailed presentation by the authors reflect the most commonly used stereotactic targets at the time. Portrait-quality black and white whole brain photographs – displayed larger than life size, one to a page – of a single brain with a 25 mm intercommissural distance and divided in the midsagittal plane included parasagittal plates at 9.0, 11.0, 13.5, 16.0, and 18.0 mm lateral to the midline cut from one hemisphere and coronal plates cut from the other hemisphere at 8.5, 10.0, 11.0, 12.5, and 15.0 mm posterior to the midcommissural point (at an angle 40 degrees frontal to the vertical plane perpendicular to the intercommissural line) (> Figure 25-10). Magnified and cropped microscopic sections from another 25-mm brain also were presented as full-page photographs to


25

. Figure 25-9 a and b. Sagittal variability maps in the 15-mm plane from Van Buren and Borke, normalized to AC and PC, respectively. Note the index mark (+) centered on AC in 9A and on PC in 9B. The solid outlines represent the regions of densest overlap of the labeled structures in specimens from 25 hemispheres. Broken outlines represent the extreme range of anatomical variability encountered among those same 25 hemisphere specimens (from [13], reproduced with permission)

identify the target sites and trajectories in each sample case corresponding to all ten sagittal and coronal whole brain sections. Finally, Tasker prepared a set of Woolsey-style [14,15] figurine diagrams to accompany each whole brain photograph and microphotograph. The site of electrical stimulation at 2-mm intervals along each trajectory was illustrated along with the type and distribu-

tion of somesthetic responses elicited during surgery (> Figure 25-11). Each three-figure suite graphically presented what a surgeon needed to know about the anatomy and physiology of the somesthetic thalamus and upper brainstem along commonly employed stereotactic trajectories.

359

360

25


. Figure 25-10 Parasagittal section 13.5 mm to the right of midline in a brain specimen having a 25 mm intercommissural distance. Short vertical hatch marks along the horizontal intercommissural line indicate AC and PC; the long vertical line indicates the midcommissural plane; the two horizontal lines above and below the IC line and the hatch marks at AC and PC approximate the boundaries of an accompanying photomicrograph that is not reproduced here. Amy, amygdala; CA, commissura anterior (AC); CC, corpus callosum; Cd, nucleus caudatus; CI, capsula interna; D, nucleus dentatus cerebelli; FH, fimbria hippocampi; GM, corpus geniculatum mediale; GP, globus pallidus; P, pes pedunculi; PCM, pedunculus cerebelli medius; TO, tractus opticus; II, ventriculus lateralis (from [16] (Tasker et al.) reproduced with permission)

Near the end of the volume, the pictures of nuclear models and homunculi representing the somesthetic thalamus, sculpted out of Styrofoam and clay, and photographed from different perspectives, appear quaint compared to the nearly instantaneous, high resolution computerassisted, three-dimensional graphic images now on display in any radiology suite or operating room (> Figure 25-12). Still, it is instructive to see and study an actual (in contrast to virtual) physiological model of the somesthetic thalamus in its totality. The value of this atlas transcends historical interest; it is the foundation for Tasker’s follow-on work, ‘‘The Thalamus and Midbrain of Man,’’ discussed in the next paragraphs. The thalamus and midbrain of man: a physiological atlas using electrical stimulation (Tasker, Organ, and Hawrylyshyn, 1982)

This 505-page 6- by 9-inch book by Tasker and colleagues does not contain a photographic stereotactic atlas like the other works described in this chapter [15]. Instead, and in light of his previous work, Tasker’s physiological atlas provides the most lucid English-language analysis of electroanatomic observations (based on 9,383 stimulation sites during 198 operations) relevant to stereotactic surgery on the thalamus and upper brain stem available to date. The 90-page miniatlas near the end of the volume depicts the results of stimulation mapping from the author’s clinical material in graphic form. The sites at which specific subjective experiences or observable phenomena were elicited are displayed on outline maps in the 2–20-mm sagittal planes based on the Schaltenbrand and Bailey atlas. Tasker’s elegant technique of anatomically and physiologically normalizing coordinates from different-sized


25

. Figure 25-11 Figurine maps of the peripheral field(s) for stimulation trajectories in six different patients superimposed on the same 13.5 mm lateral diagram. Each patient is identified in the coordinate diagram in the upper right-hand corner of the illustration by initials at the bottom of the trajectory ‘‘run’’ (Ao, Tim, Obe, F, Har, Mac). The vertical hatch mark at the left-hand side of the horizontal (intercommissural) line indicates the midpoint of the IC line. Numbers along the IC line, divided by four, equal the distance in millimeters behind the midpoint of the IC line where each trajectory crosses the IC line. The need to divide by 4 arises from the trajectories originally having been plotted on 4 life-size diagrams. ‘‘The limits of the projection field given by the patient with this initial localization of paresthesias at a relatively low stimulus intensity are outlined in solid black . . . (usually) 0.5–0.8 V with a 3 msec pulse duration at 100 Hz. After this, the stimulation intensity was increased in small steps . . . A change in the localized area was usually reported at a stimulus intensity which was twice its threshold. The area of the changed projection field is indicated by shading it with closely spaced lines’’ (Emphasis in original) (from [16] (Tasker et al.), reproduced with permission)

brains made the pooling of data possible from different patients. To perform the actual surgical procedure, he constructed sagittal brain maps for each patient, using a computer graphics program. The computer could expand or shrink selected atlas diagrams to match the patient’s intercommissural distance as determined by stereotactic

ventriculography. This was a refinement of Talairach’s system of proportionate coordinates, and a simplification of Schaltenbrand’s system of optical modeling. Later, once data from many patients (in the form of observations during electrode trajectory ‘‘runs’’ in a given sagittal plane) were available, Tasker superimposed the results on

361

362

25


. Figure 25-12 Homuncular composite sculpture of SI and SII. The open letter ‘‘B’’ indicates Plate B (from Figure 24) in the original text. Directional orienting symbols A, P, S, and I indicate anterior, posterior, superior, and inferior, respectively. Note ‘‘that the hand occupied an area that was approximately 1/4 of the total body area of the SI homunculus; whereas the shoulder of this homunculus occupied only 1/45 of the entire body . . . The relative size of the body parts of the SII homunculus was not as disproportionate as that of SI. . . . it resembled a reasonably realistic human figure’’ (from [16] (Tasker et al.), reproduced with permission)

composite maps (> Figure 25-13). In addition, during surgery or between steps in a multistage procedure, he transposed the anatomic boundaries of thalamic nuclei on the individual patient’s map according to the results of intraoperative stimulation. Consequently, important insights and generalizations emerged, greatly enhancing the surgeon’s ability to make rational decisions about what to do next during functional stereotactic procedures. Most important, Tasker emphasized that physiologically defined anatomy rather than blind obedience to atlas coordinates should determine the conduct of functional stereotactic operations. The preceding 80% of the book provides a detailed account of the remarkably stereotypical experiences that patients report and phenomena that neurosurgeons observe during stimulation mapping of the thalamus and midbrain. The reader learns how to identify responses that arise from stimulation of structures belonging to the dorsal column-lemniscus system, the spinothalamic pathway, the pyramidal and extrapyramidal systems, the auditory and vestibular systems,

. Figure 25-13 (a) Pooled data from Tasker’s early thalamotomy series, showing sites in the 13.5-mm sagittal plane at which electrical stimulation at any current threshold arrested the tremor of Parkinsonism, essential tremor, or cerebellar disease. (b) Data from Tasker’s series, showing the sites of thalamotomy lesions in the 13.5-mm sagittal plane made to relieve tremor in Parkinson’s disease and essential tremor (from [15], reproduced with permission)


the visual and oculomotor systems, and other structures. The last section provides postmortem anatomic correlations of electrode trajectories, stimulation sites, and lesion sites with records of intraoperative physiological observations in six patients with involuntary movement disorders or intractable pain. One might compare the process of functional stereotaxis to the navigation of previously charted but personally unfamiliar territory. In this sense, brain atlases are the maps, but the work of Tasker and colleagues is like the Michelin Guide. Both help the traveler to complete the journey successfully.

Traditional Stereotactic Atlases Currently in Print 1977: Schaltenbrand and Wahren Atlas for stereotaxy of the human brain with an accompanying guide (Schaltenbrand and Wahren, 1977)

This single-volume, oversized loose-leaf edition published in Germany more than 30 years ago is the last of the great stereotactic atlases of the twentieth century [6]. The ‘‘second, revised and enlarged’’ version of the Schaltenbrand and Bailey atlas represented, according to the authors, an effort to expand on the most clinically useful portions of the original work, debride the impractical or irrelevant material, and fit the finished product into a single volume. The authors were successful. They drastically reduced the number of unstained macrosections to 34 and eliminated the entire set of quadruple-foldout 20 Nissl-stained plates that occupied volume III of the first edition. Furthermore, the revised companion text that occupied volume I of the original atlas was delayed in publication until 1982, when it was released separately as a textbook [7]. The 1977 atlas reflects some of the stereotactic procedures that were in vogue at that time (analogous to the atlas by Afshar et al. [11]). For example, the myelin-stained transverse brain

25

stem and cerebellar series included 21 planes from the pontomedullary junction to the medulla (a span of 46 mm). Like all the other myelinstained microseries in this volume, each photographic section was presented at 4 magnification with a transparent overlay bearing anatomic legends and stereotactic coordinates. Surgeons interested in dentatotomy (for movement disorders), pontine spinothalamic tractotomy, mesencephalic tractotomy, and medullary trigeminal tractotomy or nucleotomy would have found this series of plates more helpful than those in the earlier edition – but still imperfect. Interest by the authors in ablative hypothalamic operations to control deviant sexual behavior led to carryover of hypothalamic Nissl-stained sections from the first edition. The ten 8 magnified sections, plus two anatomic key sections, occupy only two pages of the atlas. Other additions to the 1977 atlas include 25 color diagrams (on six pages) that summarize the radiographic and electroanatomic observations during stereotactic surgery on more than 300 patients. The Schaltenbrand and Wahren atlas contains only 34 macroseries photographs, all at 2 magnification and divided into three series as follows: 19 frontal planes from 57 mm anterior to 44 mm posterior to AC, five sagittal planes from the midline (0 plane) to 22 mm lateral to the midline, six horizontal planes from 18 mm above to 20 mm below the IC line from one brain, and four additional horizontal planes from 5 mm to 28 mm below the intercommissural line from another brain. The expanded interest in the horizontal unstained macroseries and the myelin-stained microseries was stimulated by the advent of CT and the authors’ foresight in recognizing the important role that axial imaging would play in the future. This time around, the authors cut all horizontal sections parallel to the intercommissural plane. In addition to the transverse myelin-stained brainstem series (21 planes) mentioned above, the three standard planes also were well represented, for a

363

364

25


total of 78 myelin-stained atlas photographs: 20 frontal planes from 16.5 mm anterior to 16.5 mm posterior to the midcommissural plane, 17 sagittal planes from 1.5 to 27.5 mm lateral to the midline, and 20 horizontal planes from 16 mm above to 9.5 mm below the IC line. Every neurosurgeon who performs functional stereotactic operations should have unlimited access to a stereotactic atlas. > Figure 25-14 reproduces Plate 43, brain LXXVIII, a myelinstained sagittal section 12.0 mm from the midline. This section probably holds the current world’s record – surpassing the 13.5 mm section from the 1959 edition – for having guided the most stereotactic operations (subthalamic nucleus region deep brain stimulation) for movement disorders. Although many older atlases are long out

of print, owing to constant demand, Schaltenbrand and Wahren’s 1977 volume is still available.

1988: Talairach and Tournoux Co-planar stereotaxic atlas of the human brain: three-dimensional proportional system: an approach to cerebral imaging (Talairach and Tournoux, 1988)

This volume, the next to last of Talairach’s stereotactic atlases, departs from the usual focus of such works. ‘‘In contrast to the majority of stereotaxic atlases that are primarily intended for the localization of the deep central nuclei, this atlas emphasized the interpretation of the vast cortical and subcortical spaces’’ [10]. The advances in high-resolution CT and MR imaging over the past 20–30 years and the accompanying resur-

. Figure 25-14 Plate 43, brain LXXVIII, myelin-stained sagittal section 12.0 mm from the midline, with ruled and labeled acetate overlay from the Schaltenbrand and Wahren atlas (10). In the modern era of MR image-guided deep brain stimulation of the subthalamic nucleus region (S th) for Parkinson’s disease, this atlas section probably holds the current world’s record – surpassing the 13.5 mm thalamic map from the 1959 edition – for having guided the most stereotactic operations for movement disorders (from [6], reproduced with permission)


gence of functional and anatomic stereotaxis (dealing with tumors and other structural lesions, and of course, deep brain stimulation instead of lesion creation) make such a guide through the borderland between functional and structural neurosurgery timely and informative. Mark Rayport observed in the translator’s foreword that the ‘‘images produced by these instruments [MRI and CT] have been utilized principally in the traditional manner of radiological interpretation: verbal description, identification of lesions.’’ For Talairach and coworkers, the vague, imprecise language, and frequent anatomy errors in MRI reports did not convey sufficient data for adequate neurosurgical planning and decision making. The potential pitfalls of traditional functional stereotactic operations are well known to experienced practitioners. However, nowadays many neurosurgeons with no training or experience in functional stereotaxis routinely perform anatomic stereotactic procedures such as biopsy, tumor resection, or radiosurgery. The planning and execution of such anatomic stereotactic procedures ideally should take into account possible immediate or delayed functional consequences. While some structures that surgeons wish to avoid, such as the optic chiasm and the midbrain tectum are obvious on imaging studies, other important structures, including the subcortical course of the pyramidal tract, the optic radiations, Forel’s fields, and the hypothalamic nuclei, are invisible or at least not obvious even on excellent-quality MRI. Talairach’s 1988 atlas provides neurosurgeons with a tool to help them navigate around and through such regions. In addition, by applying the lessons of this atlas to the interpretation of routine diagnostic imaging studies, one can achieve a high degree of accuracy in anatomic localization and clinical correlation. The authors begin with an exposition of Talairach’s proportional grid system in three dimensions, based on the length, height, and width of the whole brain. The orthogonal refer-

25

ence planes are based on the midline, the intercommissural plane, and the two verticofrontal planes that intersect the anterior and posterior commissures. As the authors explain, direct distance coordinates (in millimeters) vary widely from one brain to another. This is especially true the farther the point of interest lies from the IC line. Thus, Talairach and Tournoux divided the entire brain into cuboidal and rectangular prism-shaped parcels called ‘‘orthogonal parallelograms.’’ Each hemisphere is 9 major parcels in length (A-I along the axis of the IC line), 4 parcels wide (a-d along the transverse plane orthogonal to the midline and IC plane), and 12 parcels high (1–12 in vertical planes parallel to those defined by the commissures). The dimensions of each parcel are determined as follows: One-eighth of the distance between the IC line and the highest point of the parietal cortex and one-fourth of the distance between the IC line and the lowest point of the temporal cortex (parcels 1 through 12 in height); one-fourth of the distance from AC to the frontal pole, one-fourth of the distance from PC to the occipital pole, and the whole distance (subdivided into thirds) between AC and PC (nine parcels, A-I, in length); and one-fourth of the distance from the midline to the most lateral point of the parietotemporal cortex (4 parcels wide in each hemisphere). Even though this sounds complex when expressed verbally, things become clearer when one studies the diagrams (> Figures 25-15 and > 25-16) and remembers that each voxel represents a fixed proportion – not a distance – within the brain. The authors are careful to point out the limitations of this atlas, noting that ‘‘the millimetric values are valid for the brain presented here [only].’’ While this atlas is not among those one would consult before performing a procedure such as a thalamotomy for tremor, it is a valuable adjunct in planning or analyzing the effects of radiosurgery, or in planning surgical approaches to the deep hemispheric structures using stereotactic techniques. This work laid the foundation

365

366

25


. Figure 25-15 Illustration of Talairach’s space showing the brain divided into orthogonal rectangular prisms (Talairach calls them ‘‘parallelograms’’), the dimensions of which vary according to the principal axes of the brain. Each mini-volume is identified by three dimensions – indicated by a capital letter, a lowercase letter, and a number, e.g., A-d-1 for the shaded area in upper right-hand front corner) (from [10], reproduced with permission)

. Figure 25-16 Coronal (verticofrontal) section 20 mm behind AC, corresponding to plane E-3 in Talairach space. The original atlas illustration is in color. The boldface numbers refer to Brodmann’s cortical areas. GPrc, precentral gyrus; Ra, auditory radiation; Ro, optic radiation; other anatomic abbreviations are standard (from [10], reproduced with permission)


for Talairach’s final MRI and functional anatomy atlas published in 1993 [17], and inspired the present author to begin work in 1994 on an anatomical stereotactic imaging atlas of radiosurgical case studies [16].

Contemporary Stereotactic Atlases with Accompanying Digital Media 1998–2004: Mai and Colleagues Atlas of the human brain, second edition (Mai, Paxinos, and Assheuer, 2003)

The current version of the Mai atlas consists of two principal sections (in print) plus an accompanying CD and software [18]. A spiral-bound format with 9.5 13 inch (24 33 cm) pages makes this book especially convenient to carry to the operating room or radiology suite where it lies flat on the table, and folds easily to show one or two pages, and thereby occupies less space than any of the Schaltenbrand or Talairach volumes. Modest price, easy availability, portability, and the inclusion of a CD that contains digitized images and software all have contributed to the popularity of this atlas over the past 10 years. An introductory Preface and Materials and Methods section describes and illustrates in detail how the authors obtained, processed, and oriented the specimens for the anatomical and MRI preparations. Material included 17 heads – of which 11 were excluded owing to unexpected pathology, image artifacts, or other technical reasons. A ‘‘healthy, 25-year-old volunteer’’ contributed in-vivo MR images to the macroscopic imaging section of the atlas. Talairach’s method of spatial orientation and proportional boxes (voxels) was used and acknowledged. The first principal section, the Macroscopic Atlas, contains seventeen horizontal, fifteen coronal, and eight sagittal sections – each presented as a full page multi-modality plate consisting of several elements. Three pages of orienting diagrams introduce each series of macroscopic plates in the horizontal, coronal, and sagittal planes, respectively. Then, on each individual plate, small orienting

25

drawings of the whole-head (and in some cases, of the whole brain) indicate the plane and location of each individual section. These are accompanied by two small T-2 weighted MR images, and in some cases, by a small bone-windowed CT image – all in the same plane and at the same anatomic level as the anatomical section on the same page. Whole head specimens underwent post-mortem MR imaging followed by precisely oriented sectioning into approximately 1.0-cm-thick slices. Some of the CT images in the horizontal and coronal planes contain superimposed outlines of the vascular territories irrigated by major branches of the circle of Willis. Most of each page is occupied by an approximately 80% life-size (0.8) photograph of the whole-head and brain slice in-situ. On the facing page there is a comprehensively annotated artist’s tracing at the same scale. This mode of presentation helps the reader to become oriented when viewing only a single page or atlas plate at a time. Detailed identification of the extracerebral structures (muscles, orbital contents, paranasal sinuses, etc.) is a welcome addition. The largest section of the book, the ‘‘Microscopic Atlas’’ represents approximately threequarters of a century of detailed anatomical and radiographic study of one brain specimen from the Vogt collection – that of a 24-year-old man who died in 1929. This is where the Mai Atlas shines. Readers should bear in mind, however, that the Microscopic Atlas illustrates coronal sections from a single brain. The authors provide a complete bibliography and a summary of the previous and pertinent anatomical studies performed on this specimen. Four introductory pages diagram the orientation and precise location (in Talairach space) of the 69 coronal brain sections that span a distance of 60 mm anterior to the anterior commissure to 100 mm posterior to the anterior commissure – from the frontal pole to the occipital pole. Sections vary in thickness according to the anatomical complexity of each region. For example, the polar regions are sliced at 5.0–6.0 mm intervals, and the sections containing the hypothalamus are sliced at 0.6–0.7

367

368

25


mm intervals. Myelin-stained whole-mount sections are presented one per page, approximately 2.5 larger than life size. The facing page contains a richly labeled and annotated artist’s tracing of the section in Talairach space, and with ruled margins (> Figure 25-17). Each

cortical gyrus is labeled, as are deep gray matter nuclei, subnuclei, and white matter tracts. The detail presented here equals or surpasses that of any comparable work.

. Figure 25-17 Plate from the CD that accompanies the Mai et al. atlas [18]. This coronal section, in color on the CD, is at the level of the anterior commissure as indicated on the thumbnail inset in the upper right-hand corner. The horizontal and vertical 0 lines represent the intercommissural and midsagittal planes, respectively. Each numbered index line in the margins represents 1 cm; a millimeter scale is in the lower left-hand corner. Anatomical abbreviations are standard (from [18], reproduced with permission)


2007: Morel Stereotactic atlas of the human thalamus and basal ganglia (Morel, 2007)

This small ( Figure 25-18). The discussion of methods used to validate comparisons of nuclei and tracts between in vivo (or post mortem) MR imaging and measurements on preserved cut brain sections is even more valuable. This atlas-textbook provides important information to neurosurgeons regarding the limitations of other conventional atlases – and how to avoid pitfalls when ‘‘going by the book’’.

Conclusion The indications for guided brain operations have evolved over the past half-century (fewer lesions, more stimulation); and contemporary radiological techniques now provide magnificently detailed images of the brain itself, and its functional state – in contrast to the sometimes vague shadows used to navigate stereotactic procedures during the preMRI epoch. Elsewhere in this volume digital and computer-based stereotactic atlases are reviewed. Those products have long surpassed the tedious (in retrospect) printer-plotter based output of earlier generations in terms of speed, accuracy, and


utility. But for surgeons of each generation there is still something reassuring about opening one’s favorite atlas to the pages pertinent to the case at hand – and seeing the familiar myelin-stained sections and anatomical outlines right there on paper. Brain atlas books become friends – a relationship that is difficult to imagine with the computerresident atlases. This is not to diminish the value of computer-based or digital atlases, but perhaps it helps to explain the continued demand for new stereotactic atlases in print format, and for repeated press runs of the classic atlases. One barely notices the loss of computer atlas version 1.0 once version 2.0 is installed, but surgeons carefully pass their stereotactic atlas books from one generation to the next.

References 1. Spiegel EA, Wycis HT. Stereoencephalotomy, thalamotomy and related procedures. I. Methods and stereotaxic atlas of the human brain. New York: Grune and Stratton; 1952. 2. Spiegel EA, Wycis HT. Stereoencephalotomy. II. Clinical and physiological applications. New York: Grune and Stratton; 1962. 3. Baird RA, Spiegel EA, Wycis HT. Studies in stereoencephalotomy. IX. The variability in the extent and position of the amygdala. Confin Neurol 1960;20:26-36. 4. Benz RA, Wycis HT, Spiegel EA. Studies in stereoencephalotomy. XI. Variability studies of the nuclei ventralis lateralis thalami. Confin Neurol 1960;20: 366-374. 5. Schaltenbrand G, Bailey P. Introduction to Stereotaxis with an Atlas of the Human Brain. Stuttgart: Thieme; 1959.

25

6. Schaltenbrand G, Wahren W. Atlas for Stereotaxy of the Human Brain. Stuttgart: Thieme; 1977. 7. Schaltenbrand G, Walker AE. Stereotaxy of the Human Brain: Anatomical, Physiological and Clinical Applications. Stuttgart: Thieme; 1982. 8. Talairach J, Hecaen H, David M, et al. Recherche´s sur la coagulation the´rapeutique de structures sous-corticales chez L’Homme. Rev Neurol (Paris) 1949;81:1-24. 9. Talairach J, David M, Tournoux P, et al. Atlas d’Anatomie Stereotaxique. Paris: Masson; 1957. 10. Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain: 3-dimensional proportional system: an approach to cerebral imaging. Stuttgart: Thieme; 1988. 11. Afshar F, Watkins ES, Yap JC. Stereotaxic atlas of the human brainstem and cerebellar nuclei. New York: Raven Press; 1978. 12. Andrew J, Tomlinson, JDW, Watkins ES. A stereotaxic atlas of the human thalamus and adjacent structures; a variability study by J. Andrew and E. S. Watkins, with the collaboration and histological assistance of J. D. W. Tomlinson. Baltimore: Williams and Wilkins; 1969. 13. Van Buren JM, Borke RC. Variations and connections of the human thalamus. Berlin, Heidelberg: Springer; 1972. 14. Emmers R, Tasker RR. The human somesthetic thalamus, with maps for physiological target localization during stereotactic neurosurgery. New York: Raven Press; 1975 15. Tasker RR, Organ LW, Hawrylyshyn PA. The thalamus and midbrain of man: a physiological atlas using electrical stimulation. Springfield, IL: Thomas; 1982. 16. Coffey RJ, Nichols DA. A neuroimaging atlas for surgery of the brain: including radiosurgery and stereotaxis. Philadelphia: Lippincott-Raven; 1998. 17. Talairach J, Tournoux P. Referentially oriented cerebral MRI anatomy. An atlas of stereotaxic anatomical correlations for gray and white matter. Stuttgart: Thieme; 1993. 18. Mai J, Paxinos G, Assheuer J. Atlas of the human brain. 2nd ed. New York: Academic Press; 2004. 19. Morel A. Stereotactic atlas of the human thalamus and basal ganglia. New York: Informa Healthcare; 2007.

371

31 The Riechert/Mundinger Stereotactic Apparatus J. K. Krauss

The Riechert–Mundinger (RM) stereotactic system was one of the first stereotactic devices which became widely accepted and which had a major distribution, in particular in Europe. The original system was developed by Traugott Riechert in Freiburg in the late 1940s together with the physicist Wolff [1,2]. With this first device, however, the accuracy was rather limited. Soon thereafter Fritz Mundinger modified and improved the original system [3,4] which then became known worldwide as the Riechert–Mundinger system (> Figure 31-1). Over the decades Mundinger continuously modified the stereotactic device [5–11] which was manufactured and distributed until the 1980s by the Fischer Company. Incorporating always emerging technical developments such as the introduction of computer tomography or magnetic resonance imaging, considering the needs for expanded uses of stereotactic frames such as image-guided craniotomy and tumor resection, and managing the challenge with new indications, the RM system has maintained its place over the years until nowadays in contemporary stereotactic and functional neurosurgery [12–14]. One important step forward was the introduction of the Zamorano–Dujovny (ZD) arc development in the early 1990s [15,16]. This application transformed the RM system which is a translational system into a center-of-arc system using the Zamorano–Dujovny 3/8 arc. This transition implies also that the target calculation which is based primarily on a polar coordinate system with the RM system moves to target calculation primarily based on Cartesian coordinates with #


the ZD development (> Figure 31-2). For both systems the same base ring can be used. Neurosurgeons who worked for years with the frame therefore have the option to use both the advantages of the original RM system or the ZD development [14]. The RM system as well as its newer variant the ZD system are universal stereotactic systems which allow their use basically for all classical and new stereotactic procedures including tumor biopsy and interstitial curietherapy, drainage of cysts and catheter implantation, functional stereotactic surgery with radiofrequency lesioning or deep brain stimulation, image-guided craniotomy and stereotactically controlled tumor resection, stereotactic image fusion technology and stereotactic radiosurgery or radiotherapy. For decades the history and development of the RM system was firmly coupled with the F. L. Fischer company from Freiburg. Later the Leibinger company managed the development and distribution of the systems and the instruments. During the subsequent period when the RM and the ZD systems were managed by the Stryker company the major impetus was on the integration of the systems for radiosurgical and radiotherapeutic purposes. Nowadays, both systems are maintained and distributed by the Inomed company located in Teningen, Germany, a small town which is just a few kilometers away from Freiburg, the site where the original system was developed more than 50 years earlier. With the dedicated expertise of Inomed both systems have been expanded once more as exquisite tools for the performance of functional and stereotactic neurosurgery.

488

31

The riechert/mundinger stereotactic apparatus

. Figure 31-1 The classical Riechert–Mundinger system mounted on the phantom ring

. Figure 31-2 The translational principle on which the Riechert– Mundinger system is based (a) and the center-of-arc principle which is realized with the Zamorano–Dujovny system (b)

The Riechert–Mundinger Apparatus: A Translational System The Riechert–Mundinger apparatus basically consists of a circular base ring, an aiming bow and an instrument holder which sits on the aiming bow (> Figures 31-3 –31-5). The base ring has been made from different materials over the decades: steel, titanium, carbon fibre and ceramics. The most commonly used variant nowadays is made out of titanium. It is fixed to the patient’s head with four screws which are available in various lengths to accommodate for different head sizes (> Figure 31-6). Four fiducial plates are mounted on the base ring which are adapted according to the imaging technology. Using

current technology the system produces only very little artefacts with any imaging modality. As with any other stereotactic system it is important, of course, to have rectilinear alignment of the frame on the patient’s head for the purpose of stereotactic functional neurosurgery based on the anterior commissure/posterior commissure line [17], although deviations of pitch, yaw and roll can be corrected with modern planning software. The RM aiming bow is fixed along three points on the base ring which gives the system a high mechanical stability and target accuracy. The coordinates for the target point which have been defined according to polar coordinates can be checked with the help of a phantom system. This phantom system allows to detect miscalculations, however, it also allows easily to note errors in laterality and to detect even slight bendings of the guiding-cannulas or any other stereotactic instruments. The instrument holder accommodates any instruments used for stereotactic and functional stereotactic neurosurgery allowing easy readings within the submillimeter range. The main advantages of the RM systems are both its stability and its versatility, basically, any lesion in the brain can be reached from any entrance point.

The Zamorano–Dujovny Development: A Center-of-Arc System The Zamorano–Dujovny system uses the same base ring than the RM apparatus. The main difference between the two systems is the aiming bow, which along with the center-of-arc principle allows more flexibility in choosing the entry point during stereotactic surgery (> Figures 31-7 and > 31-8). Also, the instrument carrier has been modified in order to its altered functionality. Similar than it is the case with other


31

. Figure 31-3 The Riechert–Mundinger system mounted on a skull: anterior view

. Figure 31-4 The Riechert–Mundinger system mounted on a skull: lateral view

center-of-arc systems the distance from the arc to its center is 19 cm. The aiming bow is not a complete semicircle, but a 3/8 circle. With that regard, of course, it cannot be fixed at three points on the base ring as

it is the case with the RM system. Instead, the ZD aiming bow is fixed at one point either on the lateral aspect of the base ring or on its frontal aspect incorporating all three othogonal Cartesian axes. The fixation of the aiming bow also

489

490

31


allows to set and modify the coordinates in all three axes (x, y and z) during the operation which makes multifocal targeting quite comfortable. With this special construction care has to be taken to consider on which side the arc is fixed to the base ring in a given patient according to its laterality when bilateral surgery is performed. While any distance between 0 and 70 mm at an . Figure 31-5 The Riechert–Mundinger system with a contemporary instrument holder

angle between 0 and 70 can be chosen on the side where the aiming bow is fixed, the range for the x coordinate on the other side is limited to 20 mm, and the entry angle is between 0 and 20 . The instrument holder has been modified to carry any tools for functional stereotactic surgery including both electrical and mechanical microdrives. The CD development offers elegance and simplicity in its use coupled with high accuracy but with somewhat less stability than the original RM apparatus. It has expanded the scope of the RM system mainly by allowing its use as a centerof-arc system.

Tools for Stereotactic Surgery Both the RM and the ZD systems are supplied with a wide variability of hardware, targeting instruments and software. In particular, for use with magnetic resonance scanners an open stereotactic system has been developed (> Figure 31-9). The small diameter of the frame allows that it can be used even with small MR head coils. This particular frame is made out . Figure 31-6 Fixation of the base ring of the RM system to a patient’s head


31

. Figure 31-7 The Zamorano–Dujovny development demonstrating the fixation of the aiming bow to the base ring

. Figure 31-8 Set-up of the Zamorano–Dujovny development on a model patient’s head

of ceramics. A closing bow can be supplemented to provide additional stability when needed. Special interfaces are available to connect the base ring to the head rests of all CT and MR scanners. Both systems are delivered with a wide range of biopsy sets including Sedan type instruments or a mini-forceps as well as dedicated tools

for functional stereotactic neurosurgery. Microelectrode recording is possible with a microdrive allowing either exploration of a single trajectory through the central channel or exploration of five trajectories simultaneously (> Figure 31-10). It is not advisable to use heavy electric microelectrode manipulators with the ZD aiming bow

491

492

31


. Figure 31-9 The Zamorano–Dujovny system with an open base ring

. Figure 31-10 Dedicated mechanical microdrive for the RM and the ZD systems

since their weight might induce minor bending of the arc resulting in small yet relevant targeting inaccuracies. The classical radiofrequency electrodes have been developed further including monopolar and bipolar electrodes, and the unique chord

. Figure 31-11 The chord electrode which allows radiofrequency coagulation at a given distance from the target


31

. Figure 31-12 CT stereotactic control of electrode positioning with fiducial plates mounted on the base ring after implantation of four electrodes in a patient with dystonia

electrode which allows to modify and shape the target with radiofrequency lesioning without the need to reinsert the guiding-cannula and alter the target coordinates (> Figure 31-11). Radiofrequency lesions can be made with the Neuro N50 under thermal control. Both the RM and the ZD systems can be used with any planning software with minor modifications. Dedicated software which has been developed specifically for these systems is also available, the IPS software which can handle all Dicom material such as CT, MR, PET and X-ray data.

Conclusions Both the RM and the ZD system have been developed as ideal instruments for stereotactic and for functional stereotactic neurosurgery. I have used the ZD system regularly for deep brain stimulation for treatment of movement disorders and pain syndromes over the years. Some modification of instrument technology on the instrument holder has made it possible to use the same guiding-cannula for both microelectrode recording and implantation of the electrodes exactly at the sites which have been determined by the microelectrode recordings. Stereotactic control of the positioning of the electrodes is straightforward (> Figure 31-12).

Acknowledgments The support of Mrs Mattmu¨ller and Dengler from Inomed providing several of the figures is greatly appreciated.

References ¨ ber ein neues Zielgera¨t zur intrak1. Riechert T, Wolff M. U raniellen elektrischen Ableitung und Ausschaltung. Arch Psychiat Z Neurol 1951;186:225-30. 2. Spiegel EA. In memoriam, Traugott Riechert (1905– 1983). Appl Neurophysiol 1983;46:320-322. 3. Riechert T, Mundinger F. Beschreibung und Anwendung eines Zielgera¨tes fu¨r stereotaktische Hirnoperationen (II. Modell). Acta Neurochir 1955; Suppl 3:308–37. 4. Riechert T. Development of human stereotactic surgery. Confin Neurol 1975;37:399-409. 5. Birg W, Mundinger F. Computer calculations of target parameters for a stereotactic apparatus. Acta Neurochir 1973;29:123-9. 6. Mundinger F, Birg W. CT-aided stereotaxy for functional neurosurgery and deep brain implants. Acta Neurochir 1981;56:245. 7. Birg W, Mundinger F. Direct target point determination for stereotactic brain operations from CT data and the calculation of setting parameters for polar-coordinate stereotactic devices. Appl Neurophysiol 1982;34:387-95. 8. Birg W, Mundinger F. CT-guided stereotaxy with the Riechert-Mundinger apparatus for biopsy and interstitial curietherapy of intracranial processes. J Neurooncol 1984;2:280. 9. Birg W, Mundinger F, Mohadjer M, et al. X-ray and magnetic resonance stereotaxy for functional

493

494

31 10.

11.

12.

13.


and non-functional neurosurgery. Appl Neurophysiol 1985;48:22-9. Mundinger F, Birg W. The image-compatible RiechertMundinger system. In: Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus Nijhoff; 1988. p. 13-25. Mundinger F, Braus DF, Krauss JK, Birg W. Longterm outcome of 89 low-grade brain-stem gliomas after interstitial radiation therapy. J Neurosurg 1991;75:740-6. Mundinger F, Boesecke R. The Riechert/Mundinger apparatus. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. p. 73-8. Schrader B, Mehdorn HM. Operative Technik der tiefen Hirnstimulation. In: Krauss JK, Volkmann J, editors. Tiefe Hirnstimulation. Darmstadt: Steinkopff; 2004. p. 108-24.

14. Krauss JK, Grossman RG. Principles and techniques of movement disorders surgery. In: Krauss JK, Jankovic J, Grossman RG, editors. Surgery of Parkinson’s disease and movement disorders. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 74-109. 15. Zamorano L, Kadi M, Jiang Z, Diaz F. ZamoranoDujovny multipurpose neurosurgical image-guided localizing unit: experience in 866 consecutive cases of ‘‘open stereotaxis’’. Stereotact Funct Neurosurg 1994;63:45-51. 16. Zamorano L, Martinez-Coll A, Dujovny M. Transposition of image-defined trajectories into arc-quadrant centered stereotactic systems. Acta Neurochi Suppl 1989;46:109-11. 17. Krauss JK, King DE, Grossman RG. Alignment correction algorithm for transformation of stereotactic anterior commissure/posterior commissure-based coordinates into frame coordinates in image-guided functional neurosurgery. Neurosurgery 1998;42:806-12.

32 The Talairach Stereotactic System A. L. Benabid . S. Chabardes . E. Seigneuret . D. Hoffmann . J. F. LeBas

History of the Concept It is generally agreed that stereotaxy was invented in 1905 by Horsley and Clarke [1] who needed an accurate tool for electrode insertion and lesion making in laboratory animals. This laboratory equipment was later adapted for neurosurgical purposes by Spiegel and Wycis in 1947 [2], although the first human stereotactic apparatus was probably built in London around 1919 by Aubrey Mussen, who had worked with Clarke [3]. There is however a report in the French magazine L’Illustration in 1897 [4] that two operations were performed to remove head projectiles using a system very similar to the current ones, including a biorthogonal X-ray set-up of Crookes tubes and X-ray films. This surgery, reported by Marey in 1897 in front of the French Academy of Medicine could have been the first ever performed frame-based stereotaxy Since that time, a large variety of stereotactic frames have become commercially available. However, the differences between frames are more related to the industrial features than to the methodological principles, since a frame plays the role of ‘‘sugar tongs’’ holding the skull, and its content the brain, in a fixed position. All stereotactic frames share a common goal, which is to establish rigid relationships between the patients’ head and brain and the outer space, which contains surgical tools, such as cannulas, probes, electrodes, or larger systems such as X-ray tubes used either for diagnosis or for therapy, To achieve this goal, the frames are firmly anchored to the patient’s skull by several pins. Making frame positioning reproducible is easy to achieve with every type

#


of frame, at the cost of very few changes. Every frame has advantages and drawbacks. All of the specific features of each frame (such as goniometers) can be easily redesigned and adapted to the others, and. finally, all stereotactic strategies can be universal. Therefore, architectural differences between stereotactic frames, evolving with stereotaxy [5], express the main preoccupations of their designers and simultaneously, represent the solutions they considered best adapted to their needs. Most stereotactic systems are based on a polar or spherical (center-of-arch) approach, the purpose of which is solely to reach a point-like target (mainly in the basal ganglia for functional surgery), without endangering the structures encountered along the track. The first applications were in the study of abnormal movements (in order to confirm physiopathological hypotheses, either biochemical, electrical, or surgical) of pain and of epilepsy (in search of the putative pacemaker of centrencephalic epilepsies). One of Talairach’s original contributions was to try to determine the location of structures from radiological data, not only directly (such as Ammon’s horn, which makes a specific pattern in the ventricular occipital horn) but also indirectly (such as the amygdala, which is situated anterior to the temporal horn). In this perspective, during the 1950s Talairach at Sainte-Anne Hospital, in Paris, evaluated several coordinate systems, including that based on the corpus callosum, which was later readapted by Olivier [6] on the basis of MRI images. He finally chose the intercommissural line, drawn from the anterior commissure (AC) to the posterior commissure (PC).

496

32

The talairach stereotactic system

The second original contribution is undoubtedly the proportional grid system, which takes skull size into account. This provides an orthogonal system of grids, parallel or perpendicular to the AC-PC line, which respects the anatomic planes perpendicular to the X-rays used during the radiological examination. This is the rationale for the rectangular frame and the perpendicular approaches, which make possible the simultaneous exploration with a single electrode of the T2 temporal cortex and the amygdala or of the F2 frontal cortex and the cingular cortex. The theory behind this concept already appears in Talairach and Tournoux’s 1957 book on temporal epilepsy [7] which is very often quoted by Crandall [8]. In 1967, Talairach and coworkers extended these principles to the whole human cortex [9] in foresight of tridimensional imaging, long before radiologists even started to think about it. The atlas published in 1988 with Tournoux [10] is simply a colored and translated edition of the 1967 atlas, which is now commonly used as a reference by epileptologists as well as specialists in positron emission tomography (PET) and single photon emission computed tomography (SPECT). The identification of vascular patterns in 1977 by Szikla and colleagues [11] ‘‘was made possible because the proportional grid system enabled localization of the various cortical sulci. Talairach was also the first to perform stereotactic biopsies, taking a sample before introducing isotopes into lesions, in order to confirm the diagnosis. At that time, he considered biopsies of secondary interest and was essentially concerned with a tridimensional view of the brain. He devoted only two lines to biopsies at the end of his paper [12], which could explain why he is never quoted on this subject. The Talairach frame is the result of a rational attempt [9,10,13–15] to design a universal yet simple system, fulfilling these specific prerequisites: Patient placement in the frame should be reproducible in order to make it possible to divide a procedure into different stages while

still being able to take advantage of the data gathered during the previous stages Placement of the frame with respect to the X-ray system should also be reproducible for the same reasons. The X-ray system should provide a set of accurate two-dimensional (2D) projections of the cranioencephalic spatial object on films, with minimal magnification or parallax distortions, and with coherence of the Cartesian coordinates of a point in space on the two projections (which means that the vertical elevation Z of the point above the zero plane of the frame should be the same on both projections). The system should provide simple and safe means for the introduction of a tool into a designated target, without extensive calculations and with easy means of checking the safety of the chosen penetrating track, through a minimal opening of the skull limited to the size of the probes. Obviously, the recent development of computerized images and stereotactic softwares make these specificities less mandatory, as reformatting the 3D data has become easy and often transparent to the user. But the Talairach’s frame is more than a tool, it is the substrate of a methodology, if not a philosophy it lays the basis of a strategically thinking in 3D, based on 2D projections of the cranio-cerebral complex.

Description of the Talairach System The Talairach system comprises the frame itself (with the fixation pins), the double grids (which are used for calibration as well as tool introduction), and a long-distance X-Ray system. It is designed to fulfill the previously described requirements (> Figure 32-1): 1.

Reproducible placement of the patient is achieved by the use of a heavy frame base,


32

. Figure 32-1 The Talairach stereotactic system. The frame is mounted on a rotating circular support in the original Grenoble setup. X-ray images are provided by a fluoroscopic amplifier. Penetration of tools (electrodes, biopsy cannula) is made (A), using a robotized stereotactic arm (B). The various components evolve according to technological developments: the DIXI frame (C) replicates the main structure of the Talairach frame, the X-Ray amplifiers are replaced by flat digital detectors (D), and the prototype of the robot has evolved to the third generation of the Neuromate stereotactic robot (E)

which cannot be distorted by the mechanical stresses required by the procedure, and by four strong pins inserted into 2.5-mm holes twist-drilled through the full depth of the skull, held by verniers, the graduations of which arc recorded and saved, and can be replicated. These pins can be replaced in the same holes weeks or months later as long as these holes have not been obliterated by bone regrowth. It is usually easy, under local anesthesia, to make a small skin incision above the holes and to reinsert the pins. Their exact repositioning is achieved by reproducing the same vernier readings as during the initial placement. The visibility of these pins allows exact superposition of subsequent X-rays, showing a ventriculogram, angiograms, and

2.

the grids through which probes will be inserted. It is sometimes necessary to replace the patient in the same stereotactic position several times within a 2-week period. Semi permanent screws have been designed (Stevis, Sofamor, France), which are made of titanium – compatible with magnetic resonance imaging (MRI). They are inserted into a 4-mm drill hole and can receive a connecting piece with the frame vernier. Replacement is then easy, painless, and can be done in non sterile conditions, allowing the use of a compatible MRI frame. The accuracy of radiological measurements is based on perfect orthogonality of the X-ray beams in the anteroposterior and

497

498

32

4.


lateral directions, achieved by a specific installation of the frame at the center of a longdistance (3.5-m) X-ray machine set up in the operating room, using X-ray controls of the image of double grids placed on both sides of the patient’s head or using laser beam reflection on mirrors attached to the sides of the frame. The Talairach frame is positioned at the focus of a bidirectional X-ray system made of two tubes, the beams of which are orthogonal. A first vertical tube, located at the ceiling of the operating room, is used for anteroposterior views of the patient’s head in the supine position. The second horizontal tube is located along a wall of the room, for lateral X-ray views. Simple and safe introduction of tools is achieved by orthogonal approaches through the grids, placed on the sides of the frame base and perpendicular two by two. The tracks are therefore parallel to the X-ray beams used for the X-Ray examination (angiography and ventriculography) performed as a first step in the stereotactic procedure. In the absence of 3D computerized neuronavigation, this makes it possible to avoid vessels and enables high-precision positioning. The grids and grid-hole diameter (2.3 mm) are designed for specifically adapted screws, tubes for afterloading brachytherapy [16,17] or electrodes for stereoelec-troencephalography (SEEG) [18–21].

Like many other frames, the Talairach system provides the possibility of trajectory simulation and also of a double oblique track, using a multijoint probe or oblique grids. These accessories, however, were not well suited to polar approaches, which are achieved in a much more practical manner using the center-of-arch systems. Sedan’s [22] and Scerrati’s [23] goniometers were conceived and designed to provide the Talairach system with the advantages and flexibility of the center-of-arch systems, allowing easy and

precise access via oblique approaches to targets near the midline [22]. Sedan’s goniometer is made of a carrier moving back and forth and mounted on two lateral poles of the frame. This carrier holds a rod, which can rotate with a sagittal angle b and move laterally. On the medial end of this rod is mounted a sector on which the probe holder can be set up with a frontal (coronal) angle a. Correspondence between the x, y, and z cartesian coordinates of a point P; the a and y angles of the r, a, and y spherical (or polar) coordinates: the b and g angles read on the X-rays, and the b and e angles set up on the goniometer are given by the following simple equations (> Figure 32-2), where r = OP: x = r.cosa.siny y = r.cosa.cosy z = r.sina

b = arctan(tana/cosy) g = arctan(siny/tana) e = arctan [x/(x2 + z2)1/2]

Scerrati’s goniometer is a center-of-arch system [23]. Similarly, all calculations used for the center-of-arch system are applicable to this goniometer [24].

Talairach Frame Setups In all cases, long-distance X-rays contribute to the basic concept of the Talairach system in order to avoid image distortion. Computer correction of the image distortion is achieved within the neuronavigation softwares currently available, which are based on digital images, and allow using X-ray setups, with a shorter distance. The frame by itself can be mounted, as can other frames, in several different manners, as below: 1.

The Sainte Anne setup is an all-mobile system, with a motorized operating table, two X-ray tubes mounted on a motorized ceiling arch, and on a vertical lateral pole with a laser beam centering system. All patient


. Figure 32-2 Spherical and Cartesian coordinates of a point P. A point P may be represented by a set of three projections on the axes X, Y.Z (Cartesian coordinates) or by a vector OP and two polar angles (b and g)

4.

32

Like any other type of frame, the Talairach stereotactic frame can be the basis of any robotized system [25–29], which requires the head to be firmly held in a fixed position. This setup is currently used in Grenoble [25], even though the various compounds have evolved and benefited from technological progresses.

Stereotactic Practice with the Talairach System Routine Procedure

2.

3.

positions are possible, making this system especially useful for ventriculography. The Rennes setup includes an isocentric mobile seat (CGR Isocentrix), which is an ideal system for permanent centering of the frame and head whatever the position of the patient. The Grenoble setup of the frame is inexpensive and practical. The frame is mounted on a rotating holder (> Figure 32-7) coaxial to the patient, fixed on a solid-state base screwed onto the floor of the operating room at the focus point of a permanent biorthogonal X-ray system. The sitting position is not possible and complete examination of the ventricular system is obtained by rotating the patient around his or her longitudinal axis.

The patient can be placed on the frame under general or local anesthesia. Standard X-rays are taken with the double grids, which localize the central beam and display the various grid holes through which the penetrating track will be made. Angiograms and ventriculograms are performed and the ‘‘synthetic’’ diagram is then made. Computed tomography (CT) and MR1 data can be reported on this diagram, from which the final target may be determined. The penetrating track can be chosen through the grid hole that best corresponds to the target, or coordinates can be measured to direct either a Sedan’s or Scerrati’s arch or to drive a robotized system. During the penetration step, X-rays are taken that check the position of the inserted probe or of the recording or coagulating electrode, biopsy cannula, brachytherapy tubes., depth electroencephalographic (SEEG) semipermanent electrodes, ventriculoscope, and so on. These positions can, in turn, be reported on the diagram. A document summarizing all data specific to the case is progressively built up and may be used for further therapeutic steps, such as radiotherapy or cortectomy planning. Location of functional structures can be added to this diagram, using the proportional grid system.

499

500

32


Overlay of Stereotactic Neuroradiological Modalities

Data Processing in the Talairach System

The pins inserted into the skull are in a fixed and reproducible position and their appearance is the same on every X-ray picture, taken during the same or different stereotactic sessions provided that the pin-holding verniers are set at identical values each time. Therefore, by matching the pin projections on the X-ray pictures, the Talairach system has the major advantage of providing the possibility of superimposing different X-Ray modalities, such as angiograms, ventriculograms, or any other kind of image acquired in stereotactic conditions (> Figure 32-3).

Proportional Grid System Talairach has developed a proportional anamorphosis procedure [9,12,13], based on the thirdventricular landmarks (anterior and posterior commissures) and on the inner skull contours in order to normalize every individual brain on a standard diagram and eliminate individual anatomic variability (> Figure 32-4). Recognition of various structures (cortical sulci and lobes, white matter bundles, as well as basal ganglia substructures) can be done using this proportional

. Figure 32-3 X-Ray images acquired in stereotactic conditions. All the steps of the stereotactic procedure yield X-rays, showing the ventricles, vessels, and position of the probes (during biopsies or during implantation of electrodes or of cannulas). On each X-ray, certain features are always visible: pins, fiducials and even the skull bone. These can be used to match X-rays and to draw on a common diagram all the relevant data of every modality. (a) Position of the deep brain EEG electrodes. (b) Drawing of arterial vessels, superficial (continuous lines), and deep (dotted lines). (c) Superposition of vessels and sulci from MRI. (d) Planning of temporal resection


32

. Figure 32-4 Proportional grid system: the borders or the rectangle are tangent to the inner table of the skull, parallel and perpendicular to the intercommissural AC-PC line. Various structures, such as sulci, may be located using the grid parcellation. AC, PC, anterior and posterior commissures; SS, sylvian sulcus; STS, superior temporal sulcus; LFS, lower frontal sulcus; RS, rolandic sulcus; CS, calcarine sulcus; MF, motor fibers; PT, pyramidal tract; upper diagram, lateral; lower diagram, anteroposterior

anamorphosis [9,12,13]. The predicted location of cortical sulci as compared with their actual position as shown by MRI has recently been validated [30].

Data Computation in the Talairach System The Talairach system is particularly adapted to calculations developed to correct spatial distortions and provide corrected coordinates for accurate probe placements. X-ray images in stereotactic conditions provide spatial localization of points within the cerebral space, but this localization is inaccurate because of two phenomena: 1.

Magnification. Magnification depends on the respective distances to the film, of the point of the cerebral space (d) and of the X-ray tube (D). The magnification coefficient that enlarges every measured distance between two points in the space is: G = D/(Dd). Our setup, in which tubes are 3.5 m away

2.

from the center of the frame, achieves a low magnification ratio of 1.05. Correction of parallax errors. Parallax depends on the distance of a given point in the cerebral space to the axis of the X-ray beam, which is the zero point. Every point that has, with respect to this central X-ray, coordinates x, y, will actually have coordinates X, Y on the film. These are X = G x and Y = G y

This must be taken into account in calculating the setup parameters of the frame. Precise data regarding the central X-ray beam, perpendicular to the frame faces, can be determined on the doublegrid image where the central beam passes through similar holes on both grids (> Figure 32-5). The simplest situation corresponds to a central beam placed on the area of interest and centered on the target. When this central beam is centered at distance from the target, its actual position is used for exact correction of the parallax distortion for any point in the brain [30].

501

502

32


. Figure 32-5 (A) lateral X-ray of a grid with an array of 27 31 holes. (B) interference pattern of the two grids. One may recognize the central beam (cross). The next interference corresponds to a shift of one hole spacing (3 mm) from the grid on one side to the other one

Detection of Vascular Injury Along a Double Oblique Biopsy Track With any penetrating trajectory into the brain there is the risk of encountering a vascular structure; the greatest risk occurs during biopsies [12]. The Talairach grid system is mainly set up for orthogonal, frontal, or lateral approaches. Biopsy tracks performed through the grids are aligned along the X-ray axes and provide the safest procedures, since it is possible to check on the corresponding X-ray images that the projected track, which appears as a point, does not correspond to any projection of a vessel. In the case of double-oblique approaches, the problem of detecting vascular collisions is not as easy to solve, despite the fact that CT-guided stereotactic biopsies without angiographic control are popular [8,22,24,31,32]. Effective solutions must be found, and some have already been designed and used. The intrinsic features of the Talairach system has led Szikla et al. [11,17] to develop a routine procedure proven to be effective and easy to perform without any computation, but which can be easily computerized and automated. Provided that the two X-ray beams are orthogonal, a given point in the brain appears on X-rays as two pairs of coordinates (x, z) on the frontal view and (y, z) on the lateral view, z being the same in both pairs (> Figure 32-6). Therefore,

projections of the intersection of a putative track with a vessel must have the same z value (as measured on X-ray films from the base plate of the frame or from any other reference plane) on both lateral and frontal planes (> Figure 32-7). Obviously, the reciprocal is not true, and it may happen that lateral and frontal intersections having a same z value do not correspond to the same vessel: In these cases of false collision. The decision between true and false collision is made by the surgeon’s expertise. This method of collisiondetection, which can be computerized on digitized angiograms. is much easier and faster to achieve but less elegant than true collision detection without false-positive points provided by real 3D reconstruction of the vascular network. Another approach is derived from the ‘‘floating line’’ concept [33]. A specially built stereocomparator features two movable lines on transparent grids, applied onto two stereoscopic angiograms and representing the projections on these tilted angiograms of a theoretical line in brain space. Observation of this line through the stereocomparator allows the surgeon to check for eventual collision of the line with vessels and eventually to change it. The Talairach system provides another approach that has been used profitably in routine practice [17,34] to recognize the in-depth position of the vessels using small-angle double-incidence


. Figure 32-6 X-ray setup. Any point of the brain with xi, yi, zi coordinates will have Xi, Yi radiological coordinates with a magnification coefficient G(xi) that depends on the geometry of the system. When the frame (and the head) is rotated by 5 around the patient’s axis (lower figure), the spatial coordinates of point P become x0 i, y0 i, and zi remains unchanged. Its radiological coordinates become X0 i, Y0 i, and Zi

32

. Figure 32-7 Vascular collision between a vessel and a track must have the same Z altitudes on frontal and lateral X-rays. In this case B and E correspond to a probable vascular collision

. Figure 32-8 Superposition of two X-ray angiograms taken with a 5 tilt angle of the lateral X-ray beam (obtained by a 5 tilt of the frame-head ensemble). Coincidence of vessels can be obtained only for limited segments situated in the same plane. On this figure, only the pericallosal artery (midline plane) is matched, while more superficial branches of the sylvian artery are shifted. A different shift of the films relative to each other would lead to the superposition of the images of vessels situated in a plane at a different depth

angiograms (SADIA) taken under a 5 tilt angle, which corresponds to the natural binocular vision angle. One may use a stereocomparator or, with some training, it is possible to squint and obtain a 3D perception of the vascular network. One may also superimpose the two angiograms and try to make the vessels correspond. Coincidence of the two images of the vessels is possible only for those that are in the same plane perpendicular to the X-ray axis (> Figure 32-8). Slightly sliding the films one over the other will change this ‘‘coincidence plane’’, and display another array of vessels

situated within it. This technique is easily used in daily routine to evaluate the depth of vessels projecting on a proposed trajectory. Obviously, the approach described above can be formally demonstrated and could be used as a possible basis for 3D reconstruction [35]. Consider lateral views taken as SADIAs. Every point P

503

504

32


of the brain is assigned a triplet of coordinates (x. y, and z) in brain space, a pair of coordinates (Y, Z) on the regular lateral view film, and (Y, Z0 ) on the lateral view film of the 5 tilted head. Therefore, x corresponds to the ‘‘depth’’ of a point along an axis Ox perpendicular to the film plane. When the two films are superimposed with a given shift d with respect to an arbitrary reference (Y + d = Y0 ), two sets of points belonging to the two films are placed in coincidence. One may easily demonstrate [35] that there is a relationship between d and x, which is dependent on y (> Figure 32-6). When superimposition of the films is achieved, some structures, such as vessels, can be matched on both films when there is a shift equal to d. Therefore: x ¼ 11:9d 0:04Y The depth x can therefore be calculated for all points of the film that are situated at the coordinate Y and coincident to their homologous projection on the tilted film when the shift is equal to d. A paradigm can be derived from this procedure. A complete set of coordinates is therefore generated and, when displayed, provides a 3D reconstruction of the vascular network. It is clear that these methods of computations were extremely precise and useful at the time of the conception of the Talairach’s system, and are nowadays obsolete: current methods of modern imaging resident of most of the neuronavigation softwares associated to commercial stereotactic frames provide an efficient solution to these problems and allow interactive safe navigation on the MRI or CT images

Connection of a Stereotactic Frame to a Computerized Imaging System The original Talairach frame does not have localizers designed for MRI or CT examinations in stereotactic conditions. Moreover, its metallic composition makes it incompatible with MRI, and the pins verniers are too clumsy to fit easily

within the MRI gantry. Several solutions have been proposed to overcome this problem. Computed Tomography

For CT data obtained in stereotactic conditions using other MRI-compatible frames, specific adapters can be designed. Sedan has adapted the Leksell frame system in which the patient is initially set up and CT examination is performed using the localizers and software developed for this system. While still on the Leksell frame, the patient is then transferred to the Talairach frame, using a specifically designed adapter. We have adapted the Fischer-Lcibinger and then later the CRW frames and we simply transfer the vernier values from one frame to the other. When CT examination has been done under regular circumstances, several methods help in reporting the shape of the lesion in terms of the stereotactic diagram, making possible the use of this information for stereotactic procedures [35,36]. Magnetic Resonance Imaging

It has been stated that MRI cannot provide a precise spatial localization because of its nonlinearity [37,38]. This is actually partly wrong: precision is a matter of tuning the system correctly. The easiest way is to enlarge the images at the scale of the stereotactic pictures and to match them to similarly visible features and anatomic structures. Sedan et al. [30] had designed a television-based system that can pick up MRI parasagittal views and redisplay them, using a variable gain along the X and Y axes. The recent MRI systems can actually display hard copies at any desired magnification. Provided that MRI gradients are properly checked and adjusted if needed, MRI images are used by superimposition of a calibration grid positioned identically on each picture. This provides a composite picture featuring all relevant data, such as the inner contour of the skull, the coronal suture, depression of the torcular, the ventricular system and essentially the third ventricle, the aqueduct of Sylvius and the fourth ventricle, the rostrum and sple-


nium of the corpus callosum, and sometimes the siphon of the carotid artery. All these structures are visible on the stereotactic ventriculogram and angiograms, providing a stereotactic diagram that can therefore be matched to the MRI data, particularly since all these image modalities are digitized and can be therefore numerically processed and the image fusion and 3D reconstruction fully computerized and automated. Digital Radiology

Digital subtraction angiography (DSA) and ventriculography are replacing conventional X-ray films. These digital radiological images are easily processed and matched with other image modalities. Target coordinates are then precisely and quickly obtained and may be used to drive a

32

computerized or robotized system. Two flat digital detectors (Pixray, Bioscan, Geneva, Switzerland) are mounted on supports, solitary to the rotating stand holding the frame, and can be move away using revolving hinges, which allow them to give more working space around the patient’s head, particularly during SEEG procedures where implantation of the depth electrodes require a large free access on both sides of the head at the time of electrode insertion. X-ray images can be taken easily both on lateral and antero-posterior views, and immediately observable on the flat digital screen disposed close to the operating field, to be observed by the surgeon, during the introduction of probes, as well as during contrast angiography or ventriculography. Moreover, this

. Figure 32-9 The new version of a Talairach derived stereotactic frame. (on the right). (a) X-ray angiolocalizer with four plates bearing four opaque fiducials each. (b) The modified Talairach frame mounted on the rotating plate of the Grenoble set-up. (c) The MRI localizer including N shaped fiducials (being filled with copper sulphate. (d) Frame mounted on the rotating stand and equipped with the two orthogonal flat digital X-ray detectors (one for lateral X-ray pictures, and two for antero-posterior X-ray pictures.) (e) X-ray command console (1), computer screen (2) and central power unit (3) for control of the digital data acquisition and processing system

505

506

32


strongly decreases the cost of the investigation (X-ray films are no more needed) and its duration (film processing is also suppressed). The visualization of the fiducials on the angiolocalizer allows precise matching of subsequent series of images taken even at several days intervals, and is used by the robot computer to match these data with other types of digital data, such as MRI, but probably in the future, SPECT, PET and MEG data. Design of a New Talairach Frame

To circumvent the non compatibility of the original Talairach’s frame, and to make it still compatible with the basic principles of the Talairach system, we have redesigned it (Universal Talairach frame, DIXI, Besancon, France), keeping the structural design (heavy rectangular base, non deformable, able to receive grids and other mechanical systems or arches fixed on the four corner

posts), but adding new features allowing multimodal digitized imaging (CT and MRI localizers, fiducials for neuronavigation), and coregistration with the navigation software of robotized systems such as the last version of the Neuromate (Schaerer-Mayfield, Lyon, France). This frame is currently available and can be adapted to various stereotactic tool holders (from the robotized arm Neuromate to center-of-arch systems and goniometers from the industry). The fixation using transcranial pins has been kept, as well as the possibility to reposition them in bone screws previously used, allowing the repositioning of the patients along a sequence of subsequent steps (for instance ventriculography under general anesthesia, stereotactic MRI and CT in awake condition in the same evening or the day after, the implantation of deep brain stimulation electrodes 2 or 3 days

. Figure 32-10 Neuronavigation windows of the IVS Voxim software exhibiting the two orthogonal X-ray views of the target determination on the ventrieulogram (Subthalamic Nucleus), the coronal MRI image at the level of the target, with superimposition of the implantation tracks, and the 3D head reconstruction showing the image of the N shaped fiducials


32

. Figure 32-11 Pre-implantation control procedure of the accuracy of the placement by the robot of a ‘‘phantom’’ probe: (a) and (d) Ventriculographic determination of the target (coronal and lateral views). (b) and (e) Images of the ‘‘phantom’’ probe in position according to the planning. (c) and (f) Superimposition of the two sets of digital images showing the exact projection of the tip of the probe on the target

later, a second, postoperative, MRI in stereotactic conditions 2 days later, after which the screws can be removed, a few days before the implantation of the programmable stimulator under general anesthesia). One may take advantage of this easy and precise repositioning to check the accuracy of the targeting done during the preplanning session: when the targets and trajectories have been chosen and set, using the neuronavigation software, which had fed the robotized arm controller with the appropriate data of the robot to move and reach the desired position, one can perform a simulation or shame procedure just before setting the patient on the frame using the replaceable pins with same readings of the verniers as during the ventriculographic step: The robot is launched and set to reach its position, determined by the target coordinates; a ‘‘phantom tool’’, made of a rod is set in place instead of the electrode guidance system, the length of which is set a the same length than the implanted electrode; digitized X-rays images are taken,

which can be digitally superimposed to the digital scheme of the target: the correspondence must be complete, and any shift is easily detected, avoiding errors or allowing making corrections. The phantom rod is then withdrawn; the robot is retracted to a standby temporary position, the patient is reinstalled on the frame. The robot can therefore be sent again in working position and the stereotactic procedure of electrode implantation can be safely performed with minimal risks. (> Figures 32-9– > 32-11).

Conclusion The main characteristic of the Talairach system is that it renders compatible all procedures (diagnostic and therapeutic) performed on the frame during the same or during subsequent sessions, which may be separated from each other by weeks or even months. It is also designed to provide minimally distorted numerical spatial data and to allow corrections of these distortions. The

507

508

32


Talairach system has been the basis of a rational approach, taking advantage of the orthogonality of X-ray incidence to define precisely the position of the targets and vascular structures within the brain with respect to the coordinates of the frame system. These features have been used to develop methods of computation accessible to surgical teams with little or no computational means but also applicable to automated software. The stereotactic Talairach frame is suited for connection with spatially guided and computer-assisted robots, as it provides a basic spatial reference that is easy to integrate into a routine for driving a robot toward a spatially defined target.

References 1. Horsley VA, Clarke RH. On the intrinsic libels of the cerebellum, its nuclei and its effect tracts. Brain 1905;28: 12-29. 2. Spiegel EA. Wycis HT. Marks M. Lee A. stereotactic apparatus for operations on the human brain. Science 1947;57:164-7. 3. Picard C. Olivier A. Bertrand G. The first human stereotaxic apparatus: the contribution of Aubrey Mussen to the field of stereotaxis. J Neurosurg 59:67-36. 4. Remy and Contremoulins G. Le chercheur de projectiles. L’Illustration 1897;55:423. 5. Gildenberg PL. Whatever happened to stereotactic surgery? Neurosurgery 1987;20:983-7. 6. Olivier A. Extratemporal resections. In: Engel J, editor. Surgical treatment of the epilepsies. 2nd ed. New York: Raven Press; 1993. p. 489-500. 7. Talairach J, David M, Tournoux P. Exploration chirurgicale stereotaxique du lobe temporal. Paris: Manon et Cie; 1958. p. 123. 8. Crandall PH. Cortical resections. In: Engel J, editors. Surgical treatment of the epilepsies. New York: Raven Press; 1987. p. 377-404. 9. Talairach J, Ajuriaguerra JD, David M. A propos des coagulations therapeutiques sous-corticales: etude topographique du systeme ventriculaire en fonction des noyaux gris centraux. Presse Medicale 1950;58:697-701. 10. Talairach J, Tournoux P. Coplanar stereotaxic atlas of the human brain: 3-dimensional proportional system: an approach to cerebral imaging. Stuttgart: Thieme Medical Publishers; 1988. 11. Szikla G, Bouvicr G, Hori T, Petrov V. Angiography of the human brain cortex. New York: Springer-Verlag; 1977.

12. Talairach J, Ruggiero G, Aboulker J, David M. A new method of treatment of inoperable brain tumors by stereotaxic implantation of radioactive gold: a preliminary report. Br J Radiol 1955;28:62-74. 13. Talairach J, Ajuriaguerra J de, David M. Etudes stereotaxiques des structures encephaliques profondes chez I’Homme Technique, interet physiologique et therapeutique. Presse Med 1952;28:605-9. 14. Talairach J, David M, Tournoux P, et al. Atlas d’anatomie stereotaxique des noyaux gris centraux. Paris: Masson; 1957. 15. Talairach J, Szikla G, Tournoux P, et al. Atlas d’anatomie stereotaxique du telencephale. Paris: Masson; 1967. 16. Benabid AL, Chirossel JP, Mcrcier C, et al. Removable, adjustable and reusable implants for stereotactic interstitial radiosurgery of brain tumors. Appl Neurophysiol 1987;50:278-80. 17. Szikla G, Peragut JC. Irradiation interstitielle des gliomes. In: Constans JP, Schliengcr M, editors. Radiotherapie des tumeurs du systeme nerveux central. Neurochirurgie (Suppl) 1975;21:187-228. 18. Bancaud J, Talairach J, Bonis A, et al. Stereo-electroencephalographie dans l’epilepsie. Paris: Masson; 1965. 19. Bouvier G, Saint Hilaire JM, Giard N, et al. Depth electrode implantation at Notre-Dame Hospital, Montreal. In: Engel J Jr, editor. Surgical treatment of the epilepsies. New York: Raven Press; 1987. p. 589-94. 20. Olivier A. Double-headed stereotaxic carrier apparatus for insertion of depth electrodes. J Neurosurg 1986;65: 258-9. 21. Peters TM, Clark JA, Olivier A, et al. Integrated stereotaxic imaging with CT. MR imaging, and digital subtraction angiography. Radiology 1986;161:821-6. 22. Sedan R, Duparet R. Stereómetre adaptable au cadre ste´reótaxique de J Talairach. Neurochirurgie 1968;14: 577-82. 23. Scerrati M, Fiorentino A, Fiorentino M, Pola P. Stereotaxic device for polar approaches in orthogonal systems (technical note). J Neurosurg 1984;61:1146-7. 24. Colombo F, Angrilli F, Zanardo A, et al. A universal method to employ CT scanner spatial information in stereotactic surgery. Appl Neurophysiol 1982;45:352-4. 25. Benabid AL, I.avallee S, Hoffmann D, et al. Computer driven robot for stereotactic neurosurgery, In: Kelly P, Kail A, editors. Computers in stereotactic neurosurgery. Cambridge. MA: Blackwell; 1992. p. 330-42. 26. Goerss SJ, Kelly PJ, Kail BA, Alker GJ. A computed tomographic stereotactic adaptation system. Neurosurgery 1982;10:375-9. 27. Kail BA, Kelly PJ, Goerss SJ, Earnest F IV. Crossregistration of points and lesion volumes from MR and CT. Proceedings of the Seventh Annual Meeting of Frontiers of Engineering and Computing in Health Care. 1985; 935-42.


28. Kelly PJ, Kail BA, Goerss SJ. Transposition of volumetric information derived from computed tomography scanning into stereotactic space. Surg Neurol 1984;21:465-71. 29. Lavallee S. Gestes medico-chirurgicaux assiste´s par ordinateur. The`se sciences mathe´matiques. Grenoble. France: University Joseph Fourier; 1989. 30. Steinmetz H, Fiirst G, Freund HJ. Cerebral cortical localization: application and validation of the proportional grid system in MR imaging. J Comput Assist Tomogr 1989;13:10-19. 31. Brown RA. A computerized tomography-computer graphics approach to stereotactic localization. J Neurosurg 1979;50:715-20. 32. Mundinger F, Birg W, Klar M. Computer-assisted stereotactic brain operations by means including computerized axial tomography. Appl Neurophysiol 1978;41:169-82. 33. Cloutier L, Nguyen DN, Ghosh S, et al. Simulator allowing spatial viewing of cerebral probes by using a floating

34.

35.

36.

37.

38.

32

line concept. Symposium on Optical and Electro-Optical Applied Science and Engineering. Cannes, France. 1985. Szikla G, Bouvier G, Hori T. Localization of brain sulci and convolutions by arteriography: a stereotactic anatomo-radiological study. Brain Res 1975;95:497-502. Benabid AL, Lavallee S, Hoffmann D, et al. Computer support for the Talairach system. In: Kelly P, Kall A, editors. Computers in stereotactic neurosurgery. Cambridge, MA: Blackwell; 1992. p. 230-45. Nguyen JP, Van Effentere R, Fohanno D, et al. Methode pratique de reperage spatial des petites neoformations intracraniennes a partir des donnees de la tomodensitometrie. Neurochirurgie 1980;26:333-9. Schad L, Loll S, Schmitt F, et al. Correction of spatial distortion in MR imaging: a prerequisite for accurate stereotaxy. J Comput Assist Tomogr 1987;11:499-505. Wyper DJ. Turner JW. Patterson J, et al. Accuracy of stereotactic localisation using MRI and CT. J Neural Neurosurg Psychiatry 1986;49:1445-8.

509

37 BrainLab Image Guided System J. F. Fraser . T. H. Schwartz . M. G. Kaplitt

Stereotaxic neurosurgery has always been somewhat dependent upon new technologies. The ability to safely navigate to and alter the physiology of deep brain structures that are inaccessible by open surgery invariably benefits from technologies which combine maximal information regarding functional neuroanatomy with advanced three dimensional navigational tools. From the development of the first modern stereotaxic frame through incorporation of detailed anatomical targeting with CT and MRI, stereotaxic neurosurgery has benefited from early adoption of new technologies. In recent years, this has resulted in widespread use of computer-based image analysis and navigational guidance systems among even more seasoned practitioners who had long relied on homegrown methods. Although several commercial packages are currently available, and an academic manuscript is not intended to promote a particular vendor, nonetheless detailed evaluations of the major systems can provide valuable information to investigators who are considering entering this field as well as to those experienced surgeons who may be less familiar with the details of each system. Here we will review the history and current applications of the technology offered by one of the popular current vendors, BrainLab.

Neuronavigation: History, Principles, and Practice Although details of the history of stereotaxic surgery are likely to be reviewed elsewhere in this volume, understanding the tenets that guide the use and development of modern

#


neuronavigational tools such as BrainLab can benefit from a brief review of historical context. In the mid-twentieth century, stereotaxy in neurosurgery was focused upon attention to the detail of anatomical relationships among deep brain structures. For example, such anatomical references as the Schaltenbrand-Bailey stereotactic atlas provided a platform for overlaying an early neuronavigational atlas on planning of stereotactic procedures [1,2]. Use of atlases relied upon relative preservation of anatomical relationships from patient to patient, and could, at best, estimate the position of deep brain structures by applying the atlas to standard outer landmarks. However, atlas-based stereotaxy underscores the importance of mastering anatomical relationships. Stereotactic neurosurgery, therefore, requires not only a knowledge of the exact location of a target in the brain, but a thorough appreciation for important structures that surround that target in threedimensional space. To improve the accuracy of stereotactic targeting based upon more individual patient data, the first real-time image-guidance systems in stereotactic neurosurgery developed based upon ventriculography. Injection of air or contrast into the ventricular system provides an outline of the borders of the ventricles on plain radiographs and fluoroscopy, which can yield a real-time image of deep brain anatomical relationships. Prior to computed tomography and magnetic resonance, combining this imaging with atlasbased relationships of targets to standard intracranial landmarks could be used to derive the relative locations of anatomical structures [3]. More importantly, this development represented

568

37

Brainlab image guided system

a method for obtaining real-time intracranial anatomical information. Neurosurgeons could use imaging of each patient’s intracranial anatomy to guide stereotactic procedures, establishing the principle of precision in image-guidance. Although this technique was somewhat morbid, there are those who still believe that this has a role in certain procedures as a direct, real-time assessment of intracranial anatomy. An example of an ongoing application in some centers is for accurate placement of Ommaya reservoir intraventricular catheters, although rare practitioners continue to use this for functional procedures as well [4]. Computed tomography and magnetic resonance provided obvious advantages for modern neuronavigation. Rather than extrapolating the location of targeted structures from invasive radiographic studies, relevant neuroanatomy can be directly visualized. The progressively improving resolution of CT and MRI, combined with incorporation of fine-cut preoperative framebased imaging protocols, has provided a direct targeting methodology for stereotactic neurosurgery. In directly selecting the target on an MRI image, a neurosurgeon can plan trajectories, approaches, and compute relative distances and angles. Indeed, there have been multiple studies attempting to evaluate the relative accuracy of MRI-based direct targeting and ventriculography or frame-based indirect targeting [5–7]. Interestingly, much of the information obtained from earlier studies using fluoroscopy has been retained and refined in the era of CT and MRI, such as the location of functional targets in relation to landmarks such as the anterior and posterior commissures. This highlights one encouraging feature of stereotaxic neurosurgery, which is that useful methodologies are often retained in some form even as technology advances. Against this background, intraoperative neuronavigational guidance tools have developed. There have been two major advances over the years in computer-based imaging technology which now

offer novel tools for surgeons using both framebased and frameless navigational technology. For frame-based stereotaxic surgery, the use of computer planning has several advantages. There is a significant time-saving when frames can be automatically identified and registered from source images as compared with manual calculations. The ability to simultaneously visualize various image data sets as well as reconstructed probe views can facilitate optimized trajectory planning. Accurate fusion of different data sets, such as CT and MRI, can permit advanced creation of stereotactic plans and also can combine unique advantages of different modalities in a single plan to improve accuracy. Of course, the advent of frameless stereotaxy is entirely novel and dependent upon high-level computer reconstruction technology. This provides important information for general neurosurgery, radiosurgery and offers new opportunities and challenges for stereotaxic neurosurgery. In considering these roles, we present applications for each using the BrainLab neuronavigation technology. While several corporations produce neuronavigational equipment, the history and development of BrainLab as a company mirrors the advances of neuronavigational devices and principles in the last two decades. Specific features of the BrainLab system as currently configured for both frame-based and frameless procedures will also be outlined for the benefit of those will little or no experience with this particular technology.

History and Technology of BrainLab BrainLab was founded in Munich, Germany in 1989, BrainLab primarily as a software company dedicated to generating neuronavigational systems which would be more accurate and user-friendly for general neurosurgical practice. In the company timeline, 1990 saw the first use of BrainLab software at the University of Vienna, while in 1994


BrainLab USA was initiated. The first VectorVision navigation system was cleared for use by the FDA in 1997, while the first BrainSUITE was sold in 2004. Over the last 18 years, BrainLab has rapidly become a multi-national company providing neuronavigational software to 15 countries. Currently over 2,000 hospitals are using one or more BrainLab systems throughout the world. From a developmental perspective, BrainLab started by producing a mouse-controlled, menudriven software application for neurosurgical planning (> Figure 37-1). By 1998, BrainLab was also producing the hardware needed for several applications in neurosurgery; it had introduced an integrated radiosurgery system, as well as the mobile cranial planning and navigation unit (> Figure 37-1). Since the release of optical tracking mobile units, neuronavigation has . Figure 37-1 The VectorVision2 mobile optical tracking unit

37

allowed direct integration of preoperative radiographic information into the real-time operative environment [8,9]. By 2003, BrainLab had begun integrating the hardware into operating room design, introducing a comprehensive BrainSUITE, which features the cranial planning/ navigation software and hardware, as well as an incorporated intraoperative MRI. BrainLab has also helped enable new imaging applications for neurosurgical intervention. Starting in 2005, BrainLab’s systems incorporated fiber tracking into their software packages, allowing this technology to become a more user-friendly tool for preoperative and intraoperative technical planning without the need for highly specialized physicists and technicians (> Figure 37-2). Since that time, tracking of important neurological pathways preoperatively has been popularized

569

570

37


. Figure 37-2 An example of fiber tracking with DTI (diffusion tensor imaging) incorporated into the neuronavigational plan on the BrainLab screen

in the resection of intra-axial tumors [10,11]. Finally, in 2006 iPlan Net was introduced, which is a network-based system that facilitates immediate image transfer between different sites, remote preoperative planning, and direct feeds of imaging and planning information into the operative setting. In addition to preoperative image planning, BrainLab also provides hardware that allows intraoperative registration of tools to permit realtime navigation during surgery. The optical tracking system, combined with hardware attachments, allows instruments to be converted into cable-free, battery-free probes that can be seen on the intraoperative displays. The software package also allows integration and image fusion of multiple imaging modalities [12]. These hardware and software features are essential for frameless stereotactic

work; it allows the surgeon to see the trajectory and depth of an instrument relative to its anatomical target. The system also allows tracking using neurosurgical microscopes from most major manufacturers, with optional heads up displays to place navigational information directly into one of the microscope oculars if this is desired. This even uses the focal length of the microscope to identify the location of the area being visualized at any given time. BrainLab provides various modular systems which can be used for particular applications, and they can also be combined into a more comprehensive system which provides an integrative approach to planning and navigation in stereotactic and functional neurosurgery. Although certain hardware is needed for these various procedures, software development has been the major focus of


BrainLab over the years, reflecting the importance of superior software to the ultimate value of computer-based imaging technologies. Below we present several BrainLab technologies which are currently commercially available along with examples highlighting how they can be applied to the advantage of the practicing neurosurgeon.

Applications Frame-based Targeting Frame-based targeting is not a novel concept in neurosurgery; its efficacy and precision for deep brain structures has been well-established. However, pre- and peri-operative trajectory planning was previously based on hand-calculation of coordinates from either known landmarks (anterior and posterior commissures) or standardized atlases. While these techniques are still very applicable today, and can often serve as methods of verification for a selected plan, image fusion and computer-based trajectory design have added a novel dimension to functional neurosurgical preparation. iPlan stereotaxy was designed specifically for stereotactic targeting and trajectory planning for both functional neurosurgical procedures and stereotactic biopsies. Information on all major stereotactic frames is contained within the package, so registration is almost immediate once the images are loaded and the type of frame and localizer (CTor MRI) is identified for the system. After the frame is registered, the anterior and posterior commissures and midline are all identified, although for direct targeting for a biopsy or if this information is not necessary for some other reason, this step can be bypassed. The intercommisural distance is provided, which can help to evaluate the validity of the data. Multiple data sets can then be merged simultaneously, such as multiple MRI sequences as well as CT scans. Even functional MRI or diffusion tensor imaging (DTI) data can now be input for both functional as well

37

as anatomical planning. With this feature, thin cut slabs in multiple orientations can be merged with larger data sets, thereby reducing the MRI time for the patient. Small deep nuclei such as the subthalamic nucleus can now be visualized and targeted using MR technology. With the BrainLab iPlan stereotaxy planning software, a patient may have an MRI prior to the day of surgery without a frame. On the day of surgery, a frame is applied to the patient, who subsequently undergoes a thin-slice CT scan. The CT and MRI images are fused using the BrainLab planning software, and both sets of images can be used to perform the surgery. Alternatively, the coordinates for the target(s) can be obtained from the planning software, negating the need for an active neuronavigation system in the operating room. The iPlan Net can be of particular utility as an adjunct to frame-based stereotactic planning. DICOM images from any modality (MRI, CT and even PET) can be immediately ‘‘pushed’’ to the iPlan Net server by the radiology technician, in a manner identical to sending images to either a PACS or other network server. The server is simply a computer which can accept networked images, and which runs the various software packages, so that it can be placed at any remote site which is accessible to a particular network. The system is both HIPAA compliant and can be configured as a ‘‘one-way’’ server, such that images can be pushed to the iPlan Net server directly and not through the PACS system, but they cannot be sent back to the radiology PACS system. This feature can help allay any concerns by radiology information technology personnel regarding possible unauthorized entry into the PACS clinical imaging system. Since the iPlan net server can run the iPlan stereotaxy software package, image analysis and all planning can be performed online from any computer networked to the iPlan Net server. With the iPlan Net system, the surgeon is freed from the obligation to plan only at a single workstation, and planning can be completed from

571

572

37


any office or even from home (if this is in fact desirable). If planning from a preoperative MRI is completed in advance, then the image obtained after the frame is placed (MRI or CT) can be sent to the server and is available as soon as the surgeon reaches any networked computer. The images can then simply be merged, and, within a few minutes after obtaining the post-frame image, final frame coordinates are obtained to accelerate initiation of the surgical procedure. The same plan can also be projected from a networked computer in the operating room to a large screen monitor, so that the images and stereotactic information can be both displayed and modified in the operating room without the need for specific workstations or equipment. We find this feature of the iPlan Net to be a particularly useful teaching tool, as anatomical relationships and various considerations when planning surgery can be displayed anywhere that a computer can be networked to the server, such as in teaching or case conferences or in small groups within the neurosurgical offices.

Deep Brain Stimulation for Parkinson’s Disease The iPlan stereotaxy package was designed for use in deep brain stimulation (DBS) surgery, although it would be equally and readily applicable to emerging technologies such as infusion of biological agents (growth factors, gene therapy and cell-based therapies). We routinely utilize the iPlan stereotaxy with the iPlan net server for such procedures, for example targeting the subthalamic nucleus (STN) for DBS in Parkinson’s disease. In advance of surgery, we obtain an outpatient 3 Tesla MRI for pre-surgical planning. Fine cut T2-weighted axial sections are obtained in a single acquisition from just below the STN to the top of the head (usually the maximum possible in a single acquisition is just over 60 slices). Multiple acquisitions to obtain more images should be avoided, as they often are not perfectly

aligned and reconstruction results in staggered, jagged images. A second set of fine-cut, coronal T2-weighted images are obtained as a slab just around the level of the STN. Although we do not obtain contrast images, a third set of double-dose T1-weighted contrast images could be obtained and incorporated as another data set, as is the custom in some practices. These images are immediately pushed to the iPlan net server, where they are available at any time to perform target and trajectory planning. We routinely complete this from our personal office computer in advance of surgery. After merging the axial and coronal images and identifying AC, PC and midline, the STN target is then chosen. This can be determined by direct visualization of the axial and/or coronal images, indirect targeting based upon the distance in all three coordinates from the mid-commissural point (MCP) (or from AC or PC, if preferred) or based upon the position on the SchaltenbrandWharen atlas, which can be overlaid and manually adjusted to the MRI images. We routinely target initially based upon a distance from the MCP (> Figure 37-3a), then adjust that position after visually determining the location relative to nearby structures such as the red nucleus, substantia nigra and internal capsule. In general we use the atlas as a teaching tool, although occasionally an adjustment is made based on this information as well (> Figure 37-3b). The entry point can then be determined based upon visual targeting or based upon angles relative to midline and to the intercommissural line (> Figure 37-3c). This can then be adjusted manually, and two obliqued images are provided which are cut along the axis of the trajectory to visualize the entire tract in one image (> Figure 37-3d). Scrolling along the trajectory can also be performed to demonstrate the location on true axial or coronal images at each point along the tract. The trajectory can be further adjusted from this view to immediately avoid an undesirable structure without having to guess as to a change and then re-check the


37

. Figure 37-3 (a). The AC-PC View allows identification of the anterior commissure, posterior commissure, and mid-sagittal planes. From these, the mid-commissural point (MCP) may be derived. (b) The Atlas View overlays the SchaltenbrandWahren atlas, which can be adjusted to the actual anatomy of the patient, onto the MRI, and permits its use as a reference. (c) The Trajectories Overview is a 3-orientation visualization at a particular position (in this case, the target is the STN). (d) The Probe View provides two different obliqued views in the plane of the trajectory to see what is being traversed in a single image. By scrolling through the images, the user can understand the structures that the trajectory will traverse. In this case, the solid pink line represents the selected trajectory for scrolling, while the dotted green line represents another trajectory in the plan

573

574

37



plan. The final plan is saved on the server for future access. On the day of surgery, a volumetric, fine cut non-contrast CT is obtained after placement of

the stereotactic frame. While the patient is brought to the OR, the images are sent directly to the server, where they can be uploaded into the folder with the prior plan from any networked


computer either in or out of the operating room. The frame is identified and the fiducials automatically register (> Figure 37-4a). A threedimensional reconstruction of the localizing box with the fiducial bars provides added confirmation as to the accuracy of the registration. Although we prefer a CT due to the additional accuracy and minimal scan time, an MRI performed with the frame in place can also be used as the software package contains information on MRI as well as CT localizers for most popular frames. The frame-based image is then merged with the preoperative plan, and stereotaxic coordinates are generated for the particular frame (> Figure 37-4b). We utilize the Leksell frame, and the software generates coordinates for all four possible orientations of the frame, along with a picture of the frame in the orientation for the selected set of coordinates. This can be a useful feature if different orientations are routinely used, but for centers which use primarily one orientation, care should be taken to ensure that the correct set of coordinates are used since there is currently no option available to remove the irrelevant coordinates. Entry angles are also provided to facilitate passage along the exact trajectory planned to target. We project this data to a screen in the operating room from a networked computer. The data is also output into a formatted report as a PDF file for easy printing. This report also provides coordinates for all four orientations, which can create confusion, so we blacken the three irrelevant sets. The report also includes other useful information, such as the intercommissural distance and location of the target relative to MCP. This report can be printed for use in the OR and also is an excellent source for documentation in the patient chart. Intraoperatively, notes can be generated at various points along the trajectory, documenting such things as patient responses or electrophysiological data. There is the option to output data from various electrophysiological recording systems into the BrainLab system to document

37

actual recording data at different points along the trajectory. Another useful application of this technology is for post-operative analysis of electrode placement. We routinely perform a post-surgical MRI in the axial and coronal planes, however in some patients only a CT scan is possible (if there is a pacemaker in place, for example). Either of these can be merged with the preoperative data sets, and the planned trajectory and target localization can be compared with the actual electrode location. This can be particularly useful when analyzing images which have not been symmetrically cut. Electrodes which may not appear to be ideal on an image which actually obliqued often are clearly well-placed when analyzed in this manner. It should be noted that the iPlan Net server is not obligatory to utilize all of the features outlined above. Stand alone workstations are also available which run this software, and even if the server is utilized, at least one such workstation should still be obtained as a backup in case of a server failure. Images are then loaded onto the workstation using standard tapes or USB drives.

Frameless Targeting While using a fixed frame allows instruments to be steadily oriented to a selected target, hardware innovation in neuronavigation has created an alternative to frame-based stereotactic surgery. Optical tracking systems such as the VectorVision2 (BrainLab), can track any surgical instrument that is registered with attached fiducials. For example, a suction catheter or bipolar cautery instrument may be tracked by attaching a ‘‘star,’’ that maintains three fiducial markers in a fixed relationship. The instrument is established in the neuronavigation virtual environment using a mobile registration unit, commonly called ‘‘the elephant’’(> Figure 37-5). The instrument then becomes a de facto probe, allowing

575

576

37


. Figure 37-4 The planning software can accept many different frames based on both CT and MRI. (a) The Frame Localizer shows how the frame is localized and identified to the machine. (b) A CT and noncontrast T2-weighted MRI undergo image fusion. In the upper left panel, the auto-fusion of images may be checked by manipulating the window shown to verify that anatomical structures match between the two imaging modalities


. Figure 37-5 Registration of a biopsy instrument using the mobile registration unit ("the elephant")

simultaneous neuronavigation and progression of the operative plan. Through such operational flexibility, frameless stereotaxy has been successfully used for both functional (deep brain stimulation) and oncologic (tumor biopsy) neurosurgery [13–15]. With regard to frameless functional neurosurgical cases, the planning and targeting software described above is used in the same fashion. However, rather than generating frame-based coordinates, frameless navigation is used to orient either the microelectrode or permanent DBS electrode guide tubes to match the planned trajectory and then pass to the desired target. For these cases, to achieve accuracies which approach frame-based targeting, fiducials which are fixed to the skull (so-called ‘‘bone fiducials’’) are used, as compared with skinbased fiducials which usually suffice for oncological and other procedures which don’t require the same level of accuracy. However, not all instruments require ‘‘star’’ attachments and intraoperative registration. BrainLab produces particular hardware items that are pre-registered with already-attached fiducials. Such instruments are easily integrated into the neuronavigational environment, where their standardization and fixed fiducial relationships help to ensure accuracy and precision. Of particular

37

note, BrainLab produces a frameless biopsy system, consisting of a pre-calibrated biopsy needle, precalibrated biopsy alignment sheath, and a frameless biopsy arm that attaches to a standard Mayfield head holder (> Figure 37-6a). Once assembled, the system allows the surgeon to alter trajectory and entry point in real-time from virtually any angle (> Figure 37-6b). In addition to tumor biopsy, frameless stereotaxy can also be utilized for placement of intracranial monitoring electrodes in cases of intractable epilepsy. Typically, a patient will undergo a craniotomy, and strips of electrodes will be layed onto the brain in the subdural space. In addition, depth electrodes will be placed in deep temporal structures: the hippocampus and deep temporal gyri. Although the surface gyri are directly visible within the craniotomy window, the depth electrodes must be accurately placed to avoid injury to the brainstem, thalamus, and other deep brain structures. An example of this technology is demonstrated in > Figure 37-7. In this setting, frameless stereotactic guidance can be helpful in depth electrode placement [16,17].

Intraoperative MRI Intraoperative MRI technology represents a fusion of preoperative neurosurgical imaging with intraoperative planning and trajectory adjustment based. Desired as a tool for intraoperative assessment of surgical success and changes in intracranial anatomy, intraoperative MRI’s were introduced in the last decade of the twentieth century. Initial attempts to incorporate MR imaging into the operative environment were mixed, due to the low-field magnets used (0.5 Tesla) and the long transition times required. With the incorporation of neuronavigational units, the advancement of magnet strength (to 1.5 and 3.0 Tesla), and the establishment of ergonomic operative suites (set up specifically for this purpose), intraoperative MR imaging is rapidly

577

578

37


. Figure 37-6 The BrainLab stereotactic biopsy components (a) are integrated to allow intraoperative adjustment of the entry point and trajectory (b)


37

. Figure 37-7 Intraoperative screenshots of neuronavigation software demonstrating proposed trajectories for depth electrode placement for analysis of seizure foci

gaining popularity [18–21]. In the BrainSUITE setting, the intraoperatively obtained images are easily integrated into the preoperative neuronavigational plan. Logistically, either the operating table or the MR scanner move into and out of the scanning position in an ergonomicallydesigned fashion. The neuronavigation software allows fusion of intraoperative and preoperative data, which provides the advantage of real-time adjustments based upon contemporaneous intraoperative imaging data with the speed of navigation based upon previously acquired data. This

also provides feedback as to the current location relative to the original and presumably desirable plan. Currently, this technology is being applied to the resection of intra-axial brain tumors, transsphenoidal resection of pituitary tumors, placement of deep brain stimulation electrodes, and epilepsy surgery [22–26]. Objective data on the relative utility of intraoperative MRI for each of these applications is being collected, and its role in neurosurgical procedures has yet to be fully elucidated. Clearly this is a major undertaking for any center, both in terms of cost, space and logistics to

579

580

37


manage an MRI environment within an operating room setting. Nevertheless, the BrainSUITE platform provides an efficient method for obtaining and integrating intraoperative imaging data into the surgical plan. The features of the BrainSUITE that are designed to optimize neurosurgical efficiency and ease of use may be attractive to those centers with the means and space to consider intraoperative MRI imaging. In the modern neurosurgical era, neuronavigation plays a vital role in oncologic, stereotactic, and functional neurosurgery. It allows the surgeon to preoperatively plan complex operative approaches, to target deep brain nuclei for biopsy, stimulation, or lesioning, and to individualize the operative plan to each patient’s specific intracranial anatomy. Successful neuronavigation depends upon the principles of accurate and relevant imaging, precise correspondence between images and patient anatomy, and appropriate incorporation of navigational tools into the operative protocol. Neuronavigational devices will never substitute for a vital three-dimensional understanding of neuroanatomy, but can serve to aid the neurosurgeon in accurate targeting and planning. Continued advances in both software and hardware by companies such as Brainlab, with a particular focus upon the needs and constraints of the neurosurgical environment, will continue to facilitate otherwise difficult functional and stereotactic neurosurgical procedures.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

References 15. 1. Kazarnovskaya MI, Borodkin SM, Shabalov VA, Krivosheina VY, Golanov AV. 3-D computer model of subcortical structures of human brain. Comput Biol Med 1991;21(6):451-7. 2. Housepian EM. Stereotactic surgery: the early years. Neurosurgery 2004;55(5):1210-4. 3. Slaughter DG, Nashold BS, Jr. Intracranial measurements for stereotactic surgery. Confin Neurol 1970;32(2):250-4. 4. Sandberg DI, Bilsky MH, Souweidane MM, Bzdil J, Gutin PH. Ommaya reservoirs for the treatment of

16.

17.

leptomeningeal metastases. Neurosurgery 2000;47 (1):49-54; Breit S, LeBas JF, Koudsie A, et al. Pretargeting for the implantation of stimulation electrodes into the subthalamic nucleus: a comparative study of magnetic resonance imaging and ventriculography. Neurosurgery 2006;58 Suppl 1:ONS83-95. Slavin KV, Thulborn KR, Wess C, Nersesyan H. Direct visualization of the human subthalamic nucleus with 3T MR imaging. Am J Neuroradiol 2006;27(1):80-4. Schlaier J, Schoedel P, Lange M, et al. Reliability of atlas-derived coordinates in deep brain stimulation. Acta Neurochir (Wien) 2005;147(11):1175-80; discussion 1180. Gumprecht HK, Widenka DC, Lumenta CB. BrainLab vectorvision neuronavigation system: technology and clinical experiences in 131 cases. Neurosurgery 1999;44 (1):97-104; discussion 104–105. Mascott CR. In vivo accuracy of image guidance performed using optical tracking and optimized registration. J Neurosurg 2006;105(4):561-7. Berman JI, Berger MS, Chung SW, Nagarajan SS, Henry RG. Accuracy of diffusion tensor magnetic resonance imaging tractography assessed using intraoperative subcortical stimulation mapping and magnetic source imaging. J Neurosurg 2007;107(3):488-94. Bello L, Gambini A, Castellano A, et al. Motor and language DTI fiber tracking combined with intraoperative subcortical mapping for surgical removal of gliomas. Neuroimage 2008;39(1):369-82. Schlaier JR, Warnat J, Dorenbeck U, Proescholdt M, Schebesch KM, Brawanski A. Image fusion of MR images and real-time ultrasonography: evaluation of fusion accuracy combining two commercial instruments, a neuronavigation system and a ultrasound system. Acta Neurochir (Wien) 2004;146(3):271-6; discussion 276–277. Eljamel MS, Tulley M, Spillane K. A simple stereotactic method for frameless deep brain stimulation. Stereotact Funct Neurosurg 2007;85(1):6-10. Pan HC, Wang YC, Lee SD, Chen NF, Chang CS, Yang DY. A modified method to perform the frameless biopsy. J Clin Neurosci 2003;10(5):602-5. Holloway KL, Gaede SE, Starr PA, Rosenow JM, Ramakrishnan V, Henderson JM. Frameless stereotaxy using bone fiducial markers for deep brain stimulation. J Neurosurg 2005;103(3):404-13. Mehta AD, Labar D, Dean A, et al. Frameless stereotactic placement of depth electrodes in epilepsy surgery. J Neurosurg 2005;102(6):1040-5. Murphy MA, O’Brien TJ, Cook MJ. Insertion of depth electrodes with or without subdural grids using frameless stereotactic guidance systems–technique and outcome. Br J Neurosurg 2002;16(2):119-25.


18. Maciunas RJ, Dean D, Lewin J, Selman WR, Ratcheson RA. Integration of neurosurgical image guidance and an intraoperative magnetic resonance scanner. The University Hospitals of Cleveland experience. Stereotact Funct Neurosurg 2003;80(1–4):136-9. 19. Samdani A, Jallo GI. Intraoperative MRI: technology, systems, and application to pediatric brain tumors. Surg Technol Int 2007;16:236-43. 20. Yrjana SK, Tuominen J, Koivukangas J. Intraoperative magnetic resonance imaging in neurosurgery. Acta Radiol 2007;48(5):540-9. 21. Jones J, Ruge J. Intraoperative magnetic resonance imaging in pituitary macroadenoma surgery: an assessment of visual outcome. Neurosurg Focus 2007;23(5):E12. 22. Lee MW, De Salles AA, Frighetto L, Torres R, Behnke E, Bronstein JM. Deep brain stimulation in intraoperative MRI environment – comparison of imaging techniques

23.

24.

25.

26.

37

and electrode fixation methods. Minim Invasive Neurosurg 2005;48(1):1-6. Nimsky C, Ganslandt O, von Keller B, Fahlbusch R. Intraoperative high-field MRI: anatomical and functional imaging. Acta Neurochir Suppl 2006;98:87-95. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R. Intraoperative high-field magnetic resonance imaging in transsphenoidal surgery of hormonally inactive pituitary macroadenomas. Neurosurgery 2006;59(1):105-14; discussion 105–114. Nimsky C, Fujita A, Ganslandt O, von Keller B, Kohmura E, Fahlbusch R. Frameless stereotactic surgery using intraoperative high-field magnetic resonance imaging. Neurol Med Chir (Tokyo) 2004;44(10):522-33; discussion 534. Hall WA, Truwit CL. Intraoperative MR-guided neurosurgery. J Magn Reson Imaging 2008;27(2):368-75.

581

47 Comprehensive Brain Tumor Management M. Tamber . M. Bernstein

Introduction

Provision of a Histologic Diagnosis

The contemporary management of patients with primary intra-axial brain tumors is multidisciplinary. Surgery followed by radiotherapy was until fairly recently the principal treatment, but provides only palliative benefit. Historically, chemotherapy has been relegated to an adjuvant role, but recent developments have validated certain chemotherapeutic agents as important tools in our therapeutic armamentarium, providing a meaningful, albeit modest, improvement in survival. Perhaps the most exciting areas of advance lie in the genetic targeting of conventional treatments and in the expanding use of genomics and proteomics to develop novel agents specifically directed at critical molecular targets within the cell [1]. In this chapter, we summarize the current status and future prospects for each of the treatment modalities for brain tumors. We tailor our discussions in a relatively generic sense around gliomas, without in-depth discussion of more controversial and nuanced issues such as the dilemma over management of presumed lowgrade gliomas.

With relatively few exceptions, histologic diagnosis is imperative in most patients with an intracranial mass lesion, especially if a malignant intra-axial neoplasm is considered in the preoperative differential diagnosis. An accurate tissue diagnosis not only excludes the possibility of another diagnosis which may require a different treatment paradigm altogether (e.g., cerebral abscess), but offers critical information regarding the histological type and grade of the lesion which will influence decisions on further management and provide important guidance regarding prognosis. A histological diagnosis of the lesion may be achieved via craniotomy and open or ‘‘excisional’’ biopsy, or by a minimally-invasive, imageguided stereotactic procedure. The latter may be frame-based or frameless. The diagnostic accuracy of various biopsy procedures is proportional to the amount of tissue obtained and the accurate targeting of areas within the lesion of potentially high diagnostic yield. Although it provides less tissue for examination than an open excisional biopsy (with the attendant increased risk of sampling error), stereotactic/image-guided biopsy can largely replace craniotomy for the purposes of histological diagnosis with similar efficacy and comparable morbidity. In other words, craniotomy is not mandatory for the histological diagnosis of brain tumors. In fact, one can envision several circumstances where an image-guided biopsy would be the preferred means of obtaining tissue, such as

Role of Surgery Surgery has the distinct but interrelated aims of providing a histological diagnosis to guide further therapy, the relief of mass effect, cytoreduction, and facilitation of the local delivery of adjuvant therapy. Each of these roles is discussed in turn. #


736

47

Comprehensive brain tumor management

when the therapeutic benefit from craniotomy and excisional biopsy is too small to justify the risks, or, conversely, if the risk of craniotomy, regardless of the potential benefit, is felt to be prohibitive. When deciding to perform a stereotactic biopsy, one must consider both the lesion and the patient. The ideal lesion for stereotactic biopsy (versus craniotomy) is a small deep lesion or one located in eloquent cortex for which the risk of craniotomy and excisional biopsy is felt to be prohibitively high. Implicit in this is that cytoreduction is not required and/or could not be done effectively or safely due to the diffuse and indistinct nature of thee tumor (> Figure 47-1). Other characteristics of lesions suited for stereotactic biopsy are diffuseness (e.g., across corpus callosum) and multiplicity (i.e., more than one lesion in disparate parts of the brain). Concerning the patient, sometimes an individual who, under normal circumstances, would benefit from an aggressive resection is better served by a stereotactic biopsy because of significant medical illness or very advanced age. Conversely, if the patient is young and neurologically intact, the risk of neurological

morbidity from craniotomy can be avoided either at the surgeon’s or the patient’s discretion in specific cases. The important issues to examine regarding stereotactic biopsy, after patient selection, are the success rate and the complication rate. The success rate (i.e., the probability of obtaining a positive and definitive diagnosis from the procedure) varies in the literature. Based on a meta-analysis of over 4,000 stereotactic biopsies reported in the literature, one can conclude that a positive diagnosis should be attainable in excess of 90% of cases [2]. In the case of neoplastic disease, it is unusual to miss the diagnosis entirely, although it is not uncommon to miss the area of worst grade within a particular tumor due to sampling error (e.g., sampling only the grade II portion of an anaplastic astrocytoma). The reported complication rates of stereotactic biopsy vary considerably in the literature. The accepted complication rate appears to be up to 5%. The major complication is intracerebral hemorrhage resulting in neurological deterioration. In a recent large single-surgeon series from

. Figure 47-1 Diffuse glioma in eloquent cortex in a 40-year-old woman presenting with sensory seizures. A = T2. B = T1 with gadolinium enhancement. Awake image-guided biopsy revealed anaplastic oligodendroglioma


the University of Toronto, comprising almost 750 biopsies, the incidence of complications (mainly intracerebral hemorrhage) is about 5%. Two percent of these have resulted in death or neurological morbidity while the other 3% resulted in minor and/or transient deficits only. An important risk factor for hemorrhage appears to be a malignant histology [3–6]. The differential survival of patients with malignant intra-axial brain tumors diagnosed with stereotactic biopsy and then irradiated, as opposed to undergoing resection plus radiation, has been the subject of a few papers [7,8]. In general, retrospective data suggests that stereotactic biopsy plus radiation was as effective as craniotomy plus radiation for selected patients with malignant tumors in treacherous locations. However, these studies suffer from critical flaws related to selection bias, as patients were not randomly assigned to the two surgical arms. Given the present state of the evidence, one cannot conclude that for any given malignant cerebral tumor, stereotactic biopsy plus radiation is better, worse, or equal to resection plus radiation in terms of length of survival, with the assumption that debulking was not required for palliation of symptoms. At present, the treatment of brain tumor patients depends principally upon the histological signature of the lesion; in the future, it is likely to depend increasingly on tumor genotyping and the molecular phenotype of the tumor, features that can only be assessed using tissue samples. As a result, it is postulated that tumor biopsy will continue to play an important role in the future management of brain tumor patients.

Surgery and Symptom Relief Tumors exert their clinical effects by two groups of mechanisms: direct invasion of nervous tissue, and various distant actions, including high intracranial pressure, seizures, hydrocephalus, and

47

compression and distortion of adjacent nervous tissue. Clearly, no surgical intervention can reverse the consequence of direct neoplastic invasion of adjacent neural structures. Surgery, however, can relieve global symptoms caused by raised intracranial pressure, such as headache, vomiting, and a general feeling of malaise that is not recognized as abnormal until after the tumor has been removed (when the patient realizes how unwell they felt before surgery). Epilepsy secondary to intra-axial brain tumors is amenable to surgical palliation in those circumstances where seizures are multiple, intrusive, and refractory to medical management alone. In addition, some focal deficits seem to respond to surgical palliation. The mechanism of such deficits is likely some combination of distortion and local compression of adjacent nervous tissue; tumor debulking effects decompression of these distorted yet viable tissues, thereby allowing partial, if not total, recovery of function [9]. Surgically remediable symptoms typically respond to steroids. Major tumor decompression can facilitate prompt reduction of steroid medication, with the avoidance of the side effects produced by their long term use at high doses. The relative ease with which intracranial pressure can be reduced and neurological symptoms alleviated in each case is balanced against the associated risks of surgery. Drainage of a cystic lesion or resection of a lesion in a non-dominant frontal or temporal lobe, for example, are relatively low risk procedures, which will produce rapid decompression and potentially some degree of neurologic recovery. Surgical decision making in this instance is clearly more straightforward than contemplating the removal of a diffuse, ill defined lesion close to an eloquent brain area. Many tumors in more eloquent cortex are still better served by resection than biopsy, especially if they are producing neurological deficit (which might be helped by surgery) and appear relatively discrete on imaging (> Figure 47-2). In short, any surgical intervention should be tailored to the patient’s symptoms,

737

738

47


. Figure 47-2 Relatively discrete glioblastoma in eloquent cortex in a 70-year-old man. His left hemiparesis and hemisensory dysfunction resolved after awake image-guided craniotomy for gross total resection of the tumor

clinical condition and needs, prognosis, as well as to the requirements of any adjuvant treatments. In regards to this latter point, relief of mass effect related to a tumor may also serve to enhance the safety and efficacy of adjuvant treatments such as radiation, although clear evidence in support of this is lacking [10]. Although it seems intuitive that the relief of mass effect following surgery is associated with improved functional outcome (not necessarily survival), data in support of this thesis is scant. The reason behind this relative dearth of data is straightforward – it is far more methodologically challenging to measure the impact of an intervention on quality of life than it is to measure the impact of the same intervention on the same patients in terms of a ‘‘hard’’ endpoint such as prolongation of survival. Nevertheless, the observational data that does exist on the subject of palliation of symptoms following surgery seems to imply that some functional improvement after surgery may be seen in certain patients [11].

Cytoreductive Surgery and Survival Issues regarding quality of life aside, the really difficult and unresolved issue regarding the surgical treatment of malignant intra-axial brain tumors relates to the inherent value of aggressive resection in prolonging the life of the patient. Several reports have assessed the effect of surgery on survival, with most concluding that more extensive resection is associated with longer survival [12–15]. Once again, most studies are retrospective and interpretation is complicated because of selection bias. Patients presumed to have a favorable prognosis, on the basis of young age at diagnosis, absence of neurological deficits, lower grade tumor, and location remote from eloquent areas (which likely is a surrogate for the feasibility of extensive surgical resection), tend to be selected for more aggressive surgery. In studies such as these, it is difficult, if not impossible, to disentangle the effect of these important confounding variables from the effect of aggressive surgery on patient survival. Studies that use statistical modeling to account for the effect of these confounding variables tend to document more modest survival advantages than those studies that do not adjust for the effect of other important prognostic variables, with overall prolongation of survival in those patients treated with aggressive surgery ranging from none [16] to about four months [17] in adjusted analyses. Importantly, these estimates are likely optimistic, as no statistical model is able to compensate fully for other unmeasured variables that are associated with the outcome of interest. Interpretation of evidence relating the degree of resection to patient survival is further complicated by the difficulty of defining the extent of resection. The surgeon’s impression of the extent of tumor removal is unreliable beyond the gross distinction between biopsy and resection. More objective reports have sought to use


postoperative imaging to quantitatively assess resection, but even in these studies, the results depend on multiple factors, including the imaging modality used, the degree of inter-rater reliability, and the time interval between surgery and post-operative imaging [18]. On balance, more papers report improved survival than do not; this seems to influence much of current surgical practice. However, the axiom that a large quantity of poor quality data does not necessarily replace a small amount of methodologically sound information should not be forgotten. Systematic reviews of the observational literature have repeatedly found no convincing evidence for an independent benefit from surgery [19,20]. Randomized trials have been difficult to do and are conspicuously lacking. Only one randomized study has been executed to specifically examine the role of surgery for these tumors [8]; this represented an outstanding and noble effort but, in the end, was a flawed and inconclusive study. It is quite likely that no large, properly executed randomized controlled trial will ever be successfully conducted. Neurosurgeons will have to base their judgments regarding surgery for malignant brain tumor patients on available biased nonrandomized data, personal ‘‘feelings’’ based on experience, patient preference, and their individual concept of the disease.

Surgery to Facilitate Delivery of Local Therapy The blood-brain-barrier (BBB) imposes a limiting and often impervious barrier to the delivery of conventional antineoplastic agents to the local tumor microenvironment. Over the past several years, numerous attempts to circumvent this difficulty have relied on the loco-regional administration of conventional as well as novel therapeutic agents at the time of surgical treatment. Most have met with limited success, but the delivery

47

methods, the context within which they are utilized (e.g., along with cranial radiotherapy and/or systemic chemotherapy), as well as the agents themselves, are in a process of continual evolution.

Photodynamic Therapy This technique involves intravenous injection of a photosensitizing porphyrin-based dye approximately 24 h before surgery. This dye is preferentially taken up by tumor cells. At the time of surgery, the tumor bed is illuminated with light of the appropriate wavelength; this activates the dye and kills the tumor cells that have taken up the dye, but not the surrounding brain cells. Some studies of this treatment have been published, but no good evidence of benefit has been documented [21].

Gene Therapy The Herpes Simplex Virus-Thymidine Kinase (HSV-TK) viral oncolytic system relies upon the application of mouse producer cells carrying replication-incompetent retroviruses transfected with the herpes simplex thymidine kinase gene to the resection margin at the time of surgery. Once the virus is locally administered in this way, the premise is that local infection of tumor cells occurs, such that that these cells now harbor the HSV-TK gene in their DNA complement. After systemic administration of ganciclovir, the HSV-TK gene metabolizes ganciclovir to a cytotoxic nucleotide, which induces apoptosis in rapidly dividing cells. Several phase I and phase II trials have been done documenting the feasibility of the approach and some toxicity [22]. A randomized trial of 248 patients found no benefit [23], but a small randomized trial of 36 patients has since found a survival advantage with treatment (70 weeks vs. 40 weeks survival,

739

740

47


p = 0·001) [24]. New research with replication competent viruses and viruses of different types may produce more encouraging results. Other biologicals used intratumorally include interferon [25].

Brachytherapy Brachytherapy involves the direct implantation of sources of radiation into the tumor bed at the time of surgery or by stereotactic means shortly after resection. It is really a form of focused radiation therapy but is included here as it involves neurosurgical intervention. This treatment has the theoretical advantage of delivering high-dose radiotherapy to the tumor margin with concurrent minimization of radiation exposure to more distant tissue. Reports on the use of brachytherapy to treat gliomas first appeared in the late 1960s. The most contemporaneous evidence includes two randomized studies which suggest no significant benefit of brachytherapy in the setting of multimodality treatment of brain tumor patients with surgery, conventional radiation, and chemotherapy [26].

Chemotherapy Wafers Chemotherapy wafers are perhaps the best known surgical adjuvant treatment. Biodegradable wafers (Gliadel), which are placed in the tumor bed at the time of surgery have been developed. They are fabricated to release carmustine, a nitrosurea, in such a manner as to deliver high concentrations to the resection margin, where most recurrences occur, for several weeks. A large phase III trial with 240 patients with newly diagnosed malignant glioma randomized to receive either carmustine or placebo wafers at the time of primary surgical resection followed by radiation therapy demonstrated a modest survival benefit with median survival of 13.9 months

for the carmustine wafer-treated group and 11.6 months for the placebo treated group [27]. Similarly, in patients undergoing resection for recurrent GBM, placement of carmustine wafers only provided a modest prolongation of survival [28].

Enhancing Local Delivery A very promising approach for the delivery of drugs and other macromolecules to the brain explores the feasibility of using bulk flow within the brain extracellular fluid (ECF) space for the intracerebral distribution of agents [29]. This method has been called convection enhanced delivery (CED) and involves the placement within the brain parenchyma or tumor substance of one or more catheters that are subsequently connected to continuous infusion pumps. The pumps must be able to deliver the very low infusion rates that are critical for successful fluid ‘‘convection’’ within the brain; rates of infusion greater than a few ml/min will produce backflow along the catheter and loss of pressure, while too low a pressure will lead to failure of delivery altogether. An increasing body of animal and, more recently, human data shows that a much larger volume of distribution is achieved compared to previous delivery methods. CED is now the method of choice in a variety of phase I through phase III clinical trials investigating novel agents in CNS malignancy.

Surgical Adjuncts Although the clinical benefit of pursuing the goal of maximal tumor resection is unproven, the technology for achieving it is developing rapidly. Conventional surgery debulks tumors from within; resection is stopped when visually and palpably normal brain is reached, though this method of identifying the edge of a tumor is difficult and notoriously unreliable. Several technologies can


assist in extending the resection margin to its maximum. Image guided surgery involves the use of preoperative imaging data to identify the extent of tumor to be resected [30]. The equipment for this technique has become standard in neurosurgical units and consequently it is the most commonly used aid. Functional MRI data and subcortical tract location data from diffusiontensor MRI can be used with image guidance to assist in the definition of eloquent areas during surgery. The advantage of frameless stereotaxis is that the position of the pointer can be continually updated during the procedure to show the position of the pointer relative to the tumor and brain based on archival information (i.e., the MRI or CT done the morning of the procedure). The disadvantage of frameless stereotaxis is that the position is updated relative to archived (i.e., old) data and is not real-time; therefore, as the position of the tumor changes with loss of CSF, tumor resection, etc., the position of the pointer does not accurately reflect position. This makes the technique inaccurate, particularly at the end of a resection when it would potentially be of most utility in locating any residual tumor. Intraoperative imaging, with ultrasound, CT, or MRI, circumvents this problem [30]. New directions in frameless stereotaxis include pointer systems which are updated to reflect real-time position based on ultrasound and other inputted data sets. Despite relatively widespread use of technologies such as these, real data quantifying the incremental value of these imaging-based surgical adjuncts, vis-a`-vis improved patient outcomes, over the expense and other drawbacks of these systems, is still forthcoming. Awake craniotomy with intraoperative cortical mapping is an excellent treatment option as an alternative to craniotomy under general anesthesia for the routine surgical management of patients with supratentorial intra-axial tumors in whom maximal safe tumor resection is desirable. The advantages of awake craniotomy

47

with cortical mapping over standard craniotomy performed under general anesthesia include the opportunity for brain mapping to identify functional cortical areas and descending subcortical motor, sensory, and language white matter tracts, in an attempt to decrease neurological complications and maximize the extent of safe tumor resection. Awake craniotomy also avoids the use of general anesthesia with its attendant morbidity. The latter is often related to invasive monitoring techniques such as arterial lines, central venous lines, and urinary catheters as well as prolonged immobilization. A recent large series shows the benefits of awake craniotomy used non-selectively for intra-axial tumors and delineates the complication profile as a function of whether positive mapping was or was not encountered [31].

Role of Radiation Therapy Radiotherapy plays a vital role in the treatment of central nervous system malignancy, secondary only to surgery. Radiation therapy has been shown to extend survival time and improve quality of life in patients with primary intra-axial brain tumors. There have been major technological advances in both the delivery of radiotherapy and in diagnostic imaging in the last 5–10 years which have refined the delivery of radiation and have opened up new possibilities for the targeted delivery of a lethal dose of irradiation to the tumor site whilst sparing normal surrounding brain.

Conventional Radiation Post-operative external beam radiotherapy is well supported by randomized studies and remains standard therapy in the comprehensive care of brain tumor patients. Before the computed tomography (CT) and magnetic resonance imaging (MRI) era, many reports on the management of malignant

741

742

47


intra-axial brain tumors employed whole brain irradiation. However, the last 20 years have seen a definite shift away from utilizing whole brain fields to the use of regional fields with margins around enhancing disease of the order of 2 cm. Numerous factors are responsible for this shift in treatment philosophy-included amongst them are the better tumor localization afforded by CT and MR imaging, the many reports documenting that the primary cause of treatment failure was related to tumor recurrence at the original site in over 90% of cases [32], and the wish to reduce radiation related morbidity associated with whole brain irradiation. Based upon data from two randomized controlled trials [33,34], most centers have now eliminated the use of whole brain radiation for the adjuvant treatment of malignant intra-axial brain tumors following surgery in favor of local radiation fields (i.e., the high dose volume being the enhancing primary plus 2 cm margins) for the whole course of treatment, with no apparent difference in survival. The use of conformal radiotherapy techniques at most centers nowadays allows for accurate targeting of the residual tumor volume and/or the tumor bed with sparing of surrounding normal tissue. The pattern of local recurrence observed in malignant intra-axial brain tumors has led to an investigation of ways of intensifying the dose of radiation delivered by conventional means in an effort to improve local control rates. An increase in radiation dose through conventional fractionation and hyperfractionation regimens have been investigated. With respect to the results of dose escalation with conventional fractionation, no randomized studies of such protocols have shown any convincing benefit of radiation doses as high as 70 Gy compared with conventional doses in the range of 50–60 Gy [35,36]. Accordingly, the evidence would support the use of post-operative radiotherapy to a total dose in the range of 50–60 Gy utilizing conventional fractionation (60 Gy in 30 fractions over 6

weeks), particularly in view of the fact that higher doses are likely associated with higher toxicity. Hyperfractionation involves the use of a larger number of smaller sized fractions to a total dose which is higher than with conventionally administered irradiation in the same overall treatment time. Normal glial and vascular cells limit the total amount of irradiation that can be administered. These cells divide very slowly, and are better able to repair sub-lethal damage than neoplastic cells. Consequently, there might be an advantage to administering multiple smaller sized fractions to a higher total dose, the theory being that the improved repair of sub-lethal damage at lower sized fractions might allow a higher total dose to be associated with the same degree of late sequelae. Neoplastic cells are relatively rapidly dividing cells, and the increased number of daily fractions would increase the chance of radiating them at a more sensitive phase of their cell cycle. At smaller radiation doses per fraction, cell killing is less dependant on oxygen, which might be advantageous given the known areas of hypoxia in these tumors. Although investigators were able to safely escalate the dose to 72 Gy utilizing hyperfractionation, randomized studies did not demonstrate any advantage over conventionally fractionated doses in the range of 50–60 Gy [37,38]. No other modifications of the fractionation schedule, including accelerated fractionation, hypofractionation, or any combination of these, have produced any meaningful survival advantage [39].

Stereotactic Radiosurgery Stereotactic radiosurgery has similarities to surgery – a single large dose of radiation that is destructive to all tissue is given to a localized and accurately defined area. Unlike surgery, the effect of stereotactic radiosurgery is not immediate, taking weeks to months to develop. Although it has not been rigorously applied as


an alternative to surgery, several studies have investigated the role of stereotactic radiosurgery as an adjuvant to surgery plus fractionated radiotherapy. Although nonrandomized data suggested this treatment may improve survival [40], the results of a randomized trial examining the use of a radiosurgery boost to the tumor bed followed by external beam radiotherapy and BCNU chemotherapy as compared with external beam radiotherapy and BCNU chemotherapy alone found that it did not confer benefit in terms of overall survival, quality of life, or patterns of failure, and may in fact be associated with an increased incidence of radiation toxicity [41]. These results might be expected from the two older randomized brachytherapy studies – gliomas are simply not focal diseases lending themselves to cure by focal treatments. Presently, the role of stereotactic radiosurgery in the setting of malignant intra-axial brain tumors remains experimental. In order to better clarify its role in the overall management of these patients, with both newly diagnosed and recurrent tumors, several multi-center randomized controlled trials are currently underway [42].

47

unclear. By common consent, the combination of procarbazine, CCNU, and vincristine (PCV) was adopted as the ‘‘gold standard’’ in spite of there being relatively little firm evidence in support of this. The most frequent study design incorporated nitrosourea based chemotherapy as adjuvant to standard surgery and radiotherapy in comparison to standard treatment alone. Most studies were underpowered and failed to show any survival advantage [43,44]. In 2002, the glioma metaanalysis trialist group [45] carried out a systematic review of individual patient data from 12 such trials whose data were sufficiently homogeneous to allow such an analysis to be performed. Radiotherapy doses ranged from 40–60 Gy and volumes varied from whole brain to tumor only with a margin. All patients received a nitrosourea, some as single agent, and others as combination therapy. There was a statistically significant increase in median survival from 10 to 12 months, equivalent to a 6% absolute improvement in survival at 1 year (from 40 to 46%). This was just sustained at 2 years but had effectively vanished at 3 years.

Tailored Chemotherapy Chemotherapy A significant role for adjuvant chemotherapy in the majority of malignant intra-axial brain tumors is difficult to demonstrate. Until relatively recently, its major role in the treatment of most high grade intra-axial brain tumors has been as a palliative option for recurrent disease following conventional treatment, although many centers have routinely included chemotherapy as an integral part of the initial adjuvant therapy for anaplastic gliomas. The nitrosoureas (BCNU, CCNU) have historically played a central role in the chemotherapy of gliomas. However, there are very few studies in which either single agents or combinations were compared and optimum treatment is therefore

It has been intuitively recognized for some years that there exists a differential chemosensitivity among different histologic classes of malignant intra-axial brain tumors, and even differential chemosensitivity between tumors that appear histologically indistinguishable. Recently, the interrogation of brain tumors at a molecular level has been able to extract objective molecular differences between histologically similar tumors, giving a foundation to the burgeoning line of inquiry involving molecular predictors of response. Two seminal examples of this are the stratification of PCV chemotherapy response in anaplastic oligodendroglioma based upon loss of heterozygosity at chromosomes 1p and 19q, and the differential response of patients with glioblastoma multiforme

743

744

47


(GBM) to temozolamide chemotherapy predicated upon the presence or absence of epigenetic silencing of the MGMT gene.

Anaplastic Oligodendroglioma Cairncross and colleagues were the first to show that in patients with relapsed anaplastic oligodendroglioma, the deletion of the short arm of chromosome 1 and the long arm of chromosome 19 (1p-, 19q-) was a common finding, and, when present, was associated with pronounced chemosensitivity to procarbazine, lomustine (CCNU) and vincristine (PCV). Indeed, patients with loss of heterozygosity at 1p and 19q enjoyed prolonged remissions, sometimes lasting several years [46]. Based upon this initial observation in relapsed anaplastic oligodendroglioma, a large multi-center trial investigating the addition of PCV chemotherapy to radiotherapy in newly diagnosed anaplastic oligodendroglioma has been undertaken [47]. Although an increase in progression free survival in the PCV arm was observed, this did not translate into an increase in overall survival. However, patients with 1p/ 19q loss demonstrated a clearly better outcome, with median survival over 6–7 years as compared to 2–3 years in patients without 1p/19q loss. Because of the exquisite chemosensitivity of the variant of this tumor in which deletions of 1p and 19q are present, many clinicians today advocate upfront chemotherapy with either PCV or temozolamide as first-line adjuvant treatment in 1p/19q loss oligodendroglioma, whilst reserving radiotherapy for relapse, despite strong evidence in favor of this approach from a well designed and conducted randomized controlled trial.

Glioblastoma Multiforme Temozolamide first demonstrated its activity in patients with recurrent high-grade glioma. Three

pivotal phase II studies with identical entry criteria were conducted for patients with GBM and with anaplastic astrocytoma (AA). Despite disappointingly low objective response rates in GBM of 5 and 7%, respectively, but with an interesting response rate of 35% in the AA trial, these studies suggested an increase in the fraction of patients being progression-free at 6 months compared to a historical database [48]. Based on the activity in recurrent glioma, as well as in vitro data suggesting additive or supraadditive activity when temozolamide was administered concomitantly with radiotherapy, the European Organization for Research and Treatment of Cancer (EORTC) and the National Cancer Institute of Canada (NCIC) Clinical Trials Group compared the combination of temozolamide plus radiotherapy to standard radiotherapy alone in a large prospective randomized phase III trial on 573 patients in which the primary outcome was overall survival [49]. This study unequivocally demonstrated that the combination of temozolamide and radiotherapy followed by up to six cycles of adjuvant temozolamide improved survival. With combined modality treatment, the 2-year survival increased from 10 to 26%. Subgroup analyses suggested that patients of all age groups benefited from this treatment. Overall, the combined treatment was well tolerated, and the main reason for early discontinuation was disease progression. The clinical relevance of the mechanistic implication of the functional status of the DNA repair enzyme MGMT in alkylating chemotherapy was investigated within the context of this randomized EORTC/NCIC trial. Samples from 206 patients were analyzed to ascertain the methylation status of the MGMT gene promoter using methylation-specific PCR [50]. In 45% of tumor samples, the MGMT gene promoter was methylated, and the gene was silenced; this epigenetic silencing of a critical DNA repair gene greatly diminished the tumor cell’s ability to repair the DNA damage induced by the alkylating


chemotherapy. Patients with a silenced MGMT gene, which was, in essence, a predictor of chemosensitivity, indeed had longer survival. Breakdown of the data by treatment strongly suggested that the MGMT methylation status was a predictive marker for benefit from temozolamide chemotherapy. For patients who received combined radiotherapy and temozolamide, the 2-year survival rate was 46% when their tumor had a methylated MGMT promoter, in contrast to only 14% in patients with an unmethylated MGMT gene promoter. Thus, in this molecularly defined subgroup, temozolamide was found to be even more effective, increasing the median survival from 9 months to 21.7 months, while patients with an unmethylated MGMT gene promoter had little, if any, temozolamidederived benefit, with a median survival of 12.7 months. These data suggest that only patients whose tumors harbor a methylated MGMT promoter (and therefore, a silenced MGMT gene) should receive temozolamide, while for other patients, alternative strategies should be considered. Strategies aiming at overcoming MGMT-mediated treatment resistance are currently under development. Continuous administration of temozolamide, for instance, has been shown to deplete intracellular MGMT, and novel dose-dense schedules are currently being explored as adjuvant treatments in large randomized trials.

New Directions Despite significant gaps which persist in our understanding, genomic and proteomic technologies have provided a wealth of information regarding the clinical and biological behavior of malignant intra-axial brain tumors, the genetic pathways involved in their genesis, and the nature and role of prototypic alterations in these pathways. The challenge now is to integrate all of this knowledge in an interdisciplinary way,

47

involving the expertise of clinicians, epidemiologists, and basic scientists, in order to fully understand this complex disease. One of the most perplexing issues remains to comprehend, and to hopefully take advantage of, the signature molecular heterogeneity of these tumors, a feature which is in large part responsible for their resistance to therapy. On the one hand, malignant gliomas harbor a wide array of molecular defects, yielding a large number of potentially exploitable, therapeutic targets. On the other hand, given the diversity and number of molecular defects in these tumors, as well as the known redundancy and cross-talk between aberrant signal transduction pathways, one would predict that inhibition of a single target is unlikely to have a major, durable antitumor effect in most instances. Indeed, results of first generation clinical trials, conducted to evaluate a wide array of molecular targeted agents as monotherapeutics, support this prediction. Molecular profiling of malignant intra-axial tumors has uncovered some key molecules and pathways that, if perturbed, could disrupt the hallmark characteristics of a neoplastic cell, namely the propensity for uncontrolled replication, the capacity for angiogenesis, and the ability to migrate and invade adjacent normal tissue. Although a detailed review of the multiplicity of pathways and potential targets is beyond the scope of this chapter, the interested reader is referred to several review articles on this topic [1,51]. What follows is a brief discussion of a representative example of a small molecule that upsets, in a targeted way, one of the key attributes of a neoplastic cell, for which some clinical data regarding its potential efficacy is available.

Inhibition of Cellular Proliferation In malignant gliomas, the most prominent altered tyrosine kinase receptor is the epidermal growth factor receptor (EGFR), which is amplified and

745

746

47


overexpressed in approximately 50% of GBMs. The EGFR tyrosine kinase inhibitor gefitinib has been investigated in several studies in malignant glioma. The combination of gefitinib and radiotherapy was examined in a phase I/II trial with a total of 147 patients with newly diagnosed GBM [52]. Median survival for the whole population was only 11 months, comparable to that of historical controls receiving radiation treatment alone. Despite these rather disappointing preliminary results, it is possible that such a treatment strategy will work only in selected patients who have tumors that are strictly dependent on EGFR signaling for growth and survival. Alternatively, because of the redundancy of the pathway, additional targets may need to be considered in a combination therapeutic strategy. An alternative approach targeting the EGFR pathway employs a vaccination strategy. Promising results have been reported from a phase II trial of patients with newly diagnosed GBM vaccinated with the anti-EGFRvIII peptide CDX-110 in the temozolamide maintenance phase of a temozolamide/radiotherapy combination protocol [53]. Complete absence of EGFRvIII in the tumor tissues at recurrence suggested elimination of such cells by an activated immune system.

Inhibition of Angiogenesis Malignant gliomas are highly vascularized and infiltrative tumors strongly dependent on endothelial cell proliferation regulated by proangiogenic cytokines (e.g., vascular endothelial growth factor [VEGF]). Preclinical and clinical evidence suggests that antiangiogenic treatment may work best in combination with chemotherapy and radiotherapy, and a number of early phase clinical trials have been initiated. A recent clinical study with correlative imaging and biologic end points suggests that VEGF receptor inhibition by cediranib leads to normalization of tumor vasculature and restoration of the blood-brain barrier, thus reducing

contrast enhancement and edema [54]. In a recent trial, the VEGF receptor tyrosine kinase inhibitor vatalanib (which also inhibits the PDGF receptor [PDGFR]) is being evaluated in combination with radiotherapy and temozolamide chemotherapy [55]. Preclinical observations of potential increase in hypoxia and radiation resistance raise concern as to the optimal timing of antiangiogenic therapy. A further phase II trial therefore is designed to evaluate the safety and efficacy of the addition of vatalanib either concurrent with temozolamide and radiotherapy or by adding the VEGF receptor inhibitor only after completion of concomitant chemoradiotherapy [56].

Inhibition of Cell Migration Integrins are heterodimer transmembrane receptors for the extracellular matrix, regulating cell adhesion and migration. In preclinical models, inhibition of integrin function efficiently suppressed tumor cell migration and inhibited tumor progression. Cilengitide, a synthetic peptide, competitively inhibits integrin binding to extracellular matrix proteins. In a phase I study in recurrent glioma, single-agent activity of cilengitide was observed, with objective radiographic responses noted [57]. A subsequent phase II trial of cilengitide added to a standard temozolamide/ radiotherapy regimen (temozolamide with concomitant radiotherapy, followed by temozolamide with or without cilengitide) has recently been completed in patients with newly diagnosed GBM. Addition of the novel agent was associated with little or no additional toxicity, and initial results suggest efficacy in a subgroup of patients.

References 1. Furnari FB, Fenton T, Bachoo RM, Mukasa A, Stommel JM, Stegh A, Hahn WC, Ligon KL, Louis DN, Brennan C, Chin L, DePinho RA, Cavenee WK. Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev 2007;21:2683-710.


2. Greene GM, Hitchon PW, Schelper RL, Yuh W, Dyste GN. Diagnostic yield in CT-guided stereotactic biopsy of gliomas. J Neurosurg 1989;71:494-7. 3. Bernstein M, Parrent AG. Complications of CT-guided stereotactic biopsy of intra-axial brain lesions. J Neurosurg 1994;81:165-8. 4. Kulkarni AV, Guha A, Lozano A, Bernstein M. Incidence of silent hemorrhage and delayed deterioration after stereotactic brain biopsy. J Neurosurg 1998;89:31-5. 5. Soo TM, Bernstein M, Provias J, et al. Failed stereotactic biopsy ina series of 518 cases. Stereotact Funct Neurosurg 1995;64:183-96. 6. Kongkham PN, Knifed E, Tamber MS, Bernstein M. Complications in 622 cases of frame-based stereotactic brain biopsy – a decreasing procedure. Can J Neurol Sci 2008;35(1):79-84. 7. Coffey RJ, Lundsford LD, Taylor FH. Survival after stereotactic biopsy of malignant gliomas. Neurosurgery 1988;22:465-473. 8. Vuorinen V, Hinkka S, Farkkila M, Jaaskelainen J. Debulking or biopsy of malignant glioma in elderly people – a randomized study. Acta Neurochir 2003;145:5-10. 9. Whittle IR, Pringle AM, Taylor R. Effects of resective surgery for left sided intracranial tumours on language function: a prospective study. Lancet 1998;351:1014-18 10. Castro MG, Cowen R, Williamson IK, et al. Current and future strategies for the treatment of malignant brain tumors. Pharmacol Ther 2003;98:71-108. 11. Muacevic A, Kreth FW. Quality-adjusted survival after tumor esection and/or radiation therapy for elderly patients with glioblastoma multiforme. J Neurol 2003; 250:561-8. 12. Keles GE, Lamborn KR, Chang SM, Prados MD, Berger MS. Volume of residual disease as a predictor of outcome in adult patients with recurrent supratentorial glioblastomas multiforme who are undergoing chemotherapy. J Neurosurg 2004;100:41-6. 13. Keles GE, Anderson B, Berger MS. The effect of extent of resection on time to tumor progression and survival in patients with glioblastoma multiforme of the cerebral hemisphere. Surg Neurol 1999;52:371-9. 14. Laws ER, Parney IF, Huang W, et al. Survival following surgery and prognostic factors for recently diagnosed malignant glioma: data from the Glioma Outcomes Project. J Neurosurg 2003;99:467-73. 15. Lacroix M, Abi-Said D, Fourney DR, et al. A multivariate analysis of 416 patients with glioblastoma multiforme: prognosis, extent of resection, and survival. J Neurosurg 2001;95:190-8. 16. Curran WJ Jr, Scott CB, Horton J, et al. Does extent of surgery influence outcome for astrocytoma with atypical or anaplastic foci (AAF)? A report from three Radiation Therapy Oncology Group (RTOG) trials. J Neurooncol 1992;12:219-27. 17. Kreth FW, Berlis A, Spiropoulou V, et al. The role of tumor resection in the treatment of glioblastoma multiforme in adults. Cancer 1999;86:2117-21.

47

18. Albert FK, Forsting M, Sartor K, Adams HP, Kunze S. Early postoperative magnetic resonance imaging after resection of malignant glioma: objective evaluation of residual tumor and its influence on regrowth and prognosis. Neurosurgery 1994;34:45-60. 19. Grant R, Metcalfe SE. Biopsy versus resection for malignant glioma. Cochrane Database Syst Rev 2005; 2:CD002034. 20. Hess KR. Extent of resection as a prognostic variable in the treatment of gliomas. J Neurooncol 1999; 42: 227-31. 21. Rosenthal MA, Kavar B, Hill JS, et al. Phase I and pharmacokinetic study of photodynamic therapy for high-grade gliomas using a novel boronated porphyrin. J Clin Oncol 2001;19:519-24. 22. Germano IM, Fable J, Gultekin SH, Silvers A. Adenovirus/herpes simplex-thymidine kinase/ganciclovir complex: preliminary results of a phase I trial in patients with recurrent malignant gliomas. J Neurooncol 2003; 65: 279-89. 23. Rainov NG. A phase III clinical evaluation of herpes simplex virus type 1 thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical resection and radiation in adults with previously untreated glioblastoma multiforme. Hum Gene Ther 2000;11: 2389-401. 24. Immonen A, Vapalahti M, Tyynela K, et al. AdvHSV-tk gene therapy with intravenous ganciclovir improves survival in human malignant glioma: a randomised, controlled study. Mol Ther 2004;10:967-72. 25. Farkkila M, Jaaskelainen J, Kallio M, et al. Randomised controlled study of intratumoral recombinant gammainterferon treatment in newly diagnosed glioblastoma. Br J Cancer 1994;70:138-41. 26. Laperriere NJ, Leung PM, McKenzie S, et al. Randomized study of brachytherapy in the initial management of patients with malignant astrocytoma. Int J Radiat Oncol Biol Phys 1998;41:1005-11. 27. Westphal M, Hilt DC, Bortey E, Delavault P, Olivares R, Warnke PC, et al. A phase 3 trial of local chemotherapy with biodegradable carmustine (BCNU) wafers (Gliadel wafers) in patients with primary malignant glioma. Neurooncol 2003;5:79-88. 28. Brem H, Piantadosi S, Burger PC, Walker M, Selker R, Vick NA, et al. Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent malignant gliomas. The Polymer-Brain Tumor Treatment Group. Lancet 1995;345:1008-12. 29. Kioi M, Husain SR, Croteau D, Kunwar S, Puri RK. Convection-enhanced delivery of interleukin-13 receptordirected cytotoxin for malignant glioma therapy. Technol Cancer Res Treat 2006;5:239-50. 30. McDermott MW, Bernstein M. Image-guided Surgery. In: Bernstein M, Berger MS, editors. Neuro oncology. The essentials, 2nd ed. New York: Thieme; 2008. p. 112-23.

747

748

47


31. Serletis D, Bernstein M. Prospective study of awake craniotomy used routinely and nonselectively for supratentorial tumors. J Neurosurg 2007;107:1-6. 32. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glioblastoma. Neurology 1980;30:907-11. 33. Shapiro WR, Green SB, Burger PC, et al. Randomized trial of three chemotherapy regimens and two radiotherapy regimens in postoperative treatment of malignant glioma. Brain Tumor Cooperative Group Trial 8001. J Neurosurg 1989;71:1-9. 34. Kita M, Okawa T, Tanaka M, et al. Radiotherapy of malignant glioma – prospective randomized clinical study of whole brain versus local irradiation (in Japanese). Gan No Rinsho 1989;35:1289-94. 35. Bleehen NM, Stenning SP. A Medical Research Council trial of two radiotherapy doses in the treatment of grades 3 and 4 astrocytoma. The Medical Research Council Brain Tumour Working Party. Br J Cancer 1991;64:769-74. 36. Nelson DF, Diener-West M, Horton J, et al. Combined modality approach to treatment of malignant gliomas – re-evaluation of RTOG 7401/ECOG 1374 with long-term follow-up: a joint study of the Radiation Therapy Oncology Group and the Eastern Cooperative Oncology Group. NCI Monogr 1988;6:279-84. 37. Scott CB, Curran WJ, Yung WKA. Long term results of RTOG 9006: a randomized study of hyperfractionated radiotherapy (RT) to 72.0 Gy and carmustine versus standard RT and carmustine for malignant glioma patients with emphasis on anaplastic astrocytoma (AA) patients (abstract). Proc Annu Meet Am Soc Clin Oncol 1998;17:401a. 38. Shin KH, Urtasun RC, Fulton D, et al. Multiple daily fractionated radiation therapy and misonidazole in the management of malignant astrocytoma. A preliminary report. Cancer 1985;56:758-60. 39. Glinski B. Postoperative hypofractionated radiotherapy versus conventionally fractionated radiotherapy in malignant gliomas. A preliminary report on a randomized trial. J Neurooncol 1993;16:167-72. 40. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin LS. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002;50:41-6. 41. Souhami L, Seiferheld W, Brachman D, et al. Randomized comparison of stereotactic radiosurgery followed by conventional radiotherapy with carmustine to conventional radiotherapy with carmustine for patients with glioblastoma multiforme: report of Radiation Therapy Oncology Group 93-05 protocol. Int J Radiat Oncol Biol Phys 2004;60:853-60. 42. Tsao MN, Mehta MP, Whelan TJ, Morris DE, Hayman JA, Flickinger JC, et al. The American Society for Therapeutic Radiology and Oncology (ASTRO) evidence-based review of the role of radiosurgery for malignant glioma. Int J Radiat Oncol Biol Phys 2005;63:47-55. 43. Walker MD, Alexander E, Hunt WE, et al. Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

gliomas. A cooperative clinical trial. J Neurosurg 1978; 49:333-43. Walker MD, Green SB, Byar DP, et al. Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 1980;303:1323-29. Glioma Meta-Analysis Group. Chemotherapy in adult high-grade glioma: a systematic review and metaanalysis of individual patient data from 12 randomised trials. Lancet 2002;359:1011–18. Cairncross JG, Ueki K, Zlatescu MC, et al. Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst 1998;90:1473-79. Cairncross JG, Seiferheld W, Shaw E, et al. An intergroup randomized controlled clinical trial (RCT) of chemotherapy plus radiation (RT) versus RT alone for pure and mixed anaplastic oligodendrogliomas: initial report of RTOG 94– 02. Proc Am Soc Clin Oncol 2004;23:107 (abstract 1500). Brada M, Hoang-Xuan K, Rampling R, et al. Multicenter phase II trial of temozolomide in patients with glioblastoma multiforme at first relapse. Ann Oncol 2000; 12:259-66. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352:987-96. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 2005; 352:997-1003. Sathomsumetee S, Reardon DA, Desjardins A, Quinn JA, Vredenbrugh JJ, Rich JN. Molecularly targeted therapy for malignant glioma. Cancer 2007; 110:13-24. Chakravarti A, Berkey B, Robins H, et al. An update of phase II results from RTOG 0211: a phase I/II study of gefitinib with radiotherapy in newly diagnosed glioblastoma. J Clin Oncol 2006;24:64s (suppl; abstract 1527). Heimberger A, Hussain S, Aldape K, et al. Tumor-specific peptide vaccination in newly diagnosed patients with GBM. J Clin Oncol 2006;24:107s (suppl; abstract 2529). Batchelor TT, Sorensen AG, di Tomaso E, et al. AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell 2007;11:83-95. Brandes A, Stupp R, Hau P, et al. EORTC Study 26041– 22041:Phase I/II study on concomitant and adjuvant zemozolomide (TMZ) and radiotherapy with or without PTK787/ZK222584 (PTK/ZK) in newly diagnosed glioblastoma: results of Phase I. J Clin Oncol 2007; 25:81s, (suppl; abstract 2026). Nabors B, Mikkelsen T, Rosenfeld S, et al. A phase I and correlative biology study of cilengitide in patients with recurrent malignant glioma. J Clin Oncol 2007; 25:1651-7. Stupp R, Goldbrunner R, Neyns B, et al. Phase I/IIa trial of cilengitide (EMD121974) and temozolomide with concomitant radiotherapy, followed by temozolomide and cilengitide maintenance therapy in patients with newly diagnosed glioblastoma. J Clin Oncol 2007;25:75s, (suppl; abstract 2000).

40 CT in Image Guided Surgery D. Kondziolka . L. D. Lunsford

The last three decades in neurosurgery have been the era of image-guided surgery. Neurosurgeons who were trained to perform surgery without sophisticated parenchymal imaging, were used to large craniotomies, free-hand biopsies, palpation of the brain surface, and interpretation of angiograms, x-rays, or air studies. Computed tomography, as the first high-resolution imaging tool to study brain tissue, ushered in an era of tremendous possibility. Neurosurgeons are now trained in stereotactic technology, image-guided navigation, more precise lesion resection, accurate and reliable deep brain surgery, and are accustomed to seeing an imaging device within an operating room. In this chapter, we review the current use of intraoperative CT imaging. At our center, the first CT compatible stereotactic head frame, in collaboration with industry, was constructed in 1978 and utilized in 13 patients [1,2]. Virtually no brain operation in today’s era is done without image guidance. This includes craniotomy for most non trauma indications, conventional stereotactic surgery, functional surgery, and radiosurgery [3]. During this interval, the newly redesigned Leksell CT compatible stereotactic head frame [4] was used for dedicated brain biopsies under the direction of its inventor, Lars Leksell. Several groups were working on devices to allow accurate CT-based stereotactic surgery [5]. Upon returning to Pittsburgh in 1981, Dr. Lunsford began a program to obtain a dedicated CT scanner equipped operating room at a time when the regional health systems agency had allocated two diagnostic CT scanners to the entire city of Pittsburgh (> Figure 40‐1) We built a new operating room and began our first series of stereotactic #


procedures [6–15]. The scanner was updated to a GE 9800 in 1991, and a new spiral scanner in 2007. Neurosurgical technologies should be simple and practical. It must assist the surgeon to perform the operation, it should be efficient, it must promote better outcomes, and it must reduce morbidity. Intraoperative imaging using CT and then MRI technologies fulfilled these goals, although MRI has posed unique challenges to surgical efficiency.

Frame Based Stereotactic Surgery Our experience from 1977 to 2007 using frame based stereotactic surgery is shown in > Table 40‐1. Many procedures used CT imaging for ‘‘open’’ (when a surgical opening was made through the skull) stereotactic surgery. We added MRI based target localization in selected cases beginning in 1991 [16,17]. As the resolution of MR imaging improved, we increasingly relied on MRI for functional neurosurgery in patients eligible for MRI [18,19]. We also used MRI for selected biopsies, especially those in high risk locations or when (> Figure 40‐2) their imaging characteristics were best defined by MRI [19–22]. However, for brain tumor biopsy, cyst management or intracavitary irradiation, CT remains the imaging modality of choice. All of our own patients undergo stereotactic frame placement in an operating room environment followed by imaging either in our dedicated CT operating room or a nearby diagnostic 1.5 Tesla MRI unit. Calculation of coordinates is performed either using the standard software available on the scanners,

620

40

CT in image guided surgery

. Figure 40‐1 The first dedicated therapeutic CT scanner for brain surgery was installed at the University of Pittsburgh Medical Center in 1982. The room had a dedicated GE 8800 CT scanner and a Phillips C-Arm image intensifier

. Table 40‐1 Morphologic stereotactic Presbyterian; 1979–2007

brain

surgery:

UPMC-

No. patients Diagnostic biopsy Cyst aspiration Radiation implant Brain abscess Catheter and cyst reservoir insertion Hematoma aspiration Frame based craniotomy Ventriculostomy placement Total

1,664 197 145 97 19 9 10 6 2,147

or more recently using an image integrated surgical planning system (SPS Elekta, Inc., Norcross, GA). We precisely pre-plot probe trajectories to reach the target and choose the route designed to avoid as many pial or ependymal surfaces as possible (> Figure 40‐3). Brain biopsies are performed in all areas of the brain including the brain stem [21–23]. Many lesions in the deep locations of the brain

can be approached from a single trans-frontal intra-axial trajectory. In selected cases for lesions adjacent to the fourth ventricle or within the cerebellar hemisphere, a transcerebellar trajectory is performed with the patient moved to a semi-sitting position on the operating room table. For virtually all patients older than 12 years the procedures are performed under monitor assisted conscious sedation. We have performed CT stereotactic biopsy in a child as young as 5 months [24]. CT imaging is always performed with contrast-enhancement, even with non-enhancing lesions, in order to identify blood vessels along the planned trajectory. Axial images are obtained at slice intervals of 1.5–3 mm depending on the size and location of the target. The image thickness is typically 3–5 mm to provide adequate signal. For frame-based biopsy, it is important not to place the stereotactic frame pins at the plane of the target, or metallic artifact can obscure the image.


40

. Figure 40‐2 Stereotactic frame-based biopsy for precision and reliability. Certain lesions are best seen with MRI. CT imaging (left) showed poor contrast enhancement and pin artifacts compared to long relaxation time MRI (right) for this thalamic astrocytoma

Although ‘‘frameless’’ biopsy has increased in popularity, we think that the concept of general anesthesia, three point (Mayfield) rather than four point fixation, frequent use of a burrhole, a larger incision rather than a twist drill craniostomy puncture site, and other often more complex tools to hold the probe, seem counterintuitive. Reaching a final diagnosis involves a collaborative effort with the neuropathologic team, including review of the pre and postoperative images. A confirmatory intraoperative CT scan is always done, which shows a small air density at the target site. This image is used to demonstrate to the neuropathologist that sampling was accurate. In addition, immediate imaging helps rule out hemorrhage. In our experience, the current risk of a brain biopsy hemorrhage requiring evacuation using frame based techniques is less than 0.5% [25]. Frame based systems provide extreme accuracy to less than 1 mm as mandated by standards of the American Society of Testing and Manufacturing. Rigid probe fixation is critical. In addition, we are able to precisely reduplicate the

pathway of the needle previously plotted using the surgical planning system.

Brain Cyst Management CT imaging is valuable in cyst drainage. > Table 40‐2 shows our experience with stereotactic cyst aspirations. CT is used to calculate the cyst volume using available imaging software, and to choose the safest trajectory into the cyst. For cyst drainage, the target should be in the dependent portion of the cyst according to supine head position so that gravity facilitates maximum drainage. We typically allow cyst fluid to egress on its own and always collect the volume to compare to the pre-surgical calculation. For more viscous fluid, we aim to evacuate 70–80% of the contents. The alleviation of brain cysts by simple aspiration has been done in a large variety of patients, including those with glial cysts, neuroepithelial cysts, and in the management of cystic craniopharyngioma [6,26–30]. Simple aspiration of craniopharyngoma cysts

621

622

40


. Figure 40‐3 Surgical planning with SurgiPlan software assists the use of CT imaging to define the optimal probe trajectory, with angles and coordinates provided for the stereotactic frame and arc

. Table 40‐2 Stereotactic management of brain cysts: UPMC Presbyterian;1981–2007 No. patients Tumor cyst aspiration Colloid cyst aspiration Craniopharyngioma cyst aspiration Total

119 47 31 197

may be periodically needed, and we preferentially prefer this as opposed to placement of an Ommaya reservoirs drainage system. Using the fine needle (0.9 mm) puncture technique advocated by Backlund, we are able to introduce

colloidal chromic phosphate P32 into a craniopharyngioma to result in involution of the cyst over the course of time [31]. Cyst reservoir systems in our experience seem to irritate a craniopharyngioma leading to a progressively more frequent need for cyst aspiration. In contrast, A P32 injection with a dose of approximately 20,000 cGy offers an effective management for primarily monocystic craniopharyngomas >5 cc. in volume. As a pure beta emitter, the dose falloff is quite rapid, thereby keeping low the risk to surrounding critical structures such as the optic apparatus. > Table 40‐3 shows our experience with the stereotactic isotope strategies.


. Table 40‐3 Open stereotactic irradiation surgery: UPMC Presbyterian; 1981–2007 No. patients Intracavitary radiation (P-32): Craniopharyngioma 73 Glioma 13 Arachnoid cyst 4 Subtotal 90 Stereotactic interstitial radiation (I-125 Brachytherapy): Glioma 51 Ependymoma 2 Malignant meningioma 1 CNS lymphoma 1 Subtotal 55 Total 145

Image-Guided Craniotomy CT-guided, stereotactic frame-based craniotomies have been performed for more than 25 years. In the early 1980s we began to perform image guided resections using the Leksell frame in our intraoperative CT suite. Although certainly useful in finding the lesion, our initial experience failed to disclose a major benefit from intraoperative CT assisted cytoreductive glioma surgery [20,32–34]. We found little evidence that taking out the ‘‘majority’’ of a glial tumor yielded superior results. It is possible that patients eligible for extensive cytoreductive surgery, primarily those with lobar or polar tumors, may benefit from such an approach if their tumors have less infiltrative borders and are located in less critical locations. Such surgery must then be combined with additional adjuvant management strategies including radiation therapy, chemotherapy, and potentially radiosurgery. While we are confident that centers employing intraoperative MRI are making headway in terms of their ability to better resect glial neoplasms, we are, nonetheless, struck by the fact that a 99% (‘‘gross total’’) removal, still contains an enormous tumor cell population. Although subsequent adjuvant management may be better able to control such

40

tumors, this remains to be shown. In order to define such a benefit a large randomized trial would be needed to show a meaningful survival increase for gliomas.

Surgical Navigation for Resection While frame based surgical resection procedures including craniotomy were performed on our operating room CT scanner table, the introduction of frameless systems changed the paradigm to using preoperative imaging to guide most brain and many spine operations [3]. A variety of products were used successively including the early generation ISG wand and products sold by Elekta, Medtronic, and Stryker. The Surgiscope proved to be the most challenging of the image guidance products. Now almost all intracranial surgery requires precise CT or MRI guidance technologies. More recently, magnetoencephalography (MEG) and positron emission tomography (PET) imaging have been fused with MRI or CT for certain frame based procedures. We continue to believe that intraoperative CT is a practical imaging tool with high reliability. It adds to but does not modify the goal or flow of the surgical procedure. In addition, as a shared resource, intraoperative CT can be used by other surgical services. For neurosurgical procedures we fuse preoperative MRI scans with intraoperative CT to enhance further diagnostic, functional and endoscopic surgical procedures.

Abscess Management We have performed stereotactic aspiration for the diagnosis and simple or catheter assisted drainage of 97 patients with brain abscesses. Frame based stereotactic surgery is an excellent way to manage brain abscesses [9,12,35]. As in all surgical pyogenic infections, the principals include early recognition, drainage, and identification of the

623

624

40


appropriate organism where possible. A variety of techniques are available to manage brain abscesses including frame based and frameless approaches. We have treated most brain abscesses using a twist drill craniostomy, stereotactic puncture, followed by drainage of the abscess cavity. The bacteriological studies in most cases are able to define the causative organism(s). At the time of drainage, appropriate broad spectrum antibiotics are given in the operating room. We perform a full assessment of the potential sources for a pyogenic brain abscess, including heart, sinus or other septicemia sources. Rarely brain abscess cases require re-drainage if slow resolution is not forthcoming. The treatment regimen consists of appropriate antibiotic management, placement of a catheter in the abscess (for larger volumes), and gradual removal of the drainage catheter after several days. This approach has a high success rate, and virtual elimination of mortality from bacteriological brain abscesses. For patients with fungal abscesses in the context of chronic immune suppression, especially those in the transplant arena, biopsy and identification of the organism allows the potential for aggressive anti-fungal management. In our community, the need for brain biopsy in the diagnosis of HIV related conditions has declined. Patients with a progressive mass lesion within the brain in the context of acquired immunodeficiency syndrome (AIDS), normally undergo empiric antitoxoplasmosis treatment first. Biopsy is performed if the lesion progresses. On occasion, lymphoma requires brain biopsy for histological diagnosis, or for the ability to provide prognostic information to family members and patients who may have progressive multi-focal leukoencephalopathy (PML) associated with AIDS.

Functional Neurosurgery Functional neurosurgery is dependent on high resolution multiplanar imaging. In the 1980s

pain management using deep brain stimulation techniques (prior to the advent of MRI), CT based imaging proved to be quite accurate for identification of targets in the ventromedial thalamus or periaqueductal gray. We now prefer high resolution intraoperative MRI for recognition of both pallidal, thalamic, and subthalamic targets [16,18,36,37]. > Table 40‐4 provides a summary of our usage of intraoperative imaging for functional neurosurgery over the last 25 years. MRI provides superior anatomic resolution to clearly delineate white matter tracts and gray matter nuclei. Currently, stereotactic MRI is used primarily for electrode placement during movement disorder surgery, or for ablative surgery [37]. CT is used in patients who already have stimulation hardware or have a contraindication to MRI. Using fusion software we can merge preoperative MRIs with intraoperative CT for functional targeting if needed. We also have used CT-based targeting for cellular transplantation research in patients with stroke [38].

CT-based Stereotactic Endoscopic Surgery One of the most gratifying combinations of current technology has been the role of combined CT stereotactic frame-based assisted craniotomy coupled with endoscopic removal of selected . Table 40‐4 Functional stereotactic brain surgery: UPMC-Presbyterian; 1979–2007 No. patients Pallidotomy Thalamotomy Deep brain stimulation (movement disorders) Depth electrodes for seizures Deep brain stimulation (chronic pain) Cell transplantation Mesencephalotomy/Capsulotomy Total

147 72 98 95 24 20 4 460


deep seated colloid cysts and brain tumors [39,40]. Over the last 5 years, we switched to performing almost all colloid cyst removal or other intraventricular lesion removals using a small trephine craniotomy at the coronal suture region, followed by stereotactic placement of an ‘‘endoport’’ (> Figure 40‐4) [39]. Through this 11 mm. conduit, the visualization endoscope can be placed. Standard or specially constructed microsurgical instruments can be used to remove colloid cysts or brain tumors in the third or the lateral ventricles. The ability to do both intraoperative pre-planning CT imaging as well as intraoperative post-procedure CT has assisted us in our determination of adequate or complete resection of such lesions. Surgical morbidity remains low due to the limited transcortical dissection necessary, and the lack of a transcallosal section or retraction of the medial hemisphere.

Frameless CT- based Stereotactic Surgery Beginning in the early 1990s, we began to evaluate the first generation of frameless stereotactic systems to assist localization and resection of mass lesions of the brain. First using the ISG

40

wand system, a mechanically assisted arm, with image integration, we recognized early the value for lesion localization and surgical planning. Although we believe that the precision of frameless stereotactic surgery is less than frame based surgery, for many tumor craniotomies, this is sufficient. Using image-guidance, ideally placed and smaller craniotomies, using trephine or standard technique are the norm. We have used and evaluated almost every commercially available system. While each has merit, we continue to see new improvements such as image fusion (CT, MRI, fMRI, MEG) to define targets. We use imaging compatible fiducials attached to the patient’s scalp or new mask systems if the head is not rotated. At our university teaching hospital thousands of patients have undergone imageguided intracranial and skull base procedures. Various manufacturers have redesigned freehand or probe holders during image-guided MRI and CT and frameless procedures. These products are reminiscent of some of the old burr hole mounted systems first developed in the 1960s. Such devices are widely marketed for routine diagnostic or therapeutic procedures. These include brain tumor biopsies, cyst management, and now even deep brain stimulation electrode placement. Such procedures offer benefit over old

. Figure 40‐4 Stereotactic endoscopic resection of a colloid cyst is performed via an 11 mm conduit and a 22 mm trephine craniotomy

625

626

40


free-hand methods. However, in the present era, the benefit of frameless as opposed to frame-based stereotactic surgery is hard to assess. Frame based systems allow you to pre-plot precise probe placement to identify the exact target to be biopsied, and assure rigid fixation and delivery of the probe or device. Burr hole or otherwise mounted stereotactic frameless systems with trajectory recalculations during the actual probe passage may be associated with a higher risk of morbidity. The learning curve for frameless stereotactic systems may seem rapid, whereas the training and educational needs for a neurosurgeon to become proficient in frame-based stereotactic surgery may be greater. We believe that patients continue to benefit from the proper use of frame based systems, where exact planned trajectories are used and exact brain targets are manipulated. If one performs frameless stereotactic surgery, outcomes should be compared to established benchmarks using rigid frame fixation. This is especially true for deep seated lesions in the basal ganglia, thalamus, pons and medulla, and for lesions in other high risk locations of the brain such as the perisylvian area.

Development and Evolution of a Dedicated Stereotactic Operating Room In the late 1970s, our planning process for a new stereotactic operating room debated the value of bi-plane angiography versus conventional radiological imaging techniques. Fortunately, the rapid development of CT imaging clearly pointed to the need for a marriage between neurosurgical techniques and neuroimaging devices, elegantly proposed by Professor Erik-Olof Backlund. We recognized early that a dedicated operating room for stereotactic procedures was highly desirable. We were able to secure final approval for a certificate of need for a dedicated therapeutic CT scanner housed in our operating room in 1981.

A GE 8800 CT scanner was placed in a newly designed operating room. The scanner was inverted from its usual position, so that the patient and the head came through the back part of the scanner into the main area of the operating room [8,41,42]. We also integrated this with a ceiling mounted fluoroscopic system, which could be used to provide real time fluoroscopic imaging, which was done to assist certain CT guided stereotactic procedures including transsphenoidal approaches. Similarly, the C-arm fluoroscope was also used in over 1,000 patients who underwent percutaneous trigeminal neuralgia management (mostly glycerol rhizotomy) during the next 20 years. In 1991, the scanner was updated to a GE 9800. During this interval, several thousand patients underwent CT assisted stereotactic procedures. A routine stereotactic biopsy could be completed in 60 min. The entire procedure time was greatly improved by our lack of need to move the patient between surgical and diagnostic imaging sites. The life span of such an advanced technology operating room suite is only about 10 years, because of changes in the field and advances in surgical and imaging technologies. In particular, we felt that this was true related to the emerging and increasing role of endoscopic surgical management especially for intraventricular and skull base lesions. Accordingly, we set about re-designing a new operating room using new CT scan imaging technologies, eventually placing a 64 slice GE CT scanner with fluoroscopic capabilities (> Figure 40‐5). This expanded operating room suite also provides video-assisted surgical techniques for endoscopic surgery, permitting simultaneous minimally invasive procedures with almost real time imaging. Because we do not have any of the issues associated with high magnetic fields of MRI, the standard operation paradigm remained in place. No tools needed to be redesigned, no special instruments needed to be created, and anesthesia services remained the same. Again, our goal was


40

. Figure 40‐5 The new dedicated CT stereotactic operating room at UPMC Presbyterian (Courtesy of Burt Hill Architects)

to enhance the operation, not to create a difficult working environment foreign to the best current neurosurgical procedures. We evaluated the role of intraoperative MRI scan over these years, but were concerned about low resolution of the resistive 0.1 and 0.3 Tesla versions. Most of the prototype units had poor image quality, and limited field of view. Certainly new systems have been developed based on pioneering efforts done in Calgary, Minneapolis, Newark, New Jersey, and in Erlangen, Germany. Numerous investigators have shown the potential abilities to better resect certain tumors, although conclusive data is lacking. As noted above, we think the primary thrust of intraoperative MRI to facilitate aggressive resection of glial tumors is problematic. Perhaps a small minority of clinically recognized gliomas, those with circumscribed tumors in polar or lobar locations, may have better outcomes. Overall, we continue to use high resolution imaging with CT. The newest device has dramatic improvement in acquisition times, reformatting in multiple planes, fusion techniques, and

real time CT fluoroscopy. Our new operating room allows us to fuse preoperative or intraoperative MRI’s and intraoperative and postoperative CT, especially valuable in functional neurosurgical cases where electrodes are being implanted. This is facilitated by using the Surgical Planning System SPS (Elekta, Inc.) which allows image fusion of sterereotactic, non-stereotactic, and other imaging output such as MEG and PET. Image-guided neurosurgery requires technologies that assist our abilities to deal with a wide variety of brain and spine problems and to reduce patient morbidity. In today’s era, the practicing surgeon or surgeon in training should never have to ask the question, ‘‘Where is it?,’’ that was so common years ago.

References 1. Perry JH, Rosenbaum AE, Lunsford LD, Swink CA, Zorub DS. Computed tomography-guided stereotactic surgery: conception and development of a new stereotactic methodology. Neurosurgery 1980;7:376-81.

627

628

40


2. Rosenbaum AE, Lunsford LD, Perry JH. Computerized tomography guided stereotaxis: a new approach. Appl Neurophysiol 1980;43:172-3. 3. Kondziolka D, Lunsford LD. Intraoperative navigation during resection of brain metastases. Neurosurg Clin N Am 1996;7:267-77. 4. Leksell L, Jernberg B. Stereotaxis and tomography: a technical note. Acta Neurochir (Wien) 1980;52:1-7. 5. Roberts TS, Brown R. Technical and clinical aspects of CT-directed stereotaxis. Appl Neurophysiol 1980;43:170-1. 6. Coffey RJ, Lunsford LD. The role of stereotactic techniques in the management of craniopharyngiomas. Neurosurg Clin N Am 1990;1:161-72. 7. Coffey RJ, Lunsford LD. Stereotactic surgery for mass lesions of the midbrain and pons. Neurosurgery 1985;17:12-18. 8. Lunsford LD. A dedicated CT system for the stereotactic operating room. Appl Neurophysiol 1982;45:374-8. 9. Lunsford LD. Stereotactic drainage of brain abscesses. Neurol Res 1987;9:270-4. 10. Lunsford LD, Latchaw RE, Vries JK. Stereotactic implantation of deep brain electrodes using computed tomography. Neurosurgery 1983;13:280-6. 11. Lunsford LD, Leksell L, Jernberg B. Probe holder for stereotactic surgery in the CT scanner: a technical note. Acta Neurochir (Wien) 1983;69:297-304. 12. Lunsford LD, Nelson PB. Stereotactic aspiration of a brain abscess using a ‘‘therapeutic’’ CT scanner. A case report. Acta Neurochir (Wien) 1982;62:25-9. 13. Lunsford LD, Parrish R, Albright L. Intraoperative imaging with a therapeutic computed tomographic scanner. Neurosurgery 1984;15:559-61. 14. Lunsford LD, Rosenbaum AE, Perry J. Stereotactic surgery using the ‘‘therapeutic’’ CT scanner. Surg Neurol 1982;18:116-22. 15. Lunsford LD, Woodford J, Drayer BP. Cranial computed tomographic demonstration of intracranial penetration by an orbital foreign body. Neurosurgery 1977;1:57-9. 16. Kondziolka D, Dempsey PK, Lunsford LD, Kestle J, Dolan E, Kanal E, Tasker RR. A comparison between magnetic resonance imaging and computed tomography for stereotactic coordinate determination. Neurosurgery 1992;30:402-7. 17. Lunsford LD. Magnetic resonance imaging stereotactic thalamotomy: report of a case with comparison to computed tomography. Neurosurgery 1988;23:363-7. 18. Kondziolka D, Flickinger JC. Current use of magnetic resonance imaging in stereotactic surgery: an ASSFN member study. Stereotact Funct Neurosurg 1997;66:193-7. 19. Lunsford LD, Martinez AJ, Latchaw RE. Stereotaxic surgery with a magnetic resonance – and computerized tomography-compatible system. J Neurosurg 1986;64:872-8.

20. Kondziolka D, Lunsford LD. The role of stereotactic biopsy in the management of gliomas. J Neurooncol 1999;42:205-13. 21. Kondziolka D, Lunsford LD. Stereotactic biopsy for intrinsic lesions of the medulla through the long-axis of the brainstem: technical considerations. Acta Neurochir (Wien) 1994;129:89-91. 22. Kondziolka D, Lunsford LD. Results and expectations with image-integrated brainstem stereotactic biopsy. Surg Neurol 1995;43:558-62. 23. Coffey RJ, Lunsford LD. Diagnosis and treatment of brainstem mass lesions by CT-guided stereotactic surgery. Appl Neurophysiol 1985;48:467-71. 24. Kondziolka D, Adelson PD. Technique of stereotactic biopsy in a 5-month old child. Childs Nervous Syst 1996;12:615-18. 25. Field M, Witham TF, Flickinger JC, Kondziolka D, Lunsford LD. Comprehensive assessment of hemorrhage risks and outcomes after stereotactic brain biopsy. J Neurosurg 2001;94:545-51. 26. Kondziolka D, Lunsford LD. Stereotactic aspiration of colloid cysts. J Neurosurg 1993;79:965-6. 27. Kondziolka D, Lunsford LD. Stereotactic techniques for colloid cysts: roles of aspiration, endoscopy, and microsurgery. Acta Neurochir Suppl 1994;61:76-8. 28. Kondziolka D. Intracavitary irradiation with colloidal phosphorus-32 for management of arachnoid cysts. Minim Invasive Neurosurg 1997;40:55-8. 29. Lunsford LD, Pollock BE, Kondziolka DS, Levine G, Flickinger JC. Stereotactic options in the management of craniopharyngioma. Pediatr Neurosurg 1994;21 Suppl 1:90-7. 30. Niranjan A, Witham T, Kondziolka D, Lunsford LD. The role of stereotactic cyst aspiration for glial and metastatic brain tumors. Can J Neurol Sci 2000;27:229-35. 31. Pollack IF, Lunsford LD, Slamovits TL, Gumerman LW, Levine G, Robinson AG. Stereotaxic intracavitary irradiation for cystic craniopharyngiomas. J Neurosurg 1988;68:227-33. 32. Coffey RJ, Lunsford LD. Factors determining survival of patients with malignant gliomas diagnosed by stereotactic biopsy. Appl Neurophysiol 1987;50:183-7. 33. Coffey RJ, Lunsford LD, Taylor FH. Survival after stereotactic biopsy of malignant gliomas. Neurosurgery 1988;22:465-73. 34. Lunsford LD, Somaza S, Kondziolka D, Flickinger JC. Brain astrocytomas: biopsy, then irradiation. Clin Neurosurg 1995;42:464-79. 35. Lunsford LD. Stereotactic drainage of brain abscesses. J Neurosurg 1989;71:154. 36. Kondziolka D, Bonaroti E, Baser S, Brandt F, Kim YS, Lunsford LD. Outcomes after stereotactically guided pallidotomy for advanced Parkinson’s disease. J Neurosurg 1999;90:197-202.


37. Lee JYK, Kondziolka D. Thalamic deep brain stimulation for management of essential tremor. J Neurosurg 2005;103:400-3. 38. Kondziolka D, Steinberg G, Wechsler L, et al. Neurotransplantation for patients with subcortical motor stroke: a phase 2 randomized trial. J Neurosurg 2005;103:38-45. 39. Harris AE, Hadjipanayis CG, Lunsford LD, Lunsford AK, Kassam AB. Microsurgical removal of intraventricular lesions using endoscopic visualization and stereotactic guidance. Neurosurgery 2005;56:125-32.

40

40. Kondziolka D, Lunsford LD. Microsurgical resection of colloid cysts using a stereotactic transventricular approach. Surg Neurol 1996;46:485-92. 41. Lunsford LD, Kondziolka D, Bissonette DJ. Intraoperative imaging of the brain. Stereotact Funct Neurosurg 1996;66:58-64. 42. Lunsford LD, Martinez AJ. Stereotactic exploration of the brain in the era of computed tomography. Surg Neurol 1984;22:222-30.

629

Image Guided Neurosurgery

35 Engineering Aspects of Electromagnetic Localization in Image Guided Surgery E. C. Parker . P. J. Kelly

Introduction

General Principles

As the concepts of stereotactic surgery have evolved over the past few decades, so has the equipment developed and adapted for its use. Head frame systems that were initially used for point in space stereotaxis were modified to allow for volumetric procedures allowing the visualization of tumors without the need for making translations of the frame coordinates to define the borders of target structures. Over the past decade, however, use of frameless stereotactic systems has become the dominant means of performing open image guided brain surgery. Although many frameless systems allow definition of tumor volumes, the common use of these devices again implements point in space stereotaxis, but with the convenience of real time feedback and multiplanar image reconstruction. All such frameless systems make use of a digitizer system to correlate the position of the patient’s head and a pointer of some type with a stereotactic coordinate system that has been applied to an imaging database. These systems have included articulated arms with angle sensors at each joint, as well as sonic, optical, and electromagnetic digitizers. Of these, only optical and electromagnetic systems remain in widespread use, with optic devices being the more prevalent. Each of these systems has unique characteristics that determine its applicability to a given circumstance. This chapter with focus on the unique aspects of electromagnetic digitizers in their application to image guided surgery.

Electromagnetic digitizers allow identification of the position and angular orientation of a movable receiver with respect to a transmitter that is generating an electromagnetic field. Such systems were described as early as the 1970s and are now readily available [1,2]. They have been adapted to a wide range of uses including multiple neurosurgical applications [3–7]. A commercially available electromagnetic digitizer, the Flock of Birds (Ascension Technology, Burlington, VT) [8] was first adapted by Kelly et al. for use in a stereotactic system [9]. This application evolved into the Cygnus frameless stereotactic system (Compass International, Rochester, MN), a commercially available neurosurgical stereotactic system that exclusively utilizes an electromagnetic digitizer. Other navigational systems using electromagnetic digitizer systems are now available as well. All frameless stereotactic systems for neurosurgical applications consist of several key components. Unique to magnetic systems are the electromagnetic transmitter and receiver. As with any type of stereotactic system, a computer software package that relates the position and orientation data to the patient’s preoperative imaging database is necessary to transform this digital information into a graphical form that can be readily utilized by the surgeon to facilitate preoperative and intraoperative decision making. The conventional setup for a magnetic system requires that the transmitter be fixed relative

#


536

35

Engineering aspects of electromagnetic localization in image guided surgery

to the position of the patient’s head. This is usually accomplished by the use of an adaptor that rigidly secures the transmitter to a multi-pin head fixation system. This transmitter generates a magnetic field that encompasses the region of anatomical interest. A sensor can then be manipulated within this magnetic field and its relative position in space conveyed to the stereotactic computer system. Although such sensors are relatively small (2.5 cm or smaller) they can be coupled to a probe, suction tip, or other instrument of fixed size and orientation for intraoperative use, allowing the surgeon to utilize the instrument in an area too small for the magnetic sensor. Such a system is illustrated in > Figures 35-1–> 35-3. An arrangement of this type can be used in a variety of ways. It is possible to define the contours of a tumor on individual imaging slices, thereby creating a tumor volume that can be visualized along the axis of the localizing probe. More commonly, however, the position or trajectory of the probe tip will be utilized to plan the position of a scalp, bone, and dural opening and to help confirm spatial orientation during tumor resection. Electromagnetic stereotaxis can also lend itself to situations where it is desirable not to place the patient’s head in rigid fixation during surgery. This is often the case during sinus procedures performed by otolaryngologists who are not accustomed to placing the patient in pins. For this type of arrangement, a second electromagnetic sensor may be attached to the patient’s head, which can then be positioned and moved as desired. Through the use of two electromagnetic receivers, the position of both the patient’s head and the localizing probe are tracked simultaneously in relation to the transmitter, which is rigidly secured to the operating table. Alternatively, a small transmitter may be secured directly to the patient’s head, achieving the same result.

. Figure 35-1 Examples of an (a) electromagnetic transmitter, and (b) receiver

Accuracy Of prime importance in any frameless stereotactic system is accuracy. Ascension Technology, manufacturer of the Flock of Birds electromagnetic digitizer, claims positional accuracy of 1.8 mm and orientation accuracy of 0.5 with the sensor in a range of between 20.3 and 76.2 cm from the transmitter. Testing has confirmed equal or better accuracy with mean position error of 0.5 mm and maximum error of


35

. Figure 35-2 Electromagnetic transmitter secured to the head holder in preparation for surgery

. Figure 35-3 Electromagnetic receiver being used to confirm location during surgery

1 mm within an optimal operational zone of 22.5–64.0 cm [10]. The accuracy of magnetic localization when applied to neurosurgical stereotaxis has been well documented. During development of what would become the Cygnus system a magnetic digitizer was adapted for use with the COMPASS head frame (Compass International, Rochester, MN) and testing revealed average three dimensional errors of approximately 3 mm [5,9]. This testing was done using known frame reference points to minimize inaccuracy introduced during image registration. Further testing has been done under more clinically relevant circumstances. This testing has included comparison to various other digitizer types, including articulated arm and optical based systems. Measurement of spatial accuracy of the Cygnus system revealed a mean accuracy of 1.90 0.7 when tested on a stereotactic phantom and registered with standard surface fiducial markers that had been attached to the surface of the phantom [11]. Such testing takes into account not only the positional accuracy of the

537

538

35


digitizer system, but also the process of registering the position of the frame to the imaging database. This concept was taken a step further and performed intraoperatively on a series of 70 patients. Measurements of error from intracranial anatomical landmarks was 1.4 0.6 mm, again using the Cygnus system [12]. In each of these studies the accuracy of the magnetic based system compared very favorably to an optical tracking system tested in the same manner.

Magnetic Interference Metallic objects in the operating room environment may produce distortion of the magnetic field used for localization of the electromagnetic sensor. This is a chief concern regarding use of magnetic stereotactic systems. Such distortion can be quite significant. Relative sensitivity of magnetic tracking systems to various metallic objects is dependent on the type of current employed by the transmitting device. Transmitters based on alternating or direct current will react quite differently in the presence of ferromagnetic and electrically conductive material. Tracking reliability of a direct current based system is relatively unaffected by nonferromagnetic materials including aluminum. The presence of an aluminum object within a few centimeters of the sensor can cause significant deviation in both position and orientation in an alternating current system, on the other hand [13]. These deviations can be greater than 2.5 mm and 2 in position and orientation respectively when the aluminum object is in close proximity to the electromagnetic sensor. When tested in the presence of metal cylinders composed of titanium, stainless steel, cobalt chrome, aluminum, and mild steel, a pulsed direct current system was significantly affected only by the mild steel sample [10]. Therefore, given the relative prevalence of these conductive but nonferromagnetic metals in the typical operating environment, direct current magnetic tracking provides a more stable platform for stereotaxis.

Although the effects of nonferromagnetic but highly conductive materials on direct current tracking systems are minor, large aluminum objects, for example, can cause the introduction of some error. This effect can be minimized by altering the sampling rate of the electromagnetic sensor system. Aluminum’s high electrical conductivity causes the creation of eddy currents when exposed to a magnetic field. With a pulsed magnetic field, these currents form and then rapidly decay. A relatively low pulse frequency will allow more time for these currents to resolve before sampling occurs. Eddy currents in highly ferromagnetic material, on the other hand, do not decay as quickly and generate secondary magnetic fields that increase interference. Therefore, a lower sampling rate (20 Hz) will reduce error associated with aluminum, while a higher sampling rate (140 Hz) 0has been demonstrated to moderate the effects of steel [14]. It should be noted that even at this high sampling rate, the presence of ferromagnetic steel in the operating environment will still introduce significant error with the use of a direct current system and it should be avoided. Another source for potential interference in the function of magnetic tracking systems is electromagnetic background noise in the operating room. This interference may result from wires, lights, electrical equipment, etc. Comparison of background noise between a shielded room and a typical operating room, however, has demonstrated only very minimal effect from this ambient interference. Significant interference and increase in noise can be seen when electrical or electronic equipment is used in close proximity to the electromagnetic receiver. This phenomenon is diminished dramatically by positioning these devices further from the receiver. Distances as small as 30 cm can be effective in eliminating this problem [15]. Generally speaking, however, any nonessential metallic or electronic devices should be positioned away from the surgical field. While the interference resulting from metallic objects can be minimized, it cannot be


eliminated altogether. This is a result of the presence of large conducting objects (the operating table, Mayfield or Gardner head holder, retractors, etc) that will create some distortion in the magnetic field and introduce some error into the system. One method for dealing with this problem is to utilize a calibrated distortion phantom. Such a device allows the registration of a three dimensional matrix of points of know spatial relationship within the surgical field, prior to registration of the patient’s head position. Any distortion of the magnetic field will cause unexpected discrepancies in the relative positions of these points. The stereotactic computer can then use these data to calculate correction factors for all points within this calibrated area. Unfortunately, use of this method requires the registration of many points prior to surgery and is quite cumbersome. Additionally, any metallic objects brought into the field after this calibration has taken place (retractors for example) will alter the distortion, possibly reducing the accuracy of the correction factors. A more accessible approach to minimizing error associated with metallic interference is to monitor the presence of metallic objects within the magnetic field. The most recent version of the

35

Flock of Birds system tracks the error introduced by the proximity of highly electrically conductive (aluminum) or magnetically permeable (steel) to either the transmitter or sensor. Although the system cannot correct for the distortion introduced by these metals, it can report their relative presence and the likelihood of interference to the user. The latest version of the Cygnus software relays this information in a graphical form. A meter in the lower left hand corner of the display reports the relative electromagnetic interference being observed by the system (> Figure 35-4). The number and height of the meter rise as more distortion is noted within the magnetic field. A higher number serves as a warning to the user that there is an increased relative risk of error in reporting of position and orientation of the sensor and that metal objects such as retractors should be moved or taken out of the field for optimal performance. A final issue involving the magnetic interference and electromagnetic stereotaxy is interference caused by the transmitter itself. It has been our experience that the magnetic transmitter creates noise in neurophysiological monitoring equipment. This prevents the monitoring of evoked potentials while the magnet is active.

. Figure 35-4 Interference warning indicator reporting potential distortion within the magnetic field secondary to the presence of metallic objects

539

540

35


This problem is easily solved by placing the transmitter in standby mode when not in use. Since the frameless system is generally used only intermittently to confirm position or orientation, this issue has not been viewed as a serious limitation of the technology.

can be mounted to a tower in a more conventional, non-mobile, setup, it can also be stored in a small carrying case. This allows easy mobility from hospital to hospital if a surgeon covers more than one facility and dedicated systems are not available at each location.

Advantages of Electromagnetic Stereotactic Systems

Future Directions

Electromagnetic tracking systems hold some significant advantages over the more widely used optical systems. The primary advantage of magnetic systems is their unobtrusiveness during surgery [16]. In the typical arrangement, the magnetic transmitter is placed outside the sterile field under the drapes. The most convenient location for securing this transmitter is to utilize the extra starburst attachment at the base of the Gardner or Mayfield head holder. This is preferable to attaching the transmitter directly to the operating table as adjustments can be made in the position of the arm securing the head clamp to the table, and thus ˜ s head position, without altering the the patientO spatial relationship between the transmitter and patient. This arrangement requires use of an extension bracket approximately 20 cm in length. In addition to allowing room for securing the transmitter away from the head and pin apparatus, this bracket provides optimal spacing between the transmitter and the operative field, minimizing inaccuracy. The only part of the frameless system present on the sterile field is the sensor, usually attached to a blunt tipped probe. This sensor is gas sterilized and is connected to the receiver system by a thin cable allowing easy manipulation during surgery. No line of sight is necessary with this arrangement as it is with optical systems. Another advantage to electromagnetic frameless technology is its relatively small size and transportable nature. Although such a system

As with most technology, electromagnetic sensors have been introduced that are much smaller that older models. Sensors capable of providing six degrees of freedom (three positional and three rotational) are available in sizes under 2 mm in diameter, and those providing five degrees of freedom (three positional and two rotational) are as small as 0.55 mm in diameter. Sensors of this type have been found to be accurate enough for surgical applications [17]. Such small sizes allow placement of the sensor at the tip of a flexible device as opposed to the usual attachment of frameless stereotactic localizers to the handle of a rigid instrument. This has already been successfully utilized in bronchoscopy applications, allowing navigation to lesions that would otherwise be unreachable by conventional visual navigation [18,19]. Although flexible endoscopy has seen limited use within neurosurgery, these types of sensors could certainly be adapted to this application.

References 1. Kuipers JB. SPASYN – an electromagnetic relative position and orientation tracking system. IEEE Trans Instrum Meas 1980;29(4):462-6. 2. Raab FH, Blood EB, Steiner TO, Jones HR. Magnetic positioin and orientation tracking system. IEEE Trans Aerosp Electron Syst 1979;AES-15(5):709-17. 3. Fried MP, Kleefield J, Gopal H, Reardon E, Ho BT, Kuhn FA. Image-guided endoscopic surgery: results of accuracy and performance in a multicenter clinical study using an electromagnetic tracking system. Laryngoscope 1997;107:594-601.


4. Fried MP, Kleefield J, Taylor R. New armless imageguidance system for endoscopic sinus surgery. Otolaryngol Head Neck Surg 1998;119:528-32. 5. Rousu JS, Kohls PE, Kall B, Kelly PJ. Computer-assisted image-guided surgery using the regulus navigator. Stud Health Technol Inform 1998;50:103-9. 6. Sagi HC, Manos R, Benz R, Ordway NR, Connolly PJ. Electromagnetic field-based image-guided spine surgery part one: results of a cadaveric study evaluating lumbar pedicle screw placement. Spine 2003;28:2013-18. 7. Sagi HC, Manos R, Park SC, Von Jako R, Ordway NR, Connolly PJ. Electromagnetic field-based image-guided spine surgery part two: results of a cadaveric study evaluating thoracic pedicle screw placement. Spine 2003;28: E351-4. 8. Ascension Technology Corporation, http://www.ascension-tech.com/. 9. Goerss SJ, Kelly PJ, Kall B, Stiving S. A stereotactic magnetic field digitizer. Stereotact Funct Neurosurg 1994;63:89-92. 10. Milne AD, Chess DG, Johnson JA, King GJ. Accuracy of an electromagnetic tracking device: a study of the optimal range and metal interference. J Biomech 1996;29:791-3. 11. Benardete EA, Leonard MA, Weiner HL. Comparison of frameless stereotactic systems: accuracy, precision, and applications. Neurosurgery 2001;49:1409-15; discussion 1415-6.

35

12. Mascott CR. Comparison of magnetic tracking and optical tracking by simultaneous use of two independent frameless stereotactic systems. Neurosurgery 2005;57: 295-301; discussion 295-301. 13. Birkfellner W, Watzinger F, Wanschitz F, Enislidis G, Kollmann C, Rafolt D, Nowotny R, Ewers R, Bergmann H. Systematic distortions in magnetic position digitizers. Med Phys 1998;25:2242-8. 14. LaScalza S, Arico J, Hughes R. Effect of metal and sampling rate on accuracy of Flock of Birds electromagnetic tracking system. J Biomech 2003;36:141-4. 15. Poulin F, Amiot LP. Interference during the use of an electromagnetic tracking system under OR conditions. J Biomech 2002;35:733-7. 16. Mascott CR. The Cygnus PFS image-guided system. Neurosurgery 2000;46:235-8. 17. Hummel J, Figl M, Kollmann C, Bergmann H, Birkfellner W. Evaluation of a miniature electromagnetic position tracker. Med Phys 2002;29:2205-12. 18. Hautmann H,Schneider A, Pinkau T, Peltz F, Feussner H. Electromagnetic catheter navigation during bronchoscopy: validation of a novel method by conventional fluoroscopy. Chest 2005;128:382-7. 19. Schwarz Y, Greif J, Becker HD, Ernst A, Mehta A. Realtime electromagnetic navigation bronchoscopy to peripheral lung lesions using overlaid CT images: the first human study. Chest 2006;129:988-94.

541

45 Image Guided Craniotomy for Brain Tumor I. E. McCutcheon

The application of technology to the surgical resection of intracranial tumors has involved a gradual accretion of instruments and instrument systems ever since such tumors were first successfully removed in the late nineteenth century. This background of gradual improvement has been punctuated by explosions of activity in specific areas that have greatly expanded the boundaries of neurosurgical practice. These include the invention of the bipolar cautery, the introduction of angiography and ventriculography for diagnosis, the advent of the intraoperative microscope, and the development in the 1970s and 1980s of computerized imaging systems such as computed tomography (CT) and magnetic resonance imaging (MRI). The intracranial images that these modalities now provide offer a wealth of anatomical detail and delineate tumor location and tumor extent with precision. The gradual fusion of improved imaging with stereotactic localizers has yielded the latest such revolution, that of neuronavigation. This generic term covers a variety of systems designed to provide technical solutions to the basic problems of (1) Correlating a lesion seen on scan with the anatomical reality of the patient, (2) Obtaining a maximal resection of that lesion, while (3) Breaching normal brain tissue to the smallest degree compatible with an adequate resection. The original mode of image guidance during tumor surgery was through a CT- or MRIcompatible frame-based system of stereotactic localization. This has largely given way to frameless systems that permit navigation by comparison of patient anatomy during surgery with images obtained before surgery. Such systems are now #


widely available and significant experience has accrued with their use in tumor resection. Recent efforts shift the time of imaging from preoperative to intraoperative stem from the increasing availability of systems of intraoperative CT and MRI. This chapter will discuss the available technology, its application in intracranial tumor surgery, and the advantages and pitfalls attendant upon its use.

History and Methodology The stereotactic approach to cerebral localization was originally described by Horsley and Clarke in 1906 [1]. Their fairly complicated apparatus and its successors (notably Leksell’s innovative frame system designed for use with ventriculography) had relatively limited utility in tumor resection until the advent of CT technology, which allowed a shift from localizing single points within a space (i.e., the brain) to describing volumes within that larger space [2]. Such description of volume was necessary for stereotactic resection as opposed to stereotactic biopsy. Although CT scans can be rendered in a multiplanar fashion through computerized reconstruction, the advent of MRI scanning has greatly expanded the ability to display tumors in coronal, sagittal, and axial planes. With the concomitant development of relatively powerful mini-computers, such threedimensional data sets could be rendered in ways useful to a surgeon seeking guidance in tumor resection. The first practical ‘‘volumetric’’ surgery rendered a localization target as a volume rather than a point, and placed images of that volume

700

45

Image guided craniotomy for brain tumor

into a convenient and accessible visual display for use by the surgeon during surgery. The pioneer of this approach was Patrick Kelly, who developed a frame-based system called the COMPASS stereotactic system (Stereotactic Medical Systems), which he designed specifically for volumetric tumor resection [3–5]. This device allows projection of scaled planar images onto a heads up display within the operating microscope. This display projects the images onto a real-time image of the brain surface, and allows the surgeon to define the radiological abnormality precisely during surgery irrespective of any grossly visible (or invisible) abnormality. Given that the margins of a glioma are usually indistinct due to the infiltrative quality of such tumors, such technology enhanced the surgeon’s ability to achieve more complete resections of the region identified as tumor on scan. Foreshadowing the trajectory of development of more recent systems for surgical navigation, Kelly later incorporated MRI, digital subtraction angiography, and magnetoencephalography into his system [6].

The next step in the evolution of neuronavigation was the advent of the frameless stereotactic method initiated by David Roberts in a seminal paper published in 1986 [7]. He described integration and display of CT images within the operating microscope, images that were specially registered by determining the position of the microscope as its focal point was placed on three fiducial markers placed on the scalp and evident in the CT images (> Figure 45-1). This paper marked the true beginning of the avalanche of activity in surgical navigation that continues to this day. All subsequent work can be traced back to this paper which established the concept of fiducial markers on the scalp, rather than suspended in space at various points around the head as occurs with frame-based systems. In essence, stereotactic methodology and neuronavigation are used to map image space onto physical space. Many surgeons using such technology now prefer frameless systems, which are less cumbersome to the patient and more versatile, especially when volumetric resection (rather than point

. Figure 45-1 The stereotactic system of Roberts, which initiated modern frameless neuronavigation. The nonlinear array of three spark gaps, providing an acoustic localizing system, is seen protruding from the operating microscope. The optical projection system is attached to the side of the microscope (arrow), and the microphone array on the ceiling is not shown here [7]


biopsy) is being done. Such systems have four basic elements: (1) A method for co-registering images with physical space, (2) A device for intraoperative localization (such as a pointer), (3) A computer monitor that displays the images in multiplanar fashion, (4) Intraoperative feedback in real time (e.g., display of location of the localization device on the displayed images). Registration. Although displayed as slices, standard digital images of the brain obtained through CT or MRI are actually databases of three-dimensional volume. Mapping these image sets onto the physical reality of a patient’s anatomy allows them to be used for navigation. Such mapping, otherwise called registration, can use data acquired before or during surgery. The most common method of registration to date has been point-based, which defines linkage between points in the image versus points in the physical space to allow geometric linkage between the respective volumes. Typically, fiducial markers (detachable temporary tags of uniform size) placed on the surface of the patient’s face and scalp have been used. However, intrinsic anatomic landmarks can be used instead of, or as a supplement to, such registration to enhance its accuracy. Surface-based registration can also be done. This takes a number of points (usually

45

40 or more) chosen randomly on the contours of the patient’s surface anatomy, and fits them to similar contours within the previously acquired images. Accuracy is lower in surface-based methods than in point-based registration, and for that reason most systems depend upon the latter as their primary method of registration, and use the former as a fallback option should fiducial markers be displaced or absent. Intraoperative Localizers. A variety of localizing devices have been put into use for navigation in tumor resection. These have each had their proponents and each manufacturer claims superiority over its competitors’ systems. The truth is that no single system for surgical navigation has proven more accurate or more helpful than any other. Error can be introduced at any number of steps from preoperative imaging through registration, and can be further introduced by intraoperative events. However, when care is applied to the acquisition of the images, the placement and then localization of the fiducial markers, and the registration just prior to surgery, localization accuracy of 1–2 mm can generally be obtained. This accuracy degrades during surgery, largely due to the phenomenon of brain shift (> Figure 45-2). Systems that depend upon images acquired prior to surgery cannot be

. Figure 45-2 Brain shift during surgery, shown by intraoperative MRI captured (a) prior to skin incision (b) after pterional/ subfrontal craniotomy to resect a recurrent, invasive gonadotropin-secreting pituitary macroadenoma

701

702

45


corrected for such displacements without intraoperative updating of the imaging data set. Such displacements are predictably more profound at the brain surface and less evident internally, but they can still be significant at deep locations, particularly in the vicinity of critical structures like the brainstem or internal capsule. Either a surgeon must gain a sense of the degree of shift by repeated comparison of localization on the image display with specific and identifiable anatomic landmarks, or some degree of correction must be imposed by the acquisition of new realtime data. In an analysis of brain shift by Roberts et al. they found a mean displacement of 10.7 mm at the cortical surface, a deformation caused more by the effect of gravity than by other possible causes of such shift (including hyperventilation, loss of cerebrospinal fluid (CSF) and direct tissue loss as resection proceeds) [8]. Articulated Arms. The first description of a passive articulated arm for intracerebral localization came from Watanabe et al. [9,10]. They described a ‘‘neuronavigator’’ which included an

articulated arm with six joints and registration accomplished with scalp-based fiducial markers as Roberts et al. had proposed. This system gave way to the ISG viewing wand system (Elekta), which also has a six-jointed arm and electrogoniometers to detect the angle at each joint (> Figure 45-3). The arm is thus subject to electromagnetic interference, an important impediment in the operating room environment, and the physical limitations of the arm impose constraint on patient placement relative to the device. In particular, its use in conjunction with a microscope can be difficult. These systems are nonetheless quite accurate, with Golfinos et al. reporting accuracy within 2 mm in 92% of cases when localization was based on MRI [11]. While articulated arms need not be positioned in the line of site of an optical detector, they are physically limited in their reach and have given way in most centers to free-hand devices. However, they remain in use in some centers [12]. Sonic Digitizers. The system originally reported by Roberts et al. utilized spark gap generators

. Figure 45-3 ISG viewing wand. The articulated arm of this device is limited in its reach but has no line-of-sight constraint as is true for currently popular neuronavigational systems dependent on reflectance of infrared light


emitting an ultrasonic signal and attached to the operating microscope in a fixed configuration, with sound detectors placed at several locations in the operating room [7]. Similar ultrasonic emitters have since been placed on free-hand devices (probes or other surgical instruments) the location of which is determined by a nearby microphone array. Although generally reliable, such systems can be affected by noise or by physical obstruction within the line of site from the emitter to the detector. In addition, temperature fluctuation affects the speed of sound in air and temperatures are typically not constant within an operating room. For these reasons, most systems now use light emitting diode (LED) technology. Light Emitting Diode Systems. In such systems the light source can be a camera flashing a pulsed infrared light, which is reflected off coated spheres attached to the pointing probe. The reflected light is then detected by a charge coupled device. Examples of this include the Stealth (Medtronic) and Vector Vision (BrainLab) systems (> Figure 45-4). The spheres must not be fouled with blood and must be in the line of sight of the light source. An alternative concept places the LED on the probe itself which must then be attached to a power source, usually through a flexible cable. Such devices are less easy to manipulate then those using passive reflectors as described above. The LED can also be placed on the microscope, the lens of which has a known focal length and whose focal point represents the tip of virtual pointer, but such systems (e.g., the Viewscope or Zeiss MKM System) are hampered by line-of-sight issues particularly when used in operations on patients in the lateral position or when the scrub nurse’s instrument tray sits above the patient’s torso [13]. The Easy-Guide Neurosystem (Philips) is one example of a system with the LED on the pointing device. The SurgiScope (ISIS) is a combination of neuronavigation with robotics, in which the surgeon can specify a target or trajectory during preoperative planning,

45

. Figure 45-4 BrainLab vectorvision system for neuronavigation. The localizing probe reflects infrared light back to two light sources, and the resulting stereotactic data are processed in a mobile workstation with screen for displaying multiplanar images

and thereby program the microscope to act as a slave pointer. It can also be aimed actively at targets by the surgeon, converting its point of focus into a virtual probe as described above. Magnetic Field Systems. Instead of using reflected light as the mode of triangulation, some groups have used magnetic field guidance [14]. The field emanates from a transmitter near the patient’s head and is detected by a receiver on the surgical localizing probe. Although relatively simple, this technique is less accurate than other

703

704

45


methods as it is susceptible to field distortion within the surgical environment because of the abundance of metals there, and also because of the presence of electromagnetic fields from monopolar cautery and other devices. Although software algorithms have been applied to compensate for field inhomogeneity and therefore enhance the technique’s accuracy, the abundance of ferromagnetic instruments in cranial surgery (including typically the head clamp) make adoption of magnetic field localization somewhat problematic. Machine Vision (Passive Stereoscopic Video). This method requires no emitters or reflectors as it localizes the pointer’s position by determining positional differences on video images acquired by cameras placed at different angles. This technique was proposed by Heilbrun et al. and utilizes two video cameras one meter apart and eight fiducial markers in a box-like configuration of known dimensions [15]. This system has been reported to have a mean error of 2.1 mm in a series of 21 patients in whom its accuracy of localization was compared with that obtained by a BrownRoberts-Wells frame-based system (Radionics) [16]. However, it has not been brought to market, and line-of-sight considerations apply. Computer Displays. Each neurosurgical navigation system comes with an image-processing workstation preloaded with software specific for the system (> Figure 45-5). Peripherals are usually system-specific as well. Software upgrades are ongoing as clinicians detect previously unknown kinks in the various systems and manufacturers work to correct those flaws. It is typically possible for each surgeon to select from a menu of display options that allow standard multiplanar views (axial, coronal, sagittal) as well as trajectory and in-line views created from the three-dimensional dataset. Pre-contrast T1-or T2-weighted images can be loaded, as is appropriate in the case of low grade gliomas or non-enhancing tumors (> Table 45-1). Additional imaging modalities such MR spectroscopy, fMRI, and positron emission tomography (PET) can also be loaded into

the workstation but require additional software to permit cover overlay on standard images. Placement of the video display can be difficult as ideally the surgeon should not have to turn away from the operative field in order to see the monitor while holding the probe in place. With the presence of an anesthetic delivery system, instrument tables, ultrasound machine, and other bulky equipment it may be necessary to place the monitor in an inconvenient location. To get around this limitation, transfer of the images into the microscope to create a heads-up display has been helpful [17]. However, not all tumor cases are done through a microscope, thus headmounted displays have been proposed and will likely be increasingly available in the future [18]. Feedback During Surgery (Real-Time). Standard systems currently in place provide a navigational map based upon preoperative acquisition, a map which is not updated during the procedure. Because of the brain shift alluded to above, such maps can become increasingly inaccurate as the operation proceeds. Efforts at updating the preoperative map with new intraoperative data are underway and rely upon ultrasonography or upon intraoperative acquisition of new scans [19,20]. Cortical tracking by cameras mounted on a stereotactic microscope has also been used to predict deformation of the brain surface [21]. Strategies for applying intraoperative ultrasound to the problem of correction of tissue shift have been presented by several groups [8,22–25]. Algorithms for such compensation continue to be refined and there is no standard technique to achieve this. In one study, the mean displacement of cortical landmarks ranged from 0.8 to 14.3 mm, about half of which was due to reduction in tumor volume [26,27]. Shifting of superficial landmarks exceeded that of subcortical structures in all patients. Brain shift has not thus far been predictable prior to surgery, so intraoperative updating of images must be done to solidify this fourth component of surgical navigation [28].


45

. Figure 45-5 Screen view from intraoperative MRI environment (BrainSuite). Patient has an oligodendroglioma in the right frontal lobe undergoing craniotomy. Images were taken during surgery after partial resection was accomplished. Note that left-right conventions are reversed in this system. The probe is touching the lateral edge of the right lateral ventricle, medial to which significant tumor remains. The top row contains an intraoperative real-time ultrasonographic image on the left, and overlay of that image on intraoperative MRI on the right

. Table 45-1 Imaging for different tumor types Tumor type suspected

Image

Metastasis Meningioma with hyperostosis Meningioma with no hyperostosis Glioma, non-enhancing (low grade or occasional high-grade) Enhancing (high-grade) Tumors of skull base

MRI (T1, post-contrast) CT (pre- and post-contrast) MRI (T1, post-contrast) MRI (fast spin echo, no fat suppression; T2) MRI (T1, post-contrast, T2) MRI (T1, post-contrast, fat suppressed; T2) and CT (pre- and post-contrast)

705

706

45


In the subset of patients who undergo an awake craniotomy for intraoperative mapping of speech function, this issue takes on some importance. Although such operations can be done with the head fixed in a rigid clamp if a complete scalp block is performed, the technique is easier if the head is not fixed. A mobile head, however, requires updating of the registration with each movement. A reference sensor affixed to the retroauricular region has been described which can provide updating of the patient’s head position and recalculation of registration, making neuronavigation possible for an awake craniotomy without head fixation [29]. Infants and young children with deformable skulls form another subset of patients in whom rigid cranial fixation is not possible. For them a similar solution has been reported which involves screwing a reference arc to the outer table of the skull through a separate incision within the operative field [30]. Equally, innate anatomical fiducials might be used, although their localization would be subject to minor errors due to the same deformability that prevents application of a head clamp for immobilization in this age group. Ultrasonography for arterial and venous display using a color doppler is easy to acquire and can be imported into a navigational work station [31]. This adds information about hidden vessels and thereby increases safety and allows tailored surgical approaches to deep tumors near critical arteries. It also allows the vascular tree to be used as the focus for updating and provides yet another way of correcting for brain shift. Updating through intraoperative MRI scan (or in some cases CT) is the obvious solution to the dilemma. However, this poses problems of its own and discussion of this modality will be deferred to the end of the chapter.

Practical Issues Prior to Surgery The use of neuronavigation is best reserved for tumor patients undergoing surgery on an elective

basis. MRI scan provides more accuracy than CT, particularly for maximizing resection of gliomas where T2-weighted changes may reveal tumor that CT cannot show. Also, patients with pacemakers or other metal implants may not be able to undergo MRI, and CT will have to be used. Regardless of the type of scan, it is ideally performed either the day before surgery or the morning of the procedure to minimize the chance for dislodgment of fiducials. Such markers with a central indentation are affixed immediately prior to scanning and are placed abundantly so that registration can still be performed if one or two of the markers drop away in the interim. The markers are best placed on skin sites that are relatively immobile (> Table 45-2). For that reason, several are placed on the forehead, two at the vertex on the midline (in spots requiring a small amount of hair shave), two in the retroauricular area over the mastoids, and two on the right and left parietal areas (again, requiring a small shave). The total number is therefore at least nine, and two or three additional can be placed in the lateral supraorbital area or in shaved areas on the frontal scalp to enhance accuracy. The occipital area is avoided due to the folding of skin that occurs in this area with neck movement. If a patient is already bald, more freedom in placing fiducials is automatically obtained. In fixing the head with the threepoint clamp, pin placement should be carefully planned to avoid disturbing fiducials and in particular to avoid pulling the skin and shifting

. Table 45-2 Position of fiducial skin markers Position

Number

Root of mastoid (both) Midline at vertex (anterior, posterior) Forehead (lower left, lower right, middle, superior) Zygoma (both) Parietal bossa

2 2 4

a

2 2

Optional, depending on location of lesion (anterior vs. posterior within head)


them from their position on the images. A reference system, typically an array of reflective balls in the LED systems, is placed as far to the side of the patient as possible to avoid interference with the surgeon’s movements and those of the scrub nurse. Additionally, it must be placed within the line of sight of the infrared light generator. Registration then proceeds by touching the probe to the fiducial markers in sequence and importing the probe’s position into the system. The position of the reference array must similarly be imported by touching the probe to a divot placed on it for that purpose. Once registration has been completed to acceptable accuracy, the fiducial markers can be removed and shaving can commence to the degree appropriate for the operation.

Selection of Imaging The choice of imaging depends on the suspected identity of the tumor, and of course on the imaging features revealed in MRI scans done prior

45

to the surgical planning. Recommendations for imaging based on tumor type are given in > Table 45-1. Segmentation can be performed to bring out vascular structures, a maneuver that may be helpful in preserving those structures when they are displaced or encased by a bulky tumor (> Figure 45-6). In addition, overlay of image sets from fMRI or from MRI spectroscopy can also be performed, and can be very useful in extending the limits of the resection in non-eloquent areas and in identifying occult tumor [32–35]. PET has also been integrated into neuronavigation to provide metabolic information on tumor heterogeneity and extent. In one series a final target volume different from that obtained with MRI imaging alone was found in 80% of patients by FDG-PET and 88% by MET-PET [36]. The caveat here is that although it is logical to assume that application of functional or metabolic imaging maximizes tumor resection and extends survival, no proof has yet been published that this is indeed the case. In addition, fMRI is most secure in assigning hemispheric dominance rather than precise

. Figure 45-6 Segmentation of anatomic structures. (a) Venous anatomy. By subtracting the overlying layers of scalp, bone, and dura, and performing automated highlighting of slow-flow structures, display of veins can show their relation to tumor, a technique particularly helpful in resection of meningioma (b) Ventricles. Each is assigned a different color, and even the aqueduct of Sylvius can be seen. Simultaneous segmentation and coloring of neural elements can be done at the same time, as is the case here (the basal ganglia are noted just lateral to the lateral ventricles)

707

708

45


localization of cortical speech function. Intraoperative electrophysiological mapping remains the gold standard and cannot at this time be replaced by fMRI.

Surgical Planning It takes 10–20 min to register a patient at the beginning of an operation. Combining surface with fiducial registration is more time consuming, and may enhance accuracy, but in most cases simple fiducial registration is sufficient. One cadaver study has shown that fiducials give the smallest error in localizing targets of the various methods tested [37]. In checking error prior to the final step of accepting registration, it is important to test localization accuracy not just by touching fiducials but by touching skin in the proposed craniotomy zone. If this testing is off by more than a millimeter, than a ‘‘surface merge’’ is one option to enhance accuracy; another is simply to reject the registration and start again from scratch while taking care not to shift any fiducials when the probe touches them. Planning includes delineation of the incision, the underlying craniotomy opening, and the trajectories through the brain or skull base to the lesion that will least disrupt eloquent areas and vital vascular structures (> Figure 45-7). For deep lesions or tumors of the skull base, trajectory views can be very helpful in avoiding critical structures (> Figure 45-8). The surgeon should also decide whether proximity to eloquent cortical areas or important white matter tracts prevents complete resection, and if so which areas of tumor should be left in place. If multiple craniotomies are to be performed (e.g., when two or more metastases are being removed), planning includes consideration of how to create several competing incisions without compromising the neurovascular supply of the various scalp flaps. In reoperative cases, planning includes decisions on whether previous incisions and craniotomy

. Figure 45-7 Neuronavigation increases precision of localization of tumors. This patient had melanoma and three metastatic tumors in the left posterior frontal lobe. Circles have been drawn where the surgeon predicted the tumors’ locations, and additional circles after localization with StealthSystem neuronavigation (which correlated perfectly with true location once surgery proceeded). The anterior lesion was predicted correctly; however, the two posterior lesions were not, and one was placed by the surgeon across the midline, an error corrected by the stereotactic localization

openings need to be extended, and how to do that in a way that does not compromise scalp healing.

Surgical Nuances Metastasis Although radiosurgery has supplemented open craniotomy in the treatment of many smaller brain metastases, larger tumors or those that have failed radiosurgery remain candidates for surgical resection. As such tumors have sharp margins both anatomically and radiographically, they are amenable to resection through small craniotomies with the expectation of a complete removal, an outcome that is enhanced by en bloc resection. In the only report focusing on image guidance for brain metastasis, degree of resection and rates of complication were similar to those seen in series done without image guidance


45

. Figure 45-8 Transsphenoidal surgical targeting and trajectory planning. These images were acquired in the Viewscope, an earlier version of the StealthStation that married frameless surgical navigation to the operating microscope. The point of focus of the microscope was the point of a virtual localizing probe, and the trajectory shown represents the projection of the center of the field of view within the microscope

[38,39]. Neuronavigational techniques facilitate the approach to deep lesions and are particularly important when multiple lesions are present as they allow the surgeon to devise the least disruptive and most efficient trajectory to each lesion. When a metastasis is associated with a cyst, the cyst should ideally not be punctured if probe accuracy is to be maintained. In theory, decompressing the cyst facilitates tumor removal, but it also promotes brain shift, thereby making localization of margins less accurate, and it may foster seeding of tumor cells leading to later recurrence. Certain tumor types are notoriously friable due to necrosis (adenocarcinoma of lung) or intratumoral hemorrhage (melanoma) and en bloc resection including a small margin of adjacent normal brain should be pursued if the tumor’s location does not prohibit this. Image-guided placement of Ommaya reservoirs for leptomeningeal metastasis has been reported, but this is a low-risk procedure in which traditional free-hand ventricular cannulation is quite successful. Thus, the benefit of neuronavigation in such patients is evident mainly in those with slit ventricles [40].

Meningioma The main utility of neuronavigation in meningioma resection comes in planning a minimally invasive incision and in tailoring the craniotomy opening. In creating that opening, care must be taken to incorporate the full extent of the dural tail on all imaging planes, as a margin of normal dura must be achieved circumferentially around the entire tumor for an ideal Simpson grade I resection. In operations done to remove convexity meningiomas and in particular parasagittal tumors, superimposition of the venous anatomy may prove very helpful. Parasagittal meningiomas adjacent to the middle third of the superior sagittal sinus have a tendency to nestle against one or more important veins which drain into the sinus and demand preservation. Neuronavigation systems can assist in mapping out those veins ahead of time. Meningiomas are usually isolated from the brain with cottonoids and debulked internally, collapsing the tumor upon itself and therefore allowing some brain shift to occur. However, the margin is usually easy enough to detect by tactile and visual clues,

709

710

45


thus brain shift is not as important a factor in meningioma resection as it is in intraaxial tumor removal. Paleologos et al. have reported a series of 100 patients treated with image-guided surgery for meningioma [41]. These patients were compared to 170 patients treated without neuronavigation. They found shorter surgical times for the image-guided group and a shorter hospital stay likely due to the significantly lower complication rate observed when image guidance was applied. As a result of these differences, the mean cost per patient was 20% higher without image guidance, in spite of the extra costs associated with application of the technology.

Glioma The application of neuronavigation to glioma surgery is largely directed at bringing the degree of resection of the radiographic lesion as close to 100% as possible. This effort makes sense if, and only if, such incremental improvements in resection make a difference in patient outcome. This has been a matter of some controversy over the years. However, a recent detailed meta-analysis by Sanai and Berger supports the notion that improving resection improves survival [42]. This has been supported for glioblastoma by several other rigorous studies using MRI assessment of volumetric resection in all patients (as opposed to more subjective measures of resection), studies which showed the strongest benefit comes from resection exceeding 98% of the preoperative volume of contrast-enhancing tumor [43,44]. The survival advantage for low grade gliomas was even more striking with resection based on the tumor boundaries delineated on T2-weighted (not post-contrast) images, a necessary difference given that most low grade gliomas do not enhance [42]. The most difficult area of the brain for maintaining accuracy of glioma resection under neuronavigation is the occipital region. Whether the patient is placed prone or in the lateral

position, the occipital portion of the brain tends to bulge significantly once the dura is opened. Thus, a relatively immediate brain shift occurs and it is difficult to prevent or avoid that shift. As a general rule, tumors in the posterior third of the brain pose greater difficulties for accurate localization than those in the anterior third. Another compounding factor leading to lower accuracy in the posterior cerebrum (and indeed in the posterior fossa) stems from the need to concentrate fiducial markers anteriorly to take advantage of the richer variety of surfaces there. Thus, registration accuracy declines when the tumor is located farther away from that anterior cluster. Smaller tumors can be resected with better accuracy using neuronavigation, and in deep tumors error accumulates to a greater degree than in superficial tumors [45]. Retrospective analysis of 76 patients with glioblastomas undergoing resection with or without neuronavigation has been reported by Kurimoto et al. [46]. They found a greater likelihood of gross total removal in the neuronavigation group (64% vs. 38%), which correlated with increases in survival time. However, given the low incidence of aggressive removal overall, it is possible that the advantage of neuronavigation might diminish were greater effort made to achieve complete resection independent of the adjunctive technologies used. Reoperation for gliomas is encountered more frequently now given the multiplicity of approaches to tumor suppression by medical neurooncologists and the resulting longer survival times achieved. Patients with recurrent gliomas are likely to show less brain shift than those who have not previously undergone surgery. The cortical surface usually adheres to the dura around the prior craniotomy site and therefore retracts less with hyperventilation and CSF release. Indeed, surgical navigation is very helpful in such patients because the gliosis and treatment effect from prior radiation and chemotherapy tend to make the tumor’s margins less distinct and more difficult to follow. When radiation necrosis is superimposed on a


tumor, the typical hyperechogenicity seen by ultrasound converts to a bland and indistinct look that makes complete resection more difficult to achieve. Navigation is therefore of great utility in such cases. Resection of the entire extent of enhancement, including those areas of enhancement that might be interpreted as post-irradiation effect, is important because such regions often consist of a mix of tumor and treatment effect, which provides a seed for further growth if left in place. The greatest advantage of all is seen when image-guided surgery is applied to low-grade gliomas (> Figure 45-9). These tumors generally have fairly distinct margins on T2-weighted images, but a relatively bland appearance during surgery with borders that can be quite difficult to define. A highly skilled and experienced surgeon may achieve good resection of such tumors without neuronavigation, but we have found that even those with a wealth of experience still leave behind small bits of tumor that they catch only when neuronavigation is used. To minimize the effects of brain shift, the interface between tumor and adjacent brain (as defined by the navigation system) is opened by resecting along that edge and placing cottonoids to maintain the distinction. In this fashion, any brain shift that occurs as internal debulking of the tumor proceeds will be minimized. Just as is true for metastases, cysts associated with gliomas should be maintained for as long as possible in the planning and delineating stages, so that tumor outlines can be obtained while registration is still accurate. Those areas in which neuronavigation is absolutely vital in glioma surgery include tumors within the insula, those which cross the midline, those affecting the thalamus, and those brainstem tumors which by virtue of their exophytic quality are amenable at least to partial resection.

45

complications. Shirane et al. have shown that after including neuronavigation in the occipital pranstentorial approach to pineal tumors, their complication rate fell from 27% to zero [47]. Since such tumors often induce hydrocephalus by blockage of the aqueduct of Sylvius, endoscopic third ventriculostomy guided by computerassisted navigation is a useful adjunct to frameless stereotactic biopsy of the mass [48]. In a similar vein, neuronavigational endoscopy (with the endoscope enrolled as the localizing probe) can effectively be applied for septal fenestration to relieve ventricular trapping, for removal of colloid cysts, and for biopsy of tumors within the ventricle or arising from the ventricular wall [49]. Pineal endoscopy has also been proposed based on cadaver study, but it has not yet been reported clinically [50].

Stereotactic Biopsy For many years surgeons have relied on framebased methods of localization to achieve the precision required in stereotactic biopsy of deep tumors. Similar precision is now available from frameless techniques, as shown by Dammers et al. in their analysis of 227 frame-based and 164 frameless biopsies [51]. Both groups were identical in complication rates and diagnostic yield (12 and 89% respectively). This confirms the results of an earlier series whose authors made similar comparisons while achieving a lower rate (6%) of permanent morbidity and concluded that frameless biopsy was equally effective as, but more efficient than, biopsy using a frame [52].

Intraoperative Imaging Pineal Tumors and Endoscopy In the tricky, anatomically rich area of the pineal recess, image guidance can help avoid

Neuronavigation relying on images obtained during surgery might be deemed ‘‘imaginginteractive’’ as opposed to the ‘‘image-guided’’ techniques thus far described. Intraoperative

711

712

45


. Figure 45-9 Tumor resection in the BrainSuite intraoperative MRI. (a) This low grade astrocytoma in the right frontal lobe is well delineated on FLAIR images taken prior to resection (b) After resection of all visible tumor, a new image shows that removal was in fact incomplete at the deepest and most posterior part of the resulting cavity (c) Further resection eliminates the residual abnormality. Without intraoperative MRI, this patient would have had a less complete resection than was obtained using that technology

scans offer clear advantages in that they can give up-to-the-minute preoperative information if done after induction of anesthesia but before the incision has been made. The stages of the resection can be imaged in real time, extent of resection can be quantified, residual tumor identified, and reimaging to update registration

can minimize the issue of brain shift. The initial foray into intraoperative scanning was made in Pittsburgh in the early 1980s. A fixed CT scanner was placed in an operating room used for stereotactic biopsy and for craniotomies, and was used to confirm target localization for biopsies, to rule out post-resection hemorrhage, and to show


the extent of resection [53]. In the years since, mobile CT scanners have been developed for intraoperative application and produce images of sufficient quality to be usable for determining the limits of resection (> Figure 45-10). Series have been reported by two groups in Germany who were favorably impressed with the utility of the CT scan in the operative environment [54,55]. The most recent descriptions of intraoperative CT have come in pituitary surgery, specifically the use of the Arcadis Orbic which allows both conventional fluoroscopic views and multiplanar reconstructions to be acquired during surgery [56]. Fluoroscopic images match or exceed the quality of those images from standard C-arms and multiplanar reconstructions gave images equal in quality to those provided by preoperative stealth CT. These authors felt that this system provided more reliability than did registration of CT images acquired prior to surgery. The advantages of CT are its speed and lower cost. However, it sacrifices anatomic detail and defines most intracranial tumors less precisely than does MRI. Gliomas can be mapped

45

more precisely on MRI and areas of subtle involvement can be shown that CT scans overlook. In addition, extent of resection is more complete with MRI due to its better appreciation of subtle areas of residual tumor along the resection margin. For these reasons, tumor surgeons have embraced intraoperative MRI as a superior technology, particularly when resecting gliomas. The disadvantages of intraoperative MRI scanners are their lack of widespread availability and high cost. Placing an intraoperative MRI into use demands a major commitment of resources by the hospital installing it, and often pre-existing space must retrofitted with magnetic shielding and reinforced to account for the weight of the scanner. Ferromagnetic instruments and equipment cannot be used within the magnetic field, and therefore certain instruments (like the specula used in transsphenoidal surgery) must be replaced with expensive MRIcompatible equipment. For these reasons, it is likely that intraoperative CT will retain a role for many years to come, particularly in cases of contrast-enhancing gliomas or complex metastases. The ideal cases for the intraoperative MRI

. Figure 45-10 CereTom mobile CT unit which can provide intraoperative images. It is battery-powered, has a 25-cm field of view, and generates up to eight slices per revolution using a modular multi-row detector

713

714

45


are non-enhancing cerebral tumors, typically of low grade, in which a surgeon may have difficulty identifying margins precisely, and complex pituitary macroadenomas, in which lateral or superior extension of tumor may persist despite the best efforts of a skilful surgeon. Neither of those tumor types will show well on CT. Intraoperative MRI dates to the mid 1990s and was first applied to intracranial tumor surgery by Peter Black at the Brigham & Women’s Hospital [57,58]. Black’s group used a fixed machine of 0.5 T strength into which patients were placed for surgery performed within the center of the magnetic field (> Figure 45-11). Thus all instruments had to be MRI-compatible and patients did not move in and out of the machine for scanning. In the first 200 patients from this series, hyperacute hemorrhage was noted in two patients from whom the clot was then removed. Other groups reporting the use of 0.5 T MRI during surgery included Zimmerman et al. [59]. The use of lower field strengths down to 0.12 T (as in

the vertical gap systems produced by Fonar or Hitachi) allows use of standard surgical instruments within the magnetic field, but yields reduced temporal resolution and spatial resolution per unit time, as well as a higher signal-to-noise to ratio even in newer generation models [60]. For this reason, high field (1.5 T) systems have been utilized to improve image quality, which is particularly important in glioma surgery. All intraoperative MRI systems provide basic imaging capacity for T1- and T2-weighted views, but high field systems can also acquire other MRI subsets including angiography, functional MR, diffusion-weighted imaging, chemical shift imaging, and MR spectroscopy. Direct comparison by Bergsneider et al. of results obtained from surgery using 0.2 T versus 1.5 T magnets showed that a greater mean extent of resection was achieved for gliomas using intraoperative MRI versus standard imaged-guided frameless neuronavigation (91% vs. 79% respectively). Interestingly enough, there was little difference in the degree

. Figure 45-11 Intraoperative MRI at the Brigham and Women’s Hospital. In this Signa SP ‘‘double doughnut’’ unit (GE Medical Systems) with a field strength of 0.5 T, the patient remains inside the core magnetic field during surgery


of resection achieved using the 0.2 T versus the 1.5 T intraoperative MRI without updated neuronavigation; with updated registration, the mean percent resection increased from 92 to 98% [61]. Another study using the 0.5 T magnet showed no difference between two matched sets of 32 patients with high-grade gliomas, one of which underwent operation with neuronavigation and one without [62]. Accessible high-grade gliomas can be resected without neuronavigation in many patients to a near-complete level by surgeons experienced in such procedures. However, lower grade tumors pose particular challenges even to those well versed in their nuances, and we routinely perform such operations within the BrainSuite environment, an integrated operating room with a fixed magnet and rotating operating room table that

45

swings out of the MRI machine and allows performance of the procedure at a distance far enough from the magnet to allow the use of normal surgical instruments (> Figure 45-12). A detachable head coil permits the surgeon access to the patient’s head when the patient is outside the scanner and ready for the procedure to continue (> Figure 45-13). This particular system does provide updating of registration with each new scan performed and we find that feature to be particularly valuable for increasing the accuracy of localization. Nimsky et al. have published a series of 182 procedures performed using a 1.5 T magnet, and derived results favoring the use of such equipment in glioma surgery [63]. They found that the intraoperative MRI influenced the procedure in 36% of patients, in whom surgery would have

. Figure 45-12 The BrainSuite intraoperative MRI. This system places the patient on a rotationally mobile operating table which can move into the MRI for scanning, then out for the operation. This allows the use of ferromagnetic instruments during surgery, but materials that enter the magnet (e.g., intravenous catheters, endotracheal tubes, or patient implants) must still be MRI-compatible. The system includes an operating microscope and frameless neuronavigation with automatic re-registration, along with wall display of intraoperative images. Differently colored zones on the floor indicate sectors of higher or lower magnetic field strength

715

716

45


otherwise stopped but in whom instead it continued for removal of residual tumor [63]. The percent of final tumor volume was significantly reduced by intraoperative scanning in both low-grade and high-grade tumors with complete resection achieved in 57% of the lowgrade cases and in 27% of the high-grade cases. This series may be criticized for not achieving a higher degree of resection in both groups, and one might argue that more aggressive resection might have resulted in less difference between first and subsequent scans without compromising safety. Functional MRI can also be performed intraoperatively in high field machines although no one to date has performed a rigorous comparison between the results of functional MRI and those of direct cortical stimulation (for speech and motor mapping) or somatosensory evoked potentials (for identification of the central sulcus) [32,64]. It might be argued that functional MRI scans acquired prior to surgery and combined with intraoperative images by an overlay technique provide greater reliability due to their

acquisition in a more controlled environment (> Figure 45-14). This question also has yet to be studied. Precise alignment of previously obtained images with those acquired during surgery may require correction of geometric distortion inherent in intraoperative MRI [65]. The impact of neuronavigation incorporating functional MRI increases when direct subcortical stimulation is added to it, with positive stimulation indicating proximity to the corticospinal tracts within 10 mm [34]. Additional methods of identifying motor functioning include direct display by diffusion tensor imaging of motor fiber tracts overlaid on standard anatomical images (> Figure 45-15) [67]. Such imaging can be used to delineate glioma margins more crisply [68]. Tractography in intraoperative MRI has been used by Mikuni et al. to define the accuracy of motor evoked potentials [69]. These authors found that such potentials consistently provoked motor activity at distances 13 mm. When stimulation was applied in the region of the corona radiata (as

. Figure 45-13 The head-holder for the BrainSuite intraoperative MRI. The magnetic coil sits above the patient’s head during scan acquisition, and is removed to allow access to the surgical site during the operative phase


45

. Figure 45-14 Functional MRI (fMRI) showing activation of speech centers with a receptive task. This patient’s tumor was an anaplastic astrocytoma with low grade surround. Wernicke’s area had been diffused and anteriorly displaced in relation to the tumor, and was immediately adjacent to the tumor margin. This dataset was imported into the intraoperative MRI (BrainSuite) and successfully applied to neuronavigation to preserve speech function during tumor removal

717

718

45


. Figure 45-15 Diffusion tensor imaging in the intraoperative MRI. The white matter tracts in the vicinity of a cavernous hemangioma are shown in right frontal oblique (left) and left dorsal oblique (center) views, with infiltration of the left medial lemniscus shown by a segmented view showing only the tracts and the lesion (right) [66]

opposed to the corticospinal tract itself), movements were elicited at distances of 8–12 mm. Use of intraoperative MRI in this fashion can validate older techniques and provide valuable input that limits or extends resection [70]. Tractography has also been applied to patients with brainstem lesions around which the patterns of tract alteration include deviation, deformation, infiltration, and apparent tract interruption. That information can alter the ratio of benefit to risk to a favorable enough degree to make operation on a brainstem tumor actually possible with reasonable safety (> Figure 45-15) [66]. The brainstem is less subject to brain shift than are the contents of the supratentorial compartment, but the margin for error during surgery is almost nonexistent due to the anatomical density of that region. Both diffusion tensor fiber tracking and MR spectroscopy appear to be valid even in the immediate vicinity of a glioma margin. Given that such tumors are inherently invasive, it is possible that changes in neuronal connectivity at the interface between brain and tumor might impair these radiographic techniques. However, Stadlbauer et al. have found that although the number of neuronal fibers per voxel and the fractional anisotropy are both lower and that mean diffusivity is increased, these changes do

not prevent creation of useful images [33,35]. These alterations were magnified in patients with sensorimotor deficits but even in that group, they still did not prevent image creation. These authors found that overlay of proton MR spectroscopy helps in deciding whether diffusivity changes in nearby fiber tracts are due to tumor infiltration or peritumoral edema, a difference relevant for surgical decision-making. The availability over the past several years of even higher field magnets for MRI has led some groups to apply these during surgery in an effort to enhance further the image quality obtained from the intraoperative MRI. Hall et al. have reported a similar surgical environment in the 3 T magnet to that seen at 1.5 T [71]. Such minimally ferromagnetic items as surgical needles, staples, and disposable scalpels were safely controlled by the surgeon in their version of the intraoperative MRI, which involves performance of the entire procedure within the magnet. However, image quality is not sufficiently different to allow the conclusion that increasing magnet strength beyond 1.5 T further improves the extent of tumor resection, or that it reduces complications from impingement on functional fiber tracts. Intraoperative high field MRI may provide particular benefit in transsphenoidal surgery and


by extension, in craniotomy for pituitary tumor (> Figure 45-16). Although it has little role in the resection of microadenomas, many of which can be relatively occult on imaging, its ability to enhance resection of macroadenomas is relatively strong. Nimsky et al. have reported the results of intraoperative MRI in 106 patients with clinically nonfunctional pituitary macroadenomas, in whom resection continued if interim intraoperative imaging showed an accessible tumor remnant [72]. Of the 106 patients, 85 were operated with intent to achieve complete removal of tumor. In those 85, imaging showed a tumor remnant in 36 (42%), 29 (34%) of whom went on to further resection, which was then achieved completely in 21. Thus, the rate of complete tumor removal increased from 58 to 82% because of the use of intraoperative MRI. Even in the group with an intended partial removal, resection could be extended in 38% when intraoperative imaging showed accessible fragments not hitherto suspected. When combined with endoscopy, intraoperative MRI can enhance the degree of resection achieved in these often anatomically complex lesions. The difficulty with

45

such studies, whether for pituitary tumors, gliomas, or any other tumor subset, is that surgical experience may allow some surgeons to achieve similar results without intraoperative MRI; thus the degree of success of the technology may reflect the lack of experience of its user. For pituitary tumors as for gliomas and other tumors, intraoperative MRI will likely be most effectively applied to the more difficult cases, in which its benefit will be most profoundly felt. These would include patients undergoing reoperation, those whose tumors who have indistinct boundaries or which blend with adjacent structures, and those with extensions carrying them close to critical anatomical structures the integrity of which must be maintained. The great advantage of the intraoperative MRI is its ability to provide reimaging and reregistration, and thereby it reduces directly the primary impediment to accurate neuronavigation, namely brain shift. The BrainSuite system in use at our institution contains an automatic registration function that requires no external fiducial markers and no touching of a probe to the skin surface. As this feature eliminates several

. Figure 45-16 Intraoperative MRI done in the BrainSuite on the same patient with pituitary tumor shown in Figure 2. (a) Preoperative images (coronal, post-contrast) show the tumor invading the cavernous sinus, filling the sella, and compressing the optic apparatus (b) Post-resection images show nice clearance of the tumor with the exception of a small remnant adherent to the lateral wall of the right cavernous sinus, which was subsequently partially resected to preserve the function of cranial nerves to which it was adherent

719

720

45


of the potential sources of inaccuracy and cuts down on the time the surgeon spends on the registration process, automatic re-registration has been one of the particularly attractive features of this otherwise somewhat cumbersome system. Practical issues in the intraoperative MRI environment include not only the need to use non-ferromagnetic endotracheal tube components and intravenous catheter hubs, but also the absence of a simple system for maintenance of the sterile field in systems such as ours in which surgery proceeds outside the magnet with intermittent forays into the heart of the magnetic field for scanning. The drapes must be tucked up carefully around the patient before the couch is shifted into the MRI; and then recovered with an additional sterile layer when surgery recommences after each scan. Manipulation of the drapes adds to the risk of infection, and indeed we have found that our infection rate is somewhat higher in intraoperative MRI cases than in those performed in a standard operating room with simple neuronavigation. This excess infection risk has not been a factor, however, in the systems in which the operation proceeds within the magnet and in which no such draping changes are necessary.

Caveats The full utility of neuronavigation has not yet been realized, and its value has been difficult to assess objectively. Although the aficionados of technology in surgery suggest that frameless stereotactic systems are mandatory in craniotomy for tumor, it is better to conclude that they can be immensely helpful in difficult or sensitive areas of the brain, but that careful selection will maximize benefit. Bearing in mind the tendency of human beings to fall in love with their toys (and the unfortunate truth that to the man with a bat, all round objects may resemble a ball), we should take the more nuanced view that resection of

some tumors will be greatly enhanced by neuronavigation, whereas others may be well removed without image guidance. The degree to which neuronavigation is applied in any given surgeon’s practice will depend on many factors, including the nature of the cases operated, the experience of that surgeon with more traditional methods of resection, and the availability of equipment of proven accuracy. Such technology will certainly augment the reach of a skilled surgeon, but cannot substitute for the fundamentals of surgical skill and good intraoperative judgment. Efforts at proving the value of neuronavigation have been confounded by selection bias: in the Glioma Outcomes Project, it was used more for younger patients and for those with smaller low grade tumors, and multivariate analysis to eliminate those variables showed no survival advantage [73]. Willems et al. randomized 45 patients with solitary contrast-enhancing tumors to surgery with or without neuronavigation [74]. They found no difference between the two groups in either survival, extent of resection, or induction of postoperative neurological deficit, and concluded that there was no rationale for routine use of the technology. However, this study’s numbers were small, and it is possible (although certainly not proven) that greater statistical power achieved by a larger sample size would allow detection of a difference that a smaller sample size cannot show. The ongoing evolution of the available navigation systems has advanced this field tremendously, but rigorous analysis of outcomes is difficult, as no one system is accepted as ‘‘best’’ and large series of patients cannot be accrued quickly enough to prevent one or more changes in methodology during the accrual.

Future Directions Practical advances in neuronavigation will likely come from three areas. The first is imaging, which is undergoing a revolution of its own


with the advent of methods for creating visual maps of metabolic and even gene activity. Current systems already have multimodality capability including fMRI, tractography, and MR spectroscopy in addition to standard images. As newer imaging based on molecular events becomes more robust and allows detection of tumor at earlier stages than now possible, or even permits detection of pretumoral tissue or of stem cells, such images will certainly be converted to use for intraoperative image guidance. The degree of advantage conferred by surgical navigation depends on the quality of the images that drive it. The second area is virtual reality. Gildenberg and Labuz have used and described a system that superimposes a computer-generated rendering of the target volume on a real-time video image, updated during surgery to show only the part of the tumor actually being resected [75]. Kockro et al. described preoperative planning in a stereoscopic virtual reality environment for tumors and arteriovenous malformations in areas with difficult access, and found that it allowed efficient assembly of surgically relevant spatial information from multiple modalities, and predictably that its utility depends on accurate co-registration of the imaging datasets and on its rapidity of real-time interaction [76]. Such systems have been described for training residents in ventriculostomy and for planning aneurysm clipping, but they have not yet been much applied to tumor work [77–79]. The third predictable area of future activity is robotics. Already in use in laparoscopic procedures, its applications in neurosurgery have barely been touched but show clear logic for a marriage between robotic and neuronavigational technology. Navigated robotic endoscopic ventriculostomy has been reported [80]. In addition, instrument tracking during surgery has established criteria for useable neurosurgical robotic systems including translational speeds of up to 12.7 cm/s and rotational speeds of up to 40 /s [81]. Telemanipulation within the context of frameless

45

stereotactic localization will undoubtedly be applied to tumor resection as robotic technology develops further, and it represents the next frontier for stereotactic surgery.

References 1. Horsley V, Clarke RH. ‘‘The structure and function of the cerebellum examined by a new method.’’ Brain 1906;31:45-124. 2. Leksell L. ‘‘A stereotactic apparatus for intracerebral surgery.’’ Acta Chir Scan 1949;99:229-33. 3. Kelly PJ, Kall BA, et al. ‘‘Computer-assisted stereotaxic laser resection of intra-axial brain neoplasms.’’ J Neurosurg 1986;64(3):427-39. 4. Morita A, Kelly PJ. ‘‘Resection of intraventricular tumors via a computer-assisted volumetric stereotactic approach.’’ Neurosurgery 1993;32:920-6. 5. Moshel YA, Link MJ, et al. ‘‘Stereotactic volumetric resection of thalamic pilocytic astrocytomas.’’ Neurosurgery 2007;61:66-75. 6. Rezai AR. ‘‘Integration of functional brain mapping in image-guided neurosurgery.’’ Acta Neurochir Suppl 1997;68:85-9. 7. Roberts DW, Strohbehn JW, et al. ‘‘A frameless stereotaxic integration of computerized tomographic imaging and the operating microscope.’’ J Neurosurg 1986;65:545-9. 8. Roberts DW, Hartov A, et al. ‘‘Intraoperative brain shift and deformation: a quantitative analysis of cortical displacement in 28 cases.’’ Neurosurgery 1998;43:749-58. 9. Watanabe E, Watanabe T, et al. ‘‘Three-dimensional digitizer (neuronavigator): new equipment for computed tomography-guided stereotaxic surgery.’’ Surg Neurol 1987;27:543-7. 10. Watanabe E, Mayanagi Y, et al. ‘‘Open surgery assisted by the neuronavigator, a stereotactic, articulated, sensitive arm.’’ Neurosurgery 1991;28:792-9. 11. Golfinos JG, Fitzpatrick BC, et al. ‘‘Clinical use of a frameless stereotactic arm: results of 325 cases.’’ J Neurosurg 1995;83:197-205. 12. Kleinpeter G, Lothaller C. ‘‘Frameless neuronavigation using the ISG-system in practice: from craniotomy to delineation of lesion.’’ Minim Invasive Neurosurg 2003;46:257-64. 13. Roessler K, Ungersboeck K, et al. ‘‘Frameless stereotactic lesion contour-guided surgery using a computer-navigated microscope.’’ Surg Neurol 1998;49:282-8. 14. Kato A, Yoshimoto T, et al. ‘‘A frameless, armless navigational system for computer-assisted neurosurgery. Technical note.’’ J Neurosurg 1991;74(5):845-9. 15. Heilbrun MP, McDonald P, et al. ‘‘Stereotactic localization and guidance using a machine vision technique.’’ Stereotact Funct Neurosurg 1992;58:94-8.

721

722

45


16. Heilbrun MP, Koehler S, et al. ‘‘Preliminary experience using an optimized three-point transformation algorithm for spatial registration of coordinate systems: a method of noninvasive localization using framebased stereotactic guidance systems.’’ J Neurosurg 1994; 81(5):676-82. 17. Roberts DW, Strohbehn JW, et al. ‘‘The stereotactic operating microscope: accuracy refinement and clinical experience.’’ Acta Neurochir Suppl 1989;46:112-4. 18. Chen JCT, Moffit K, et al. ‘‘Head-mounted display system for microneurosurgery.’’ Stereotact Funct Neurosurg 1997;68:25-32. 19. Ray WZ, Barua M, et al. ‘‘Anatomic visualization with ultrasound-assisted intracranial image guidance in neurosurgery: a report of 30 patients.’’ J Am Coll Surg 2004;199(2):338-43. 20. Wu Z, Hartov A, et al. ‘‘Multimodal image re-registration via mutual information to account for initial tissue motion during image-guided neurosurgery.’’ Conf Proc IEEE Eng Med Biol Soc 2004;3:1675-8. 21. Sun H, Roberts DW, et al. ‘‘Cortical surface tracking using a stereotactic operating microscope.’’ Neurosurgery 2005;56:86-97. 22. Miga MI, Paulsen KD, et al. ‘‘Model-updated image guidance: initial clinical experiences with gravity-induced brain deformation.’’ IEEE Trans Med Imaging 1999;18:866-74. 23. Comeau RM, Sadikot AF, et al. ‘‘Intraoperative ultrasound for guidance and tissue shift correction in imageguided neurosurgery.’’ Med Phys 2000;27:787-800. 24. Roberts DW, Lunn K, et al. ‘‘Intra-operative image updating.’’ Stereotact Funct Neurosurg 2001;76:148-50. 25. Roth J, Biyani N, et al. ‘‘Real-time neuronavigation with high-quality 3D ultrasound sonowand in pediatric neurosurgery.’’ Pediatr Neurosurg 2007;43:185-91. 26. Nimsky C, Ganslandt O, et al. ‘‘Quantification of, visualization of, and compensation for brain shift using intraoperative magnetic resonance imaging.’’ Neurosurgery 2000;47:1070-9. 27. Reinges MH, Nguyen HH, et al. ‘‘Course of brain shift during microsurgical resection of supratentorial cerebral lesions: limits of conventional neuronavigation.’’ Acta Neurochir 2004;146:369-77. 28. Wittek A, Miller K, et al. ‘‘Patient-specific model of brain deformation: application to medical image registration.’’ J Biomech 2007;40:919-29. 29. Suess O, Picht T, et al. ‘‘Neuronavigation without rigid pin fixation of the head in left frontotemporal tumor surgery with intraoperative speech mapping.’’ Neurosurgery 2007;60:330-8. 30. Mangano FT, Limbrick DDJ, et al. ‘‘Simultaneous imageguided and endoscopic navigation without rigid cranial fixation: application in infants: technical case report.’’ Neurosurgery 2006;58 4 Suppl 2:ONS-E377. 31. Rygh OM, Nagelhus Hernes TA, et al. ‘‘Intraoperative navigated 3-dimensional ultrasound angiography in tumor surgery.’’ Surg Neurol 2006;66:581-92.

32. Liu H, Hall WA, et al. ‘‘The roles of functional MRI in MR-guided neurosurgery in a combined 1.5 Tesla MRoperating room.’’ Acta Neurochir Suppl 2003;85:127-35. 33. Stadlbauer A, Moser E, et al. ‘‘Integration of biochemical images of a tumor into frameless stereotaxy achieved using a magnetic resonance imaging/magnetic resonance spectroscopy hybrid data set.’’ J Neurosurg 2004;101: 287-94. 34. Mikuni N, Okada T, et al. ‘‘Clinical impact of integrated functional neuronavigation and subcortical electrical stimulation to preserve motor function during resection of brain tumors.’’ J Neurosurg 2007;106(4):593-8. 35. Stadlbauer A, Nimsky C, et al. ‘‘Changes in fiber integrity, diffusivity, and metabolism of the pyramidal tract adjacent to gliomas: a quantitative diffusion tensor fiber tracking and MR spectroscopic imaging study.’’ AJNR Am J Neuroradiol 2007;28:462-9. 36. Pirotte B, Goldman S, et al. ‘‘Integrated positron emission tomography and magnetic resonance imaging-guided resection of brain tumors: a report of 103 consecutive procedures.’’ J Neurosurg 2006;104:238-53. 37. Helm PA, Eckel TS. ‘‘Accuracy of registration methods in frameless stereotaxis.’’ Comput Aided Surg 1998;3 (2):51-6. 38. Sawaya R, Hammoud M, et al. ‘‘Neurosurgical outcomes in a modern series of 400 craniotomies for treatment of parenchymal tumors.’’ Neurosurgery 1998; 42(5):1044-55. 39. Tan TC, McL Black P. ‘‘Image-guided craniotomy for cerebral metastases: techniques and outcomes.’’ Neurosurgery 2003;53(1):82-9. 40. Takahashi M, Yamada R, et al. ‘‘Navigation-guided ommaya reservoir placement: implications for the treatment of leptomeningeal metastases.’’ Minim Invasive Neurosurg 2007;50(6):340-5. 41. Paleologos TS, Wadley JP, et al. ‘‘Clinical utility and costeffectiveness of interactive image-guided craniotomy: clinical comparison between conventional and image-guided meningioma surgery.’’ Neurosurgery 2000;47:40-7. 42. Sanai N, Berger MS. ‘‘Glioma extent of resection and its impact on patient outcome.’’ Neurosurgery 2008;62:753-64. 43. Lacroix M, Abi-Said D, et al. ‘‘A multivariate analysis of 416 patients with glioblastoma multiforme: prognosis, extent of resection, and survival.’’ J Neurosurg 2001;95:190-8. 44. Stummer W, Reulen HJ, et al. ‘‘Extent of resection and survival in glioblastoma multiforme: identification of and adjustment for bias.’’ Neurosurgery 2008;62:564-76. 45. Benveniste R, Germano IM. ‘‘Evaluation of factors predicting accurate resection of high-grade gliomas by using frameless image-guided stereotactic guidance’’. Neurosurg Focus 2003;14(2):E5. 46. Kurimoto M, Hayashi N, et al. ‘‘Impact of neuronavigation and image-guided extensive resection for adult patients with supratentorial malignant astrocytomas: a


47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

single-institution retrospective study.’’ Minim Invasive Neurosurg 2004;47:278-83. Shirane R, Shamoto H, et al. ‘‘Surgical treatment of pineal region tumours through the occipital transtentorial approach: evaluation of the effectiveness of intra-operative micro-endoscopy combined with neuronavigation.’’ Acta Neurochir 1999;141:801-8. Kim IY, Jung S, et al. ‘‘Neuronavigation-guided endoscopic surgery for pineal tumors with hydrocephalus.’’ Minim Invasive Neurosurg 2004;47(6):365-8. Schroeder HW, Wagner W, et al. ‘‘Frameless neuronavigation in intracranial endoscopic neurosurgery.’’ J Neurosurg 2001;94(1):72-9. Youssef AS, Keller JT, et al. ‘‘Novel application of computer-assisted cisternal endoscopy for the biopsy of pineal region tumors: cadaveric study.’’ Acta Neurochir 2007;149(4):399-406. Dammers R, Haitsma IK, et al. ‘‘Safety and efficacy of frameless and frame-based intracranial biopsy techniques.’’ Acta Neurochir 2008;150(1):23-9. Woodworth GF, McGirt MJ, et al. ‘‘Frameless imageguided stereotactic brain biopsy procedure: diagnostic yield, surgical morbidity, and comparison with the frame-based technique.’’ J Neurosurg 2006;104 (2):233-7. Engle DJ, Lunsford LD. ‘‘Brain tumor resection guided by intraoperative computed tomography.’’ J Neuro-Oncol 1987;4:361-70. Grunert P, Muller-Forell W, et al. ‘‘Basic principles and clinical applications of neuronavigation and intraoperative computed tomography.’’ Comput Aided Surg 1998;3:166-73. Gumprecht H, Lumenta CB. ‘‘Intraoperative imaging using a mobile computed tomography scanner.’’ Minim Invasive Neurosurg 2003;46:317-22. Fox WC, Wawrzyniak S, et al. ‘‘Intraoperative acquisition of three-dimensional imaging for frameless stereotactic guidance during transsphenoidal pituitary surgery using the Arcadis Orbic System.’’ J Neurosurg 2008;108:746-50. Martin C, Alexander EI, et al. ‘‘Surgical treatment of lowgrade gliomas in the intraoperative magnetic resonance imager.’’ Neurosurg Focus 1998;4(4):E8. Schwartz R, Shamoto H, et al. ‘‘Intraoperative MR imaging guidance for intracranial neurosurgery: experience with the first 200 cases.’’ Radiology 1999; 211(2):477-88. Zimmermann M, Seifert V, et al. ‘‘Open MRI-guided microsurgery of intracranial tumours in or near eloquent brain areas.’’ Acta Neurochir 2001;143:327-37. Schulder M, Salas S, et al. ‘‘Cranial surgery with an expanded compact intraoperative magnetic resonance images.’’ J Neurosurg 2006;104:611-7. Bergsneider M, Sehati N, et al. ‘‘Extent of glioma resection using low-field (0.2 T) versus high-field (1.5 T) intraoperative MRI and image-guided frameless neuronavigation.’’ Clin Neurosurg 2005;52:389-99.

45

62. Hirschberg H, Samset E, et al. ‘‘Impact of intraoperative MRI on the surgical results for high-grade gliomas.’’ Minim Invasive Neurosurg 2005;48:77-84. 63. Nimsky C, Fujita A, et al. ‘‘Volumetric assessment of glioma removal by intraoperative high-field magnetic resonance imaging.’’ Neurosurgery 2004;55:358-70. 64. Nimsky C, Ganslandt O, et al. ‘‘Intraoperative visualization of the pyramidal tract by diffusion-tensor-imagingbased fiber tracking.’’ Neuroimage 2006; 30(4):1219-29. 65. Archip N, Clatz O, et al. ‘‘Compensation of geometric distortion effects on intraoperative magnetic resonance imaging for enhanced visualization in image-guided neurosurgery.’’ Neurosurgery 2008; 62 3 Suppl 1:209-15. 66. Chen X, Weigel D, et al. ‘‘Diffusion tensor imaging and white matter tractography in patients with brainstem lesions.’’ Acta Neurochir 2007;149(11):1117-31. 67. Nimsky C, Ganslandt O, et al. ‘‘Intraoperative high-field MRI: anatomical and functional imaging.’’ Acta Neurochir Suppl 2006;98:87-95. 68. Price SJ, Jena R, et al. ‘‘Improved delineation of glioma margins and regions of infiltration with the use of diffusion tensor imaging: an image-guided biopsy study.’’ AJNR Am J Neuroradiol 2006;27(9):1969-74. 69. Mikuni N, Okada T, et al. ‘‘Clinical significance of preoperative fibre-tracking to preserve the affected pyramidal tracts during resection of brain tumours in patients with preoperative motor weakness.’’ J Neurol Neurosurg Psychiatry 2007;78(7):716-21. 70. Mikuni N, Okada T, et al. ‘‘Comparison between motor evoked potential recording and fiber tracking for estimating pyramidal tracts near brain tumors.’’ J Neurosurg 2007;106(1):128-33. 71. Hall WA, Galicich W, et al. ‘‘3-Tesla intraoperative MR imaging for neurosurgery.’’ J Neuro-Oncol 2006;77: 297-303. 72. Nimsky C, von Keller B, et al. ‘‘Intraoperative high-field magnetic resonance imaging in transsphenoidal surgery of hormonally inactive pituitary macroadenomas.’’ Neurosurgery 2006;59(1):105-14. 73. Litofsky NS, Bauer AM, et al. ‘‘Image-guided resection of high-grade glioma: patient selection factors and outcome.’’ Neurosurg Focus 2006;20(4):E16. 74. Willems PW, Taphoorn MJ, et al. ‘‘Effectiveness of neuronavigation in resecting solitary intracerebral contrast-enhancing tumors: a randomized controlled trial.’’ J Neurosurg 2006;104:360-8. 75. Gildenberg PL, Labuz J. ‘‘Use of a volumetric target for image-guided surgery.’’ Neurosurgery 2006;59:651-9. 76. Kockro RA, Serra L, et al. ‘‘Planning and simulation of neurosurgery in a virtual reality environment.’’ Neurosurgery 2000;46:118-35. 77. Anil SM, Kato Y, et al. ‘‘Virtual 3-dimensional preoperative planning with the dextroscope for excision of a 4th ventricular ependymoma.’’ Minim Invasive Neurosurg 2007;50(2):65-70.

723

724

45


78. Banerjee PP, Luciano CJ, et al. ‘‘Accuracy of ventriculostomy catheter placement using a head- and hand-tracked high-resolution virtual reality simulator with haptic feedback.’’ J Neurosurg 2007;107(3):515-21. 79. Wong GK, Zhu CX, et al. ‘‘Craniotomy and clipping of intracranial aneurysm in a stereotactic virtual reality environment.’’ Neurosurgery 2007;61(3):564-8.

80. Zimmermann M, Krishnan R, et al. ‘‘Robot-assisted navigated endoscopic ventriculostomy: implementation of a new technology and first clinical results.’’ Acta Neurochir 2004;146(7):697-704. 81. Woerdeman PA, Willems PW, et al. ‘‘The analysis of intraoperative neurosurgical instrument movement using a navigation log-file.’’ Int J Med Robot 2006; 2(2):139-45.

55 Image Guided Management of Cerebral Metastases P. Kongkham . M. Bernstein

Introduction Cerebral metastases represent the most common type of brain tumor seen in adults in clinical practice, with an annual incidence outnumbering primary brain tumors by approximately 10:1 [1]. Over 1 million people per year in the United States are diagnosed with cancer, and of these, up to 170,000 will develop brain metastases [1–3]. In fact, between 10 and 15% of cancer patients are ultimately diagnosed with metastatic brain disease during their lifetime, making cerebral metastasis the most common neurologic complication of systemic malignancy [1,4]. Up to 20–40% of patients with metastatic disease will have evidence of cerebral involvement at autopsy [1]. The most common sources of metastases to the brain include tumors from lung (40–60% of metastases), followed by breast, melanoma, and less often colon or kidney primary sites [1,5]. The site of the primary tumor may be unknown in up to 10% of cases [1,5]. The distribution of metastatic disease follows that of the cerebral volume and blood flow, with 80% occurring in the cerebral hemispheres, 15% in the cerebellum, and 5% in the brainstem [1,6]. Up to 75% of patients with cerebral involvement will harbor multiple lesions, based on modern imaging series using contrast-enhanced MRI [7–9]. Specific tumor histologies appear to be more likely to result in multiple metastatic lesions, including melanoma, colon, breast, and lung primaries. In contrast, renal cell carcinoma metastases are more likely to be single lesions. #


In recent years the incidence of brain metastasis appears to have been on the rise, attributable at least in part to an aging population, improved therapy for systemic malignancy resulting in longer survival of cancer patients, an increasing incidence of lung cancer and melanoma, and improvements in neuroimaging offering the ability to detect smaller lesions [4,10]. Due to this rising incidence, the management of patients with cerebral metastatic disease represents an increasingly significant clinical as well as economic challenge. Fortunately, along with improvements in our ability to detect cerebral metastases, innovations in neuroimaging and stereotaxy have expanded the therapeutic armamentarium available to the clinician to treat this growing patient population. This chapter will summarize the evidence in the medical literature regarding current treatment options available for patients with cerebral metastatic disease.

Diagnostic Work-up and Patient Selection for Therapy A contrast-enhanced MRI scan of the brain remains the most sensitive diagnostic tool for the detection and follow-up surveillance of cerebral metastatic disease [11]. On MRI, cerebral metastases typically appear as contrast-enhancing lesions located at the grey-white matter junction, with abundant peritumoral edema. In addition to disclosing the number of lesions present, T1-weighted contrast-enhanced MRI of the brain may identify leptomeningeal involvement, as

832

55

Image guided management of cerebral metastases

evidenced by an irregular brightly enhancing pial surface, with or without involvement of the arachnoid, dura, or ependymal surfaces. Furthermore, tumor histology may be suggested by MRI, with metastatic melanoma often appearing bright on non-contrast T1-weighted images due to the presence of blood and melanin within the lesions. Finally, associated peritumoral edema is demonstrated clearly on either T2-weighted or fluid-attenuated inversion recovery (FLAIR) sequences. From a practical standpoint however, many physicians typically perform a CT scan of the brain initially to confirm the diagnosis of suspected intracranial lesions including cerebral metastases. CT studies demonstrate single brain lesions in up to 50% of patients with cerebral metastases, while MRI examination reveal a single metastasis in up to 33%, with the remaining patients harboring multiple lesions [6,7]. Therefore, patients with single lesions identified by CT should have this finding confirmed by contrastenhanced MRI studies prior to initiating focal therapy targeting a single lesion, such as surgical resection or radiosurgery. For patients presenting with a cerebral mass suspected of being a metastasis, without a history of a known systemic primary lesion, it is important to identify the site of origin of the brain lesion. Lung cancer should be suspected as the most likely primary source, due to the high incidence with which it metastasizes to the brain. The diagnostic work-up should include chest X-ray or CTscan. In addition, an abdominopelvic CTscan may identify gastrointestinal or renal primary tumors. Mammography may be performed to rule out breast cancer as the primary lesion, although brain metastases from breast cancer prior to the primary disease declaring itself are rare. Finally, a positive radionuclide bone scan may suggest primary tumors that have a tendency to spread to bone. If an extracranial lesion is identified, it is typically biopsied first to confirm the diagnosis due to lower biopsy-associated risk in comparison to the brain lesion. If no extracranial site is identified, one

should consider either biopsy or open resection of the cerebral lesion for tissue diagnosis in order to direct further management. In patients with a cerebral lesion which appears typical for a cerebral metastasis, without a history of malignancy or identifiable systemic lesion on examination, the brain lesion will ultimately prove to be a metastasis in as little as 15% of cases [12,13]. Furthermore, some argue for the importance of tissue diagnosis for the cerebral lesion itself, as even brain lesions in patients with a known history of systemic malignancy may not be metastatic in up to 11% of cases [14]. Several factors play a significant role in the treatment decision-making process for patients with cerebral metastases. Among these are the number, size, and location of the cerebral lesions, as well as the presence of leptomeningeal disease [15–18]. Additional factors include the primary tumor histology, status of systemic disease activity, and presence or absence of extracranial metastases [19]. The patient’s neurologic status or Karnofsky Performance Status (KPS) and the disease-free interval prior to the diagnosis of cerebral metastasis are also prognostically important [5,20–22]. Among these factors, the most important appear to be the status of the primary cancer, and the patient’s neurologic status, with patients suffering from uncontrolled systemic malignancy and demonstrating significant neurologic dysfunction due to their intracranial involvement carrying a poor prognosis despite neurosurgical intervention [19]. Tumor histology factors significantly in the decision-making process for patients with cerebral metastases. Patients with brain metastases from renal cell carcinoma or malignant melanoma often exhibit poor survival, in comparison to patients with cerebral breast metastases [16,17]. Melanoma, renal cell carcinoma, and non-small cell lung cancer (NSCLC) are traditionally regarded as being chemoresistant, while melanoma, renal cell carcinoma, and sarcoma are considered resistant to standard fractionated


radiotherapy. Conversely, small cell lung cancer typically responds dramatically to radiotherapy. Similarly, tumors such as testicular tumors, choriocarcinoma, and secondary lymphoma typically respond to fractionated radiotherapy, focal radiation, or systemic chemotherapy. When the diagnosis of a ‘‘non-surgical’’ tumor type is made, a potentially unnecessary craniotomy may be avoided. In addition, studies have shown renal cell carcinoma and melanoma to be more susceptible to radiosurgery in comparison to fractionated radiotherapy [18]. In order to identify subgroups of patients with cerebral metastases whose overall prognosis warranted more aggressive therapy, the Radiation Therapy Oncology Group (RTOG) devised a recursive partitioning analysis (RPA) method of classifying patients with cerebral metastases into three subgroups according to their KPS score, age, and the status and extent of extracranial disease [23]. RPA class I consists of patients age 65 years or less, with a KPS of 70 or greater, with good control of their systemic disease and absence of any extracranial metastases. Patients in this subgroup have the best prognosis, and are considered optimal candidates for aggressive treatment of their disease. RPA class III includes all patients with a KPS of Figure 55-1).

Surgery Versus Radiosurgery for the Treatment of Cerebral Metastases The available evidence comparing microsurgical resection versus SRS for the treatment of brain metastases is currently limited to retrospective case series. Two such series suggest a clinical superiority for surgical resection in comparison with SRS. Bindal et al. reported their results for a matched case-control study, examining the efficacy of SRS versus open surgery [95]. The SRS group was comprised of 13 patients, who received a median dose of 20 Gy, while the surgical group consisted of 62 patients. Groups were matched based on their primary tumor type, extent of systemic disease, KPS, time to brain metastasis, number of brain metastases, age, and sex. This study identified a significant difference in median overall survival between the SRS group (7.5 months) and the surgical group (16.4 months), and this difference in survival was attributed to progression of the treated lesion in the SRS group. These authors concluded that indications for radiosurgery should be restricted to surgically inaccessible lesions or for patients unfit to undergo surgical resection. A criticism of this study is that the dosing regimen used in the SRS group resulted in a lower prescribed dose to the tumor margins than is considered standard. Therefore, underdosing may have hindered the efficacy of SRS at effecting tumor control in comparison with microsurgery. Also, a selection bias may have favored the surgical group, as only lesions amenable to surgery were included in this study. Another study suggesting improved outcome following surgery for cerebral metastasis in comparison with SRS was reported by Shinoura et al. [96]. This group compared recurrence rates of metastatic brain tumors following Linac SRS


55

. Figure 55-1 MRI of young woman with four metastatic tumors from rectal carcinoma, all growing 3 months following WBRT. The large left frontal tumor (a) was removed using awake image-guided craniotomy, sparing speech which was found overlying the tumor. General anesthesia was then immediately induced and the large cerebellar metastasis (b) was removed with the aid of surgical navigation based on a fresh registration. Five days later she underwent Gamma Knife radiosurgery for the right frontal tumor (c) and the right parietal tumor (d)

versus surgery plus WBRT. They found that the time to recurrence was 25 months in the surgical group but only 7.2 months in SRS group (p = 0.0199). Baseline patient characteristics describing extent and activity of extracranial disease was not provided, however. In addition, the SRS group had a greater number of patients with multiple metastases than the surgical group, which may have biased the study in favor of the surgical arm. Two retrospective studies support the conclusion that surgical resection and SRS offer equivalent clinical benefit for the treatment of cerebral metastases amenable to both therapies.

Auchter et al. reported the results from a multiinstitutional retrospective series looking at a group of patients with newly diagnosed brain metastasis treated with SRS plus WBRT, and also met the study inclusion criteria used by Patchell et al. previously [14,97]. In total, 122 patients from 4 institutions were identified. All patients except five received WBRT following SRS treatment. An overall local tumor control rate of 86% was observed in this group of patients. In addition, a median overall survival of 56 weeks and a median duration of functional independence (KPS > 70) of 44 weeks were observed. These authors concluded that SRS with WBRT

843

844

55


for the treatment of a single brain metastasis resulted in functional survival comparable with surgical resection followed by WBRT. Muacevic et al. retrospectively examined their experience comparing their results of surgery plus adjuvant WBRT with gamma knife SRS alone for the treatment of solitary brain metastases deemed suitable for radiosurgery [25]. All tumors were 3.5 cm in diameter or smaller. In total, 52 patients in the microsurgery plus WBRT group were compared with 56 patients in the GK SRS group. No statistically significant difference in median overall survival was identified. Local tumor control rates were 75 and 83% for the surgery versus radiosurgery groups, respectively (p = 0.49). There was also no difference observed in neurologic death rates at 1 year. The authors concluded that SRS alone results in local tumor control rates equal to surgery plus WBRT in selected patients, and therefore adjuvant WBRT need not be combined with SRS in this population. The remaining two studies provide data that supports a conclusion that SRS provides better local tumor control rates in comparison with microsurgery. Schoggl et al. reported results from their retrospective case-control study comparing SRS (67 patients) with microsurgical resection (66 patients) for a single cerebral metastasis [98]. Patients were treated between August 1992 and October 1996. All patients received adjuvant WBRT. Groups differed in their baseline characteristics, with the SRS group having on average smaller lesions (median size 7,800 ml) compared with surgically treated lesions (median size 12,500 ml). No significant difference in overall survival was identified. The SRS group, however, demonstrated a lower rate of local recurrence (5% vs. 17%), attributed to a better response rate of metastases traditionally considered ‘‘radioresistant’’ to WBRT. Based on their results, these authors advocate for the use of SRS as a first line therapy for the treatment of single cerebral metastases, unless the lesion is greater than 3 cm in diameter, or requires debulking due to symptomatic mass effect.

O’Neil et al. retrospectively reviewed their experience with newly diagnosed solitary brain metastases treated between 1991 and 1999 [99]. All patients had surgically accessible lesions, less than 3.5 cm diameter, and without evidence of obstructive hydrocephalus. A total of 74 patients made up the surgical arm, with 23 patients in the GK SRS arm. This study found no difference in overall survival between the two treatment arms, but did identify better local tumor control rates associated with GK SRS (p = 0.020). To date, no prospective RCT directly comparing surgery with radiosurgery has been completed. An international phase III trial (EORTC 22952) comparing surgery and radiosurgery with or without adjuvant WBRT for patients with 1–3 brain metastases has been accruing patients since 1996 [94]. Enrolment in this trial has been completed. The results from this study are eagerly anticipated, as they will shed important light on the roles each of the currently available therapeutic modalities has to play in the management of patients with cerebral metastases. Until the results of EORTC 22952 are available, the clinician must take into account several important considerations when choosing between surgical resection, SRS, or a combination of the two for the treatment of cerebral metastases. Among these considerations are the location, size and number of lesions, the presence of significant mass effect or edema, and the need for tissue diagnosis. In addition, one must consider the extent of extracranial disease and the patient’s performance status. Small, superficial tumors that are resectable and associated with minimal edema or mass effect are candidates for either modality of treatment. In this scenario, the decision regarding surgery versus SRS depends on factors such as the expertise of the treating physician, access to care (e.g., radiosurgical facilities), and patient preference. Deeply located, surgically inaccessible lesions may be better treated with SRS. Large lesions greater than approximately 3–3.5 cm may preclude SRS. Lesions causing significant symptomatology or posing an immediate threat to


life due to mass effect or exuberant edema may necessitate surgical resection. Patients with poor performance status and uncontrolled systemic disease are unlikely to benefit from aggressive management of their intracranial disease, even if they possess only a single cerebral lesion. In these cases, minimally invasive approaches may be more appropriate, whether this be with WBRT, fractionated stereotactic radiotherapy, or SRS. One must be careful to discern, however, whether the patient’s poor KPS is due to tumorrelated mass effect or edema, which may be improved with aggressive treatment of the cerebral disease.

55

Treatment Algorithm for Patients with Cerebral Metastatic Disease Choosing among the numerous treatment options available to the patient with cerebral metastatic disease requires that the treating clinician consider a multitude of patient and tumor-related variables in order to prescribe the most appropriate therapeutic plan. > Figure 55-2 provides a simplified outline of a treatment algorithm for this patient population. The most important prognostic factors appear to be the presence or absence of active systemic malignancy, and the patient’s neurologic and performance (KPS) status.

. Figure 55-2 A simplified treatment algorithm for the management of patients with cerebral metastatic disease. SRS, stereotactic radiosurgery; FSR, fractionated stereotactic radiotherapy; WBRT, whole-brain radiotherapy; Sx, surgical resection

845

846

55


Patients with aggressive systemic cancer or a poor KPS in general have few appropriate treatment options available to them. For this group, in patients with 1–3 cerebral lesions, supportive care plus WBRT, with or without SRS to all lesions is appropriate. For patients with four or more lesions, WBRT is typically offered, unless the primary histology is considered ‘‘radioresistant’’ (renal cell carcinoma, sarcoma, melanoma) in which circumstance one may elect to add SRS to the treatment plan. For patients with absent or controlled systemic disease and good performance status, one must then consider the number of cerebral lesions identified on contrast-enhanced MRI. For patients with a single lesion, surgical resection or SRS, followed by adjuvant WBRT is appropriate if the lesion size is under 3 cm in diameter. If the lesion is greater than 3 cm, or causing significant symptomatology or mass effect, surgical resection should be considered up front. For poor surgical candidates, fractionated stereotactic radiotherapy (FSR) may be used, followed by SRS if the lesion’s size reduces in response to FSR. For patients with 2–3 lesions, one may offer surgical resection of all lesions, SRS to all lesions, or a combination of surgery plus SRS, followed by adjuvant WBRT. Again, for lesions greater than 3 cm or highly symptomatic, one should consider surgical resection of this lesion up front. Patients with four or more lesions are typically offered WBRT and supportive care. In those with good systemic control and performance status one may consider offering combinations of surgical resection and SRS for progressive disease following WBRT.

Conclusions Cerebral metastasis represents in increasingly common complication in the cancer patient, owing in part to improved therapy for and survival associated with systemic cancers. Advances

in neuroimaging, microsurgical technique, stereotactic-based surgical adjuncts and noninvasive treatment modalities such as stereotactic radiosurgery have increased our ability to treat brain metastases. At the same time, the availability of numerous treatment options raises many questions regarding which strategy is most appropriate for the individual patient with metastatic brain disease. Ongoing and future studies will provide some insight into these management decisions.

References 1. Posner JB. Management of brain metastases. Rev Neurol (Paris) 1992;148(6–7):477-87. 2. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J Clin 1999;49(1):1,8‐31. 3. Chang SD, Lee E, Sakamoto GT, Brown NP, Adler JR Jr. Stereotactic radiosurgery in patients with multiple brain metastases. Neurosurg Focus 2000;9(2):e3. 4. Wen PY, Loeffler JS. Management of brain metastases. Oncology (Williston Park), 1999. 13(7):941-54, 957‐61; discussion 961‐2, 969. 5. Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM, Hanssens PE, Schmitz PI. Identification of prognostic factors in patients with brain metastases: a review of 1292 patients. Int J Radiat Oncol Biol Phys 1999; 43(4):795-803. 6. Delattre JY, Krol G, Thaler HT, Posner JB. Distribution of brain metastases. Arch Neurol 1988;45(7): 741-4. 7. Sze G, Milano E, Johnson C, Heier L. Detection of brain metastases: comparison of contrast-enhanced MR with unenhanced MR and enhanced CT. AJNR Am J Neuroradiol 1990;11(4):785-91. 8. Mintz AP, Cairncross JG. Treatment of a single brain metastasis: the role of radiation following surgical resection. JAMA 1998;280(17):1527-9. 9. Davis PC, Hudgins PA, Peterman SB, Hoffman JC Jr. Diagnosis of cerebral metastases: double-dose delayed CT vs contrast-enhanced MR imaging. AJNR Am J Neuroradiol 1991;12(2):293-300. 10. Patchell RA. The management of brain metastases. Cancer Treat Rev 2003;29(6):533-40. 11. Healy ME, Hesselink JR, Press GA, Middleton MS. Increased detection of intracranial metastases with intravenous Gd-DTPA. Radiology 1987;165(3): 619-24. 12. Lang FF, Sawaya R. Surgical management of cerebral metastases. Neurosurg Clin N Am 1996;7(3):459-84.


13. Lang FF, Sawaya R. Surgical treatment of metastatic brain tumors. Semin Surg Oncol 1998;14(1):53-63. 14. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, Kryscio RJ, Markesbery WR, Macdonald JS, Young B. A randomized trial of surgery in the treatment of single metastases to the brain. N Engl J Med 1990; 322(8):494-500. 15. Wronski M, Arbit E, McCormick B. Surgical treatment of 70 patients with brain metastases from breast carcinoma. Cancer 1997;80(9):1746-54. 16. Korinth MC, Delonge C, Hutter BO, Gilsbach JM. Prognostic factors for patients with microsurgically resected brain metastases. Onkologie 2002;25(5):420-5. 17. Pollock BE, Brown PD, Foote RL, Stafford SL, Schomberg PJ. Properly selected patients with multiple brain metastases may benefit from aggressive treatment of their intracranial disease. J Neurooncol 2003; 61(1):73-80. 18. Brown PD, Brown CA, Pollock BE, Gorman DA, Foote RL. Stereotactic radiosurgery for patients with ‘‘radioresistant’’ brain metastases. Neurosurgery 2002;51(3): 656-65; discussion 665‐7. 19. Vogelbaum MA, Suh JH. Resectable brain metastases. J Clin Oncol 2006;24(8):1289-94. 20. Pieper DR, Hess KR, Sawaya RE. Role of surgery in the treatment of brain metastases in patients with breast cancer. Ann Surg Oncol 1997;4(6):481-90. 21. Saitoh Y, Fujisawa T, Shiba M, Yoshida S, Sekine Y, Baba M, Iizasa T, Kubota M. Prognostic factors in surgical treatment of solitary brain metastasis after resection of non-small-cell lung cancer. Lung Cancer 1999;24(2): 99-106. 22. Salvati M, Capoccia G, Orlando ER, Fiorenza F, Gagliardi FM. Single brain metastases from breast cancer: remarks on clinical pattern and treatment. Tumori 1992;78(2):115-7. 23. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, McKenna WG, Byhardt R. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997;37(4): 745-51. 24. Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999;45:41-7. 25. Muacevic A, Kreth FW, Horstmann GA, SchmidElsaesser R, Wowra B, Steiger HJ, Reulen HJ. Surgery and radiotherapy compared with gamma knife radiosurgery in the treatment of solitary cerebral metastases of small diameter. J Neurosurg 1999;91(1):35-43. 26. Black PM, Johnson MD. Surgical resection for patients with solid brain metastases: current status. J Neurooncol 2004;69(1–3):119-24. 27. Tan TC, Mc LBP. Image-guided craniotomy for cerebral metastases: techniques and outcomes. Neurosurgery 2003;53(1):82-9; discussion 89‐90.

55

28. Vecht CJ, Haaxma-Reiche H, Noordijk EM, Padberg GW, Voormolen JH, Hoekstra FH, Tans JT, Lambooij N, Metsaars JA, Wattendorff AR, et al. Treatment of single brain metastasis: radiotherapy alone or combined with neurosurgery? Ann Neurol 1993;33(6):583-90. 29. Mintz AH, Kestle J, Rathbone MP, Gaspar L, Hugenholtz H, Fisher B, Duncan G, Skingley P, Foster G, Levine M. A randomized trial to assess the efficacy of surgery in addition to radiotherapy in patients with a single cerebral metastasis. Cancer 1996;78(7): 1470-6. 30. Smalley SR, Schray MF, Laws Jr, ER O’Fallon JR. Adjuvant radiation therapy after surgical resection of solitary brain metastasis: association with pattern of failure and survival. Int J Radiat Oncol Biol Phys 1987;13(11): 1611-16. 31. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, Kryscio RJ, Markesbery WR, Foon KA, Young B. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998;280(17):1485-9. 32. Simpson JR, Mendenhall WM, Schupak KD, Larson D, Bloomer WD, Buckley JA, Gaspar LE, Gibbs FA, Lewin AA, Loeffler JS, Malcolm AW, Schneider JF, Shaw EG, Wharam MD Jr, Gutin PH, Rogers L, Leibel S. Follow-up and retreatment of brain metastasis. ACR appropriateness criteria. Radiology 2000;215 Suppl:1129-35. 33. Ransohoff J. Surgical management of metastatic tumors. Semin Oncol 1975;2(1):21-7. 34. Bindal RK, Sawaya R, Leavens ME, Lee JJ. Surgical treatment of multiple brain metastases. J Neurosurg 1993;79(2):210-16. 35. Iwadate Y, Namba H, Yamaura A. Significance of surgical resection for the treatment of multiple brain metastases. Anticancer Res 2000;20(1B):573-7. 36. Nieder C, Andratschke N, Grosu AL, Molls M. Recursive partitioning analysis (RPA) class does not predict survival in patients with four or more brain metastases. Strahlenther Onkol 2003;179(1):16-20. 37. Bindal RK, Sawaya R, Leavens ME, Hess KR, Taylor SH. Reoperation for recurrent metastatic brain tumors. J Neurosurg 1995;83(4):600-4. 38. Arbit E, Wronski M, Burt M, Galicich JH. The treatment of patients with recurrent brain metastases. A retrospective analysis of 109 patients with nonsmall cell lung cancer. Cancer 1995;76(5):765-73. 39. Linskey ME. The changing role of stereotaxis in surgical neuro-oncology. J Neurooncol 2004;69(1–3):35-54. 40. Golfinos JG, Fitzpatrick BC, Smith LR, Spetzler RF. Clinical use of a frameless stereotactic arm: results of 325 cases. J Neurosurg 1995;83(2):197-205. 41. Wadley J, Dorward N, Kitchen N, Thomas D. Preoperative planning and intra-operative guidance in modern neurosurgery: a review of 300 cases. Ann R Coll Surg Engl 1999;81(4):217-25.

847

848

55


42. Maciunas RJ. Intraoperative cranial navigation. Clin Neurosurg 1996;43:353-81. 43. Bernstein M. Outpatient craniotomy for brain tumor: a pilot feasibility study in 46 patients. Can J Neurol Sci 2001;28(2):120-4. 44. Nimsky C, Ganslandt O, Fahlbusch R. Implementation of fiber tract navigation. Neurosurgery 2006;58 4 Suppl 2:ONS-292-303; discussion ONS-303‐4. 45. Ganslandt O, Fahlbusch R, Nimsky C, Kober H, Moller M, Steinmeier R, Romstock J, Vieth J. Functional neuronavigation with magnetoencephalography: outcome in 50 patients with lesions around the motor cortex. J Neurosurg 1999;91(1):73-9. 46. Wu JS, Zhou LF, Hong XN, Mao Y, Du GH. Role of diffusion tensor imaging in neuronavigation surgery of brain tumors involving pyramidal tracts. Zhonghua Wai Ke Za Zhi 2003;41(9):662-6. 47. Clark CA, Barrick TR, Murphy MM, Bell BA. White matter fiber tracking in patients with space-occupying lesions of the brain: a new technique for neurosurgical planning? Neuroimage 2003;20(3):1601-8. 48. Holodny AI, Schwartz TH, Ollenschleger M, Liu WC, Schulder M. Tumor involvement of the corticospinal tract: diffusion magnetic resonance tractography with intraoperative correlation. J Neurosurg 2001;95(6): 1082. 49. Hlatky R, Jackson EF, Weinberg JS, McCutcheon IE. Intraoperative neuronavigation using diffusion tensor MR tractography for the resection of a deep tumor adjacent to the corticospinal tract. Stereotact Funct Neurosurg 2005;83(5–6):228-32. 50. Smits M, Vernooij MW, Wielopolski PA, Vincent AJ, Houston GC, van der Lugt A. Incorporating functional MR imaging into diffusion tensor tractography in the preoperative assessment of the corticospinal tract in patients with brain tumors. AJNR Am J Neuroradiol 2007;28(7):1354-61. 51. Lang FF, Olansen NE, DeMonte F, Gokaslan ZL, Holland EC, Kalhorn C, Sawaya R. Surgical resection of intrinsic insular tumors: complication avoidance. J Neurosurg 2001;95(4):638-50. 52. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R. Intraoperative compensation for brain shift. Surg Neurol 2001;56(6):357-64; discussion 364‐5. 53. Nauta HJ. Error assessment during ‘‘image guided’’ and ‘‘imaging interactive’’ stereotactic surgery. Comput Med Imaging Graph 1994;18(4):279-87. 54. Lindseth F, Kaspersen JH, Ommedal S, Lango T, Bang J, Hokland J, Unsgaard G, Hernes TA. Multimodal image fusion in ultrasound-based neuronavigation: improving overview and interpretation by integrating preoperative MRI with intraoperative 3D ultrasound. Comput Aided Surg 2003;8(2):49-69. 55. Bernstein M, Al-Anazi AR, Kucharczyk W, Manninen P, Bronskill M, Henkelman M. Brain tumor surgery with the Toronto open magnetic resonance imaging system: preliminary results for 36 patients and analysis of

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

advantages, disadvantages, and future prospects. Neurosurgery 2000;46(4):900-7; discussion 907‐9. Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ, Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA. Development and implementation of intraoperative magnetic resonance imaging and its neurosurgical applications. Neurosurgery 1997;41(4):831-42; discussion 842‐5. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M, Kaus M, Heigl T, Lenz G, Kuth R, Huk W. Intraoperative magnetic resonance imaging with the magnetom open scanner: concepts, neurosurgical indications, and procedures: a preliminary report. Neurosurgery 1998;43(4):739-47; discussion 747‐8. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, Bonsanto MM, Albert FK, Kuth R, Staubert A, Schlegel W, Sartor K, Kunze S. Intraoperative diagnostic and interventional magnetic resonance imaging in neurosurgery. Neurosurgery 1997;40(5):891-900; discussion 900‐2. Pamir MN, Peker S, Ozek MM, Dincer A. Intraoperative MR imaging: preliminary results with 3 tesla MR system. Acta Neurochir Suppl 2006;98:97-100. Yrjana SK, Katisko JP, Ojala RO, Tervonen O, Schiffbauer H, Koivukangas J. Versatile intraoperative MRI in neurosurgery and radiology. Acta Neurochir (Wien) 2002;144(3):271-8; discussion 278. Unsgaard G, Rygh OM, Selbekk T, Muller TB, Kolstad F, Lindseth F, Hernes TA. Intra-operative 3D ultrasound in neurosurgery. Acta Neurochir (Wien) 2006;148(3): 235-53; discussion 253. Unsgaard G, Gronningsaeter A, Ommedal S, Nagelhus Hernes TA. Brain operations guided by real-time twodimensional ultrasound: new possibilities as a result of improved image quality. Neurosurgery 2002;51(2): 402-11; discussion 411‐2. Unsgaard G, Ommedal S, Muller T, Gronningsaeter A, Nagelhus Hernes TA. Neuronavigation by intraoperative three-dimensional ultrasound: initial experience during brain tumor resection. Neurosurgery 2002;50(4): 804-12; discussion 812. Roberts DW, Miga MI, Hartov A, Eisner S, Lemery JM, Kennedy FE, Paulsen KD. Intraoperatively updated neuroimaging using brain modeling and sparse data. Neurosurgery 1999;45(5):1199-206; discussion 1206‐7. Lindner D, Trantakis C, Renner C, Arnold S, Schmitgen A, Schneider J, Meixensberger J. Application of intraoperative 3D ultrasound during navigated tumor resection. Minim Invasive Neurosurg 2006;49(4):197-202. Pallatroni H, Hartov A, McInerney J, Platenik LA, Miga MI, Kennedy FE, Paulsen KD, Roberts DW. Coregistered ultrasound as a neurosurgical guide. Stereotact Funct Neurosurg 1999;73(1–4):143-7. Rasmussen IA Jr, Lindseth F, Rygh OM, Berntsen EM, Selbekk T, Xu J, Nagelhus Hernes TA, Harg E, Haberg A,


68.

69.

70.

71.

72.

73. 74.

75.

76.

77.

78.

Unsgaard G. Functional neuronavigation combined with intra-operative 3D ultrasound: initial experiences during surgical resections close to eloquent brain areas and future directions in automatic brain shift compensation of preoperative data. Acta Neurochir (Wien) 2007;149(4):365-78. Nagelhus Hernes TA, Lindseth F, Selbekk T, Wollf A, Solberg OV, Harg E, Rygh OM, Tangen GA, Rasmussen I, Augdal S, Couweleers F, Unsgaard G. Computer-assisted 3D ultrasound-guided neurosurgery: technological contributions, including multimodal registration and advanced display, demonstrating future perspectives. Int J Med Robot 2006;2(1):45-59. Bonsanto MM, Staubert A, Wirtz CR, Tronnier V, Kunze S. Initial experience with an ultrasound-integrated single-RACK neuronavigation system. Acta Neurochir (Wien) 2001;143(11):1127-32. Gronningsaeter A, Kleven A, Ommedal S, Aarseth TE, Lie T, Lindseth F, Lango T, Unsgard G. SonoWand, an ultrasound-based neuronavigation system. Neurosurgery 2000;47(6):1373-9; discussion 1379‐80. Tronnier VM, Bonsanto MM, Staubert A, Knauth M, Kunze S, Wirtz CR. Comparison of intraoperative MR imaging and 3D-navigated ultrasonography in the detection and resection control of lesions. Neurosurg Focus 2001;10(2):E3. Lindseth F, Lango T, Bang J, Nagelhus Hernes TA. Accuracy evaluation of a 3D ultrasound-based neuronavigation system. Comput Aided Surg 2002;7(4):197-222. Bradley KA, Mehta MP. Management of brain metastases. Semin Oncol 2004;31(5):693-701. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami L, Rotman M, Mehta MP, Curran WJ Jr. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363(9422): 1665-72. Flickinger JC, Kondziolka D, Lunsford LD, Coffey RJ, Goodman ML, Shaw EG, Hudgins WR, Weiner R, Harsh GRT, Sneed PK, et al. A multi-institutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994;28 (4):797-802. Alexander E, 3rd, Moriarty TM, Davis RB, Wen PY, Fine HA, Black PM, Kooy HM, Loeffler JS. Stereotactic radiosurgery for the definitive, noninvasive treatment of brain metastases. J Natl Cancer Inst 1995;87(1):34-40. Joseph J, Adler JR, Cox RS, Hancock SL. Linear accelerator-based stereotaxic radiosurgery for brain metastases:the influence of number of lesions on survival. J Clin Oncol 1996;14(4):1085-92. Chen JC, Petrovich Z, O’Day S, Morton D, Essner R, Giannotta SL, Yu C, Apuzzo ML. Stereotactic radiosurgery in the treatment of metastatic disease to the brain. Neurosurgery 2000;47(2):268-79; discussion 279‐81.

55

79. Hoffman R, Sneed PK, McDermott MW, Chang S, Lamborn KR, Park E, Wara WM, Larson DA. Radiosurgery for brain metastases from primary lung carcinoma. Cancer J 2001;7(2):121-31. 80. Gerosa M, Nicolato A, Foroni R, Zanotti B, Tomazzoli L, Miscusi M, Alessandrini F, Bricolo A. Gamma knife radiosurgery for brain metastases: a primary therapeutic option. J Neurosurg 2002;97 Suppl 5:515-24. 81. Petrovich Z, Yu C, Giannotta SL, O’Day S, Apuzzo ML. Survival and pattern of failure in brain metastasis treated with stereotactic gamma knife radiosurgery. J Neurosurg 2002;97 Suppl 5:499-506. 82. Gerosa M, Nicolato A, Severi F, Ferraresi P, Masotto B, Barone G, Foroni R, Piovan E, Pasoli A, Bricolo A. Gamma Knife radiosurgery for intracranial metastases: from local tumor control to increased survival. Stereotact Funct Neurosurg 1996;66 Suppl 1:184-92. 83. Pirzkall A, Debus J, Lohr F, Fuss M, Rhein B, EngenhartCabillic R, Wannenmacher M. Radiosurgery alone or in combination with whole-brain radiotherapy for brain metastases. J Clin Oncol 1998;16(11):3563-9. 84. Lutterbach J, Cyron D, Henne K, Ostertag CB. Radiosurgery followed by planned observation in patients with one to three brain metastases. Neurosurgery 2003; 52(5):1066-73; discussion 1073‐4. 85. Hasegawa T, Kondziolka D, Flickinger JC, Germanwala A, Lunsford LD. Brain metastases treated with radiosurgery alone: an alternative to whole brain radiotherapy? Neurosurgery 2003;52(6):1318-26; discussion 1326. 86. Muacevic A, Kreth FW, Tonn JC, Wowra B. Stereotactic radiosurgery for multiple brain metastases from breast carcinoma. Cancer 2004;100(8):1705-11. 87. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger JC. Stereotactic radiosurgery plus whole brain radiotherapy versus radiotherapy alone for patients with multiple brain metastases. Int J Radiat Oncol Biol Phys 1999;45(2):427-34. 88. Sneed PK, Lamborn KR, Forstner JM, McDermott MW, Chang S, Park E, Gutin PH, Phillips TL, Wara WM, Larson DA. Radiosurgery for brain metastases: is whole brain radiotherapy necessary? Int J Radiat Oncol Biol Phys 1999;43(3):549-58. 89. Sneed PK, Suh JH, Goetsch SJ, Sanghavi SN, Chappell R, Buatti JM, Regine WF, Weltman E, King VJ, Breneman JC, Sperduto PW, Mehta MP. A multiinstitutional review of radiosurgery alone vs. radiosurgery with whole brain radiotherapy as the initial management of brain metastases. Int J Radiat Oncol Biol Phys 2002; 53(3): 519-26. 90. Chidel MA, Suh JH, Reddy CA, Chao ST, Lundbeck MF, Barnett GH. Application of recursive partitioning analysis and evaluation of the use of whole brain radiation among patients treated with stereotactic radiosurgery for newly diagnosed brain metastases. Int J Radiat Oncol Biol Phys 2000;47(4):993-9.

849

850

55


91. Sheehan JP, Sun MH, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery in patients with renal cell carcinoma metastasis to the brain: long-term outcomes and prognostic factors influencing survival and local tumor control. J Neurosurg 2003;98(2):342-9. 92. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K, Kenjyo M, Oya N, Hirota S, Shioura H, Kunieda E, Inomata T, Hayakawa K, Katoh N, Kobashi G. Stereotactic radiosurgery plus whole-brain radiation therapy vs stereotactic radiosurgery alone for treatment of brain metastases: a randomized controlled trial. JAMA 2006;295(21):2483-91. 93. Regine WF, Huhn JL, Patchell RA, St Clair WH, Strottmann J, Meigooni A, Sanders M, Young AB. Risk of symptomatic brain tumor recurrence and neurologic deficit after radiosurgery alone in patients with newly diagnosed brain metastases: results and implications. Int J Radiat Oncol Biol Phys 2002;52(2):333-8. 94. Mueller RP. European Organisation for Research and Treatment of Cancer: EORTC protocol 22952: No radiotherapy versus whole brain radiotherapy for 1 to 3 brain metastases from solid tumor after surgical resection or radiosurgery – A randomized phase III trial. http://www.eortc.be/protoc/details.asp?protocol=22952.

95. Bindal AK, Bindal RK, Hess KR, Shiu A, Hassenbusch SJ, Shi WM, Sawaya R. Surgery versus radiosurgery in the treatment of brain metastasis. J Neurosurg 1996;84(5):748-54. 96. Shinoura N, Yamada R, Okamoto K, Nakamura O, Shitara N. Local recurrence of metastatic brain tumor after stereotactic radiosurgery or surgery plus radiation. J Neurooncol 2002;60(1):71-7. 97. Auchter RM, Lamond JP, Alexander E, Buatti JM, Chappell R, Friedman WA, Kinsella TJ, Levin AB, Noyes WR, Schultz CJ, Loeffler JS, Mehta MP. A multiinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastasis. Int J Radiat Oncol Biol Phys 1996;35(1):27-35. 98. Schoggl A, Kitz K, Reddy M, Wolfsberger S, Schneider B, Dieckmann K, Ungersbock K. Defining the role of stereotactic radiosurgery versus microsurgery in the treatment of single brain metastases. Acta Neurochir (Wien) 2000;142(6):621-6. 99. O’Neill BP, Iturria NJ, Link MJ, Pollock BE, Ballman KV, O’Fallon JR. A comparison of surgical resection and stereotactic radiosurgery in the treatment of solitary brain metastases. Int J Radiat Oncol Biol Phys 2003;55(5): 1169-76.

52 Image Guided Management of Intracerebral Hematoma A. Losiniecki . G. Mandybur

Introduction Spontaneous intracerebral hemorrhage (ICH) remains a major cause of hemorrhagic strokes. Mortality from ICH far outpaces that from subarachnoid hemorrhage or cerebral infarction. Incidence of spontaneous ICH ranges from 13 to 35 per population of 100,000 [1–4]. As the prevalence of mortality and severity of morbidity associated with spontaneous ICH is much higher in comparison with other types of stroke, identifying potential surgical candidates is of utmost importance. To address patient selection, randomized studies comparing best medical management to open surgical evacuation might reduce the risk of death without improving functional outcome. With open surgical techniques providing less efficacious outcomes than best medical management, stereotactic evacuation of ICH may provide a less-is-more scenario. Widespread use of computed tomography (CT) scanning has made the identification of spontaneous ICH quick and simple and allows for identification of location and quantification of hemorrhage volumes. In the CT era, early diagnosis also makes earlier treatment possible. Management of spontaneous ICH to this point has had two main pathways – medical and surgical. Surgical intervention has been undertaken only when patients were no longer able to be managed medically as evidenced by either progression of symptoms or increased intracranial pressure (ICP). During these circumstances, patient outcomes after surgery are usually poor, characterized by significant mortalities and morbidities [5,6]. However, stereotactic #


evacuation of spontaneous ICH provides a quick, minimally invasive approach to this difficult problem. Although some practitioners believe that less collateral injury during stereotactic aspirations may improve patient outcomes, guidelines for management of spontaneous ICH continue to evolve and Class I data are minimal [7]. A landmark study by McKissock et al. in 1961 is often cited to argue against the routine use of surgical evacuation of intracranial hemorrhages [8]. Yet the study was completed in an era before CT scans and without the advantages of many contemporary microsurgical techniques and practices of postoperative intensive care, all of which can potentially improve surgical outcomes.

Etiology of ICH Hypertension is an important risk factor for spontaneous ICH that is found in 40–60% of affected patients. Other identified causes include aneurysms, vascular malformations, coagulopathies, tumors, conversion of ischemic to hemorrhagic infarctions, amyloid angiopathy, post-traumatic reactions, and reactions to drugs, both legal and illegal. Given this long list, identification of patient-specific causes is important because treatment can vary depending on the cause. CT, which provides the diagnosis in the vast majority of ICH patients, is the acute study of choice. It provides quick identification of the location, determination of the amount of midline shift, detection of hydrocephalus, and volume estimation of the clot (> Figure 52-1). Once the patient is stabilized, additional studies

798

52

Image guided management of intracerebral hematoma

. Figure 52-1 CT is the initial study of choice in the identification of ICH and its location, extent of midline shift, hydrocephalus and volume. a, Right-sided lobar hemorrhage with extension to the cortex and accompanying midline shift b, Midline cerebellar ICH. c, Left-sided hemorrhage of the basal ganglia with extension into the ventricle and associated with hydrocephalus (with permission from Mayfield Clinic)

can be attempted to identify the root cause of the hemorrhage; such studies that may be helpful include magnetic resonance imaging (MRI) and cerebral angiography. MRI is effective in characterizing the age of the hemorrhage and, with addition of contrast, may help to identify underlying tumor as the source. Cerebral angiography can be preformed if a vascular lesion is highly suspected based on hemorrhage characteristics and location. Cardiac ultrasound can be performed to evaluate for possible valvupopathies if an embolic or infectious cause is suspected. Without completion of these studies during the evaluation of spontaneous ICH, eliminating a structural, vascular, or distant lesion as a cause can be difficult. One major issue with ICH is that, after hemorrhage, much brain damage may already have occurred. Evacuation of the hematoma may only prevent further brain injury, either by physical compression or cytotoxic by-products. Thus initial presentation and progression of neurologic symptoms are paramount in the evaluation of a patient for any kind of surgical evacuation.

Medical Management Any discussion of surgical management of spontaneous ICH must also include some basics of medical management. Initial management should be directed at the basics of airway, breathing, and circulation. The acute management of patients with spontaneous ICH involves admittance to an intensive care unit (ICU) setting, followed by frequent neurologic and medical monitoring. Neurologic exams (i.e., Glasgow Coma Score, type quick assessments) are performed often soon after stabilization. Intubation is not routinely used in all patients with spontaneous ICH but is used in those who show signs of insufficient ventilation (i.e., pO2 > 60 mm Hg or pCO2 > 50 mm Hg) or those who are obtunded and unable to protect their airway. In the setting of suspected elevated ICP, hyperventilation is appropriate yet is only a temporary measure; its use should not be prolonged. Prevention of increase in size of spontaneous ICH and maintenance of the surrounding unaffected brain is the goal of most if not all of the available medical interventions. ICH expansion


52

. Table 52-1 Medical management goals and study benefits related to spontaneous intracerebral hemorrhage (with permission from the Mayfield Clinic) Intervention

Goal range

Potential benefit

Study results

Control of blood pressure

Undefined, suggestions toward normotensive N/A

Decrease hemorrhage expansion, maintain cerebral perfusion

SBP 140 suggest improvement, yet recent studies show no correlation [10,11]

Reduction of cerebral edema

None, suggests worse outcome [12]

CPP >70 mm Hg, ICP 70 [13,14]

N/A

Prevention of status epilepticus and associated increase in cerebral metabolism Normoglycemia provides for ideal cerebral metabolism

High incidence of seizures with lobar hemorrhages, seizures should always be treated [15,16] Elevated blood glucose is associated with increased mortality [17]

Administration of corticosteroids Monitoring of intracranial pressure Use of antiseizure medications Control of hyperglycemia

Unknown

after initial presentation has been shown to be as high as 28% within the first 24 h [9]. Although the optimal blood pressure in management of spontaneous ICH has not yet been determined, studies do exist that suggest mean arterial pressure (MAP) and cerebral perfusion pressure (CPP) should be kept within a nominal range [10,11]. Spontaneous ICH often occurs outside the setting of large-vessel vasculopathy. Therefore, risk of hemorrhagic expansion with mild blood pressure elevation may actually be less than anticipated. A clear target blood pressure in the setting of spontaneous ICH does not exist – rather physicians must balance the risk of hemorrhage expansion with maintenance of CPP. Many controversies exist regarding the management of blood pressure when spontaneous ICH presents and as such no definitive guidelines are available. Underlying coagulopathies should be addressed with correction of prothrombin time (PT) and partial thromboplastin time (PTT) and use of appropriate blood products as necessary. The routine use of steroids has not been shown to improve outcome in spontaneous ICH in comparison to placebo [12] and thus is not included as a first-line

management. Some of the medical management parameters discussed in treating spontaneous ICH are shown in > Table 52-1. The use of intraventricular catheters (IVCs) is preferred because of their effectiveness to monitor ICP and provide cerebrospinal fluid (CSF) drainage. Treatment regimens used for normalization of ICP include elevation of the head of bed, CSF drainage, pain medication and sedation, and osmotic therapy (3% saline and mannitol). When most of these treatments fail, barbiturate-induced coma can be used. These strategies can be used alone or in combination, and may provide sufficient effects to avoid surgical decompression. Much of the data that exist about the treatment of elevated cerebral hypertension come from the traumatic brain injury (TBI) literature. In TBI studies that specifically assess ICP control for spontaneous ICH patients [13,14], normalization of ICP and maintenance of CPP are indicated. Seizure activity should be monitored closely because EEG criteria show seizure activity in nearly 30% of patients; however, only 5% of these patients exhibit seizures clinically [15,16]. Any evidence of seizure activity should be aggressively

799

800

52


treated: Prophylactic administration of antiepileptic medications may reduce the risk of developing seizures, especially in patients with lobar hemorrhages [15]. In addition, blood glucose should be tightly controlled; use of an insulin drip in refractory cases is appropriate. Although the exact parameters for blood glucose do not exist, high-admission levels of blood glucose correlate with poor outcomes [17]. The medical management paradigm for spontaneous ICH may change over the next few years. Preliminary studies have examined with varying degrees of success the utility of hypothermia, recombinant activated factor VII, and hyperosmolar therapy. However, further studies are required before their universal acceptance.

Surgical Evacuation of Intracerebral Hemorrhage Questions regarding timing and which patients are most likely to benefit from surgery have yet to be scientifically answered. Some of the following randomized studies still leave questions, thus providing motivation to further evaluate new minimally invasive techniques.

Techniques for Evacuation Surgical techniques for evacuation of intracranial hemorrhage are numerous, ranging from small, minimally invasive procedures to large, decompressive evacuation techniques. The location and size of hemorrhage and medical comorbidities dictate which and when surgical procedures should be performed. A cerebellar location is a site where reasonable evidence exists for the early evacuation of hemorrhages >3 cm. Studies have shown that surgical evacuation can be superior to medical management alone, especially in patients with hemorrhage to the cortical surface

and a GCS >5 [7]. The goals of surgery must be defined before the selection of the ultimate surgical techniques. Minimally invasive/stereotactic techniques for evacuation of spontaneous ICH carry some theoretical benefits. These include reduced time under anesthesia, decreased anesthetic load with less violation of uninjured brain tissue, and the ability to evacuate deep-seated lesions (e.g., in the thalamus or pons) [18–20]. This chapter culminates with a review of the pros and cons of both craniotomy and stereotactic aspiration techniques for the treatment of spontaneous ICH.

Craniotomy The traditional craniotomy approach for evacuation of ICH involves the creation of bony window that must be sufficiently sized over the hemorrhagic site to allow for its evacuation and enough space for visualization for hemostasis. At the University of Cincinnati, a keyhole craniotomy is performed in the region of noneloquent cortex to access hematomas that lie close to the cortex (> Figure 52-2). After opening the dura mater, a small cortical tunnel is typically made to enter the clot by using bipolar forceps and suction. Once sufficiently evacuated, the ICH cavity is explored. Using bipolar forceps, the surgeon obtains hemostasis with or without use of additional synthetic hemostatic agents. Intraventricular catheters are only inserted in the setting of symptomatic hydrocephalus or large intraventricular clot burden. Some surgeons have advocated creation of large craniotomies (e.g., as used for TBIs) to relieve ICP and allow surgical hematoma evacuation. However, the true benefits for such a procedure still require study. A multicenter study, the International Surgical Trial in Intracerebral Hemorrhage, which randomized more than 1,000 patients to undergo best medical management or surgical evacuation,


52

. Figure 52-2 Surgical technique for removal of an intracerebral hemorrhage (ICH). (a) Operative set-up for frameless imageguided stereotactic trajectory to a right-sided frontal lesion. (b) Top figure, dura is opened and a cortical entry point is selected. (c) Bottom figure, blunt dissection is performed with bipolar until the clot is entered and evacuated with suction (with permission from Mayfield Clinic)

showed that the patients most likely to benefit from surgical evacuation had a moderate GCS and a clot within 1 cm of the cortical surface [5]. However, this study also suggested that some patients may actually worsen after surgery. Although the above-mentioned group was suggested to benefit from surgical evacuation, no significant differences were obtained between the two major groups. Therefore further trials are required.

Timing of Surgery Early diagnosis of an ICH that can potentially undergo surgical treatment is soon followed by a question of optimal timing. Surgery performed within 12 h has been shown to be only modestly

effective [6]. At 1-month evaluation in this study, Morgenstern et al. reported a nearly 15% mortality and poor initial functional outcomes. As only one patient had undergone surgery within 4 h of symptom onset, these findings beg the question – Is this early enough? [6]. In a retrospective review of 100 patients who underwent surgery within 7 h of symptom onset, Kaneko et al. reported good functional outcome in 35% of patients and mortality only slightly higher than 5%; these data suggest that early surgical treatment results in improved outcomes [21]. Studies have shown that enlargement of ICH occurs most dramatically within the first few hours after initial symptom onset [22,23]. Angiographically, contrast extravasation into the ICH cavity appears to slow after 6 h from onset [24]. How this relates to ideal surgical timing has not been convincingly shown.

801

802

52


Minimally Invasive/Stereotactic Evacuation of ICH The idea of using a minimally invasive technique to aide in evacuation of ICH has been studied since the late 1970s. One of the numerous techniques evaluated involved the insertion of a cannula and an Archimedes screw aspirator to aid in the break-up of blood clot and evacuation of ICH. Initial results were disappointing with a mortality rate exceeding 80% [20]. Other advances included insertion of a modified nucleotome [25], a double-track aspiration system [26], and ultrasonic aspirator [27], a waterjet irrigation system [28], and others. Although the results were not particularly encouraging, the future provides hope for improvements in imaging capabilities and surgical techniques. On the basis of the STICH trial findings, the authors concluded that all non-craniotomy techniques had worse outcome when compared with conservative management. Yet the data were not overwhelming. Many patients who were selected to undergo minimally invasive evacuations had deep-seated lesions seemingly destined to do poorly [5]. This highlights one of many questions regarding the optimal patient selection that has yet to be completely understood. Other factors that affect patient selection involve lesion size, location, and ideal surgical timing. Minimally invasive techniques offer significant theoretical advantages over traditional craniotomy techniques: use of local anesthesia, shorter operating times, and less damage to normal surrounding brain. One key to success of minimally invasive techniques lies in the accuracy of localization. Techniques for localization first included ultrasound, then CT scans, and finally MRI, all of which evolved with potential intraoperative application that could improve the precision of localization. After accurate localization of the ICH, the next step involves its evacuation, for which various techniques are available or remain under study. The ideal technique can be performed

quickly, effect minimal trauma to surrounding tissues, allow complete hematoma removal, and provide means for improved hemostasis. Endoscopic techniques described for ICH evacuation consist of insertion of an endoscope through a burr hole directly into the ICH cavity and evacuation of contents under direct visualization (> Figure 52-3). The trajectory and planning of the burr-hole placement can be simplified with stereotactic guidance. Placement of the entry point and planned trajectory is dependent on ICH location, with all attempts made to avoid eloquent tissue and known vascular structures. Endoscopic techniques using lavage followed by suction of the ICH have been reported to evacuate nearly 90% of clot burden [18]. In a randomized study of 100 patients who either underwent endoscopic techniques or were managed medically, the endoscopic group had prolonged survival, especially those with large ICHS of >50 cc. Patients younger than 60 years with lobar hemorrhages showed a significant benefit in quality of life assessments after endoscopic management versus the best medical management [18]. Another purported advantage of endoscopic evacuation is speed of surgery. In fact, some evacuations have been completed in less than 60 min [19]. The concept of stereotactic aspiration of ICH is similar to that of endoscopic evacuation. Instead of direct endoscopic visualization, an aspirating needle is placed into the middle of the hemorrhage cavity. After the evacuation of the ICH clot, a drain placed directly into the cavity then encourages further drainage. If a CT scan of the patient’s head still shows residual clot, injection of a thrombolytic agent can provide additional evacuation [29]. Although injection of a thrombolytic agent into an ICH would seem to be counterintuitive to maintenance of hemostasis, studies have shown that rebleeding rates are in the range of those encountered in open craniotomy [30]. Urokinase, which is no longer available in the U.S., was used in initial studies that involved injection of thrombolytic


52

. Figure 52-3 Endoscopic technique for removal of an intracerebral hemorrhage (ICH). A large burr hole is made and the dura is opened. In one pass, a 14 French peel-away introducer is placed into the central core of the hematoma to a depth at least two thirds of the overall hematoma diameter. The inner stylet is carefully removed while the cannula remains within the clot. Hematoma is aspirated using a syringe until there is no longer a fluid component of the clot. The endoscope is inserted and the hematoma cavity is inspected. Additional techniques, such as lavage and piece-meal removal with a nucleotome (inset) or forceps, may be used to remove the solid components of the clot (with permission from Mayfield Clinic)

agents. More recent studies have shown tPA to be a viable alternative [31,32]. Currently the MISTIE study funded by the National Institute of Health is underway to better answer this question: This study hypothesizes that tPA injection is safe, reduces clot size, and improves clinical outcomes when compared with best medical management. > Figure 52-4 shows one case example by imaging studies in which tPA was administered in conjunction with clot evacuation.

There are differences and similarities between complications of the minimally invasive techniques and those of traditional craniotomy evacuation. Rates of rebleeding are about 3% after both procedures [19,33–35]. As the use of indwelling catheter would seem to increase the risk of infection, the prophylactic use of antibiotics needs to be researched. Use of tPA introduces a rebleeding risk that is not inherently present for traditional craniotomy. In studies in

803

804

52


. Figure 52-4 CT imaging studies showing ICH before and after evacuation with instillation of tissue plasmogen activator (t-PA). (a) initial scan. (b) after one instillation of t-PA at a rate of 1 mg/10cm3 of clot. (c) after 30 days (with permission from Mayfield Clinic)

which tPA was administered, patients experienced an increase in rebleeding when aspiration was performed within 6 h, with rebleeding rates of nearly 40% if performed within the first 4 h; of note, these studies did not include a thrombolytic agent [6,36]. However, maximal benefit appears to occur no later than 12–24 h after hemorrhage [6,37], seemingly providing an ideal window for evacuation in the 6–24 h range. For example, should a structural lesion (i.e., aneurysm or arteriovenous malformation) be the underlying cause of hemorrhage, a minimally invasive operation may provide inadequate exposure to control bleeding and the result would be conversion to an open craniotomy. Although not statistically significant, Cho et al. suggested that minimally invasive techniques (especially endoscopic techniques) are more cost effective and entail shorter ICU stays than traditional open craniotomy techniques [19].

Conclusions In patients with spontaneous ICH for whom medical management is no longer tolerated and who experience a progression of neurological

deficits and/or increasing intracranial pressures, surgical evacuation of the hemorrhage can be performed. The surgical goal is to remove the blood clot related to the hemorrhagic stroke, with the aim to reduce further brain damage and thus improve the patient’s chances of survival and return to independent living. Stereotactic surgical techniques for evacuation of ICH provide an alternative to traditional open surgical evacuation. Craniotomy (in its current form) appears to be relatively ineffective or maybe even worse than medical treatment in certain patients. Less invasive techniques, including stereotactic evacuation, may improve the chances of making a meaningful recovery. The numbers of available trials that compare medical management to minimally invasive techniques are limited. Current trials have looked at routine use rather than selective use of surgery in patients who would most likely benefit. As imaging techniques continue to advance and become more available in the community setting and understanding of the pathophysiology of spontaneous ICH improves, minimally invasive evacuation may become available earlier and eventually offer long-term benefits over that of traditional open craniotomy.


References 1. Fogelholm R, Nuutila M, Vuorela AL. Primary intracerebral hemorrhage in the Jyva¨skyla¨ region, central Finland, 1985‐89: incidence, case fatality rate and functional outcome. J Neurol Neurosurg Psychiatry 1992;55:546-52. 2. Giroud M, Gras P, Chadan N, et al. Cerebral hemorrhage in a French prospective population study. J Neurol Neurosurg Psychiatry 1991;54:595-8. 3. Nilsson OG, Lindgren A, Stohl N. Incidence of intracerebral and subarachnoid hemorrhage in Southern Sweden. J Neurol Neurosurg Psychiatry 2000;69:601-7. 4. Ojemann RG, Heros RC. Spontaneous brain hemorrhage. Stroke 1983;14:468-75. 5. Mendelow AD, Gregson BA, Fernandes HM, et al. Early surgery versus initial conservative treatment in patients with spontaneous supratentorial intracerebral haematomas in the International Surgical Trial in Intracerebral Haemorrhage (STICH): a randomised trial. Lancet 2005;365:387-97. 6. Morgenstern LB, Frankowski RF, Shedden P, et al. Surgical treatment for intracerebral hemorrhage (STICH): a single center, randomized clinical trial. Neurology 1998;51:1359-63. 7. Broderick JP, Adams HP, Jr, 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;30:905-15. 8. McKissock W, Richardson A, Taylor J. Primary intracerebral hemorrhage: a controlled trial of surgical and conservative treatment. Lancet 1961;2:221-6. 9. Jauch EC, Lindsell CJ, Adeoye O, et al. Lack of evidence for an association between hemodynamic variables and hematoma growth in spontaneous intracerebral hemorrhage. Stroke 2006;37:2061-5. 10. Flaherty ML, Woo D, Haverbusch M, et al. Racial variations in location and risk of intracerebral hemorrhage. Stroke 2005;36:934-7. 11. Flaherty ML, Haverbusch M, Sekar P, et al. Long-term mortality after intracerebral hemorrhage. Neurology 2006;66:1182-6. 12. Italian Acute Stroke Study Group. Haemodilution in acute stroke: results of the Italian haemodilution trial. Lancet 1988;1:318–21. 13. Chambers IR, Banister K, Mendelow AD. Intracranial pressure within a developing intracerebral haemorrhage. Br J Neurosurg 2001;15:140-1. 14. Fernandes HM, Siddique S, Banister K, et al. Continuous monitoring of ICP and CPP following ICH and its relationship to clinical, radiological and surgical parameters. Acta Neurochir Suppl 2000;76:463-6. 15. Passero S, Rocchi R, Rossi S, et al. Seizures after spontaneous supratentorial intracerebral hemorrhage. Epilepsia 2002;43:1175-80.

52

16. Vespa PM, O’Phelan K, Shah M, et al. Acute seizures after intracerebral hemorrhage: a factor in progressive midline shift and outcome. Neurology 2003;60:1441-6. 17. Fogelholm R, Murros K, Rissanen A, et al. Admission blood glucose and short term survival in primary intracerebral haemorrhage: a population based study. J Neurol Neurosurg Psychiatry 2005;76:349-53. 18. Auer LM, Deinsberger W, Niederkorn K, et al. Endoscopic surgery versus medical treatment for spontaneous intracerebral hematoma: a randomized study. J Neurosurg 1989;70:530-5. 19. Cho DY, Chen CC, Cheng CS, et al. Endoscopic surgery for spontaneous basal ganglia hemorrhage: comparing endoscopic surgery, sterotactic aspiration, and craniotomy in noncomatose patients. Surg Neurol 2006;65:547-56. 20. Broseta J, Gonzalez-Darder J, Barcia-Salorio JL. Stereotactic evacuation of intracerebral hematomas. Appl Neurophysiol 1982;45:443-8. 21. Kaneko M, Tanaka K, Shimada T, et al. Long-term evaluation of ultra-early operation for hypertensive intracerebral hemorrhage in 100 cases. J Neurosurg 1983;58:838-42. 22. Brott T, Broderick J, Kothari R, et al. Early hemorrhage growth in patients with intracerebral hemorrhage. Stroke 1997;28:1-5. 23. Kazui S, Naritomi H, Yamamoto H, et al. Enlargement of spontaneous intracerebral hemorrhage. Stroke 1996;27:1783-7. 24. Takada I. On the phenomena of extravasation of contrast media in cerebral angiogram of the case of hypertensive intracerebral hematoma and their clinical significance-analysis of 14 cases 1976;4:471‐8. 25. 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;31:330-4. 26. Tanikawa T, Amano K, Kawamura H, et al. CT-guided stereotactic surgery for evacuation of hypertensive intracerebral hematoma. Appl Neurophysiol 1985;48:431-9. 27. Donauer E, Faubert C. Management of spontaneous intracerebral and cerebellar hemorrhage. In: Kaufman HH, editor. Intracerebral hematomas. New York: Raven Press; 1992. p. 211-27. 28. Mukai H, Yamashita J, Kitamura A . et al. Stereotactic aqua-stream and aspirator in the treatment of intracerebral hematoma. An experimental study. Stereotact Funct Neurosurg 1991;57(4):221-7. 29. Zuccarello M, Brott T, Derex L, et al. Early surgical treatment for supratentorial intracerebral hemorrhage: a randomized feasibility study. Stroke 1999;30:1833-9. 30. Fujii Y, Tanaka R, Takeuchi S, et al. Hematoma enlargement in spontaneous intracerebral hemorrhage. J Neurosurg 1994;80:51-7. 31. Lippitz BE, Mayfrank L, Spetzger U, et al. Lysis of basal ganglia haematoma with recombinant tissue

805

806

52 32.

33.

34.

35.

36.

37.


plasminogen activator (rtPA) after stereotactic aspiration: initial results. Acta Neurochir (Wien) 1994; 127:157-60. Schaller C, Rohde V, Meyer B, et al. Stereotactic puncture and lysis of spontaneous intracerebral hemorrhage using recombinant tissue plasminogen activator. Neurosurgery 1995;36:328-33. Broderick JP, Brott T, Zuccarello M. Management of intracerebral hemorrhage. In: Batjer HH, editor. Cerebrovascular disease. Philadelphia, PA: LippincottRaven; 1997. p. 611-27. Kanaya H, Kuroda K. Development in neurosurgical approaches to hypertensive intracerebral hemorrhage in Japan. In: Kaufman HH, editor. Intracerebral hematomas. New York: Raven Press; 1997. p. 197-210. Kaufman HH. Stereotactic aspiration with fibrinolytic and mechanical assistance. In: Kaufman HH, editor. Intracerebral hematoma. New York: Raven Press; 1992. p. 182-5. Niizuma H, Shimizu Y, Yonemitsu T, et al. Results of stereotactic aspiration in 175 cases of putaminal hemorrhage. Neurosurgery 1989;24:814-9. Lee JI, Nam do H, Kim JS, et al. Stereotactic aspiration of intracerebral hematoma: significance of surgical timing and hematoma volume reduction. J Clin Neurosci 2003;10:439-43.

49 Image-Guided Management of Brain Abscess E. Taub . A. M. Lozano

"

In uncomplicated abscess of the brain, operated on at a fairly early period, recovery ought to be the rule. —William Macewen, 1893 [1]

Incidence The incidence of brain abscesses has been estimated at 3 per 1 million people per year in Northern Ireland [2] and 0.8 per 1 million people per year in the region around Lund, Sweden [3]. The incidence in other developed areas of the world is probably similar. These figures include only cases treated by neurosurgeons. Brain abscesses are thus much less common than intracranial neoplasms but are encountered occasionally at every major neurosurgical center. This entity poses both the diagnostic challenge of early recognition to achieve the best possible outcome [2–5], and the therapeutic challenge of permanent eradication of infection while doing the least harm to the surrounding brain. In recent decades, a major decrease in mortality has been achieved through the use of computed tomography (CT), magnetic resonance imaging (MRI), image-guided stereotaxy, and improved antimicrobial agents [4].

Etiology and Pathogenesis Many organisms can give rise to brain abscesses. Streptococcus and Staphylococcus species are the most common pathogens, although anaerobes, #


especially Bacteroides, are also common [3–7]. Anaerobic abscesses occur particularly frequently in children [8]. Multiple organisms are cultured in about 20% of cases, and no organisms are cultured in up to 25% [3,4,6,9]. It has been observed that even when a presumptive extracerebral source of infection can be identified, the brain abscess may nevertheless be due to a different organism or organisms [10]; the implications for treatment are clear. Pathogenic microorganisms can reach the brain parenchyma in three ways: by extension from contiguous structures, by hematogenous spread, and by direct inoculation. Some of the more common sources of infection are listed in > Table 49-1. Infection can extend contiguously from the paranasal sinuses, middle ear, mastoid air cells, or teeth. Hematogenous spread can arise from infections of the heart, lung, or other organs; intravenous drug abuse and congenital cyanotic heart disease are predisposing factors. Infectious material can also be introduced directly into the brain by trauma or by neurosurgical procedures. In civilian life, post-traumatic brain abscesses are often the result of open, depressed skull fractures, especially when these are unrecognized or inadequately treated [11]. Brain abscesses arising in military settings are often due to metallic splinters lodged in the brain and may not become clinically apparent till decades after the initial injury [12,13]. The source of brain abscesses is undetermined (presumably hematogenous) in 20–50% of cases [3–5]. Immunosuppressed persons are susceptible to brain abscesses, especially of fungal and mycobacterial origin. The acquired immune deficiency

770

49

Image-guided management of brain abscess

. Table 49-1 Sources of brain abscesses Extension of infection from contiguous structures Paranasal sinusitis Otitis media Mastoiditis Dental abscess Hematogenous spread Endocarditis Pulmonary infection Intravenous drug use Congenital cyanotic heart disease Direct inoculation Trauma (civilian, military) Neurosurgical procedures Unknown sources

syndrome (AIDS) is a common cause of immunosuppression. Stereotactic procedures of particular relevance to the care of AIDS patients are discussed elsewhere in this book.

Clinical Presentation and Differential Diagnosis Brain abscesses may come to medical attention because of a focal neurological deficit, symptoms and signs of intracranial hypertension (headache, nausea, vomiting, papilledema, impairment of consciousness), seizures, fever and other systemic signs of infection, or none of these manifestations. The diagnosis is sometimes suspected from the presence of one or more contributing historical factors (> Table 49‐1). Unfortunately, none of the typical clinical and laboratory findings are highly specific. The characteristic ring enhancement with central clearing seen on CT and MRI may be seen with primary and metastatic neoplasms as well, and neoplasms are much more common than brain abscesses. Thirty-five percent of patients with brain abscesses have a normal erythrocyte sedimentation rate, more than half have a normal leukocyte count, and more than 60% are afebrile [3]. Intraventricular rupture of a brain abscess can have rapid and

devastating effects; only by the early recognition and prompt treatment of brain abscesses can this complication be prevented [14]. In the pre-CT era, patients with suspected brain abscess sometimes were subjected to lumbar puncture to obtain cerebrospinal fluid for culture. This practice frequently fails to yield an organism [3] and has been found to lead to clinical deterioration in as many as 25% of patients [5]. A brain abscess may be a surprise finding at surgery when there is a different presumptive diagnosis. When 54 patients with known systemic cancer underwent resection or biopsy of presumed brain metastases, 2 proved to have brain abscesses [15]. In other reported series, 11 of 67 cases of brain abscess were initially misdiagnosed [9], as were a majority of 12 cases of fungal brain abscess [16]. It is prudent neurosurgical practice to send portions of brain biopsy specimens for bacterial, fungal, and mycobacterial culture whenever an abscess is considered in the differential diagnosis. Advances in diagnostic imaging in the last 10 years have made it easier to identify a brain abscess correctly as such before the diagnosis is confirmed by biopsy. Brain abscesses have been shown to have markedly different signal characteristics from cystic and/or necrotic brain tumors on diffusion-weighted MRI: the signal intensity and apparent diffusion coefficient of the diseased tissue provide highly indicative, though not absolutely reliable, clues to the underlying pathology [17–20]. When proton magnetic resonance spectroscopy is performed in addition to diffusion-weighted MRI, it seems that even higher diagnostic specificity and selectivity can be achieved, although only a small number of cases have been reported to date [21].

Nonstereotactic Methods of Treatment The currently practiced methods of treatment for brain abscesses are listed in > Table 49‐2.


. Table 49‐2 Current methods of treatment for brain abscesses Non-stereotactic methods Nonsurgical treatment (antimicrobial agents alone) Excision Open evacuation of pus Freehand aspiration Stereotactic methods Conventional stereotactic aspiration Frameless stereotactic aspiration Endoscopic stereotactic aspiration Stereotactic aspiration with real-time imaging: Ultrasound-guided stereotactic aspiration Interactive MRI-guided stereotactic aspiration

Although this chapter deals primarily with stereotactic treatment, a review of other methods will provide the proper context for rational clinical decision making.

49

In one series, lesions that responded to antimicrobial agents alone had a mean diameter of 1.7 cm, while lesions that did not respond had a mean diameter of 4.2 cm [25]. Yang and Zhao [5] obtained good results by restricting nonsurgical treatment to lesions smaller than 2 cm in diameter. We consider nonsurgical treatment a reasonable option if the diagnosis of abscess is nearly certain on clinical grounds, the presumed extracerebral source of the infection is known and drug-sensitive organisms have been cultured from it, and the lesion is less than 2.5 cm in diameter. In such cases, broad-spectrum antimicrobial coverage should be combined with specific treatment of the presumed etiologic organism.

Excision Nonsurgical Treatment (Antimicrobial Agents Alone) Six clinical series from 1971 to 1993 reported on the use of antimicrobial agents alone in a total of 50 patients [5,9,22–25]. Five patients (10%) died, and the rest recovered. Most of the mortality was encountered in only one of the six series (4 of 10 patients) [23]. The small number of patients in each of these series precludes any definitive conclusions, and the good outcomes may be at least partly a result of ‘‘publication bias.’’ It does seem that, for carefully selected patients, the results of nonsurgical treatment may be as good as those of stereotactic aspiration, as is discussed below. Nonetheless, nonsurgical treatment generally should be avoided. All modes of surgical treatment accomplish at least three purposes: confirmation of the diagnosis of abscess, reduction of the infective load, and acquisition of tissue for culture. None of these things are possible without surgery. The size of the lesion is an important determinant of the success of nonsurgical treatment.

Large series of primary excisions of brain abscesses via craniotomy have generally had excellent results. Some representative mortality figures are 1 of 16 patients (6%) [26], 3 of 50 patients (6%) [27], 5 of 56 patients (9%) [5], and 7 of 36 patients (19%) [9]. Less invasive methods of treatment obviously can do no better than excision with regard to reduction of the infective load and acquisition of tissue for culture. Excision is more likely than stereotactic aspiration to require a general anesthetic and involves more extensive disruption of the normal brain tissue surrounding the abscess, although these factors seem to add little morbidity. The outcome data for excision have been matched in recent years by those for stereotactic aspiration. Thus, stereotactic aspiration is generally preferred as the initial treatment, with excision reserved for its occasional failures. There remain a number of specific situations in which excision is indicated as the initial treatment. An abscess containing a foreign body should be excised, not aspirated, so that the foreign body can be removed [28]. It has been suggested that

771

772

49


posttraumatic abscesses secondary to contaminated wounds or to a communication with the paranasal sinuses almost always require excision [29]. The presence of gas in an abscess cavity demonstrates the presence of an extracranial communication or gas-forming bacteria; excision has been recommended in this situation [30]. Nocardial brain abscesses have high mortality, are usually multiloculated, and may more suitable for excision than for aspiration [31,32], although the mortality in a recent series of patients with nocardial abscesses treated by aspiration alone was zero [33]. Some neurosurgeons prefer to treat large abscesses in the posterior fossa by excision because of the greater risk associated with failed aspiration [34,35]. Finally, multi-locularity of an abscess may be an indication for excision if the neurosurgeon thinks that aspirating one or more loculi probably will not achieve an adequate removal of infective material.

only when a lesion is so large that it is unlikely to be missed on the first pass of the aspiration probe. To date, there have been at least three published series of brain abscesses treated by freehand CT-guided aspiration [37–39].

Stereotactic Methods of Treatment Conventional Stereotactic Aspiration Except in the particular situations listed above, excision has given way to stereotactic aspiration of brain abscesses as the initial treatment of choice. Typical mortality figures in recent large series of aspiration (at least 20 patients) are 8% [5], 7% [40], 5% [41], 4% [42], and 0% [43]. These results compare favorably with those of excision.

Operative Technique

Open Evacuation of PUS Maurice-Williams [36] originated a variation on excision of brain abscesses in which pus is thoroughly removed from the interior of the abscess via craniotomy but the capsule is left intact. The theoretical advantage is that the surrounding brain is disturbed less than it would be by excision; the theoretical disadvantage is that some infective material is likely to be left behind. Excellent results were obtained in MauriceWilliams’s hands: Only 1 of 27 patients (4%) died, and 24 (89%) recovered free of disability. To our knowledge, there are no reports of others using this technique.

Freehand Aspiration Freehand technique offers no advantage over stereotaxy in terms of safety or efficacy and probably should be discarded. It is a reasonable option

Our experience and that reported in the literature suggest that the safety and efficacy of stereotactic aspiration are maximized by adhering to a few technical principles. The procedure is performed under CT or MRI guidance with any of several commercially available stereotactic systems. We have used the Brown-Roberts-Wells, Leksell, and Fischer-Leibinger systems. Local anesthesia alone is sufficient in all patients except those too anxious or agitated to stay still for the procedure. Antimicrobial agents should be withheld until after surgery if possible to preserve the best chance of obtaining a positive culture. The stereotactic target should be chosen on minimally thin (1.5 mm) CT or MRI slices to minimize error caused by the partial volume effect. If the abscess contains a large volume of fluid, one should select a target that will be in the lowest part of the abscess when the patient is positioned on the operating table. In this way, a maximum volume of aspirate will be obtained. On the other


hand, if the abscess is expected to be solid (i.e., in the cerebritis stage) or if an abscess is only one of the diagnostic possibilities, the probability of a diagnostic specimen should be maximized by targeting both the center of the lesion and an area near the contrast-enhancing rim. The trajectory to the target should be as short as possible and should avoid ventricles, cisterns, sulci, major vessels, and vital brain areas such as the primary motor cortex. The surgeon should advance the aspiration probe slowly toward the target while being aware of the mechanical resistance of the tissue. An increase and then a decrease in resistance are generally felt as the probe traverses the abscess capsule. Electrical impedance monitoring is an optional technique for further intraoperative confirmation of the position of the probe [44]. After positioning, the probe should be connected to a 10-ml syringe partially filled with normal saline, and the plunger should be pulled out 1 ml to exert negative pressure on the abscess. Excessive suction should be avoided, as a hemorrhage may result. Whatever volume of fluid emerges should be sent for pathological examination, Gram staining, aerobic and anaerobic bacterial cultures, and fungal and mycobacterial cultures. The volume of the aspirate should be compared to the estimated volume of the abscess as calculated from the preoperative imaging study. For a spherical unilocular abscess, the relevant formula is V = 4/3 pr3, where r is the radius of the sphere.

Antimicrobial Agents Aspiration of abscesses is intended to remove most, but not all, of the infective organisms and must be combined with appropriate antimicrobial chemotherapy to be effective. As was stated above, we prefer to withhold antimicrobial agents before surgery if possible. Broadspectrum coverage is begun immediately after the procedure and is tailored to the causative

49

organism or organisms once the species and drug sensitivities are known or is given for a full course if the cultures are negative. Antimicrobial agents generally are continued for 4–6 weeks after surgery. The abscess is monitored with follow-up imaging studies until it has disappeared. The optimal choice of antimicrobial agents for broad-spectrum coverage has not, to our knowledge, been defined in a controlled study. Probably no single regimen can be recommended as being the best, as the prevalence of organisms resistant to particular agents varies among communities and in general increases over time. However, a few principles of agent selection can be derived from the published data. Grampositive aerobes may be covered with penicillin [6] or a related agent. In many hospitals, nafcillin or vancomycin may be the preferred agent because of the likely presence of resistant organisms. An increasing incidence of methicillinresistant Staphylococcus aureus (MRSA) brain abscesses has been documented in recent years [45]. Gram-negative aerobes may be covered with an aminoglycoside or a third-generation cephalosporin [4,5,34]. Anaerobes may be sensitive to penicillin but may require another agent, such as metronidazole [6,8,34]. In the absence of universally valid recommendations, it is best for each neurosurgical service to determine an appropriate regimen in consultation with specialists in infectious disease. Some neurosurgeons have instilled antimicrobial agents directly into the abscess cavity either at the time of aspiration or afterward via indwelling catheters [9,42,46,47]. To our knowledge, there are no clinical data regarding the possible additional benefit of this practice.

Other Medications Dexamethasone

The well-known efficacy of dexamethasone in reducing peritumoral brain edema has led

773

774

49


neurosurgeons to ask whether it has a role to play in the treatment of brain abscesses. The consensus in experimental studies of dexamethasone in brain abscess is that it reduces perilesional edema but also slows the process of abscess encapsulation [48–51]. The clinical significance of the latter finding is unclear. Thus, dexamethasone should be used in all patients with brain abscess in whom perilesional edema is severe enough to contribute to neurological impairment. Its use in other patients is optional. Anticonvulsants

Twenty-five to forty-five percent of patients with brain abscesses have seizures as a component of the presentation [27,52]. Other patients develop a seizure disorder weeks, months, or even years after eradication of the lesion or lesions. We give anticonvulsants to all patients before surgery. We do not recommend the long-term use of anticonvulsants as prophylaxis in patients who have never had a seizure.

Possible Need for Reaspiration Yang and Zhao [5] in Tienjin, China, reported the largest series of brain abscesses treated by aspiration in the CT era. In the 69 patients so treated, ‘‘two or three procedures were usually sufficient,’’ and 12 patients went on to require excision when repeated aspiration failed. Five of sixteen patients in another series required excision when repeated aspiration failed; these were unusually large abscesses that were aspirated with the freehand technique [37]. In contrast, only 1 of 20 patients in the Toronto Hospital series required more than a single aspiration procedure, and none required excision [43].

Drainage Catheters In a number of series, external drainage catheters were placed in some [9] or all [42,47]

brain abscesses that were treated by aspiration. Kondziolka and coworkers [41] placed external drainage catheters only in large abscesses (>3 cm). Broggi and associates [46] implanted an intracavitary catheter connected to a subcutaneous Rickham reservoir, which could then be tapped to gain access to the abscess cavity. Indwelling catheters allow continued drainage of pus in the days after surgery and thus may reduce the need for repeated aspiration (which, however, is rare in some series; see above). They also may be used for repeated intracavitary instillation of antimicrobial agents; the possible benefit of this has not been determined. The outcome data for these techniques do not appear to differ significantly from those for simple aspiration, although the numbers are small.

Other Techniques Frameless Stereotactic Aspiration Frameless stereotaxy is an alternative method of directing the aspiration probe to a target chosen on preoperative CT or MRI. The results obtainable with this technique would be expected a priori to be the same as those of conventional stereotactic aspiration, as long as the frameless targeting is accurate. Laborde and associates [53] reported good results in two patients.

Endoscopic Stereotactic Aspiration Hellwig and colleagues [54] reported on the stereotactic aspiration of brain abscesses through an endoscope in seven patients, with good results. There have been at least two further small series [55,56]. Endoscopy adds to the stereotactic technique a direct visual confirmation of the adequate evacuation of pus, as well as an opportunity to see and electrocoagulate bleeding points on the inside of the abscess


capsule. Whether these small advantages confer a better outcome is unknown.

Ultrasound-Guided Stereotactic Aspiration Berger [57] described a skull-mounted apparatus for ultrasound-guided stereotactic biopsy of brain lesions through a burr hole in awake or anesthetized patients. Five patients underwent stereotactic aspiration of brain abscesses with this device in the original study [57] and a subsequent report [58]. A further nine cases were reported by a third group of neurosurgeons [59], with good results. This technique offers a real-time intraoperative view of the advancing needle as it enters the abscess cavity and of the abscess before and after aspiration. These features are unavailable with conventional or frameless stereotaxy.

Interactive MRI-guided Stereotactic Aspiration Kollias and Bernays [60] recently described the stereotactic aspiration of brain abscesses under intraoperative MRI guidance, with good results. Like intraoperative ultrasound, this technique provides real-time images of the procedure as it is being performed.

Multiple Brain Abscesses Multiple brain abscesses are seen in 10–50% of patients with brain abscesses, depending on the series [61]. There may be as few as 2 abscesses or as many as 20 or more. The general principles of treatment are the same as those for solitary brain abscesses: tissue diagnosis, reduction of the infective load, and directed antimicrobial

49

chemotherapy. It is obviously logistically impractical as well as unsafe to operate on each of many brain abscesses in a single patient. Fortunately, it is also unnecessary, as multiple abscesses generally arise by hematogenous spread and are likely to be due to a single organism or a single mix of organisms. Basit and coauthors [52] in 1989 and Kratimenos and Crockard [62] in 1991 aspirated or excised only the largest of multiple abscesses and then gave tailored antimicrobial therapy. The mortality figures were 5 of 21 patients (24%) and 2 of 11 patients (18%), respectively. Mamelak and coauthors [61] operated on all abscesses that were larger than 2.5 cm in diameter, were situated in critical areas of the brain, or caused significant mass effect. This more aggressive policy resulted in the performance of 43 procedures on 13 patients, while 3 further patients were treated with antimicrobial agents alone. The results were excellent: 15 of 16 patients recovered and 1 (6%) died. Rousseaux et al. [63] used antimicrobial agents alone in 10 of 12 patients with multiple brain abscesses. One patient (10%) died, and the rest made a good recovery. These authors recommended operating on all abscesses larger than 3 cm in diameter.

Conclusion Brain abscesses are curable lesions that should be diagnosed and treated rapidly to prevent progression and permanent neurological sequelae. The mainstay of treatment for brain abscesses is stereotactic aspiration. This should be followed by antimicrobial chemotherapy directed at the causative organism or organisms, which is continued for 4–6 weeks. For multiple brain abscesses, good results have been obtained by aspirating or excising only those greater than 2.5 cm in diameter, those situated in critical brain areas, and those causing significant mass effect.

775

776

49


References 1. Macewen W. Pyogenic infective diseases of the brain and spinal cord. Glasgow: James Maclehose and Sons; 1893. 2. McClelland CJ, Craig EF, Crockard HA. Brain abscesses in Northern Ireland: a 30 year community review. J Neurol Neurosurg Psychiatry 1978;41:1043. 3. Svanteson B, Nordstrom CH, Rausing A. Non-traumatic brain abscess: epidemiology, clinical symptoms and therapeutic results. Acta Neurochir (Wien) 1988;94:57. 4. Mampalam TJ, Rosenblum ML. Trends in the management of bacterial brain abscesses: a review of 102 cases over 17 years. Neurosurgery 1988;23:451. 5. Yang S, Zhao C. Review of 140 patients with brain abscess. Surg Neurol 1993;39:290. 6. Gortvai P, De Louvois J, Hurley R. The bacteriology and chemotherapy of acute pyogenic brain abscess. Br J Neurosurg 1987;1:189. 7. Stapleton SR, Bell BA, Uttley D. Stereotactic aspiration of brain abscesses: is this the treatment of choice? Acta Neurochir (Wien) 1993;121:15. 8. Brook I. Aerobic and anaerobic bacteriology of intracranial abscesses. Pediatr Neurol 1992;8:210. 9. Bidzifiski J, Koszewski W. The value of different methods of treatment of brain abscess in the CT era. Acta Neurochir (Wien) 1990;105:117. 10. Loftus CM, Osenbach RK, Biller J. Diagnosis and management of brain abscess. In: Wilkins RH, Rengachary SS, editors. Neurosurgery. 2nd ed. New York: McGraw-Hill; 1996. p. 3285-98. 11. Stephanov S. Brain abscesses from neglected open head injuries: experience with 17 cases over 20 years. Swiss Surg 1999;5(6):288-92. 12. Wegner-Kempf L, Tornow K, Schmiedek P. Intrazerebraler Abszess 48 Jahre nach Granatsplitterverletzung [Intracerebral abscess 48 years after grenade splinter injury]. Der Radiologe 1994;34(11):671-3. 13. Lee JH, Kim DG. Brain abscess related to metal fragments 47 years after head injury. Case report. J Neurosurg 2000;93(3):477-9. 14. Takeshita M, Kagawa M, Yato S, Izawa M, Onda H, Takakura K, Momma K. Current treatment of brain abscess in patients with congenital cyanotic heart disease. Neurosurgery 1997;41(6):1270-8. 15. 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. 16. Young RF, Gade G, Grinnell V. Surgical treatment for fungal infections in the central nervous system. J Neurosurg 1985;63:371. 17. Noguchi K, Watanabe N, Nagayoshi T, Kanazawa T, Toyoshima S, Shimizu M, Seto H. Role of diffusionweighted echo-planar MRI in distinguishing between brain abscess and tumour: a preliminary report. Neuroradiology 1999;41(3):171-4.

18. Chang SC, Lai PH, Chen WL, Weng HH, Ho JT, Wang JS, Chang CY, Pan HB, Yang CF. Diffusion-weighted MRI features of brain abscess and cystic or necrotic brain tumors: comparison with conventional MRI. Clin Imaging 2002;26(4):227-36. 19. Guzman R, Barth A, Lo¨vblad KO, El-Koussy M, Weis J, Schroth G, Seiler RW. Use of diffusion-weighted magnetic resonance imaging in differentiating purulent brain processes from cystic brain tumors. J Neurosurg 2002;97(5):1101-7. 20. Fertikh D, Krejza J, Cunqueiro A, Danish S, Alokaili R, Melhem ER. Discrimination of capsular stage brain abscesses from necrotic or cystic neoplasms using diffusionweighted magnetic resonance imaging. J Neurosurg 2007;106(1):76-81. 21. Lai PH, Hsu SS, Ding SW, Ko CW, Fu JH, Weng MJ, Yeh LR, Wu MT, Liang HL, Chen CK, Pan HB. Proton magnetic resonance spectroscopy and diffusion-weighted imaging in intracranial cystic mass lesions. Surg Neurol 2007;68 Suppl 1:S25-36. 22. Berg B, Franklin G, Cuneo R, et al. Nonsurgical cure of brain abscess: early diagnosis and follow-up with computerized tomography. Ann Neurol 1978;3:474. 23. Chun CH, Johnson JD, Hofstetter M, Raff MJ. Brain abscess: a study of 45 consecutive cases. Medicine (Baltimore) 1986;65:415. 24. Heineman HS, Braude AL, Osterholm JI. Intracranial suppurative disease: early presumptive diagnosis and successful treatment without surgery. JAMA 1971; 218:1542. 25. Rosenblum ML, Hoff JT, Norman D, et al. Nonoperative treatment of brain abscesses in selected high-risk patients. J Neurosurg 1980;52:217. 26. Choudhury AR, Taylor JC, Whitaker R. Primary excision of brain abscess. BMJ 1977;2:1119 27. Taylor JC. The case for excision in the treatment of brain abscess. Br J Neurosurg 1987;1:173. 28. Emery E, Redondo A, Berthelot JL, Bouali I, Ouahes O, Rey A. Abce`s et empye`mes intracraˆniens. prise en charge neurochirurgicale [Intracranial abscess and empyema. neurosurgical management]. Ann Fr Anesth Reanim 1999;18(5):567-73. 29. Patir R, Sood S, Bhatia R. Post-traumatic brain abscess: experience of 36 patients. Br J Neurosurg 1995;9:29. 30. Young RF, Frazee J. Gas within intracranial abscess cavities: an indication for surgical excision. Ann Neurol 1984;16:35. 31. Hall WA, Martinez AJ, Dummer JS, Lunsford LD. Nocardial brain abscess: diagnostic and therapeutic use of stereotactic aspiration. Surg Neurol 1987;28:114. 32. Mamelak AN, Obana WG, Flaherty JF, Rosenblum ML. Nocardial brain abscess: treatment strategies and factors influencing outcome. Neurosurgery 1994;35:622. 33. Lee GY, Daniel RT, Brophy BP, Reilly PL. Surgical treatment of nocardial brain abscesses. Neurosurgery 2002;51(3):668-71.


34. Gormley WB, del Busto R, Saravolatz LD, Rosenblum ML. Cranial and intracranial bacterial infections. In: Youmans JR, editor. Neurological surgery. 4th ed. Philadelphia: Saunders; 1996. p. 3191-220. 35. Agrawal D, Suri A, Mahapatra AK. Primary excision of pediatric posterior fossa abscesses—towards zero mortality? A series of nine cases and review. Pediatr Neurosurg 2003; 38(2):63-7. 36. Maurice-Williams RS. Experience with ‘‘open evacuation of pus’’ in the treatment of intracerebral abscess. Br J Neurosurg 1987;1:343. 37. Stroobandt G, Zech F, Thauvoy C, et al. Treatment by aspiration of brain abscesses. Acta Neurochir (Wien) 1987;85:138. 38. Savitz MH. CT-guided needle procedures for brain lesions: 20 years’ experience. Mt Sinai J Med 2000;67 (4):318-21. 39. Seliem RM, Assaad MW, Gorombey SJ, Moral LA, Kirkwood JR, Otis CN. Fine-needle aspiration biopsy of the central nervous system performed freehand under computed tomography guidance without stereotactic instrumentation. Cancer 2003;99(5):277-84. 40. Hsieh PC, Pan HC, Chung WY, Lee LS. Computerized tomography-guided stereotactic aspiration of brain abscesses: experience with 28 cases. Zhonghua Yi Xue Za Zhi (Taipei) 1999;62(6):341-9. 41. Kondziolka D, Duma CM, Lunsford LD. Factors that enhance the likelihood of successful stereotactic treatment of brain abscesses. Acta Neurochir (Wien) 1996;127:85. 42. Hasdemir MG, Ebeling U. CT-guided stereotactic aspiration and treatment of brain abscesses: an experience with 24 cases. Acta Neurochir (Wien) 1993;125:58. 43. Shahzadi S, Lozano AM, Bernstein M, et al. Stereotactic management of bacterial brain abscesses. Can J Neurol Sci 1996;23:34. 44. Organ LW, Tasker RR, Moody NF. Brain tumor localization using an electrical impedance technique. J Neurosurg 1968;28:35. 45. Roche M, Humphreys H, Smyth E, Phillips J, Cunney R, McNamara E, O’Brien D, McArdle O. A twelve-year review of central nervous system bacterial abscesses; presentation and aetiology. Clin Microbiol Infect 2003;9(8):803-9. 46. Broggi G, Franzini A, Peluchetti D, Servello D. Treatment of deep brain abscesses by stereotactic implantation of an intracavitary device for evacuation and local application of antibiotics. Acta Neurochir (Wien) 1985;76:94. 47. Itakura T, Yokote H, Ozaki F, et al. Stereotactic operation for brain abscess. Surg Neurol 1987;28:196.

49

48. Bohl I, Wallenfang T, Bothe H, Schiumami K. The effect of glucocorticoids in the combined treatment of experimental brain abscess in rats. Adv Neurosurg 1981;9:125. 49. Quartey GRC, Johnston JA, Rozdilsky B. Decadron in the treatment of cerebral abscess: an experimental study. J Neurosurg 1976;45:301. 50. Schroeder K, McKeever PE, Schaberg D, Hoff JT. Effect of dexamethasone on experimental brain abscess. J Neurosurg 1987;66:264. 51. Yildizhan A, Pas¸aog˘lu A, Kandemir B. Effect of dexamethasone on various stages of experimental brain abscess. Acta Neurochir (Wien) 1989;96:141. 52. Basit AS, Ravi B, Banerji AK, Tandon PN. Multiple pyogenic brain abscesses: an analysis of 21 patients. J Neurol Neurosurg Psychiatry 1989;52:591. 53. Laborde G, Klimek L, Harders A, Gilsbach J. Frameless stereotactic drainage of intracranial abscesses. Surg Neurol 1993;40:16. 54. Hellwig D, Bauer BL, Dauch WA. Endoscopic stereotactic treatment of brain abscesses. Acta Neurochir Suppl (Wien) 1994;61:102. 55. Fritsch M, Manwaring KH. Endoscopic treatment of brain abscess in children. Minim Invasive Neurosurg 1997;40(3):103-6. 56. Longatti P, Perin A, Ettorre F, Fiorindi A, Baratto V. Endoscopic treatment of brain abscesses. Childs Nerv Syst 2006;22(11):1447-50. 57. Berger MS. Ultrasound-guided stereotaxic biopsy using a new apparatus. J Neurosurg 1986;65:550. 58. Borgstein RL, Moxon RA, Hately W, et al. Preliminary experience with the Berger neurobiopsy device for ultrasound guided aspiration and biopsy of intracranial lesions. Clin Radiol 1991;44:98. 59. Strowitzki M, Moringlane JR, Steudel W. Ultrasoundbased navigation during intracranial burr hole procedures: experience in a series of 100 cases. Surg Neurol 2000;54(2):134-44 60. Kollias SS, Bernays RL. Interactive magnetic resonance imaging-guided management of intracranial cystic lesions by using an open magnetic resonance imaging system. J Neurosurg 2001;95(1):15-23 61. Mamelak AN, Mampalam TJ, Obana WG, Rosenblum ML. Improved management of multiple brain abscesses: a combined surgical and medical approach. Neurosurgery 1995;36:76. 62. Kratimenos G, Crockard HA. Multiple brain abscesses: a review of fourteen cases. Br J Neurosurg 1991;5:153. 63. Rousseaux M, Lesoin F, Destee A, et al. Developments in the treatment and prognosis of multiple cerebral abscesses. Neurosurgery 1985;16:304.

777

50 Image-Guided Management of Brain Stem Lesions M. Levivier

The brain stem, which comprises the mesencephalon, pons, and medulla, is a highly complex and functional structure. It is packed with cranial nerves nuclei, many neuronal fascicules and pathways, as well as reticular formations. Its surface has intimate relationships with afferent and efferent cranial nerves, and is surrounded by delicate vasculature. The brain stem is embedded deep in the cranium, located almost entirely in the posterior fossa, protected by the clivus and the petrous bone anteriorly and laterally, and covered by the cerebellum posteriorly. Because of the complex neurosurgical access and the functional importance of the structure, the management of mass lesions of the brain stem remains difficult and controversial. The advances in imaging modalities have improved many aspects of the diagnostic and therapeutic techniques currently available for patients with intracranial pathology. Investigations such as computed tomography, magnetic resonance (MR), and angiography, have become more sensitive and provide finer details. New imaging techniques based on MR technology, such as spectroscopy and diffusion tensor imaging (DTI), or based on the use of radioactive tracers, such as positron emission tomography (PET), can provide additional information on both the nature of the lesion and the functional aspects of the surrounding neural structures. Current imaging modalities, combined with improved computational speed and advanced display technology, allow optimized management of mass lesions of the brain stem, not only at the diagnostic level, but also in using image-guided neurosurgery for their treatment. Indeed, an #


increasing number of imaging techniques can be incorporated in software used for framebased stereotactic biopsy, conventional neuronavigation, or radiosurgery, and are all applied for the management of brain stem lesions.

Imaging of Brain Stem Tumors and Management Strategy MR represents the current imaging of choice for the diagnostic work-up of mass lesions of the brain stem. MR has high-contrast resolution, is devoid of artifacts, and provides multiplanar, multimodal (e.g., T1-weighted, T2-weighted, FLAIR, contrast-enhancement) information that are invaluable on the topography, anatomical relationships, and characteristics of lesions involving the brain stem. It allows evaluating if the tumor is focal, diffuse, or infiltrating. In many instances, it will determine if the origin of the tumor is intraaxial or extra-axial. These characteristics are of importance to better approach the diagnosis and the therapeutic approach. In some cases, however, it is difficult to establish the origin of the tumor, as intra-axial tumors may be associated with an exophytic extension, and extra-axial tumors may present with secondary infiltration of the brain stem through one of the foramina of Luschka. MR is therefore the current single most important imaging technique to decide on which neurosurgical approach can be offered to patients with brain stem lesions. A typical intrinsic lesion, which is suggestive of a brain stem glioma, will need further evaluation on its operability. Based on its MR characteristics, the status of the patient,

780

50

Image-guided management of brain stem lesions

and the therapies available, stereotactic biopsy is often indicated; in some selected cases, direct resection may be performed. Pilocytic astrocytoma may be suggested by its typical marked contrast enhancement. It may be managed by stereotactic biopsy first, or by direct approach for which, neuronavigation is of great help. Residual or recurrent pilocytic astrocytoma can be treated by radiosurgery. Extrinsic tumors, such as ependymoma or medulloblastoma, may invade the brain stem secondarily. A careful imaged-based planning and preparation of the surgery allows optimal resection of such tumors under neuronavigation. Metastasis spread into the brainstem is relatively rare. However, typical images of brain stem metastases can be found, especially in patients with melanoma, lung and breast carcinoma. Such lesions can benefit from radiosurgical therapy. Bleeding in the brain stem deserve detailed MR work-up to evaluate the exact location of the hematoma with respect to the brain structure, especially the surface of the brain stem, as well as to identify an underlying cause. Cavernoma is the most common cause of bleeding in the brain stem and its surgical excision, when possible, is the treatment of choice. Radiosurgery represent an alternative when surgery in contraindicated. These image-guided approaches can also be used in symptomatic cavernomas with progressive brain stem dysfunction due to growing of the lesion without bleeding. Other vascular lesions that may bleed in the brain stem include arteriovenous malformations, for which radiosurgery is indicated. In some instances, other image modalities are useful for the diagnosis and management of lesions of the brain stem. Typically, when MR cannot identify a cause of bleeding, or when an arteriovenous malformation is suspected, angiography is mandatory. This will be useful not only to establish the diagnosis, but also to determine the treatment approach, which may include radiosurgery. In tumors, PET is useful to reveal areas of hypermetabolism, which correlates with tumors of higher grade. Also, the PET characteristics,

especially when heterogeneous, are helpful to guide a stereotactic biopsy, or to focus a partial resection under neuronavigation towards the most aggressive part of the tumor.

Stereotactic Biopsy for Brain Stem Lesions Although the use of image-guided stereotactic brain biopsy is regarded as a safe and reliable approach in supratentorial lesions, its application to the management of infratentorial lesions, especially those involving the brain stem, has remained limited. Moreover, since the direct surgical approach of brainstem lesions for their removal is often associated with a high morbidity, it has favored a nonsurgical management of brain stem tumors in many centers, leading to the prescription of empiric adjuvant treatment, with radiation therapy and/or chemotherapy. Modern neuroimaging techniques, especially MR, have increased our knowledge on brain stem tumors. The more precise assessment of their location and extension, and more specific characteristics of their nature has also predispose to base their management of a tentative diagnosis based on the clinical history and image characteristics, especially in the pediatric population [1]. However, it is well established that the tumor diagnosis and grading is a significant prognostic factor, including in adult brain stem gliomas [2,3]. Moreover, tumor sampling helps in establishing more refined prognostic factors based on biological and genomic characteristics. More information on the characteristics of the tumor give access to more treatment possibilities, especially regarding new approaches with targeted therapies. Similarly, in the pediatric population, the use of systematic biopsy of brain stem tumors allow to include those young patients in new therapeutic protocols [4]. Actually, in many centers, stereotactic biopsy of lesions of the brain stem is performed whenever possible, with an


acceptable low surgical morbidity and mortality, which is no greater than that associated with lesions in other brain locations, including in the pediatric population [5–10]. Stereotactic biopsy of lesions of the brain stem can be performed using either a transfrontal approach or a transcerebellar approach. Location of the lesion, as defined on the preoperative MR work-up or during the analysis of the stereotactic images in the planning software, will determine the best approach. Ideally however, the approach should be defined before the intervention is scheduled, as it will influence frame placement position, as well as the anesthesia setting and the position of the patient on the operating table. MR images, whether obtained in stereotactic conditions, or co-registered with a stereotactic CT, are mandatory for optimal planning of stereotactic biopsy in the brain stem. It allows to visualize the anatomy of brain stem and surrounding structures accurately, as well as to define precisely the target for biopsy within the

50

brain stem lesion. However, the diagnostic yield of this procedure culminates at 95–96%. Other imaging modalities may help to improve the accuracy of targeting lesions of the brain stem. The author has integrated PET data information in the planning of stereotactic biopsy, routinely since 1990 [11,12]. Its systematic prospective use in brain stem lesions suggest that PET guidance improves the diagnostic yield allowing to reduce the sampling procedure, including in the pediatric population [13,14] (> Figure 50-1). The transfrontal approach provides a direct route to all parts of the brain stem, as it follows its longitudinal axis. It is the only available option for the biopsy of tegmental lesions of the midbrain. Lesions located close to the midline and lower in the brain stem, such as in the pons or medulla, can also be approached using the same route. Actually, this stereotactic approach is a well-known route used under local anesthesia in functional procedures, such as the stereotactic electrode implantation for deep brain stimulation,

. Figure 50-1 Examples of combined MR- and PET-guided targeting for stereotactic biopsy in gliomas of the brain stem, illustrating the two approaches. (a) The lesion is located lower in the pons, and the PET-defined target is lateral in the lesion; a transcerebellar approach has been planned. (b) The lesion is located high in the midbrain and the PET-defined target is on the midline; a transfrontal approach has been planned

781

782

50


with a long record of safety. Usually, a parasagittal, anterior coronal, entry point is used. After traversing the frontal cortex, the trajectory will run either through the frontal horn of the ventricle or lateral to it, before entering the anterior thalamus, the cerebral peduncle, and then the target in the brain stem. The planning software, which allows to visualize simultaneously the three orthogonal planes (axial, coronal, and sagittal) and/or reformatted images, is used to calculate an optimal trajectory and to adjust the entry point, based on the target definition and analysis of the safety of the route through the cerebrum. This will lessen the risk of hemorrhage from the pial vessels, especially at the level of the midbrain and cerebellum, by staying within the brain and not crossing the subarachnoid space. Typically, this leads to a burr-hole located about 2 cm from the midline, avoiding the sagittal sinus, entering the brain through a midpoint in a gyrus, and compatible with a trajectory close to the midline. Technically, the procedure is conducted as for stereotactic biopsies in other brain regions. In most frame systems, the coordinate system requires that the plane of the base ring be positioned below the level of the target. Thus, careful frame placement is needed in order to position the base ring low enough, while ensuring that this will not limit the fixation of the fiducial systems for image acquisition, or limit the proper placement of the arc system during surgery. In any case, the base ring must be placed far enough from the target, to ensure the acquisition of artifact-free stereotactic images. The biopsy procedure is performed with the patient in prone position, under local or general anesthesia, depending on the individual practice and setting. A side-cutting canulla (Sedan biopsy needle) is the most commonly used tool for tissue sampling; most manufacturers provide smaller diameter canullas (2.1 mm or less), which should be used for stereotactic biopsy in that area. Staged biopsies along the trajectory are not recommended when performing biopsy in the brain stem; limited suction-aspiration

(usually, with a maximum of 1-ml suction) is obtained at the level of the target. The suboccipital trancerebellar approach is used to reach lesions located laterally in the pons. The trajectory is located entirely in the posterior fossa; it traverses the homolateral cerebellar hemisphere and middle cerebellar peduncle. As such, this approach is also indicated for lesions of the cerebellar peduncles and of the cerebellar hemispheres. Based on the target definition, the planning software is used to define the shortest and safest route through the cerebellar peduncle and hemisphere, while avoiding the tentorium. The planning software then allows to define the entry point and burr-hole position in the lateral occipital bone. The technical aspects described above for the transfrontal approach also apply here. Frame placement may differ however, and there may be some physical limitation to access the entry point and to allow proper setting of the coordinates on the stereotactic apparatus. Frames, such as those used in the BRW/CRW or Leksell systems, have to be placed low, below the level of the lesion, even when planning a trancerebellar approach. Care must taken to anticipate correct access to the occipital bone for the time of the surgical procedures. In some instances, if the biopsy access is rendered difficult by the position of the homolateral posterior post, the latter may be removed at the time of surgery. Some other systems, such as the Fischer–Leibinger ZD, or the Laitinen system, use of an inverted frame placement (i.e., high base ring, with low inverted posts) for posterior fossa lesions, with the plane of the base ring placed above the level of the lesion; the entry point and biopsy trajectory are entirely located below the base ring. In most instances, the biopsy is performed under general anesthesia with the patient in prone position. However, local anesthesia with the patient in a sitting position can be considered if there is contraindication to the general anesthesia. The approach to lesions of the tectum of the midbrain differs however. They should be


biopsied similarly to lesions of the pineal region. The major theoretical risk is bleeding from injury of the internal cerebral veins situated above the tumor, the basilar veins of Rosenthal laterally, as well as the precentral cerebellar vein, the great vein of Galen, and the posterior medial choroidal arteries posteriorly. Thus, tectal lesions should be approached from a more anterior and lateral entry point, in order to pass under the internal cerebral veins, above the basilar vein of Rosenthal, and beneath the precentral cerebellar vein. Current planning software allow to simulate accurately such trajectories on MR. Actually, stereotactic biopsy in the pineal region has mortality, morbidity, and diagnostic rates that are not different from stereotactic biopsy in other regions of the brain [15].

Other Image-guided Procedures for Brain Stem Lesions The evolution of imaging technology allows an earlier diagnosis and a better anatomical understanding of lesions of the brain stem. Parallely, the advent of image-guidance has modified our neurosurgical approaches to the brain stem and open new therapeutic avenues. Some of them are illustrated below, as examples of the developments in image-guided neurosurgery for lesions of the brain stem.

Neuronavigation for Surgery of the Brain Stem The evolution of skull base surgery and a better understanding of the surgical anatomy of the brain stem have allowed neurosurgeons to develop new avenues in brain stem surgery. These advances reflect technological improvements, which include the use of frameless neuronavigation systems, both for the planning of the approach to the brain stem and for the surgical guidance during the procedure. This is especially

50

true for discrete lesions, for which more aggressive approaches are now recommended. The reports of successful resection of cavernomas of the brain stem during the recent years represent typical examples of the evolution of the management of brain stem lesions, with that respect [16–19]. The selection of an optimal and safe approach for the resection of discrete lesions of the brain stem remains a critical issue, however. Using the two-point method described by the group of R. Spetzler at the Barrow Neurological Institute [20,21] different surgical approaches can be defined. A point is placed in the center of the lesion and a second point is placed where the lesion is the closest to the surface of the brain stem. A straight line connecting these two points dictates the optimal trajectory and surgical approach. This method helps to define the choice between the different surgical approaches, which allow access to a corresponding area of the brain stem. These approaches usually represent major surgical procedures, and include the orbitozygomatic, subtemporal, petrosal, retrosigmoid, supracerebellar infratentorial, retrosigmoid suboccipital, and far lateral approaches. Simulating the two-point method in a navigation planning software using high quality MR is of great help to determine the angle of attack and to analyze the surgical approaches that are possible (> Figure 50-2). The integration of this information in the surgical navigation system is also helpful during surgery, including during the skull base approach [22]. Moreover, because surgery for brain stem lesions is at high risk the analysis of tensor imaging–based fiber tracking adds essential information to the preoperative planning. Hence, the integration of the tractography into the neuronavigation system will make possible to anticipate the location of the major fiber tracts during surgery. Thus, this image-guided method may also increase the likelihood of total resection of tumors adjacent to, or involving, eloquent fiber tracts in the brain stem, in order to avoid new neurological deficits after surgery. This strategy

783

784

50


. Figure 50-2 Composite picture of two different planning performed to study the possibility to perform open surgical approaches in an hemorrhagic cavernoma of the brain stem, comparing a retrosigmoid approach (a), and a supracerebellar infratentorial approach (b), which are both simulated in 3-D (c)

has already been applied to the resection of a deeply located brain stem cavernous angioma [23] and will certainly benefit to other surgical approaches to the brain stem. Although the management of intrinsic lesions of the brain stem remains more controversial, the surgical management of intrinsic gliomas has evolved towards a more aggressive surgical treatment during the two last decades. Indeed, direct resection can be performed with relatively low morbidity in subgroups of lesions previously managed conservatively [24–26]. Based on MR imaging, brain stem gliomas can be categorized as focal, cervicomedullary, dorsally exophytic, or diffuse. A focal brain stem glioma grows as an expanding

mass, which usually dislocates the neighboring nervous structures without invading them. Also, as it grows, the tumor tends to move towards the surface of the brain stem, which will influence the feasibility of the resection and choice of surgical approach. It is the experience of many neurosurgeons that brain stem glioma with a focal growth pattern are often benign and totally respectable without neurological worsening, which suggest that these tumors do not directly invade the surrounding neurological structures [24,26– 28]. As such, the image-guided neurosurgical techniques should be used to optimize their approach, as described above for discrete lesions of the brain stem.


Frameless Image-Guidance in the Brain Stem Although frameless navigation systems are now routinely used for the biopsy of supratentorial tumors, most neurosurgeons recognize that frame-based procedures are still mandatory for lesions of the brain stem. As frameless stereotaxy has already been used for deep brain stimulation

50

comparable accuracy than frame-based systems [29], it is not far fetched that frameless systems may be used in the future to biopsy brain stem lesions. Along the same line, the placement of an infusion cannula for the convectionenhanced delivery of therapeutic agents in the brain stem via a transfrontal approach, as already been performed using a frameless navigation system [30].

. Figure 50-3 Radiosurgery using combined MR- and PET-target definition in a patient with a recurrent pilocytic astrocytoma of the brain stem. The hypermetabolic area that was visible in the tumor at the time of radiosugery (LGK Leksell Gamma Knife) disappeared 9 months after radiosurgery; at that time the tumor was still visible but had started to disclose a necrotic center. The latter was even more visible between 12 and 24 months after radiosurgery. At 36 months, the lesion has disappeared on MR. During that time, there was recurrence of increased metabolism in the lesion area

785

786

50


Radiosurgery for Brain Stem Lesions Radiosurgery is used an alternative or as a combined treatment for numerous lesions of the brain. It represents one the least invasive image-guided procedures in neurosurgery, as the stereotactic delivery of convergent beams is performed through the intact skull with minimal constraints for the patient. Radiosurgery is often used to treat tumors in the brain stem. One of the most common indications is metastases of the brain stem. Our group and several others have recently shown that radiosurgery for brain stem metastases provide a high tumoral control (up to 95%) and prolonged survival, with a relatively low complication rate related to the treatment [31–35]. This confirms that, owing to the high risk of surgical resection of brain stem metastases and low efficacy of medical treatment, radiosurgery should be proposed upfront when the size of the lesion is compatible with this approach. Other tumors of the brain stem, especially focal gliomas, may also benefit from radiosurgery [36]. Pilocytic astrocytomas have a high tumoral control with radiosurgery, and this technique is very useful when a multimodality approach is necessary [37]. In order to better define the target volume on stereotactic images, we have integrated the metabolic information of PET in the planning procedure for radiosurgery of brain tumors [38,39]. When combined with PET metabolic data during the follow-up after radiosurgery, this approach provides valuable information on the response and prognosis of intrinsic tumors of the brain stem treated with radiosurgery (> Figure 50-3). Vascular lesions of the brain stem are also treated with radiosurgery. Arteriovenous malformation (AVM) of the brain stem are difficult to treat. Radiosurgery can be performed as a single treatment in small lesions, or as part of a multimodality approach in larger AVM. However, the rate of obliteration is lower, and the risk of new neurological deficit is higher than for AVM of

other locations [40–42]. This emphasizes the difficulty in treating patients with deeply located AVM, the majority of whom are also poor candidates for resection or embolization. In symptomatic cavernomas, with imaging-confirmed hemorrhages, and for which resection is considered to be of too high risk, radiosurgery represents an interesting alternative, as it confers a reduction in the risk of new hemorrhage [43].

References 1. Albright AL, Packer RJ, Zimmerman R, Rorke LB, Boyett J, Hammond GD. Magnetic resonance scans should replace biopsies for the diagnosis of diffuse brain stem gliomas: a report from the Children’s Cancer Group. Neurosurgery 1993;33:1026-9. 2. Kesari S, Kim RS, Markos V, Drappatz J, Wen PY, Pruitt AA. Prognostic factors in adult brainstem gliomas: a multicenter, retrospective analysis of 101 cases. J Neurooncol 2008;88:175-83. 3. Rosenthal MA, Ashley DM, Drummond KJ, Dally M, Murphy M, Cher L, Thursfield V, Giles GG. Brain stem gliomas: patterns of care in Victoria from 1998–2000. J Clin Neurosci 2008;15:237-40. 4. Roujeau T, Machado G, Garnett MR, Miquel C, Puget S, Geoerger B, Grill J, Boddaert N, Di Rocco F, Zerah M, Sainte-Rose C. Stereotactic biopsy of diffuse pontine lesions in children. J Neurosurg 2007;107:1 Suppl Pediatrics:1-4. 5. Kondziolka D and Lunsford LD. Results and expectations with image-integrated brainstem stereotactic biopsy. Surg Neurol 1995;43:558-62. 6. Kratimenos GP, Nouby RM, Bradford R, Pell MF, Thomas DG. Image directed stereotactic surgery for brain stem lesions. Acta Neurochir (Wien) 1992;116: 164-70. 7. Hall WA. The safety and efficacy of stereotactic biopsy for intracranial lesions. Cancer 1998;82:1749-55. 8. Pincus DW, Richter EO, Yachnis AT, Bennett J, Bhatti MT, Smith A. Brainstem stereotactic biopsy sampling in children. J Neurosurg 2006;104: 2 Suppl Pediatrics:108–14. 9. Samadani U, Stein S, Moonis G, Sonnad SS, Bonura P, Judy KD. Stereotactic biopsy of brain stem masses: decision analysis and literature review. Surg Neurol. 2006;66:484-9. 10. St George EJ, Walsh AR, Sgouros S. Stereotactic biopsy of brain tumours in the paediatric population. Childs Nerv Syst 2004;20:163-7. 11. Levivier M, Goldman S, Bidaut LM, Luxen A, Stanus E, Przedborski S, Bale´riaux D, Hildebrand J, Brotchi J.


12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Positron emission tomography-guided stereotaxic brain biopsy. Neurosurgery 1992;31:792-7. Levivier M, Goldman S, Pirotte B, Brucher J-M, Bale´riaux D, Luxen A, Hildebrand J, Brotchi J. Diagnostic yield of stereotactic brain biopsy guided by positron emission tomography with [18F]fluorodeoxyglucose. J Neurosurg 1995;82:445-52. Massager N, David P, Goldman S, Pirotte B, Wikler D, Salmon I, Nagy N, Brotchi J, Levivier M. Combined MR and PET imaging in brain mass lesions: diagnostic yield in a series of 30 stereotactically biopsied patients. J Neurosurg 2000;93:951-7. Pirotte B, Lubansu A, Massager N, Wikler D, Goldman S, Levivier M. Results of positron emission tomography guidance and reassessment of the utility of and indications for stereotactic biopsy in children with infiltrative brainstem tumors. J Neurosurg 2007;107:5 Suppl Pediatrics: 392-9. Regis J, Bouillot P, Rouby-Volot F, Figarella-Branger D, Dufour H, Peragut JC. Pineal region tumors and the role of stereotactic biopsy: review of the mortality, morbidity, and diagnostic rates in 370 cases. Neurosurgery 1996;39: 907-12. Bruneau M, Bijlenga P, Reverdin A, Rilliet B, Regli L, Villemure JG, Porchet F, de Tribolet N. Early surgery for brainstem cavernomas. Acta Neurochir (Wien) 2006;148:405-14. Ferroli P, Sinisi M, Franzini A, Giombini S, Solero CL, Broggi G. Brainstem cavernomas: long-term results of microsurgical resection in 52 patients. Neurosurgery 2005;56:1203-12. 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. Samii M, Eghbal R, Carvalho GA, Matthies C. Surgical management of brainstem cavernomas. J Neurosurg 2001;95:825-32. Brown AP, Thompson BG, Spetzler RF. The two-point method: evaluating brain stem lesions. BNI Q 1996;12:20-4. Porter RW, Detwiler PW, Spetzler RF. Surgical approaches to the brain stem. Op Tech Neurosurg 2000;3:114-30. Brinker T, Arango G, Kaminsky J, Samii A, Thorns U, Vorkapic P, Samii M. An experimental approach to image guided skull base surgery employing a microscope-based neuronavigation system. Acta Neurochir (Wien) 1998;140:883-9. Chen X, Weigel D, Ganslandt O, Fahlbusch R, Buchfelder M, Nimsky C. Diffusion tensor-based fiber tracking and intraoperative neuronavigation for the resection of a brainstem cavernous angioma. Surg Neurol 2007;68:285-91. Bricolo A, Turazzi S. Surgery for gliomas and other mass lesions of the brainstem. Adv Tech Stand Neurosurg 1995;22:261-341.

50

25. Le´vesque MF, Parker F. MKM-guided resection of diffuse brainstem neoplasms. Stereotact Funct Neurosurg 1999;73:15-8. 26. Teo C, Siu TL. Radical resection of focal brainstem gliomas: is it worth doing? Childs Nerv Syst 2008;24: 1307-14. 27. Constantini S, Epstein F. Surgical indication and technical considerations in the management of benign brain stem gliomas. J Neurooncol 1996;28:193-205. 28. Hoffman HJ, Becker L, Craven MA. A clinically and pathologically distinct group of benign brain stem gliomas. Neurosurgery 1980;7:243-8. 29. Holloway KL, Gaede SE, Starr PA, Rosenow JM, Ramakrishnan V, Henderson JM. Frameless stereotaxy using bone fiducial markers for deep brain stimulation. J Neurosurg 2005;103:404-13. 30. Lonser RR, Warren KE, Butman JA, Quezado Z, Robison RA, Walbridge S, Schiffman R, Merrill M, Walker ML, Park DM, Croteau D, Brady RO, Oldfield EH. Real-time image-guided direct convective perfusion of intrinsic brainstem lesions. Technical note. J Neurosurg 2007;107:190-7. 31. Fuentes S, Delsanti C, Metellus P, Peragut JC, Grisoli F, Regis J. Brainstem metastases: management using gamma knife radiosurgery. Neurosurgery 2006;58:37-42. 32. Hussain A, Brown PD, Stafford SL, Pollock BE. Stereotactic radiosurgery for brainstem metastases: survival, tumor control, and patient outcomes. Int J Radiat Oncol Biol Phys 2007;67:521-4. 33. Kased N, Huang K, Nakamura JL, Sahgal A, Larson DA, McDermott MW, Sneed PK. Gamma knife radiosurgery for brainstem metastases: the UCSF experience. J Neurooncol 2008;86:195-205. 34. Lorenzoni JG, Devriendt D, Massager N, Desmedt F, Simon S, Van Houtte P, Brotchi J, Levivier M. Brain stem metastases treated with radiosurgery: prognostic factors of survival and life expectancy estimation. Surg Neurol 2008;(in press). 35. Yen CP, Sheehan J, Patterson G, Steiner L. Gamma knife surgery for metastatic brainstem tumors. J Neurosurg 2006;105:213-9. 36. Yen CP, Sheehan J, Steiner M, Patterson G, Steiner L. Gamma knife surgery for focal brainstem gliomas. J Neurosurg 2007;106:8-17. 37. Hadjipanayis CG, Kondziolka D, Gardner P, Niranjan A, Dagam S, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for pilocytic astrocytomas when multimodal therapy is necessary. J Neurosurg 2002;97:56-64. 38. Levivier M, Wikler D Jr, Massager N, David P, Devriendt D, Lorenzoni J, Pirotte B, Desmedt F, Simon S Jr, Goldman S, Van Houtte P, Brotchi J. The integration of metabolic imaging in stereotactic procedures including radiosurgery: a review. J Neurosurg 2002;97 5 Suppl:542-50. 39. Levivier M, Massager N, Wikler D, Lorenzoni J, Ruiz S, Devriendt D, David P, Desmedt F, Simon S, Van Houtte P,

787

788

50


Brotchi J, Goldman S. Use of stereotactic PET images in dosimetry planning of radiosurgery for brain tumors: clinical experience and proposed classification. J Nucl Med 2004;45:1146-54. 40. Massager N, Re´gis J, Kondziolka D, Njee T, Levivier M. Gamma knife radiosurgery for brainstem arteriovenous malformations: preliminary results. J Neurosurg 2000; 93:102-3. 41. Maruyama K, Kondziolka D, Niranjan A, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for brainstem

arteriovenous malformations: factors affecting outcome. J Neurosurg 2004;100:407-13. 42. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg 2004;100: 210-4. 43. Hasegawa T, McInerney J, Kondziolka D, Lee JY, Flickinger JC, Lunsford LD. Long-term results after stereotactic radiosurgery for patients with cavernous malformations. Neurosurgery 2002;50:1190-7.

41 Impedance Recording in Central Nervous System Surgery R. J. Andrews . J. Li . S. A. Kuhn . J. Walter . R. Reichart

Introduction and Historical Aspects Introduction This chapter will first describe electrical bioimpedance and its relevance to neurosurgeons, and provide a brief historical review. Current and near-future techniques for central nervous system (CNS) tissue identification (e.g., brain vs. tumor) are then considered. A brief review of impedance ‘‘imaging’’ – electrical impedance tomography EIT) – is then presented, given EIT may assume a significant clinical role in the future. The final section presents the new field of charge transfer at the neuronal/subneuronal level, made possible in the past decade by advances in nanoelectrode techniques, and contrasts such a neural-electrical interface (NEI) with the traditional macro- and micro-electrodes for neuromodulation or deep brain stimulation (DBS).

Electrical Bioimpedance The conduction of electric current through biological tissues depends on the tissue’s composition. This is important not only in the use of impedance for localization within the CNS (e.g., gray vs. white matter, specific nuclei), but also for tissue identification (e.g., tumor vs. brain) and tissue status (e.g., normal vs. edematous brain). Although the clinical use of impedance monitoring in neurosurgery was greater prior #


to the advent of computed tomography (CT) and magnetic resonance imaging (MRI), there continue to be important applications of the principles of brain impedance in neuromodulation (deep brain stimulation – DBS – in particular) and also the evolving field of brain imaging based on impedance (electrical impedance tomography – EIT). However, as technological advances allow the development of a NEI at the neuronal or sub-neuronal level (e.g., axon, dendrite, intracellular), the need to understand the properties of charge monitoring and transfer at the neuronal level become paramount. Ohm’s Law describes the relationship between voltage, current, and resistance to current flow in a direct current (DC) situation: Rðor Oresistance; in ohmsÞ ¼ Vðvoltage; in voltsÞ=Iðcurrent; in amperesÞ In an alternating current (AC) situation, one must add another term – reactance – in addition to resistance, because of the phase changes in voltage and current in the AC situation: Z ðimpedance; in ohmsÞ ¼ R ðresistance; in ohmsÞ þ X ðreactance; in ohmsÞ

If X = 0, the impedance is purely resistive. If X < 0, the reactance is capacitive; if X > 0, the reactance is inductive. A capacitor can be used to store electric charge; an inductor can be used – in the form of electromagnetic coils – as a transformer. Conductance or conductivity is the reciprocal of resistance.

632

41

Impedance recording in central nervous system surgery

Bioimpedance in brain depends only in part on characteristics of the brain or spinal cord tissue: gray versus white matter, cell composition (percentage of neurons vs. glial cells), extent of myelination, extracellular tissue fluid composition, etc. Characteristics of the electrode are relevant: composition (e.g., noble metal vs. ceramic), contact area with CNS tissue, orientation within CNS tissue, proximity to capillaries or other blood vessels, corrosion, etc. Also relevant is monopolar versus bipolar recording. Perhaps the major source of impedance in deep brain stimulation is the layer of gliosis or scar tissue which forms around the electrode [1,2]. The importance of impedance is clear to neurosurgeons performing DBS surgery: the greatly reduced ‘‘life’’ of a pulse generator driving a deep brain electrode with low impedance (2 KΩ) is clear – one pulse generator may function for more than 5 years, the other for only a year or less (not to mention differences in clinical efficacy and volume of tissue activated (VTA) [1]).

Historical Aspects Bioimpedance was first studied by Hoeber nearly 100 years ago, who investigated the conductivity of the erythrocyte membrane and the cell interior [3]. It was not until the middle of the century, however, that Hodgkin and Huxley conducted their Nobel prize-winning research that marked the era of membrane biophysics. The phenomenon of spreading depression was also described in the early 1950s, and noted to be accompained by a significant increase in corticial impedance. The conductance (reciprocal of impedance) of the rabbit cerebal cortex was studied in a series of experiments in the mid-1950s, with various factors affecting impedance being noted: brain cooling, cerebrospinal fluid (CSF) drainage, exsanguinations, and cirulatory arrest all resulted in increased

brain impedance-but intrestingly ether anesthesia did not have a consistent effect on impedance [4]. A brief review of bioimpedance research has been written by one of the pioneers in the field [5].

Impedance Monitoring for CNS Tissue Identification CNS Impedance Prior to CT/MRI Prior to CT and MRI imaging for tissue identification and localization in the CNS, impedance monitoring played a significant role in several neurosurgical procedures. These will be reviewed briefly to place the current interest in impedance measurement in perspective. One early study of impedance differences between normal brain and malignant gliomas studied formalin-fixed autopsy specimens [6]. Malignant gliomas typically had impedance measurements less than one-half that of normal brain. A clinical study of 14 patients with intracranial tumors reported shortly prior to the CT era found that brain tumors softer than normal brain tended to have lower impedance than normal brain, whereas firmer tumors (meningiomas) had higher impedance; penetration of a tumor capsule with the recording electrode demonstrated a transient increase in impedance [7]. Impedance monitoring has also been used in conjunction with image-guided stereotactic brain biopsies [8,9]. General observations from such studies include: 1.

2.

Tissue impedance tends to vary with tissue density (i.e., increased impedance in cyst walls, tumor capsules, and firm tumors such as certain meningiomas). Cerebrospinal fluid (CSF), edematous brain, and necrotic tissue all have lower impedance than normal brain.


3.

Given variations in electrode characteristics and monitoring systems, changes in impedance are probably of more value than absolute impedance values (in ohms).

Apart from brain tumor and cyst localization, the other major use of impedance monitoring has been in percutaneous cordotomies. Localization is determined by the drop in impedance as the neck tissues are penetrated and the CSF entered, followed by a rise in impedance as the spinal cord is penetrated once the CSF space has been traversed [10,11].

Impedance Monitoring During CT/MRI Image Guidance Brain Biopsy – Current Status The advent of DBS for movement disorders over the past 20 years has brought focus on electrical stimulation of the brain as a substitute for, and extension of, ablative procedures. The importance of impedance in the efficacy of DBS has become increasingly apparent, as noted in A. II. above and in the literature [1]. Additionally, microrecording of spontaneous electrical activity from specific brain nuclei (e.g., the subthalamic nucleus (STN) in DBS for Parkinson’s disease) has become an important aspect of precise localization – an adjunct to high-resolution MRI localization for functional neurosurgery. Thus a platform has been established that offers an opportunity for refinements in intraoperative impedance monitoring over techniques of previous decades. The group at Friedrich-Schiller-University in Jena, Germany, has recently begun state-ofthe-art impedance measurements during intracranial procedures with CT/MRI guidance, in particular stereotactic brain biopsy. The technique is similar to that customarily used for DBS electrode placement: following high-resolution CT and MRI scans with a stereotactic frame attached

41

to the patient’s head, a burr hole is made and the dura cauterized and opened. A standard microdrive is used to introduce the microelectrode with a platin tip (1 mm thickness) into the brain along the trajectory to the tumor (> Figure 41-1). Continuous photo documentation is made of the electrode depth and recording data to facilitate correlation with histopathology and molecular biology at each point along the trajectory. Following the recording of extracellular potentials along the trajectory, serial biopsies are taken for standard neurohistopathology as well as mRNA extraction. Assays include ion channels (sodium, potassium, chloride, calcium) and neurotransmitter receptors. This multimodality technique promises to yield correlations between imaging, electrical patterns, histopathology, and molecular biology (> Figure 41-2).

Electrical Impedance Tomography Background Electrical impedance tomography (EIT) is the term adopted over 20 years ago for the technique of stimulating, electrode by electrode, a volume of tissue with multiple electrodes applied (e.g., the head) in order to obtain data on the impedance of the tissue at any point within the volume queried (> Figure 41-3). In practice, for EIT of the brain EEG-type electrodes are affixed to the scalp and recordings made (> Figure 41-4). The concept of mapping the impedance of a volume of material was first utilized more than 75 years ago in geological studies, and more recently in industry to detect air bubbles, etc [12]. The term ‘‘tomography’’ – in comparison with computed tomography (CT) – is a misnomer since, unlike the x-rays used in CT, electrical current injected into a volume of tissue cannot be confined to a plane. The resulting EIT ‘‘images’’ are values for

633

634

41


. Figure 41-1 Intraoperative deep brain microrecording in a brain tumor patient (prior to collection of biopsy specimens). The monitor shows the recorded potentials. The platin electrode is clamped into the microdrive device, which in turn is fixed to the Leksell® stereotactic frame. Total electromagnetic quiescence is required during microrecording [Reichart R, et al., unpublished results]

the impedance at various points (more accurately volumes or voxels) within the volume investigated. Because of variations in the contact impedances of the electrodes, absolute values are problematic; thus relative impedances at the various points or voxels are usually recorded (so-called ‘‘difference imaging,’’ in contrast to ‘‘absolute imaging’’). Additional data are obtained by varying the frequency of the stimulation (usually from the kHz to MHz range), multi-frequency electrical impedance tomography (MFEIT). Several recent reviews detail the issues involved in the development of EIT for clinical use [12,13].

Applications of EIT Several clinical applications of EIT are currently being pursued, although none has achieved clinical status: (1) breast cancer detection;

(2) pulmonary emboli detection; (3) gastric emptying time (in the assessment of gastrointestinal disorders); (4) brain disorders [12]. The use of EIT in brain disorders has concentrated on epilepsy focus localization and the early detection of ischemic versus hemorrhagic stroke. The latter application, stroke identification, is seen as valuable in emergency departments and clinics where CT is not readily available; EIT could provide rapid information that a stroke is ischemic and not hemorrhage – thus increasing the number of patients who might benefit from early hemolytic therapy in acute stroke (where the treatment must be instituted within a few hours of stroke onset). The most extensive experience to date with MFEIT for brain lesions (stroke, brain tumors, and arteriovenous malformations) has yet to document sufficient specificity to allow MFEIT to assume a clinical role at present [14].


41

. Figure 41-2 Correlation between intraoperative photo documentation, brain microrecording, and postoperative histopathology in a brain tumor patient [Reichart R, et al., unpublished results]

EIT in the Future Advances in computational processing have made EIT and MFEIT feasible over the past

10–15 years, but the issue of contact (electrode) impedance variability in particular has made clinical usefulness elusive. Thus recent techniques have sought to develop noncontact methods for

635

636

41 . Figure 41-2


(Continued)

inducing the current and measuring the resulting electrical field (impedance), i.e., the use of coils for induction and recording [12]. One such technique is magnetic induction tomography (MIT), which has the possibility of resolution referred to as MIT-spectroscopy (MITS).

Another technique is combining EITwith MRI, i.e., magnetic resonance electrical impedance tomography (MREIT). One variation of MREIT obtains the MRI with a current injection through surface electrodes to obtain a current density/conductivity image of the tissue under investigation [12,15].


. Figure 41-3 Schematic of the application of a current around the surface electrodes on a biological tissue (e.g., the human scalp) to obtain an impedance tomogram of tissue conductivity

Electrical Charge Monitoring and Transfer at the Neuronal Level – Contrasts Between Macro/MicroElectrodes and Nano-Electrodes

spatial complexities introduced by the extracellular matrix and the non-neuronal supporting cells (e.g., astrocytes). As CNS recording and stimulation become more fine tuned – potentially down to the sub-neuronal level – the need to construct a neural-electrical interface (NEI) that mimics the communication within the CNS becomes evident. As we move from simple ‘‘brain stimulation’’ to true neuroprosthetics, the NEI must become bidirectional and multifunctional: electrical and neurochemical (neurotransmitter) information must be monitored by the neuroprosthetic device, and in turn the device must be able to ‘‘sculpt’’ the local CNS landscape toward normal functioning (e.g., for epilepsy, movement disorders, spinal cord injury). Until recently, the major advances in neuromonitoring and neurostimulation have been refinements in metal microelectrodes – mostly miniaturization. Limitations of such electrodes for NEI include the following: 1.

Limitations of Current CNS Electrodes Neuromodulation of brain tissue (deep brain stimulation – DBS) with macro- or microelectrodes functionally appears to be a form of ‘‘reversible ablation,’’ i.e., the effects on the CNS function as a whole is quite similar to irreversibly ablating a similar volume of tissue by, e.g., thermocoagulation. Our understanding of the mechanisms of action of DBS is increasing rapidly, with computer modeling being a powerful technique [16]. Even a microelectrode, being tens of microns or larger in diameter, does not interface intimately with individual neurons in the way that axons and dendrites interact. The 2-D nature of the microelectrode – nervous tissue contact surface does not resemble the 3-D nature of axo-dendritic interactions with neuron al cell bodies or processes, not to mention the

41

2.

3.

The electrode surface is 2-D, interacting with a 3-D complex of neurons, neuronal processes, supporting cells such as astrocytes, etc. Impedance is greatly increased in noble metal microelectrodes in comparison with alternatives such as composite materials and carbon nanotube arrays [17,18]. The stiffness (Young’s modulus) of neural tissue is roughly six orders of magnitude less than metal microelectrodes (~2.5 kPa vs. >10 GPa, respectively).

The problems with noble metals for CNS microelectrodes have been discussed recently [19]. Considerable effort has been spent recently in modifying the surface of traditional metal microelectrodes to lower impedance and thus improve electrical charge transfer for neurostimulation, most notably through the use of electrically conductive polymer (ECP) coatings [20]. However,

637

638

41


. Figure 41-4 Multi-frequency electrical impedance tomography (MFEIT). Thirty-one EEG scalp electrodes are placed using a modified 10–20 EEG scheme, and contact impedance checked using two-terminal impedance measurements between electrode 1 and the other electrodes. Each image data set is made from 258 impedance measurements obtained from combinations of the 31 EEG electrodes

carbon nanotubes (CNTs), appropriately configured, have been shown to possess properties highly desirable for precise recording and stimulation of neural tissue: impedance can be greatly reduced, and capatinance greatly increased, in comparison with platinum or other noble metal electrodes, especially when coated with an ECP such as polypyrrole (PPy) – as measured by electrochemical impedance spectroscopy [17–19]. In addition to the three issues noted above, neurons must ‘‘cohabit’’ successfully with their ‘‘electrical’’ counterparts in the NEI, i.e., long-term toxicity must not be a problem. This has been demonstrated so far for periods of days to weeks [18,19].

Fabrication and Advantages of Carbon Nanotube Electrodes As an example of some of the issues involved in creating a 3-D NEI, we here summarize recently

reported findings with CNT (or carbon nanofiber, CNF) arrays [18]. > Figure 41-5 is a schematic of the steps needed to create a CNF or CNT nanoelectrode array that will support the growth of PC12 cell networks. PC12 (rat pheochromocytoma) cells are neuron-like cells that under appropriate conditions can release dopamine, and thus are of interest in models for movement disorders such as Parkinson’s disease and mood disorders such as severe refractory depression (both Parkinson’s disease and severe depression involve abnormalities of dopamine). Not only does the polypyrrole (PPy) coating greatly improve the electrical properties of the nanoelectrode arrays (as noted above), but it also prevents the clumping of the vertically aligned CNFs when immersed in physiologic solutions. Additional steps required for successful PC12 cell neural network growth include coating the PPy-treated CNF array with a thin (~3 nm) layer of type IV collagen to promote PC12 cell adhesion, and addition of nerve growth


41

. Figure 41-5 Schematic of the sample preparation procedure of CNF arrays for PC12 cell culture. CNFs collapse into microbundles when dried after submersion in solution. With a thin conformal coating of PPy, the vertical alignment of the CNFs is preserved in biological solutions. Coating with type IV collagen (an ECM protein) improves adhesion of the PC12 cells. NGF facilitates the extension of neurites to form a neural network that interfaces directly with the CNFs

factor (NGF) to the cell culture medium to promote the formation of mature neurites (over 100 microns in length). > Figure 41-6 illustrates the effect of the PPy coating on the neural network of PC12 cells that grow on the CNF array. > Figure 41-6a shows the PC12 cells grown on CNF array without PPy coating – note the abundance of neural nanofibrils (~100 nm diameter, similar to the diameter of the CNFs) bridging between the neurite branches over the clumped CNFs. These neural nanofibrils are likely a local stress response resulting from adhesion to a stiff substrate (the clumped CNFs); similar nanoscale filopodia have been demonstrated extending from human corneal epithelial cells grown on parallel nanoridges of silican substrates [21]. > Figure 41-6b shows the PC12 cells grown on the CNF array with PPy coating – note the absence of the neural nanofibrils, presumably because of the bending of the individual CNFs supporting the PC12 cells and neurites (which is not possible with the clumped CNFs). > Figure 41-6c is a highmagnification image which illustrates the flexibility

of the individual CNFs, as well as occasional penetration of a CNF through the cell membrane. The flexibility of the CNF array can be modified quite readily by changing the outer diameter of the individual CNFs – a property which is likely to be important for NEIs in different regions of the CNS with differing tissue characteristics (e.g., gray vs. white matter). An additional benefit of coating electrode arrays with an ECP such as PPy (beyond improved electric charge transfer and the anti-clumping of CNFs noted above) is that such ECPs allow the controlled release of drugs that were preloaded in the ECP at the time of nanoarray manufacture [22,23]. Examples of drugs likely to be of benefit in neuromodulation include anti-inflammatory drugs such as dexamethasone, NGF, and various neurotrophic factors (e.g., brain derived neurotrophic factor, BDNF) [22]. By fine-tuning the depth of ECP film deposition on the CNF nanoelectrode array, one can optimize the array’s electrochemical characteristics for the specific purpose (recording vs. stimulation vs. drug release) and tissue location (e.g., gray vs. white matter) [18].

639

640

41


. Figure 41-6 Scanning electron microscopy images of PC12 neurons in a network on a CNF array. (a) CNFs without PPy coating are collapsed into microbundles, and the PC12 neurons demonstrate bridging neural nanofibrils, likely a stress response. (b) PPy-coated CNFs remain vertically aligned, supporting the PC12 neural network so that bridging neural nanofibrils are not seen. (c) PPy-coated CNFs are sufficiently flexible to bend with the weight of the cell body

Micro- and Nano-level Neurotransmitter Monitoring/ Modulating: CNS Electrochemistry Neurons in the CNS communicate with each other through a combination of electrical and chemical means. Brain electrical activity and electrical characteristics such as impedance have been studied with increasing precision over many

decades, but the study of neurotransmitters in vivo is much more recent. Microdialysis has been used to determine neurotransmitter levels in specific regions of the brain, but due to its large size (typically >100 mm) and slow response time (typically >1 min) the microdialysis probe is of limited value for study dynamic brain activity. The most important technique for in vivo neurotransmitter monitoring currently is fast-scan


cyclic voltammetry (FSCV). Although first described about 20 years ago, it was not until a decade later that FSCV was used to follow changes in dopamine levels during behavior [24]. Although not small enough to permit neurotransmitter monitoring within the synaptic cleft (~100 nm), the 5 mm diameter carbon-fiber microelectrodes used for FSCV make very localized monitoring of neurotransmitters, e.g., dopamine, possible. FSCV is illustrated in > Figure 41-7: dopamine is rapidly oxidized to dopamine-oquinone, then reduced back to dopamine by ramping the electrode potential from 0.4 to +1.0 V, and back again, at 300 V/s – typically at 10 times per second (10 Hz) repetition rate. This technique detects subsecond dopamine concentration changes [25]. By changing the

41

parameters to 0.6 to +1.4 V and 450 V/s, sensitivity can be increased 10-fold, but the response time lengthens by >0.5 s. As shown in > Figure 41-7, the cyclic voltammogram for dopamine is found by subtracting the large background charging current from the current obtained when dopamine is present. FSCV can be quite accurate at detecting changes in dopamine levels with subsecond resolution, but the subtraction process makes basal level determination questionable [25]. A recent report documents dopamine release within the nucleus accumbens of the rat in reward-seeking behavior that is spatially and temporally heterogeneous [26]. CNF nanoelectrode arrays can improve significantly on the neurochemistry monitoring with carbon-fiber microelectrodes. An individual

. Figure 41-7 Fast-Scan Cyclic Voltammetry. (a) Electrode potential is scanned from 0.4 to +1.0 V and back every 100 ms at 300 V/s. (b) Dopamine is oxidized to dopamine-o-quinone and then reduced back to dopamine. (c) Black line: large background charging current of the electrode. Red line: small changes in the presence of dopamine. (d) Subtracting the black line form the red line in (c) produces the cyclic voltammogram “fingerprint’’ for dopamine. (e) The current at the oxidation potential is converted to concentration (using an in vitro calibration value) which is plotted versus time to monitor dopamine concentration changes after brief electrical stimulation (4 pulses at 100 Hz, indicated by hash marks) Reprinted with permission from Analytical Chemistry 2003, 75, 414A–417A. Copyright 2003 American Chemical Society

641

642

41


CNF tip is Figure 41-8 demonstrates the measurements of the electrochemically active neurotransmitter dopamine in phosphate buffered saline with an inlaid CNF nanoelectrode array. The detection limit can readily reach down to 60 nM. With further optimization, this technique may provide the capability to detect dopamine at the 10 nM level (Nguyen-Vu TD, Mandikian D, Cassell AM, Andrews RJ, Meyyappan M, Kawagoi K, and Li J – unpublished results 2008).

Nanoelectrode Arrays for Neuromonitoring and Neuromodulation As described above, and detailed in the references in the previous section, nanoelectrode arrays are already a laboratory research tool for monitoring and modulating both neurotransmitter and electrical

. Figure 41-8 Dopamine detection (in phosphate buffered saline solution) by FSCV using CNF nanoarrays. (a) A small dopamine peak is detectable at 64 nM. (b) Calibration curve of measured signal versus dopamine concentration over 5 orders of magnitude (13 nM–1.0 mM)

activity in neuronal systems. Much work remains, however, on optimizing the fabrication details, miniaturization of the connectors between the nanoarrays and the external recording/stimulating device, and biocompatibility issues. Although trials of nanoelectrodes arrays in small animal models (e.g., rodent models of epilepsy or Parkinson’s disease) may be expected within the next year or two, it will likely be 5 years or more before nanoelectrode arrays are in clinical practice in humans. An important step will be comparing nanoelectrode arrays with the current ‘‘gold standard’’ DBS macroelectrode array in a large animal model (e.g., a primate model of Parkinson’s disease). > Figure 41-9 illustrates a configuration of CNF


41

. Figure 41-9 Macro-size nanoarray for DBS (comparison with macroelectrode). (a) Schematic of 1.5 mm diameter electrode. (b) 3 3 CNF arrays, each with independent lead. (c) Uninsulated CNF array for electrical stimulation/recording (large contact area). (d) Insulated (e.g., silicon dioxide) CNF array for electrochemical (neurotransmitter) recording. Scale bars: (b) 200 mm; (c) 1 mm; (d) 2 mm

nanoarrays for both electrical and chemical (neurotransmitter) recording and stimulating that can be compared with the current DBS macroelectrode. The CNF nanoarray would be much larger than necessary, but would allow for physical similarity with the current DBS macroelectrode: the left hemisphere might be implanted with the standard DBS macroelectrode, the right with the CNF nanoarray (of similar size/configuration). Such a ‘‘head-tohead’’ comparison would not take advantage of most of the benefits of the CNF nanoelectrode array (noted above), but would permit baseline comparisons of both baseline recording sensitivity and energy requirements for stimulation.

References 1. Butson CR, Maks CB, McIntyre CC. Sources and effects of electrode impedance during deep brain stimulation. Clin Neurophysiol 2006;117:447-54.

2. Moss J, Ryder T, Aziz TZ, Graeber MB, Bain PG. Electron microscopy of tissue adherent to explanted electrodes in dystoniz and Parkinson’s disease. Brain 2004;127:2755-63. 3. Hoeber R. Eine methode die elektrische leitfaehigkeit im inner von zellen zu messen. Arch Ges Physiol 1910;133:237-59. 4. Van Harreveld A, Ochs S. Cerebral impedance changes after circulatory arrest. Am J Physiol 1956;187:180-92. 5. Schwan HP. The practical success of impedance techniquest from an historical perspective. In: Riu PJ, Rosell J, Bragos R, Casa O, editors. Electrical bioimpedance methods: applications to medicine and biotechnology. Ann NY Acad Sci 1999;873:1-12. 6. Grant FC. Localization of brain tumors by determination of the electrical resistance of the growth. JAMA 1923;81:2169-71. 7. Organ LW, Tasker RR, Moody NF. Brain tumor localization using an electrical impedance technique. J Neurosurg 1968;28:35-44. 8. Bullard DE, Makachinas TT. Measurement of tissue impedance in conjunction with computed tomographyguided stereotaxic biopsies. J Neurol Neurosurg Psychiatry 1987;50:397-401. 9. Rajshekhar V. Continuous impedance monitoring during CT-guided stereotactic surgery: relative value in cystic and solid lesions. Br J Neurosurg 1992;6:439-44.

643

644

41


10. Gildenberg PL, Zanes C, Flitter M, et al. Impedance measuring device for detection of penetration of the spinal cord in anterior percutaneous cervical cordotomy. J Neurosurg 1969;30:87-92. 11. Taren JA, Davis R, Crosby EC. Target physiologic corroboration in stereotaxic cervical cordotomy. J Neurosurg 1969;30:569-84. 12. Bayford RH. Bioimpedance tomography (electrical impedance tomography). Ann Rev Biomed Eng 2006;8:63-91. 13. McEwan A, Cusick G, Holder DS. A review of errors in multi-frequency EIT instrumentation. Physiol Meas 2007;28:S197-215. 14. Romsauerova A, McEwan A, Horesh L, Yerworth R, Bayford RH, Holder DS. Multi-frequency electrical impedance tomography (EIT) of the adult head: initial findings in brain tumours, arteriovenous malformations and chronic stroke, development of an analysis method and calibration. Physiol Meas 2006;27:S147-61. 15. Oh SH, Lee BI, Woo EJ, et al. Electrical conductivity images of biological tissue phantoms in MREIT. Physiol Meas 2005;26:S279-88. 16. McIntyre CC, Miocinovic S, Butson CR. Computational analysis of deep brain stimulation. Expert Rev Med Devices 2007;4:615-22. 17. Nguyen-Vu TD, Chen H, Cassell AM, Andrews R, Meyyappan M, Li J. Vertically aligned carbon nanofiber arrays: an advance toward electrical-neural interfaces. Small 2006;2:89-94. 18. Nguyen-Vu TD, Chen H, Cassell AM, Andrews RJ, Meyyappan M, Li J. Vertically aligned carbon nanofiber architecture as a multifunctional 3-D neural electrical interface. IEEE Trans Biomed Eng 2007;54:1121-9. 19. Wang K, Fishman HA, Dai H, Harris JS. Neural stimulation with a carbon nanotube microelectrode array. Nano Lett 2006;6:2043-8.

20. Llinas RR, Walton KD, Nakao M. Hunter I, Anquetil PA. Neuro - vascular central neruous recording/stimulating system: Using nonotechnology probes. J Nanopart Res 2005;7:111-127. 21. Karuri NW, Liliensiek S, Teixeira AI. Biological length scale topography enhances cell-substratum adhesion of human corneal epithelial cells. J Cell Sci 2004;117:3153-64. 22. Wadhwa R, Lagenaur CF, Cui XT. Electrochemically controlled release of dexamethasone from conducting polymer polypyrrole coated electrode. J Control Release 2005;110:531-41. 23. Abidian M, Kim DH, Martin DC. Conducting polymer nanotubes for controlled drug release. Adv Mater 2006;18:405-9. 24. Rebec GV, Christensen JR, Guerra C, Bardo MT. Regional and temporal differences in real-time dopamine efflux in the nucleus accumbens during free-choice novelty. Brain Res 1997;776:61-7. 25. Venton BJ, Wightman RM. Psychoanalytical electrochemistry: dopamine and behavior. Anal Chem 2003;75:414A-421A. 26. Wightman RM, Heien MLAV, Wassum KM, et al. Dopamine release is heterogeneous within microenvironments of the rat nucleus accumbens. Eur J Neurosci 2007;26:2046-54. 27. Li J, Ng HT, Cassell A, et al. Carbon nanotube nanoelectrode array for ultrasensitive DNA detection. Nano Lett 2003;3:597-602. 28. Li J, Koehne JE, Cassell AM, et al. Inlaid multi-walled carbon nanotube nanoelectrode arrays for electroanalysis. Electroanalysis 2005;17:15-27. 29. Koehne JE, Chen H, Cassell AM, et al. Miniaturized multiplex label-free electronic chip for rapid nucleic acid analysis based on carbon nanotube nanoelectrode arrays. Clin Chem 2004;50:1886-93.

54 Intraoperative Image Guidance in Skull Base Tumors D. Omahen . F. Doglietto . D. Mukherjee . F. Gentili

Introduction With the advent of modern brain imaging techniques, neurosurgeons came into possession of powerful diagnostic tools. Intraoperative image guidance systems (IGS) attempt to apply this technology to therapeutic surgical interventions, creating the field of image-guided therapy (IGT). It is hoped that widespread utilization of these techniques will translate into improved patient outcomes. Examples of ways in which preoperative or realtime image guidance can aid the surgeon abound. Image guidance systems can help to optimally place skin incisions and bone flaps, and can help minimize their size. Their ability to precisely localize vital anatomic structures can help to safely guide the approach to surgical targets. The extent of lesions can also be gauged, which is invaluable in the resection of lesions with grossly indistinct borders. Image guidance systems fall into two broad categories: those that use images acquired preoperatively, and those in which imaging is updated during the procedure. Each system has its own advantages and disadvantages. They may utilize many different imaging modalities including fluoroscopy, ultrasound, computerized tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), angiography, single photon emission CT (SPECT) [1,2]. Images obtained using these varied techniques may be fused together, taking advantage of the slightly different information captured by these disparate methods. Additionally, some systems allow these images to be superimposed upon real time images obtained with the operating microscope or endoscope [3]. #


In this chapter we will survey common methods in current use, stressing their advantages and disadvantages. An overview of specific applications in skull base surgery will then be provided. Technology is changing at a rapid pace, and it is anticipated that many new advances and improvements will be forthcoming in short order.

Methods Based on Preoperative Imaging At the most basic level, the majority of modern neurosurgery is guided by some form of imaging. Preoperative imaging is studied by the neurosurgeon and the information gleaned is used in formulating a surgical plan. An advantage of using preoperative imaging is that optimal image quality is possible under controlled conditions [1]. A giant leap forward was provided by the advent of stereotaxy. The Greek words for ‘‘three dimensional’’ (stereo) and ‘‘arrangement’’ (taxis) were combined to create the term ‘‘stereotaxic’’. Use of the Latin term tactus (‘‘to touch’’) gave rise to the synonymous term stereotactic [4]. The Russian anatomist Zernov [5] was the first to use a rudimentary frame, but it was the team of Horsley and Clarke [6] who created the first true stereotactic system designed to create precise lesions in the cerebellum of monkeys. The first stereotactic surgery performed on human beings was reported by Spiegel and Wycis [7], who used X rays and ventriculograms to image the brain. The use of imaging as opposed

816

54

Intraoperative image guidance in skull base tumors

to atlas-derived coordinates ushered in the era of image-guided surgery. Localization procedures fall into two broad categories [4]. The first is a linked system, in which angle-sensing devices are employed to determine the position of a point localized by a mechanical linkage arm. This is the method used by frame-based systems. The second method, that of unlinked systems, uses devices such as cameras, magnetic sensors, or microphones which receive positional information from a pointer, which is not contiguous with the pointing device.

Frame-based, or Linked Systems Frame-based systems utilize a mechanical arm, which is affixed to the patient’s head prior to the procedure. Most arms are moved passively by the surgeon, although robotic systems with active control of arm movement have also been devised [8]. This process involves the localization of target tissue and anatomical structures relative to each other in space utilizing imaging technology. Pioneering stereotactic techniques utilized frames, which were rigidly fixed to the patient’s head prior to imaging. From the images thus obtained, the position of a target relative to the frame could be fixed in space using a Cartesian or polar coordinate system [4]. Since the relationship between the head and frame was rigidly maintained, this allowed for vary precise spatial accuracy, on the order of 1–3 mm of error [9,10]. Such systems are still widely used for performing biopsies, for electrode insertion or lesion localization in functional neurosurgical procedures, and in targeting for radiosurgical procedures. Examples of well-known stereotactic frames which use interlocking arcs, include the Brown-RobertsWells (BRW) frame [11], and the target-centered Cosman-Roberts-Wells (CRW) frame [4]. Advantages of frame-based systems include a proven track record, reliability and documented accuracy.

Disadvantages include the fact that frame application can be uncomfortable and time consuming [4]. Since the frame covers most of the patient’s skull and the frame must be left in place to be used, applications in skull base surgery are limited.

Frameless Stereotaxy In an effort to overcome the limitations of framebased stereotactic systems, frameless approaches have been devised. They are categorized as unlinked systems, since position sensing devices are not in physical continuity with pointing devices. They are easier to use, provide full access to the patient’s head, and more comfortable for the patient. Studies have shown a similar degree of reliability and spatial accuracy as frame-based systems [12,13], especially using a ‘‘probe’s eye view’’ for targeting [12] (> Figure 54-2). The basic principles of frameless stereotaxy are based on those of frame-based stereotaxy. Simplistically, an image of the patient is taken, sometimes with fiducial markers in place to define set points. The patient’s head is placed in a pin-based surgical head holder, which is affixed to the operating table. A reference array, which defines points on a spatial plane in physical space, is used. It is paramount that this array is held in a fixed spatial relationship to the patient’s head. Newer systems do not necessitate a rigid head fixation and are based on systems which are able to keep a fixed spatial relationship with the head even when this is moved. Several methods have been employed for localization. Some groups have experimented with magnetic field sensor-based localization [14] and initial attempts were abandoned due to problems with magnetic interference; newer systems are now becoming available with markedly increased results [15]: theoretically problems hich are present in optical based system, such as the line-of-sight problem and missing


tracking of the tip of flexible instruments, should be solved by the new magnetic field sensor-based tracking systems [15]. An early system used a spark gap-generated ultrasonic emission, which was used by an array of detecting microphones to triangulate the position of the emitter [16]. This method suffered from susceptibility to echoes and non-linear variations with changes in temperature [4], and has largely been supplanted by optical methods of localization. Most optical systems use two infrared cameras, which are set a fixed distance apart and used to triangulate the distance to markers on the reference array, allowing the position of the array relative to the cameras to be calculated by the computer. To use IGS, two main processes must be carried out: Segmentation

The first step is known as segmentation [17]. In this step the tissue of anatomic or pathological interest is delineated and selected. This is often done on a computer workstation prior to initia-

54

tion of the procedure (> Figure 54-1). This step is not essential if the lesion is easily recognizable in the imaging that is used and the simple images in the three planes (sagittal, axial, coronal) can be used (> Figure 54-2). Registration

A registration process must be performed to relate coordinates in physical space to coordinates in the virtual image space [4] (i.e., matching points on the patient with corresponding points in the imaging data set(s)) [18]. There are two main methods of carrying this out. The first, known as point-based transformation, uses intrinsic anatomic landmarks, or extrinsically applied fiducial markers (> Figure 54-1). The positions of the markers are identified on the preoperative imaging. The pointing device is then used to identify these points on the patient using the infraredsensing camera system. Calculations show that there is an inverse relationship between the targeting error and the square root of the number of fiducials. Thus, to double the accuracy of such

. Figure 54-1 Frameless stereotactic optic navigation system. (a) The computer work station and video monitor are visible and are together with the optic reading system (arrow). (b) The head is rigidly fixed. The reference frame (encircled in blue) is attached to the head pin fixation system. The fiducials are visible on the patient’s forehead (StealthStation TREON – Medtronic)

817

818

54


. Figure 54-2 Neuronavigation and repeat surgery. Neuronavigation during an extended transsphenoidal approach for a recurrent suprasellar craniopharyngioma (segmentation has not been performed as the lesion is easily recognized in the MRI). In this case neuronavigation was essential to determine the midline and optimize the sellar and suprasellar opening in a patient with an intercarotid distance in the suprasellar area of only 10 mm (coronal post-contrast MRI in the upper left corner). In the lower right corner the so called ‘‘trajectory view’’ shows to the surgeon the direction of the probe, which is depicted in blue. The coronal, sagittal and axial images show where the tip of the pointer has been positioned by the surgeon (at the level of the sella, in the midline) (StealthStation TREON – Medtronic)

a system, four times as many fiducial points must be utilized [4]. A second method, known as surface matching, uses points obtained by scanning over the patient’s facial features. A surface rendering of the patient’s face and scalp is created from the imaging data set by the computer. The set of points acquired during scanning is then fitted to the corresponding set of points obtained from the image data surface rendering using a least-squares-based transformation [4]. The location of the surgical target can then be identified from the area of interest selected from the imaging set (via the process of segmentation). The position of the target can be measured relative to the fiducial points and/or patient’s fixed

anatomical points identified on the imaging set. Since the target position is known relative to the fiducial points (on the imaging data set), and the location of the fiducial points are known relative to the reference array (in physical space), it is a straightforward calculation to determine the position of the target relative to the reference frame. Basic Components of a Frameless IGS

There are currently many commercial frameless tereotaxy systems available, however they share in common the same basic components (> Figure 54-1). Computer work station and video monitor

Most systems employ a trolley-mounted com-


54

Reference frame An array containing a variable

puter system onto which the image set is loaded prior to the procedure. In the process of segmentation anatomic areas with pathological or physiological importance are selected [1]. After the registration process, information from the camera system is compared with information contained in the image data set, and data is displayed interactively during the course of the procedure.

number of fixed points marked by reflective spheres is used to create a reference plane. Mathematically, three points are needed to define a plane, but most systems use addition points to increase accuracy and reliability. The position of this reference array is maintained in a fixed spatial relationship to the patient’s head, usually by attaching it to the Mayfield or Sugita head clamp.

Several different localization systems have been devised, as outlined above. At present, most systems use optical methods. An infrared beam is produced by light-emitting diodes (LEDs). After it is emitted, it bounces off reflective markers and is picked up by two infrared-sensing CCD cameras located a fixed distance from each other. An alternative method is to place LEDs on the reference frames and probes [19]. Infrared radiation emitted from the probe and reference frame is then picked up by the cameras and used to triangulate their positions in physical space.

A variety of pointing devices are available. Commonly, a blunt tipped pointer with an attached array of reflective spheres is used. The array can be detected by the camera and is located a predefined distance from the tip of the pointer. Hence, the position of the pointer is known relative to its array. Additionally, most systems are equipped with mobile arrays, which can be attached to surgical instruments. Through a registration process in which the position of the instrument’s tip is measured in relation to the attached array, almost any instrument can be used as a pointing device.

Camera system

Many systems use small markers, which are affixed to the skin of the patient’s face and scalp prior to imaging. These serve to provide readily identifiable points to aid in the registration process. Fiducials, which screw into bone, have also been developed to diminish the possibility of them changing position in the time between imaging and registration; they are used much less frequently, due to their invasiveness.

Pointing device

Skin fiducials

Infrared radiation emitted from the localization system bounces off reflective markers placed on components of the localizing system and is detected by the camera system. The markers are usually passive spherical glass beads impregnated with aluminium [4]. These beads can be gas sterilized, and can be fitted into sterile adaptors on pointing devices and reference arrays.

Reflective markers

Sources of Registration and Navigation Errors There are a multitude of potential errors that can degrade navigational accuracy [4]. Scanner slice thickness is one variable that can be manipulated. Voxel size places a limit on the ultimate accuracy obtainable, as localization is only as good as the data set used. The position of skin fiducials can shift considerable, especially when placed on mobile areas of skin. Bone-anchored fiducials help overcome this problem, but are invasive [20]. Data sets from CT scans may suffer from error due to shear, which can be compensated for if the gantry angle is known. MRI scans suffer from image distortions due to magnetic field inhomogeneity. Several methods designed to correct for such problems exist, and aim to create a more accurate image known as the

819

820

54


rectified image. Unfortunately, automated registration algorithms can find a transformation solution which meets criteria for a global acceptable fit, but which has wide degrees of local mismatch error. Another potential source of error is inadvertent movement of the patient’s head relative to the reference array.

Real-time Intraoperative Imaging A major disadvantage of using imaging obtained preoperatively is that the position of intracranial structures can shift during the course of an operation [21]. Removing a bone flap, draining cerebrospinal fluid, administering mannitol, and manipulating or resecting tissue can change the relationship between cranial structures and reference points, resulting in erroneous localization. Even prior to tumor resection, surface deformation greater than one centimeter has been documented after dural opening in over half of patients [22]. Fortunately for skull base surgeons, the rigid bony base of the skull is less susceptible to this problem of brain shift [23]. In an attempt to overcome this limitation, methods of updating the imaging set during the course of the procedure have been devised, producing real-time or near real-time images. Intraoperative imaging provides not only surgical guidance, but also an ability to monitor the progress of an operation through repeated updating of images. The goal is 3D imaging in real time — a concept which has been labeled ‘‘4D-imaging’’ [1]. Unlike diagnostic imaging, where a greater emphasis is placed on specificity, for therapeutic applications, sensitivity is of primary importance [1]. Although intraoperative MRI has been the most widely publicized, several other imaging modalities have also been used. The main advantage of intraoperative imaging is the ability to update images in essentially real time. Since tissue shift and deformation of anatomic and pathologic structures during the

course of surgery is inevitable, this poses clear advantages for the neurosurgeon. This also allows for early detection and treatment of operative complications, such as hemorrhage, or unsuspected residual tumor [1]. The disadvantages of this approach are mainly technical in nature, related to adapting imaging technology to the constraints imposed by the operating room environment. Ready availability of technical support and input from experienced radiologists is also required.

Ultrasound The use of intraoperative ultrasound in neurosurgery is not recent [24,25]. Its ability to differentiate tissues of different densities has long been recognized. It has been used for tumor biopsy, catheter placement, and cyst aspiration [26]. Experience is required to interpret ultrasound images, and blood and air must be scrupulously cleared from the operative site to improve image quality [27]. Structures such as the falx or ventricles aid in image interpretation. Positioning the patient in such a way that tumor resection cavities can be filled with water aids in visualization. Ultrasound provides an excellent method of localizing major vascular structures [28]. Recently, systems which can update preoperative MRI data sets using intraoperative ultrasound, have been developed [27,29]. An ultrasound probe with a range of frequencies from 5 to 7.5 MHz allows tissue penetration to 120 mm. This allows correction for tissue shift and distortion, which ranges up to 1.5 cm. A recent study has documented that adequate fusion with MRI data can be obtained in 95% of cases [29]. Imaging proved better with metastases and meningiomas compared to gliomas [29]. As visualized with ultrasound, high-grade tumors appeared larger than expected based on CT/MRI imaging [27]. Difficulty differentiating tumor mass from edematous brain appears to be


the cause. In half of cases, the size of the craniotomy flap limited visualization. Over 15% of the time additional resection was carried out as a result of ultrasound imaging [29]. Error as low as 2 mm has been described [30].

Intraoperative MRI One of the most eagerly anticipated advances in image-guided surgery was the development of intraoperative magnetic resonance imaging (iMRI) in the early 1990s [31]. At present, the field has yet to reach its full potential, yet its role in providing immediate quality control appeals to many [32]. The main application to date has been in glioma surgery, and more complete resection has been reported as a result, but an improvement in clinical outcomes has yet to be definitively proven [1,33]. One group has reported that unsuspected residual tumor was discovered in approximately one third of cases in which iMRI was utilized [34]. Others report numbers as high as 57% [35]. Studies on low and high grade gliomas seem to indicate potential for a greater degree of resection using iMRI [35]: this may be dependent upon advances in imaging that allow better differentiation of the interface between invasive glioma tissue and normal brain. Working in the MR environment requires physical adaptation of the operating room. Installation, shielding and maintenance can be costly. Use of these devices prohibits the presence of ferromagnetic instruments in the vicinity due to the magnetic fields utilized. Specially designed head coils must also be used, specific for the iMRI system being used [35]. Using iMRI also increases the length of surgical procedures, although a deleterious effect on patient outcome has not been demonstrated [36]. Design of iMRI requires a tradeoff between field strength and image quality on one hand, and convenience, cost, and the surgeon’s access to the patient on the other hand [1,37].

54

Three main categories of iMRI exist: ‘‘Double donut’’ Configuration

General Electric introduced one of the first designs, the so-called ‘‘double donut’’ configuration, in 1991. This device used two magnets with a 54 cm gap between them to allow surgical access to the patient. The magnetic fields of the two magnets overlap and work in concert, producing an overall field strength of 0.5 Tesla. Advantages of this type of design include almost continual patient access by the surgeon and assistant, and the ability to integrate an operating microscope [1]. Surgeon and patient positioning options are physically limited by this design, but the patient does not leave the imaging space during the procedure facilitating frequent imaging updates [1]. An optical 3D frameless stereotactic system has been designed to be used in the bore of this iMRI [38]. In addition, since the patient remains in the imaging field, the need to reregister guidance systems after each scan is obviated. The design does limit field strength, and suffers from increased magnetic field inhomogeneites [39]. Biplanar Magnet Design

A second iMRI design paradigm, known as biplanar design, allows virtually unlimited access to the patient at the expense of extra time and effort required to update imaging. In this type of iMRI, mobile horizontally or vertically oriented magnets can be moved in or out of the operative field. When the magnet is out of the operative field surgeon access, comfort and freedom of movement is maximized. The cost of such freedom of movement is a relatively lower field strength [39]. Cylindrical Superconducting Magnets

A third approach uses cylindrical superconducting magnets in an effort to maximize field strength and homogeneity [39]. The patient and/or magnet is moved in and out of the field when imaging is required. There are fewer restrictions on ferromagnetic instruments used for the surgery, as long as they are removed prior

821

822

54


to imaging. When not in use for an operation, some designs allow for use of the iMRI unit for diagnostic purposes by movement along a ceiling track-mounted magnet to an imaging suite adjacent to the operating room [40]. The disadvantage is that each time updated imaging is required the magnet must be moved into position, causing delays in the surgery. Complex local RF shielding, such as covering the patient with a copper-impregnated Plexiglas tent may be required [40]. During scanning access to the patient is quite limited [39]. Recent work with higher field magnets has focused on the application of diffusion tensor imaging, MR spectroscopy, MR angiography, and functional MRI to aid in operative neurosurgical procedures [1,32]. Magnets with field strengths up to 3 Tesla are now in use [41]. At the moment, the main drawback to widespread use of higher field iMRI technology is cost [1]. As with frameless stereotactic systems, the position of instruments or pointers in space can be determined, but with real-time updating of tissue deformation [1]. Real-time 3-dimensional reformatting capability is provided by technologies such as the 3D Slicer [42]. Methods of compensating for image degradation caused by radiofrequency pulses from bipolar cautery or surgeon hand motion have been developed [1]. Several authors have reported on the use of temperature-sensitive iMRI protocols (e.g., chemical shift, diffusion imaging) to guide the creation of lesions by thermal ablation, focused ultrasound treatment, cryoablation, or interstitial laser therapy [1]. The temporal resolution of iMRI is currently around one second with contemporary units [1]. The ability to fuse MRI images with other modalities such as CT, PET, SPECT, and MEG is also under intense investigation, but generally requires field strengths of 1 Tesla or higher. Field strength up to 3 T may be required for optimal neurovascular imaging [1]. One day routine use of diffusion imaging may

aid in the early detection of ischemic changes during the course of the operation [1]. Fusion with high field MRI images obtained preoperatively is one way to augment the inferior image quality of low field iMRI without incurring the higher financial costs of upgrading to more powerful magnets. Algorithms are in development, which allow the higher quality preoperative images to be warped to match the tissue deformation resulting from surgical manipulations. This process is known as ‘‘single modality image augmented fusion’’ [1]. Combination of iMRI and surgical endoscopy to guide minimally invasive surgical approaches has sparked some interest as of late. Work on optimal head coil design for iMRI is also in progress [32]. Finally, further integration with new robotic designs may one day allow for real-time image-guided robotic assisted surgery [1].

Fluoroscopy The utility of intraoperative fluoroscopic guidance for spinal surgery applications such as pedicle screw insertion and kyphoplasty is well-documented [19,43–45]. As described below, this was a standard intraoperative imaging technique for pituitary surgery for many years [46]. Intraoperative angiography has expanded the role of fluoroscopy in the operating room. Disadvantages of fluoroscopy include being limited to viewing in a single plane, and radiation exposure [19]. The combination of C-arm fluoroscopy and image guidance has been referred to as ‘‘virtual fluoroscopy’’ [19]. This allows for updating of preoperatively obtained image sets and reduction of radiation exposure.

CT Scanning Limitations of CT-based intraoperative imaging include exposure to ionizing radiation, poor multiplanar imaging capability, and poor tissue detail and resolution compared with MRI [1].


Advantages include superior bony visualization and affordability, portability, and ‘‘near real time’’ imaging [47]. Interest in this technique have led to the development of nascent spinal and CT angiographic [48] intraoperative techniques. Using a mobile cone beam CT unit, Rafferty et al. [47] were able to achieve sub-millimeter spatial resolution with about one tenth the dose of ionizing radiation of a conventional CT scan. Cone beam CT scanners differ from conventional scanners in that they acquire all needed images in a single rotation or less about the patient, without translation of the patient [47]. Three-dimensional reconstruction is performed using a back projection modified Feldkamp algorithm. Imaging radiation dose ranges from 0.5 to 3 mGy [49].

Angiography Intraoperative angiography has been utilized as an adjunct to vascular neurosurgery, allowing assessment of aneurysm or AVM obliteration and vessel patency [50,51]. Using road-mapping techniques, intraoperative angiograms can be used for lesion localization [52].

Application of Image Guidance to Skull Base Surgery Skull base surgeons have been quick to adopt image guidance methods. The bony structures of the skull base do not suffer the same degree of shift and deformation that brain and other soft tissues do. Thus, the drawbacks of guidance systems that use static preoperative imaging are minimized. Image guidance systems provide important information regarding the position and extent of lesions, as well as about the location of vital vascular and nervous structures.

54

Sellar Region and Anterior Skull Base An early application of image guidance to skull base procedures was the adoption of fluoroscopy to aid in transsphenoidal pituitary surgery. Guiot was a pioneer in this application [23]; Jules Hardy [53] introduced its use in North America. Its modern application has been recently described by Jane et al. [72]. Imaging is limited to a single plane at one time, however. Experience with intraoperative CT scanning has been described [54], but not widely adopted. Similarly, ultrasound applications to pituitary surgery have also been limited [18,23,28]: ecographic probes are still too big or have low specificity for application in transsphenoidal surgery; Atkinson et al. [55] described a method to ultrasonically monitor pituitary surgery via a burr hole placed at the coronal suture. On the other hand, doppler ultrasound is an excellent way to locate vascular structures [28], such as the internal carotid artery in the cavernous sinus (> Figure 54-5). The introduction of the pure endoscopic transsphenoidal approach has led to a decrease in the use of fluoroscopy, due mostly to the wider view and the use of a natural pathway, which allows a prompt recognition of anatomical landmarks during surgery (choana, ostium sphenoidale, sella, clivus, carotid protuberances, optic-carotid recess). Some Authors actually recommend the use of the endoscope, in microscopic surgery, as an ‘‘intraoperative navigator’’ due to its wider visualization. Anatomical landmarks are not though always easily recognized, as in the case of a conchal sphenoid sinus: the incomplete pneumatization does not allow the recognition of anatomical landmarks, such as the clivus, the carotid prominence and the optic carotid recess; anatomical landmarks might not be recognizable in repeat surgery [23] (> Figure 54-2): a navigation system is mandatory in these cases.

823

824

54


The use of preoperative images to guide pituitary surgery has indeed gained widespread popularity, even as an adjunct to endoscopic approaches [56]. During the approach to the sellar region, landmarks along the bony skull base are of interest to the surgeon, and this is best imaged using CT scanning. Once the sella is reached however, soft tissue detail provided by MRI imaging has greater utility. Early work on CT–MRI data set fusion shows this is a safe and effective method to maximize navigational safety [20] (> Figure 54-4). The utility of intraoperative stereotacic navigation is also obvious in complex approaches, as in extended endonasal approaches to the suprasellar area (> Figure 54-2), to the anterior cranial fossa (> Figure 54-3) and to the cavernous sinus (> Figure 54-5), as it provides assurance to the

surgeon on the complex anatomy which needs to be mastered for these approaches. iMRI has met with much interest in anterior skull base surgery, as it may have the capability to identify residual tumor not initially identified by the surgeon. Identification of bony landmarks is difficult [23]. Darakchiev et al. [57] provide a comprehensive review of this topic and elaborate on details of their own iMRI setup for pituitary surgery . Others have suggested that MR-angiography may aid in localizing vascular structures such as the carotid or sphenopalatine arteries [20]. Operating in the pre-endoscopic era, Walker and Black have reported that in 7 of 19 patients with macroadenomas additional unsuspected, but ultimately resectable tumor was identified [46]. SPGR (spoiled gradient recalled sequence) and sagittal T1-weighted images were used for

. Figure 54-3 Neuronavigation and extended endoscopic anterior skull base surgery. Extended transsphenoidal approach for an olfactory groove meningioma. The neuronavigation probe is positioned over the anterior skull base, after an ethmoidectomy has been performed, to confirm the exposure of the right anterolateral margin of the olfactory groove meningioma. The three MRI images document the position of the probe in respect to the meningioma; the endoscopic intraoperative picture of the probe position (lower right quadrant) can be transferred to the neuronavigation system and recorded (iNtellect Cranial Navigation System – Stryker)


54

. Figure 54-4 CT–MRI fusion technique. Endoscopic approach for a giant pituitary adenoma. Intraoperative determination of the medial optic carotid recess before the dural opening: both CT (a) and MRI (b) reconstructions can be used for intraoperative image guidance, as well as fusion images that incorporate variable percentages of data images from the two image sets. The position of the probe (asterisk) over the medial OCR is confirmed, both on CT (a) and MRI (b) images (iNtellect Cranial Navigation System – Stryker)

. Figure 54-5 Stereotactic neuronavigation and doppler in cavernous sinus surgery. Endoscopic transsphenoidal surgery for a pituitary adenoma extending in the cavernous sinus. (a) Neuronavigation is used to confirm the position of the probe inside the right cavernous sinus, at the level of the internal carotid artery (ICA – visible in all its intracavernous portion in the sagittal post-contrast MRI reconstructions – upper right quadrant). (iNtellect Cranial Navigation System – Stryker). (b) After further pituitary adenoma removal, a doppler ultrasound probe (asterisk) is positioned over the posterior portion of the right ICA to confirm its exposure in the cavernous sinus. A suction tube is positioned over the anterior portion of the cavernous ICA

navigation purposes. A SPGR sequence takes about 1.5 min to perform [23]. They describe a protocol for intraoperative dynamic contrast imaging to take advantage of the delayed contrast

uptake exhibited by pituitary adenomas. The use of specialized heme-sensitive gradient echo [46] sequences have made differentiation between blood and residual tumor easier [23,58].

825

826

54


Contrast-enhanced scans are less useful than initially anticipated due to blood pooling in the sella [24], and enhancing areas created by electrocautery and surgical manipulation [35,59]. More recently, an iMRI ‘‘road mapping’’ technique to augment endoscopic sinus surgery has been described [60]. In general, it is felt that field strength of at least 0.2 T is required for pituitary surgery [37]. This allows for detecting tumor residual, hemorrhage, and cavernous sinus invasion. A group from the University of Cincinnati has reported on a series of 115 patients undergoing resection of pituitary macroadenomas using 0.3 T iMRI [57]. They utilized 3 mm coronal slices (TR = 450 ms, TE = 20 ms). At the very end of the procedure, these images are followed by standard dose postcontrast scans. iMRI revealed unsuspected, but ultimately resectable tumor in 56% of patients, after what was initially felt to be gross total resection. Intraoperative images were felt to be of superior quality to postoperative images due to the absence of artifact from fat grafting and sellar reconstruction. Operating on acromegalic patients in a 1.5 T magnet, Fahlbusch et al. [61], reported that iMRI increased the rate of total tumor resection and normalization of postoperative serum growth hormone from 33% to 44%, with another 17% experiencing ‘‘near-normalization’’. Not all reports of iMRI for sellar lesions are favorable, however. Nimsky et al. [62] reported disappointing results in a series of 21 procedures for craniopharyngioma using a 0.2 T iMRI unit. Complete resection, as indicated by iMRI did not preclude tumor recurrence. This probably stresses the importance of the biological behavior of the tumor and the intrinsic limitations of surgery.

Surgery of the Temporal Bone CT scanning may be the ideal technique to image the temporal bone intraoperatiavely [47]. Studies of the use of cone-beam CT in temporal bone surgery have demonstrated sub-millimeter accuracy with low radiation dose, as described

above. Special utility was found in demonstrating the degree of bony coverage during skeletonization of the facial nerve [47]. Stereotactic neuronavigation has proved useful in lateral skull base surgery, providing greater assurance in avoiding neurovascular damage during complex skull base procedures [63,64].

Posterior Fossa Surgery Recent papers have proved the utility of stereotactic neuronavigation in the retrosigmoid approach (> Figure 54-6): classic anatomical landmarks (for example: the asterion for localizing the transverse-sigmoid sinus transition (TST) complex) can be individualized to the patient (the asterion was located from 2 mm medial to 7 mm lateral and from 10 mm inferior to 17 mm superior to the TST, respectively) [65]. A recent retrospective study analyzed the impact of image guidance on complication rates (venous sinus injury, venous air embolism, postoperative morbidity caused by venous air embolism) and operation times for the lateral suboccipital craniotomies performed with the patient in the semi-sitting position: a significant increased speed and safety in lateral suboccipital approaches was documented when stereotactic neuronavigation was used [66]. Doppler sonography is useful in confirming the position of the vertebral artery in the farlateral approach. Recent applications of stereotactic neurosurgery in skull base surgery include also the quantitative evaluation of new surgical approaches: with the aid of neuronavigation quantitative data about surgical exposure, depth and trajectory can be obtained, allowing an objective comparison of different surgical approaches [67].

Future Advances and Applications Technological advances will continue to drive progress in the field of intraoperative image


54

. Figure 54-6 Stereotactic neuronavigation in posterior fossa surgery. Definition of the transverse and sigmoid sinus during a right retrosigmoid approach: (a) The head is fixed and the reference frame (encircled in blue) attached to the Mayfield; the pointer is used to define the margins of the transverse and sigmoid sinus. (b) The superior and inferior margins of the sinus is defined as well as the projection of the tumor on the skin. (c) After bone exposure the position of the sinuses is checked on the bony surface before its drilling

guidance. As computing power increases, more automation, less computational time, and more user-friendly formats should be expected. Digital deformation models [68] will allow updating of preoperative data sets using inexpensive, user friendly technology [27,29]. Improvements in coil design and field strength will make iMRI a more powerful technique. In an attempt to overcome problems with non tumor-specific intraoperative contrast enhancement, work is being done on monocrystalline iron oxide nanoparticles (MIONs), which appear to have selective uptake by malignant glioma cells [35]. Coupling of intraoperative imaging advancements with robotic technology may help bring the promise of robotic surgery to fruition. iMRIguided robotic devices have already been developed [69,70]. Finally, work is already in progress to create stereoscopic 3-dimensional images to facilitate visualization of complex pathological and/or anatomical structures [71]. These might one day be integrated with images from an

endoscope or operating microscope in a ‘‘heads up display’’. In conclusion, image-guided surgery has been applied to skull base surgery since Guiot introduced the fluoroscope in transsphenoidal surgery in 1950s. Since then, tremendous technological advances have provided tools that aid skull base neurosurgeons. Further technological development are awaited in image-guided skull base surgery in the near future, possibly making it even safer and more effective.

References 1. Jolesz FA. Future perspectives for intraoperative MRI. Neurosurg Clin N Am 2005;16:201-13. 2. Peters TM, Henri CJ, Munger P, Takahashi AM, Evans AC, Davey B, Oliverier A. Integration of stereoscopic DSA and 3D MRI for image-guided neurosurgery. Comput Med Imaging Graph 1994;18:289-99. 3. Gleason PL, Kikinis R, Altobelli D, et al. Video registration virtual reality for nonlinkage stereotactic surgery. Stereotact Funct Neurosurg 1994;63:139-43. 4. Bucholz RD LA, Marzouk S. Stereotactic and imageguided techniques. In: Kaye AH, Black PMcLB, editor.

827

828

54 5.

6. 7.

8.

9.

10.

11. 12.

13.

14.

15.

16.

17.

18.

19.


Operative neurosurgery. Toronto: Harcourt Publishers; 2000. p. 99-116. Kandel EI, Schavinsky YV. Stereotaxic apparatus and operations in Russia in the 19th century. J Neurosurg 1972;37:407-11. Horsley V CR. The structure and function of the cerebellum examined by a new method. Brain 1908;31:45-124. Spiegel EA, Wycis HT, Thur C. The stereoencephalotome (model III of our stereotaxic apparatus for operations on the human brain). J Neurosurg 1951;8:452-3. Kwoh Y. Special stereotactic techniques: robotic methods applied to stereotactic surgery. In: Heilbrun MP, editor. Stereotactic neurosurgery. Baltimore: Williams & Wilkins; 1988. p. 219-32. Maciunas RJ, Galloway RL, Jr, Latimer JW. The application accuracy of stereotactic frames. Neurosurgery 1994;35:682-94; discussion 94-5. Galloway RL, Jr, Maciunas RJ, Latimer JW. The accuracies of four stereotactic frame systems: an independent assessment. Biomed Instrum Technol 1991;25: 457-60. Brown RA. A stereotactic head frame for use with CT body scanners. Invest Radiol 1979;14:300-4. Quinones-Hinojosa A, Ware ML, Sanai N, McDermott MW. Assessment of image guided accuracy in a skull model: comparison of frameless stereotaxy techniques vs. frame-based localization. J Neurooncol 2006;76:65-70. Dorward NL, Alberti O, Palmer JD, Kitchen ND, Thomas DG. Accuracy of true frameless stereotaxy: in vivo measurement and laboratory phantom studies. Technical note. J Neurosurg 1999;90:160-8. Kato A, Yoshimine T, Hayakawa T, et al. A frameless, armless navigational system for computer-assisted neurosurgery. Technical note. J Neurosurg 1991;74:845-9. Schichor C, Witte J, Scho¨ller K, Tanner P, Uhl E, Goldbrunner R, Tonn JC. Magnetically guided neuronavigation of flexible instruments in shunt placement, transsphenoidal procedures, and craniotomies. Neurosurgery 2008;63 1 Suppl 1:ONS121-7. Roberts DW FE, Strohbehn JW. The sonic digitizing microscope. In: Maciunas RJ, editor. Interactive imageguided neurosurgery. Chicago: AANS Publications Committee; 1993. p. 179-200. Cline HE, Lorensen WE, Kikinis R, Jolesz F. Threedimensional segmentation of MR images of the head using probability and connectivity. J Comput Assist Tomogr 1990;14:1037-45. Ram Z, Shawker TH, Bradford MH, Doppman JL, Oldfield EH. Intraoperative ultrasound-directed resection of pituitary tumors. J Neurosurg 1995;83: 225-30. Nakagawa H, Kamimura M, Uchiyama S, Takahara K, Itsubo T, Miyasaka T. The accuracy and safety of image-guidance system using intraoperative fluoroscopic images: an in vitro feasibility study. J Clin Neurosci 2003;10:226-30.

20. Sindwani R, Bucholz RD. The next generation of navigational technology. Otolaryngol Clin North Am 2005;38: 551-62. 21. Kelly PJ, Kall BA, Goerss SJ. Results of computed tomography-based computer-assisted stereotactic resection of metastatic intracranial tumors. Neurosurgery 1988;22: 7-17. 22. Hill DL, Maurer CR, Jr, Maciunas RJ, Barwise JA, Fitzpatrick JM, Wang MY. Measurement of intraoperative brain surface deformation under a craniotomy. Neurosurgery 1998;43:514-26; discussion 27-8. 23. Walker DG, Kaye AH. Image guidance and transsphenoidal surgery: past, present and future. J Clin Neurosci 2003;10:289-92. 24. Hwang KP, Lim J, Wendt M, Merkle E, Lewin JS, Duerk JL. Improved device definition in interventional magnetic resonance imaging using a rotated stripes keyhole acquisition. Magn Reson Med 1999;42:554-60. 25. Chandler WF, Knake JE, McGillicuddy JE, Lillehei KO, Silver TM. Intraoperative use of real-time ultrasonography in neurosurgery. J Neurosurg 1982;57:157-63. 26. Sutcliffe JC. The value of intraoperative ultrasound in neurosurgery. Br J Neurosurg 1991;5:169-78. 27. Miller D, Heinze S, Tirakotai W, et al. Is the image guidance of ultrasonography beneficial for neurosurgical routine? Surg Neurol 2007;67:579-87; discussion 87-8. 28. Arita K, Kurisu K, Tominaga A, et al. Trans-sellar color Doppler ultrasonography during transsphenoidal surgery. Neurosurgery 1998;42:81-5; discussion 6. 29. LIndner D, Trantakis TC, Arnold S, Schmitgen A, Schneider J, Meixensberger J. Neuronavigation based on intraoperative 3D-ultrasound during tumor resection. International Congress Series 2005;1281:81-82. 30. Comeau RM, Sadikot AF, Fenster A, Peters TM. Intraoperative ultrasound for guidance and tissue shift correction in image-guided neurosurgery. Med Phys 2000;27: 787-800. 31. Schenck JF, Jolesz FA, Roemer PB, et al. Superconducting open-configuration MR imaging system for image-guided therapy. Radiology 1995;195:805-14. 32. Fahlbusch R, Nimsky C. Intraoperative MRI developments. Neurosurg Clin N Am 2005;16:xi-xiii. 33. Schneider JP, Schulz T, Schmidt F, et al. Gross-total surgery of supratentorial low-grade gliomas under intraoperative MR guidance. AJNR Am J Neuroradiol 2001;22: 89-98. 34. Black PM, Alexander E, 3rd, Martin C, et al. Craniotomy for tumor treatment in an intraoperative magnetic resonance imaging unit. Neurosurgery 1999;45:423-31; discussion 31-3. 35. Yrjana SK, Tuominen J, Koivukangas J. Intraoperative magnetic resonance imaging in neurosurgery. Acta Radiol 2007;48:540-9. 36. Schulder M, Carmel PW. Intraoperative magnetic resonance imaging: impact on brain tumor surgery. Cancer Control 2003;10:115-24.


37. Lewin JS, Metzger A, Selman WR. Intraoperative magnetic resonance image guidance in neurosurgery. J Magn Reson Imaging 2000;12:512-24. 38. Moriarty TM, Quinones-Hinojosa A, Larson PS, et al. Frameless stereotactic neurosurgery using intraoperative magnetic resonance imaging: stereotactic brain biopsy. Neurosurgery 2000;47:1138-45; discussion 45-6. 39. Hinks RS, Bronskill MJ, Kucharczyk W, Bernstein M, Collick BD, Henkelman RM. MR systems for imageguided therapy. J Magn Reson Imaging 1998;8:19-25. 40. Sutherland GR, KaibaraT, LouwD, HoultDI, TomanekB, Saunders J. A mobile high-field magnetic resonance system for neurosurgery. J Neurosurg 1999;91:804-13. 41. Hall WA, Galicich W, Bergman T, Truwit CL. 3-Tesla intraoperative MR imaging for neurosurgery. J Neurooncol 2006;77:297-303. 42. Nabavi A GD, Kacher DF, Talos IF, Wells WM, Kikinis R, et al. Surgical navigation in the open MRI. Acta Neurochir Suppl (Wein) 2003;85:121-5. 43. Acosta FL, Jr, Thompson TL, Campbell S, Weinstein PR, Ames CP. Use of intraoperative isocentric C-arm 3D fluoroscopy for sextant percutaneous pedicle screw placement: case report and review of the literature. Spine J 2005;5:339-43. 44. Rajasekaran S, Vidyadhara S, Shetty AP. Iso-C3D fluoroscopy-based navigation in direct pedicle screw fixation of Hangman fracture: a case report. J Spinal Disord Tech 2007;20:616-9. 45. Villavicencio AT, Burneikiene S, Bulsara KR, Thramann JJ. Intraoperative three-dimensional fluoroscopy-based computerized tomography guidance for percutaneous kyphoplasty. Neurosurg Focus 2005;18:e3. 46. Walker DG BP. Use of intraoperative MRI in pituitary surgery. Operative techniques in neurosurgery 2002;5: 231-8. 47. Rafferty MA, Siewerdsen JH, Chan Y, et al. Intraoperative cone-beam CT for guidance of temporal bone surgery. Otolaryngol Head Neck Surg 2006;134:801-8. 48. Chibbaro S TL. Image-guided microneurosurgical management of vascular lesions using navigated computed tomography angiography. An advanced IGS technology application. The International Journal of Medical Robotics and Computer Assisted Surgery: MRCAS 2006;2. 49. Rafferty MA, Siewerdsen JH, Chan Y, et al. Investigation of C-arm cone-beam CT-guided surgery of the frontal recess. Laryngoscope 2005;115:2138-43. 50. Klopfenstein JD, Spetzler RF, Kim LJ, et al. Comparison of routine and selective use of intraoperative angiography during aneurysm surgery: a prospective assessment. J Neurosurg 2004;100:230-5. 51. Schievink WI, Vishteh AG, McDougall CG, Spetzler RF. Intraoperative spinal angiography. J Neurosurg 1999;90: 48-51. 52. Ayad M, Ulm AJ, Yao T, Eskioglu E, Mericle RA. Realtime image guidance for open vascular neurosurgery

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

54

using digital angiographic roadmapping. Neurosurgery 2007;61:55-61; discussion 2. Hardy J, Wigser SM. Trans-sphenoidal surgery of pituitary fossa tumors with televised radiofluoroscopic control. J Neurosurg 1965;23:612-9. Okudera H, Takemae T, Kobayashi S. Intraoperative computed tomographic scanning during transsphenoidal surgery: technical note. Neurosurgery 1993;32:1041-3. Atkinson JL, Kasperbauer JL, James EM, Lane JI, Nippoldt TB. Transcranial-transdural real-time ultrasonography during transsphenoidal resection of a large pituitary tumor. Case report. J Neurosurg 2000;93:129-31. Gong J, Mohr G, Vezina JL. Endoscopic pituitary surgery with and without image guidance: an experimental comparison. Surg Neurol 2007;67:572-8; discussion 8. Darakchiev BJ, Tew JM, Jr, Bohinski RJ, Warnick RE. Adaptation of a standard low-field (0.3-T) system to the operating room: focus on pituitary adenomas. Neurosurg Clin N Am 2005;16:155-64. Pergolizzi RS, Jr, Nabavi A, Schwartz RB, et al. Intraoperative MR guidance during trans-sphenoidal pituitary resection: preliminary results. J Magn Reson Imaging 2001;13:136-41. Knauth M, Aras N, Wirtz CR, Dorfler A, Engelhorn T, Sartor K. Surgically induced intracranial contrast enhancement: potential source of diagnostic error in intraoperative MR imaging. AJNR Am J Neuroradiol 1999;20:1547-53. Fried MP, Topulos G, Hsu L, et al. Endoscopic sinus surgery with magnetic resonance imaging guidance: initial patient experience. Otolaryngol Head Neck Surg 1998;119:374-80. Fahlbusch R, Keller B, Ganslandt O, Kreutzer J, Nimsky C. Transsphenoidal surgery in acromegaly investigated by intraoperative high-field magnetic resonance imaging. Eur J Endocrinol 2005;153:239-48. Nimsky C, Ganslandt O, Hofmann B, Fahlbusch R. Limited benefit of intraoperative low-field magnetic resonance imaging in craniopharyngioma surgery. Neurosurgery 2003;53:72-80; discussion 1. Staecker H, O’malley BW, Eisenberg H, Yoder BE. Use of the LandmarXtrade mark surgical navigation system in lateral skull base and temporal bone surgery. Skull Base 2001;11:245-55. Van Havenbergh T, Koekelkoren E, De Ridder D, Van De Heyning P, Verlooy J. Image guided surgery for petrous apex lesions. Acta Neurochir 2003; 145:737-42. Gharabaghi A, Rosahl SK, Feigl GC, Liebig T, Mirzayan JM, Heckl S, Shahidi R, Tatagiba M, Samii M. Imageguided lateral suboccipital approach: part 1-individualized landmarks for surgical planning. Neurosurgery 2008;62 3 Suppl 1:18-22. Gharabaghi A, Rosahl SK, Feigl GC, Safavi-Abbasi S, Mirzayan JM, Heckl S, Shahidi R, Tatagiba M, Samii M.

829

830

54


Image-guided lateral suboccipital approach: part 2-impact on complication rates and operation times. Neurosurgery. 2008;62 3 Suppl 1:24-9. 67. Chatrath P, Nouraei SA, De Cordova J, Patel M, Saleh HA. Endonasal endoscopic approach to the petrous apex: an image-guided quantitative anatomical study. Clin Otolaryngol 2007 Aug;32(4):255-60. 68. Roberts DW, Miga MI, Hartov A, et al. Intraoperatively updated neuroimaging using brain modeling and sparse data. Neurosurgery 1999;45:1199-206; discussion 206-7. 69. Chinzei K, Miller K. Towards MRI guided surgical manipulator. Med Sci Monit 2001;7:153-63.

70. Bumm K, Wurm J, Rachinger J, et al. An automated robotic approach with redundant navigation for minimal invasive extended transsphenoidal skull base surgery. Minim Invasive Neurosurg 2005;48:159-64. 71. Hernes TA, Ommedal S, Lie T, Lindseth F, Lango T, Unsgaard G. Stereoscopic navigation-controlled display of preoperative MRI and intraoperative 3D ultrasound in planning and guidance of neurosurgery: new technology for minimally invasive image-guided surgery approaches. Minim Invasive Neurosurg 2003;46:129-37. 72. Jane JA, Jr, Thapar K, Alden TD, Laws ER, Jr. Fluoroscopic frameless stereotaxy for transsphenoidal surgery. Neurosurgery 2001;48:1302-7; discussion 7-8.

39 MRI in Image Guided Surgery M. Schulder . L. Jarchin

Introduction Magnetic resonance imaging (MRI) has transformed the field of neurosurgery by improving the accuracy of diagnostic imaging, integration into presurgical planning, and moving into the operating room (OR) itself. For surgical applications, most innovations have occurred within the realm of stereotactic neurosurgery. There has been some controversy regarding credit for the discovery that the physical properties of nuclear magnetic resonance could be used to image living tissue. In 2003 the Nobel Prize in Medicine was awarded to Drs. Paul Lauterbur and Peter Mansfield for this work. Dr. Raymond Damadian publicly objected to his exclusion despite his pioneering efforts in the field [1,2]. Be that as it may, by the mid 1980s MRI had become a commercially available technology whose benefits were obvious. These included much better soft tissue imaging of the central nervous system, multiplanar views, and the avoidance of ionizing radiation. Within short order MRI was adapted to image guided surgery and is now the mainstay of this approach in the developed world.

Stereotactic Biopsy with MRI The initial use of MRI for image guided surgery was in frame-based stereotactic brain biopsy. The imaging advantages of MRI over CT were readily apparent, especially improved target definition [3]. Several authors described their experience in patients whose lesions could be seen only on MRI [4,5]. This was particularly so for stereotactic targeting of lesions in the brainstem [6]. It is #


worth noting that as early as 1985 Kelly described the integration of MRI in his pioneering volumetric stereotactic system [7]. For most neurosurgeons that sort of technology belonged to the future, and ‘‘conventional’’ frames were used, including the Leksell (Elekta, Norcross GA) [8] and Brown/Cosman-Roberts-Wells (Radionics, Burlington MA) [9] systems (> Figure 39-1). Much effort was expended so CT-compatible frames would be made suitable for MRI. The most obvious difference was the need to avoid ferromagnetic materials in the frame [9]. Besides the frame construct itself, the rods on the localizer box needed to be visible on MRI. Some of the early solutions to this problem, such as filling empty tubes with copper sulfate solution, were quite cumbersome, leading the neurosurgeon to choose CT based guidance when possible. The accuracy of stereotactic frames was well established, and quantified in 1992 by Maciunas et al [10]. Using CT scans with a variety of commercially available frames, they demonstrated that an error of less than 2 mm could be reasonably expected. This error increased with greater imaging slice thickness, scanner gantry angle, and arc reapplication. Early clinical use of MRI made plain the presence of image distortions, from chemical shifts and magnetic susceptibility artifacts. Kondziolka et al compared the accuracy of stereotactic targeting using MRI and CT [11]. They found discrepancies of about 2 mm overall on average, increasing with distance from the center of the patient’s head. They concluded that MRI targeting for frame-based stereotaxy was sufficiently accurate for most procedures (with obvious implications for radiosurgery and for functional procedures).

600

39

MRI in image guided surgery

. Figure 39-1 Cosman-Roberts-Wells frame on phantom base in OR

latter technique eliminating the need to subject patients to prolonged MRI scans with a frame applied. Still, most patients needing biopsies today will have them done using a ‘‘frameless’’ technique or with intraoperative MRI (see below).

Frameless Stereotactic Surgery and MRI

The safety and accuracy of stereotactic biopsy had been established in the 1980s with the emergence of CT-guided techniques [12,13]. There remained, then as now, a risk of nondiagnostic biopsy of about 8% [14]. This remains a risk with frame-based stereotaxis, although it can be lowered to about 3% with the use of intraoperative pathological examination [15]. MRI-guided frame-based stereotactic biopsies remain an important part of operative neurosurgery, but there are limitations with this technology. Conceptually, frames provide registration of single points in space. Multiple points can indeed be targeted but this requires a separate calculation for trajectory and distance, with resetting of the frame coordinates each time. While stereotactic craniotomy was described early in the era of CT guided stereotaxis, the ergonomic difficulty of doing an open surgical procedure with a frame in place have relegated this approach to historical interest [12]. The need to image with the frame in place also prolongs the procedure and requires patient transport through the hospital. The use of frame-based stereotaxis has greatly improved with the incorporation of volumetric navigation [16] and image fusion [17], with the

The benefits of MRI in image guided surgery were perhaps best realized with the advent of frameless stereotaxy, also termed surgical navigation or image guided surgery (IGS) [18]. Kelly described the concept of stereotactic volumetric resection with 3-dimensional computer modeling in 1981 [18]. This method used a modified Todd-Wells stereotactic frame as its platform. The concept of a ‘‘frameless’’ approach was proposed by Roberts et al. who envisioned using the operating microscope as a pointer [19]. This technique was marketed as part of the Zeiss microscope system [20]. However, it proved more practical to uncouple the operating microscope from the IGS technology. Various systems became commercially available in the early to mid 1990s, employing different technologies. The first device, the IGS viewing wand, tracked an articulated arm in space [21]. Due to the bulk and ergonomic interference of this articulated arm, other authors created a digitizing platform with ultrasound [22], magnetic vectors [23,24] or infrared light [25,26]. The latter technology for now has become the one most commonly used in IGS. The most obvious advantage of IGS is that the entire volumes of interest are registered, for instance, the patient’s head and his entire MRI scan. Typically this is done by scanning the patient with fiducial markers applied to the scalp. The image may be obtained as far as in advance of surgery as is convenient. As alluded to above, this uncoupling of imaging and surgery allows for prolonged scanning sessions, in particular anatomical MRI and functional imaging as well.


Multiple datasets may be acquired for registration in the OR, and loaded onto the planning workstation before surgery, resulting in further saving of time. These different images can be viewed separately or as a blended view after registration. The other images may include CT [27], PET [28], and MR angiography [29]. Particularly useful has been the incorporation of functional MRI in IGS, allowing for navigation based on more than anatomical information alone [30,31]. McDonald et al. found that fMRI, registered for use in IGS, was as accurate as magnetoencephalography for localizing the primary cortex [32]. This may be combined additionally with tractography derived from diffusion tensor imaging, to truly provide a comprehensive anatomical and functional map of the brain [33] (> Figure 39-2). In the OR, after general anesthesia is induced and the patient’s head secured in a pin fixation clamp, a probe or ‘‘wand’’ with active infrared light emitting diodes (IRLEDs) or passive infrared reflecting spheres is touched to at least four fiducials that have been marked on the preoperative image. When enough spatial information has been gathered to allow for a transformation that matches image and surgical spaces, an accurate registration has been achieved. At this time, unlike with a frame-based procedure,

39

the surgeon has theoretically unlimited entry points, targets, and trajectories from which to choose. In practice, of course, the surgical plan will be constrained by different considerations such as cosmetics, eloquent brain areas, previous surgery, etc. At first, IGS was mainly used to guide craniotomies and tumor resections. The technology was useful for planning incisions and craniotomies [23–26]. These benefits may be lost after dural opening and tumor resection is begun, because of brain shift [34,35]. This phenomenon, in part, spurred the development of intraoperative MRI (see below). In any event, intracranial surgery with IGS using preoperative anatomical MRI has become a routine part of contemporary neurosurgery; for surgery at the skull base, concerns regarding brain shift matter relatively little (> Figure 39-3). Application of IGS to such ‘‘purely stereotactic’’ applications as biopsy and functional stereotaxis required evidence that the accuracy of frameless approaches equaled that of framebased surgery. Dorward et al demonstrated that this was the case [36]. They found that CT-based navigation was slightly more accurate than that based on MRI, with a mean error of 1.1 mm for CT and 1.4 mm for MRI; however, this did not have clinical implications in their series of

. Figure 39-2 fMRI and DTI in a patient with right frontal low grade glioma. Images confirm that the right primary motor cortex is posterior to the tumor (a) and that the corticospinal tract is also behind and inferior to the lesion (b)

601

602

39


. Figure 39-3 Image guided surgery: transsphenoidal surgery with StealthStation images. (a), navigation with CT. The pituitary gland and tumor are poorly seen. (b), navigation with MRI, clearly showing the gland and microadenoma


21 stereotactic biopsies using IGS. Several years later, the same group demonstrated advantages of IGS guidance over frame-based stereotaxis for biopsy [37]. OR time was shorter, the complication rate was lower, and as a result hospital stays and costs were reduced. It is worth noting that other authors have found that for the localization of very small cavernous angiomas a stereotactic frame was more accurate than a surgical navigation system [38]. As the accuracy of ‘‘frameless’’ localization for functional indications and/or for radiosurgery has been shown elsewhere to be comparable to that of stereotactic frames (see below), the significance of this finding is unclear. Another group examined the effect of different MRI sequences on stereotactic accuracy. They assumed that CT was the standard against which accuracy should be measured, and that infrared IGS technology was inherently accurate. Although the findings were that CT was indeed more accurate, T1 weighted MRI was only 23% less so, and T2 weighted images 37% less accurate. These results indicated that even lesions seen only on T2W scans could be approached with adequate precision using IGS [39]. With this knowledge in hand, the obvious question relates to the desirability of using IGS to perform stereotactic biopsy. IGS biopsies have been done with freehand guidance or with an articulated and rigidly fixed arm holding the biopsy instrument. A skull-mounted device (Navigus, Medtronic Navigation, Louisville CO), invented as a guide for intraoperative MRI based surgery, has been adapted for biopsies and functional stereotaxis using ‘‘conventional’’ IGS. One group found that targeting with the Navigus system was slightly more accurate than with a stereotactic frame [40]. The mean error from target was 0.33 mm +/ 0.16 mm with the probe’s eye planning method using the Navigus and StealthStation software, versus 1.03 mm +/ 0.19 mm with the Cosman-Roberts-Wells frame.

39

The conceptual advantages of IGS for stereotactic biopsy are in part similar to those achieved with craniotomy. These include uncoupling of imaging from surgery, the ability to preplan the procedure, and registration of multiple datasets. For instance, the use of PET has been described as a way to maximize the diagnostic yield of biopsy by targeting the metabolically most active area [41]. IGS also brings to stereotactic biopsy certain specific advantages. Perhaps the most important is the ability to pick multiple targets with ease, and without a separate registration for each, as is needed for ‘‘traditional’’ frame-based biopsies. Of course, IGS computers are compatible with the most commonly used stereotactic frames (Leksell and Radionics), in which case the planning software can be used to choose biopsy sites; however, coordinates will still need to be set separately for each target, with the attendant error and possibility of missing the area of interest. Image-guided biopsy with a frame allows for surgery to be done easily under local anesthesia. However, if the patient’s head is secured in three-point fixation (such as the Mayfield clamp, Integra Radionics), positioning may be easier than with a frame regardless of anesthetic technique. This is so for lateral and especially prone positions, which may be desirable based on lesion location and the safest trajectory to the target. Several authors employing different technologies have established the safety and reliability of IGS guided biopsy [42–45]. These studies confirmed that the diagnostic yield and complication rate were similar whether biopsies were done using a frame-based or ‘‘frameless’’ technique. The complication rate was on the order of 2–3%, and the chance of a nondiagnostic biopsy about 8% overall. As in the case of frame-based biopsies, the odds of making a diagnosis are improved by confirmation, via intraoperative pathological examination, that lesional tissue has been obtained [37,43].

603

604

39


Stereotactic Functional Surgery with MRI The use of IGS to target lesions visible on MRI may be termed ‘‘morphological stereotaxis’’, as opposed to ‘‘functional stereotaxis’’ aimed at nuclei and/or tracts that are not necessarily seen on contemporary imaging. The small volumes involved, and the great precision required to achieve a good clinical result, allow for very little error in IGS. The history of clinically applied functional stereotaxis dates to the late 1940s, when Spiegel and Wycis applied the concept of Cartesian coordinates put forward by Horsley and Clarke in 1908 (but used only for physiological animal experimentation by Horsley) [46]. Air or positive contrast ventriculography were the standard method of functional target localization for nearly 40 years, before the advent of CT-guided stereotaxis [47,48]. Measurements derived from the coordinates of the anterior and posterior commisures were used to identify target locations in the basal ganglia, and thalamus according to the method described by Talairach [49,50]. The advent of CT held out the possibility of functional targeting without the discomfort, morbidity, and time needed for ventriculography-guided surgery. CT has the advantage of high spatial accuracy, with little distortion. However, it is mainly useful in functional stereotaxis as a source of indirect targeting derived from ventricular landmarks. MRI, on the other hand, holds out the potential of detailed anatomic visualization and accurate direct targeting of deep brain structures (> Figure 39-4). As in morphological stereotaxis, the accuracy of digital, sectional imaging compared to the proven technique of ventriculograpy was required. This was done first with CT [51]. Hariz showed how distortion from air injection caused third ventricular widening and anterior displacement of the midbrain; physiologically determined targets were found equally well with both techniques [52]. Not long afterwards, similar studies were done comparing the accuracy of

MRI-guided surgery to ventriculography, a more sensitive task given the possibility of spatial inaccuracies in MRI; in fact, one study found that ventriculography was significantly more accurate, with errors of up to 4.6 mm with the MRI derived coordinates [53]. However, the consensus emerging from various studies comparing the two methods was that MRI accuracy was adequate for functional neurosurgery [49,54,55]. As a result, ventriculography has largely become a technique of historical interest. In fact, a relatively recent study showed that even this time-tested technique was inadequate for accurate target definition in the globus pallidus, and that intraoperative electrophysiology found differences as high as 10 mm compared to the target predicted by imaging alone. There is little disagreement now that even with the best MRI-derived direct targeting, intraoperative electrophysiology with some combination of stimulation and microelectrode recording will help to achieve the best results in functional stereotactic surgery, as is discussed in detail elsewhere in this book [56,57]. There is a reasonable consensus now regarding current technique for functional stereotaxis. Most patients will undergo a high resolution stereotactic MRI in advance of the surgery or as part of the procedure, with a stereotactic frame applied. Acquisition with 3T MRI may provide more anatomical detail without loss of spatial accuracy compared to 1.5 T scanners [58,59] On various sequences the anterior and posterior commisures are easily identified for indirect targeting. On coronal T2W images the red nucleus can be identified as a landmark for the STN; ideally the STN, globus pallidus or other targets can be seen, allowing for true direct targeting. Imaging data is transferred to the IGS computer in the OR. If a stereotactic CT is done as part of the procedure, image fusion is performed to register this scan to the MRI. A combination of targeting methods is done to define the target – indirect off the commisures, indirect off the red nucleus


39

. Figure 39-4 IGS for functional neurosurgery. (a), CT image for targeting (on imaging console) of left Vim nucleus in patient with essential tremor. (b), MRI on StealthStation with targeting of left Vim

605

606

39


(if the STN is the target), direct based on the MRI, and possibly an atlas that is included with the IGS software. The ultimate site for DBS electrode placement or lesioning is refined using intraoperative electrophysiology [60–62]. Zonenshayn et al. found that no one imaging technique clearly was the single best choice but that in fact a ‘‘combination’ correlated best with electrophysiology [55]. It is worth noting that direct targeting of the STN was the least reliable method in their series. The reemergence of surgery for patients with refractory psychiatric conditions including obsessive-compulsive disorder and depression has been supported in large part by fMRI studies that have demonstrated a measurable organic substrate in these disorders. MRI-targeted lesioning, or more frequently DBS, has been applied based on this new physiologic information with promising results [63,64]. While DBS and in some cases lesioning are at this point the most common techniques of modulating the brain via functional stereotaxis, the future holds the prospect of biological therapies for movement disorders and other conditions [65,66]. These treatments still require precise insertion of catheters, implanted cells, etc. into specific and small areas of the brain. MRI guided IGS will remain a critical part of functional stereotactic neurosurgery. Technical advances in image guidance may make the stereotactic frame obsolete, but these new methods likewise will rely on MRI as the basis for targeting [67]. MRI is an important part of IGS in functional neurosurgery beyond targeting the diencephalon, basal ganglia or brainstem with ‘‘purely stereotactic’’ technique. Epidural motor cortex stimulation is a method proposed as a treatment for neuropathic pain [68] or movement disorders [69]. Placement of the stimulating electrode strip over the primary motor cortex is greatly facilitated by using fMRI and IGS [70]. Ablative surgery for patients with refractory seizure disorders is facilitated by the use of

stereotactic MRI for evaluation and definitive surgery. This includes accurate insertion of depth electrodes [71] and stereotactic lesioning of the medial temporal lobe [72].

Intraoperative MRI All of the above uses of MRI refer to preoperative scans used for image guidance. The next stage in the evolution of MRI in IGS has been the introduction of intraoperative MRI (iMRI), which has been applied to practically all of the indications described above. The impetus behind this development derived from the desire for resection control [73,74] and compensation for brain shift [34,75], primarily in tumor surgery. Inevitably, as the possibilities of this new technology became apparent, different groups have used it for other types of stereotactic neurosurgery. The first iMRI was conceived and built in the early to mid 1990s at the Brigham and Women’s Hospital in Boston, as a collaboration between the Departments of Neurosurgery and Radiology and General Electric Medical Systems (GEMS, Milwaukee, WI) [76,77]. Based on a 0.5 T magnet, it was named the GE Signa. This remains the most ‘‘purely stereotactic’’ system to date, as the Cartesian coordinate space and the surgical space are identical, with surgery performed via a gap between the two vertical magnet poles. Scans can be done as often as needed, even without interrupting surgery, and an integrated infrared navigation device made this a truly IGS system. However, certain factors limited the acceptance of the GE Signa, including the narrow (56 cm) gap between the magnet poles, difficulties in patient positioning, the location of the magnet outside of the main OR area, and the need for all instruments to be completely MRI compatible. As other iMRI solutions have gained popularity, GEMS no longer markets the Signa. Conceptually, however, it remains the iMRI against which other devices should be measured.


The search for the ideal iMRI has involved an attempt to reach various compromises with the two main conflicting goals of intraoperative imaging: image quality and overall functionality versus user-friendliness. iMRIs have been classified by the way that patients or images are moved for scanning [78]. Most neurosurgeons, though, refer to systems as being either ‘‘high field’’ or ‘‘low field’’, referring to the magnet strength and hence to how close the iMRI approaches the utility of a diagnostic system [79]. Not longer after the advent of the GE Signa, several authors described their experience using a 0.2 T–0.3 T iMRI with a 25 cm horizontal gap [80–82]. Practically this meant that surgery had to be done outside the ‘‘fringe fields’’ of the magnet, with the patient then moved into the iMRI for imaging [83]. The desire to achieve diagnostic image quality, with the ability to acquire functional imaging and MR spectroscopy, led the Erlangen group to develop a 1.5 T iMRI [84]. This system is implemented in a regular OR and includes integrated navigation. The rapid falloff the magnetic field allows for the use of routine instrumentation. When imaging is done, the patient is rotated on the specially designed table into the magnet. Nimsky et al have described true intraoperative, updated diffusion tensor imaging (DTI) tractograpy with this system [85]. Another method of acquiring high field iMRI in an otherwise routine OR environment was developed in Calgary, and uses a 1.5 T magnet that sits in a shielded alcove between imaging sessions. When needed, the imager moves on a track into scanning position [86]. Further work with this system has allowed for the incorporation of tractography (G Sutherland, personal communication). At the University of Minnesota a 1.5 T diagnostic magnet was used for surgery in the radiology department [87]. Doing so meant revisiting many of the ergonomic obstacles of the Signa system, but full functional imaging capability was available including fMRI. This group was the first to describe surgery in a 3 T

39

environment [88]. Two main benefits have come from their work. First is the feasibility and utility of a shared resource iMRI that is used for diagnostic imaging when (as is true most of the time) it is not being used for surgery. Second, Hall et al. developed the abovementioned Navigus guide, an ingenious skull-mounted tool that can be used for any procedure requiring stereotactic guidance [89]. More on 3 T iMRI later on. The above projects are efforts to bring diagnostic MRI – the more the better – to the neurosurgical OR. To some extent they require adapting the OR to the MRI. There has been one iMRI system that has taken the opposite approach, i.e., to adapt the MRI to the OR. In 2000 Hadani et al. reported the development of a 0.12 T iMRI, meant to be used in a regular neurosurgical OR [90]. Other groups also reported their early experience with this system, named the PoleStar N-10 (Medtronic Navigation, Louisville CO) [78,79,91]. This compact iMRI is stored in a shielded compartment and rolled out as needed for surgery, and is parked under a regular OR table. The low field strength allows for the use of regular instrumentation (except when operating directly through the magnet poles). The major limitations of this system were the limited field of view (FOV), 25 cm vertical magnet pole gap, and the variable image quality. This spurred an upgrade that enlarged the device somewhat but increased the magnet field strength to 0.15 T and the gap to 27 cm, increased the FOV, and in general improved the image quality [92]. fMRI was acquired with this system, although not during surgery [93]. The above and other publications, unsurprisingly, have mostly focused on the use of iMRI for tumor resection control. Across all platforms there were fairly consistent conclusions: that iMRI is particularly useful for surgery in patients with gliomas (especially low grade tumors) [94,95] and pituitary macroadenomas [96,97]. Overall, intraoperative imaging has led to the resection of otherwise unvisualized tumor

607

608

39


in about 1/3 of cases. Compared to ‘‘conventional’’ IGS, surgery with iMRI was more likely to achieve a 90% tumor resection in patients with low grade gliomas (M Schulder, unpublished data). A similar trend in patients with high grade tumors did not reach statistical significance. Of course, surgical navigation updated with new imaging has avoided the problem of brain shift [98]. iMRI has been used for indications other than tumor resection. Perhaps the most obvious is for stereotactic biopsy. The theoretical advantages of doing so are that imaging can be used to confirm that specimens are indeed being taken from the desired target, thereby improving diagnostic yield. The other main benefit is to rule out a procedurerelated hematoma or other surgical complication. Bernays et al described their experience with stereotactic biopsy in a GE Signa iMRI [99] in 114 patients. Pathological tissue was obtained from all. One patient required a delayed emergency craniotomy because of edema. Stereotactic biopsy in the PoleStar iMRI has been done by this chapter’s primay author. All imaging was done in the OR, with a Navigus guide used to direct an MRI compatible cannula to the target. Biopsies were taken after imaging with the cannula in place. In 30 patients, a diagnosis was obtained in all. Intraoperative imaging led to a repositioning of the cannula to ensure ideal placement in the lesion in 8/30 patients (> Figure 39-5). There were no complications in this series. In patients with refractory epilepsy iMRI use has been similar to that used in tumor resection – namely, for navigation and resection control of nonlesional epileptic foci [100,101]. Applications that have used iMRI for more ‘‘stereotactic uses’’ have included management of complex hydrocephalus in children, when intraoperative verification of catheter placement is desired [102]. iMRI for DBS electrode placement has been described, and with the increasing indications for this procedure we may see greater of iMRI use in the coming decade. The functional neurosurgery group at UCSF has shown that DBS can be done

in a high field MRI without increased risk, with the ability to target the STN without intraoperative electrophysiology, and with excellent placement confirmed using near real-time imaging [103]. Of course, more work will need to be done to ensure that clinical results are as good or better using this method as with frame-based or ‘‘frameless stereotaxis’’. Recently, the move to 3 T iMRI has begun, with reports of new systems appearing [104]. These and other iMRI implementations have been proposed as part of a comprehensive imaging suite, to include CT, digital angiography, and possibly PET [105]. The advantages of these projects have been touted as being able to serve multiple operating rooms and services, and to facilitate shared-resource imaging that make intraoperative imaging a profit center and not a financial drain. These are worthy goals, of course. But the question is raised, at what point do these systems cease to provide actual intraoperative imaging? If the patient has to be moved from the OR to another site for imaging, even if it is next door, is that different in concept than moving to a completely separate radiology department? At a minimum, such moves will not encourage routine use of iMRI and will allow for enough ongoing change, in the time between imaging and return to the OR, that the benefits of updated navigation also will be unclear. The jury will remain out on this concept of iMRI. A move back towards the original Signa concept, where stereotactic, image, and surgical space are one, is the best possible alternative.

Stereotactic Radiosurgery with MRI A field that has benefited greatly from the use of MRI has been stereotactic radiosurgery (SRS). Invented and named by Leksell in 1951, this minimally invasive surgical technique almost by definition relies on image guidance to find its target and ensure a safe and effective result [106].


39

. Figure 39-5 Stereotactic biopsy with iMRI. (a), patient positioned prone with MRI receive coil in place. (b), preoperative iMRI with entry and target point chosen for biopsy. Field of view does not encompass entire cranium. (c), Compare function shows (left) initial cannula pass and (right) improved placement in the lesion

Until the mid 1980s SRS was done with predigital imaging – angiography, ventriculography, or skull X-rays (or a version thereof such as polytomography) [107]. The availability of CT-guided

stereotaxis began to open up SRS to treat a greater variety of lesions, and smaller ones, than had been possible before. This in large part encouraged the development of linear accelerator (Linac) based

609

610

39


treatments [108,109] and the wider use of the gamma knife (GK) SRS [110]. This was true as well for the use of heavy particle irradiation, but the number of such centers is limited by the cost and complexity of the systems [111]. CT not only allowed for the targeting of directly visualized lesions, but also for treatment planning that would show the dose to be delivered to adjacent normal structures. The advent of MRI-guided SRS has greatly improved the efficacy and safety of SRS. Radiobiology suggests that SRS is most potent for treating benign lesions, such that patients with schwannomas and meningiomas of the skull base are often ideal candidates for SRS [112]. These tumors are seen far better on MRI than CT thanks to the elimination on MRI of the bonehardening artifact (> Figure 39-6). The same is true for patients with lesions in close proximity to the skull elsewhere, such as parasagittal meningiomas [113]. Furthermore, the proximity of critical structures such as the optic chiasm that are poorly seen on CT makes the risk of visual loss from SRS potentially high. Using MRI, even individual cranial nerves can be seen and the

treatment plan adjusted accordingly [114]. This is so whether direct MRI targeting is used [115] or image fusion between MRI and CT [116]. The Cyberknife (Accuray, Sunnyvale CA), an innovative, robotic, Linac SRS tool (discussed in more detail elsewhere in this volume) has been used to treat patients with medically refractory trigeminal neuralgia (TN). As recently as several years ago, CT cisternography was recommended as the imaging technique of choice to target the trigeminal nerve [117]. This reflected the vendor’s (reasonable) insistence on CT as the reference image, for maximum spatial accuracy. Fortunately, excellent image fusion software now allows for the use of MRI to treat TN patients with the Cyberknife [118], as is done with other Linacs [119] and the GK [120]. Functional neuroimaging has been evaluated as an adjunct to SRS. The tolerances to irradiation of certain cranial nerves and the brainstem, which can be imaged with standard morphological MRI, are reasonably well understood [121,122]. fMRI can be registered for use in SRS planning software, and identification of ‘‘eloquent’’ cortical areas can be used to redirect

. Figure 39-6 Cyberknife SRS plan in patient with a right vestibular schwannomas. (a), CT based plan shows only tumor in cerebellopontine angle. (b), MRI (fused to CT) shows intracanalicular component and edge of tumor with greater clarity


dose curves in a way that will limit the chance of a radiation-induced deficit (> Figure 39-7) [123,124]. Maruyama et al. have incorporated DTI tractography into GKRS planning, and have reported data suggesting that the dose of single session SRS to the optic radiation should be limited to eight Gy or less [125] whereas the corticospinal tract can tolerate up to 23 Gy [126]. More work will need to be done to confirm these results and to determine to what extent functional neuroimaging can affect SRS planning.

MRI for Functional Stereotactic Radiosurgery with MRI An irony of SRS is that while Leksell envisioned and invented the technique primarily as a tool for functional neurosurgery, the explosive interest in SRS that began some 20 years ago has been almost all directed at treating lesions – . Figure 39-7 fMRI for SRS. Treatment plan in patient with parasagittal meningiomas. Isodose lines confirm low dose is delivered to the left primary motor cortex (activation area colored blue)

39

mostly patients with neoplasms, and some with arteriovenous malformations. More recently, however, neurosurgeons have again begun to employ SRS to treat patients with functional disorders, besides TN. The latter condition has the advantage of a target that can usually be seen with relative ease on MRI [120]. This is not the same for functional targets in the diencephalon, basal ganglia, or brainstem. For instance, defining thalamic, pallidal, or subthalamic targets in patients with tremor or Parkinson’s disease who are to undergo DBS implantation requires a combination of imaging and physiological methods [127]. For decades, before making a surgical lesion for movement disorders, macrostimulation at a minimum was done to ensure the safety of this irreversible step. The main skepticism regarding functional SRS, therefore, has been related to the lack of such feedback in the non-invasive SRS [128]. Several authors have countered this argument [129], in particular Sato et al. who performed stereotactic radiofrequency thalamotomy in four patients using MRI-derived coordinates that would be used for radiosurgical lesioning [130]. The safety and efficacy of radiosurgical thalamotomy and pallidotomy, for patients with essential tremor, Parkinson’s disease, and multiple sclerosis tremor has been reported over the last decade [131–133]. Even the STN has been proposed as a radiosurgical target [134]. Needless to say, these procedures could not be considered without the most accurate MR imaging possible. Further evaluation of clinical and imaging results after functional SRS will still need to be done before this technique can be considered as an equivalent alternative treatment for patients with movement disorders.

Conclusions Contemporary stereotactic and functional neurosurgery is barely conceivable without image guided surgery based on MRI. The near future will see refinements based on magnets with 3 T or even

611

612

39


higher field strengths [135]. The role of neurosurgeons in confirming the spatial accuracy of these devices will be critical in allowing them to be used as the imaging source for open and radiosurgical treatments of functional disorders. Stereotactic neurosurgeons will also be at the forefront of refinements in IGS and iMRI that will bring these methods even more into the neurosurgical mainstream. The next edition of this textbook will have many exciting new developments to report.

References 1. Macchia RJ, Termine JE, Buchen CD, Raymond V. Damadian, MD.: magnetic resonance imaging and the controversy of the 2003 Nobel Prize in Physiology or Medicine. J Urol 2007;178(3 Pt 1):783-5. 2. Damadian R, Minkoff L, Goldsmith M, Stanford M, Koutcher J. Tumor imaging in a live animal by focusing NMR (FONAR). Physiol Chem Phys 1976;8(1):61-5. 3. Lee JY, Lunsford LD, Subach BR, Jho HD, Bissonette DJ, Kondziolka D. Brain surgery with image guidance: current recommendations based on a 20-year assessment. Stereotact Funct Neurosurg 2000;75(1):35-48. 4. Bradford R, Thomas DG, Bydder GM. MRI-directed stereotactic biopsy of cerebral lesions. Acta Neurochir Suppl (Wien) 1987;39:25-7. 5. Mercier C, Le Bas JF, Benabid AL, Pasquier B. MRI contribution to the stereotactic management of cerebral tumours. Appl Neurophysiol 1987;50(1–6):155-8. 6. Frank F, Fabrizi AP, Frank-Ricci R, Gaist G, Sedan R, Peragut JC. Stereotactic biopsy and treatment of brain stem lesions: combined study of 33 cases (Bologna-Marseille). Acta Neurochir Suppl (Wien) 1988;42:177-81. 7. Kelly PJ, Kall BA, Goerss S, Earnest F. Present and future developments of stereotactic technology. Appl Neurophysiol 1985;48(1–6):1-6. 8. Ferry M, Cabanis E, Cosnard G, Bocquet M, Desgeorges M. Stereotactic biopsy of the brain with the Leksell frame in the context of modern radiology. J Neuroradiol 1988;15(1):63-75. 9. Heilbrun MP, Sunderland PM, McDonald PR, Wells TH, Jr., Cosman E, Ganz E. Brown-Roberts-Wells stereotactic frame modifications to accomplish magnetic resonance imaging guidance in three planes. Appl Neurophysiol 1987;50(1–6):143-52. 10. Maciunas R, Galloway R, Lattimer J. The application accuracy of stereotactic frames. Neurosurgery 1994;35: 682-95. 11. Altschuler E, Lunsford LD, Kondziolka D, Wu A, Maitz AH, Sclabassi R, Martinez AJ, Flickinger JC.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Radiobiologic models for radiosurgery. Neurosurg Clin N Am 1992;3(1):61-77. Apuzzo ML, Sabshin JK. Computed tomographic guidance stereotaxis in the management of intracranial mass lesions. Neurosurgery 1983;12(3):277-85. Couldwell WT, Apuzzo ML. Initial experience related to the use of the Cosman-Roberts-Wells stereotactic instrument. Technical note. J neurosurg 1990;72(1): 145-8. Soo T, Bernstein M, Provias J, Tasker R, Lozano A, Guha A. Failed stereotactic biopsy in a series of 518 cases. Stereotact Funct Neurosurg 1995;64(4):183-96. Chandrasoma PT, Smith MM, Apuzzo ML. Stereotactic biopsy in the diagnosis of brain masses: comparison of results of biopsy and resected surgical specimen. Neurosurgery 1989;24(2):160-5. Winkler D, Trantakis C, Lindner D, Richter A, Schober J, Meixensberger J. Improving planning procedure in brain biopsy: coupling frame-based stereotaxy with navigational device STP 4.0. Minim Invasive Neurosurg 2003;46(1):37-40. Alexander EI, Kooy M, van Herk M, Shwartz M, Barnes D, Tarbell N, Mulkern R, Holupka E, Loeffler J. Magnetic resonance image-directed sterotactic neurosurgery: use of image fusion with computerized tomography to enhance spatial accuracy. J Neurosurg 1995;83 (August,1995):271-6. Kelly PJ, Alker GJ, Jr. A stereotactic approach to deepseated central nervous system neoplasms using the carbon dioxide laser. Surg Neurol 1981;15(5):331-4. Roberts DW, Strohbehn JW, Hatch JF, Murray W, Kettenberger H. A frameless stereotaxic integration of computerized tomographic imaging and the operating microscope. J Neurosurg 1986;65(4):545-9. Roessler K, Ungersboeck K, Aichholzer M, Dietrich W, Goerzer H, Matula C, Czech T, Koos WT. Frameless stereotactic lesion contour-guided surgery using a computernavigated microscope. Surg Neurol 1998;49(3):282-8; discussion 8-9. Golfinos J, Fitzpatrick B, Smith L, Spetzler R. Clinical use of a frameless stereotactic arm: results of 325 cases. J Neurosurg 1995;83:197-205. Barnett G, Kormos D, Steiner C, Weisenberger J. Intraoperative localization using an armless, frameless stereotactic wand. J Neurosurg 1993;78:510-4. Benardete EA, Leonard MA, Weiner HL. Comparison of frameless stereotactic systems: accuracy, precision, and applications. Neurosurgery 2001;49(6):1409-15. Mascott CR. Comparison of magnetic tracking and optical tracking by simultaneous use of two independent frameless stereotactic systems. Neurosurgery 2005; 57(4 Suppl):295-301; discussion 295-301. Smith KR, Frank KJ, Bucholz RD. The Neurostation – a highly accurate, minimally invasive solution to frameless stereotactic neurosurgery. Comput Med Imaging Graph 1994;18:247-56.


26. Germano I. The NeuroStation system for image-guided frameless stereotaxy. Neurosurgery 1995;37:348-50. 27. Chen JC, Petrovich Z, Giannotta SL, Yu C, Apuzzo ML. Radiosurgical salvage therapy for patients presenting with recurrence of metastatic disease to the brain. Neurosurgery 2000;46(4):860-6; discussion 6-7. 28. Pirotte B, Goldman S, Dewitte O, Massager N, Wikler D, Lefranc F, Ben Taib NO, Rorive S, David P, Brotchi J, Levivier M. Integrated positron emission tomography and magnetic resonance imaging-guided resection of brain tumors: a report of 103 consecutive procedures. J Neurosurg 2006;104(2):238-53. 29. Vougioukas VI, Coulin CJ, Shah M, Berlis A, Hubbe U, Van Velthoven V. Benefits and limitations of image guidance in the surgical treatment of intracranial dural arteriovenous fistulas. Acta Neurochir (Wien) 2006; 148(2):145-53; discussion 53. 30. Schulder M, Maldjian JA, Liu WC, Holodny AI, Kalnin AT, Mun IK, Carmel PW. Functional imageguided surgery of intracranial tumors located in or near the sensorimotor cortex. J Neurosurg 1998;89(3): 412-8. 31. Atlas S, Howard RI, Maldjian J, Alsop D, Detre J, Listerud J, D’Esposito M, Judy K, Zager E, Stecker M. Functional MRI of regional brain activity in patientws with intracerebral gliomas: findings and implications for clinical management. Neurosurgery 1996;38:329-38. 32. McDonald J, Chong B, Lewine J, Jones G, Burr R, McDonald P, Koehler S, Tsuruda J, Orrison W, Heilbrun M. Integration of preoperative and intraoperative functional brain mapping in a frameless stereotactic environment for lesions near eloquent cortex. J Neurosurg 1999;90:591-8. 33. Holodny AI, Schwartz TH, Ollenschleger M, Liu WC, Schulder M. Tumor involvement of the corticospinal tract: diffusion magnetic resonance tractography with intraoperative correlation. J Neurosurg 2001;95(6):1082. 34. Dorward N, Alberti O, Velani B, Gerritsen F, Harkness W, Kitchen N, Thomas D. Postimaging brain distortion: magnitude, correlates, and impact on neuronavigation. J Neurosurg 1998;88(4):656-62. 35. Roberts D, Hartov A, Kennedy F, Miga M, Paulsen K. Intraoperative brain shift and deformation: a quantitative analysis of cortical displacement in 28 cases. Neurosurgery 1998;43(4):749-58; discussion 58-60. 36. Dorward NL, Alberti O, Palmer JD, Kitchen ND, Thomas DG. Accuracy of true frameless stereotaxy: in vivo measurement and laboratory phantom studies. Technical note. J Neurosurg 1999;90(1):160-8. 37. Dorward NL, Paleologos TS, Alberti O, Thomas DG. The advantages of frameless stereotactic biopsy over frame-based biopsy. Br J Neurosurg 2002;16(2):110-8. 38. Grunert P, Charalampaki K, Kassem M, BoecherSchwarz H, Filippi R, Grunert P, Jr. Frame-based and frameless stereotaxy in the localization of cavernous angiomas. Neurosurg Rev 2003;26(1):53-61.

39

39. Schulder M, Fontana P, Lavenhar MA, Carmel PW. The relationship of imaging techniques to the accuracy of frameless stereotaxy [In Process Citation]. Stereotact Funct Neurosurg 1999;72(2–4):136-41. 40. Quinones-Hinojosa A, Ware ML, Sanai N, McDermott MW. Assessment of image guided accuracy in a skull model: comparison of frameless stereotaxy techniques vs. frame-based localization. J Neurooncol 2006; 76(1):65-70. 41. Levivier M, Goldman S, Pirotte B, Brucher J-M, Baleriaux D, Brotchi J. Diagnostic yield of stereotactic brain biopsy guided by positron emission tomography with [18-F] fluorodeoxyglucose. J Neurosurg 1995;83: 445-52. 42. Paleologos TS, Dorward NL, Wadley JP, Thomas DG. Clinical validation of true frameless stereotactic biopsy: analysis of the first 125 consecutive cases. Neurosurgery 2001;49(4):830-5; discussion 5-7. 43. Barnett GH, Miller DW, Weisenberger J. Frameless stereotaxy with scalp-applied fiducial markers for brain biopsy procedures: experience in 218 cases. J Neurosurg 1999;91(4):569-76. 44. Woodworth GF, McGirt MJ, Samdani A, Garonzik I, Olivi A, Weingart JD. Frameless image-guided stereotactic brain biopsy procedure: diagnostic yield, surgical morbidity, and comparison with the frame-based technique. J Neurosurg 2006;104(2):233-7. 45. Dammers R, Haitsma IK, Schouten JW, Kros JM, Avezaat CJ, Vincent AJ. Safety and efficacy of frameless and frame-based intracranial biopsy techniques. Acta Neurochir (Wien) 2008;150(1):23-9. 46. Gildenberg P. The history of stereotactic neurosurgery. Neurosurg Clin No Am 1990;1:765-81. 47. David M, Ruggiero G, Talairach J. Comparison between encephalography and ventriculography. Acta radiol 1954;42(1):37-42. 48. Tasker RR. One man’s recollection of 50 years of functional and stereotactic neurosurgery. Neurosurgery 2004; 55(4):968-74; discussion 74-6. 49. Levesque MF, Wilson CL, Behnke EJ, Zhang JX. Accuracy of MR-guided stereotactic electrode implantation. 1. Preimplantation correlation with the Talairach system. Stereotact Funct Neurosurg 1990;54–55:51-5. 50. Hawrylyshyn PA, Tasker RR, Organ LW. Third ventricular width and the thalamocapsular border. Appl Neurophysiol 1976;39(1):34-42. 51. Narabayashi H. Recent status of stereotaxic surgery. Surg Neurol 1983;19(6):493-6. 52. Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computerized tomography-guided determinations of brain targets in functional stereotaxis. J Neurosurg 1990;73(4):565-71. 53. Mandybur G, Morenski J, Kuniyoshi S, Iacono RP. Comparison of MRI and ventriculographic target acquisition for posteroventral pallidotomy. Stereotact Funct Neurosurg 1995;65(1–4):54-9.

613

614

39


54. Schuurman PR, de Bie RM, Majoie CB, Speelman JD, Bosch DA. A prospective comparison between threedimensional magnetic resonance imaging and ventriculography for target-coordinate determination in frame-based functional stereotactic neurosurgery. J Neurosurg 1999;91(6):911-4. 55. Zonenshayn M, Rezai AR, Mogilner AY, Beric A, Sterio D, Kelly PJ. Comparison of anatomic and neurophysiological methods for subthalamic nucleus targeting. Neurosurgery 2000;47(2):282-92; discussion 92-4. 56. Hutchison WD, Allan RJ, Opitz H, Levy R, Dostrovsky JO, Lang AE, Lozano AM. Neurophysiological identification of the subthalamic nucleus in surgery for Parkinson’s disease. Ann Neurol 1998;44(4):622-8. 57. Weinberger M, Hamani C, Hutchison WD, Moro E, Lozano AM, Dostrovsky JO. Pedunculopontine nucleus microelectrode recordings in movement disorder patients. Exp Brain Res 2008;188(2):165-74. 58. Spiegelmann R, Nissim O, Daniels D, Ocherashvilli A, Mardor Y. Stereotactic targeting of the ventrointermediate nucleus of the thalamus by direct visualization with highfield MRI. Stereotact funct Neurosurg 2006;84(1):19-23. 59. Azmi H, Schulder M. Stereotactic accuracy of a 3-tesla magnetic resonance unit. Stereotact funct Neurosurg 2003;80(1–4):140-5. 60. Starr PA, Christine CW, Theodosopoulos PV, Lindsey N, Byrd D, Mosley A, Marks WJ, Jr. Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imagingverified lead locations. J Neurosurg 2002;97(2):370-87. 61. Machado A, Rezai AR, Kopell BH, Gross RE, Sharan AD, Benabid AL. Deep brain stimulation for Parkinson’s disease: surgical technique and perioperative management. Mov Disord 2006;21 Suppl 14:S247-58. 62. diPierro CG, Francel PC, Jackson TR, Kamiryo T, Laws ER, Jr. Optimizing accuracy in magnetic resonance imaging-guided stereotaxis: a technique with validation based on the anterior commissure-posterior commissure line. J Neurosurg 1999;90(1):94-100. 63. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, Schwalb JM, Kennedy SH. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45(5):651-60. 64. Greenberg BD, Malone DA, Friehs GM, Rezai AR, Kubu CS, Malloy PF, Salloway SP, Okun MS, Goodman WK, Rasmussen SA. Three-year outcomes in deep brain stimulation for highly resistant obsessivecompulsive disorder. Neuropsychopharmacology 2006; 31(11):2384-93. 65. Feigin A, Kaplitt MG, Tang C, Lin T, Mattis P, Dhawan V, During MJ, Eidelberg D. Modulation of metabolic brain networks after subthalamic gene therapy for Parkinson’s disease. Proc Natl Acad Sci USA 2007;104(49): 19559-64. 66. Marks WJ, Jr., Ostrem JL, Verhagen L, Starr PA, Larson PS, Bakay RA, Taylor R, Cahn-Weiner DA,

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

Stoessl AJ, Olanow CW, Bartus RT. Safety and tolerability of intraputaminal delivery of CERE-120 (adenoassociated virus serotype 2-neurturin) to patients with idiopathic Parkinson’s disease: an open-label, phase I trial. Lancet Neurol 2008;7(5):400-8. Holloway KL, Gaede SE, Starr PA, Rosenow JM, Ramakrishnan V, Henderson JM. Frameless stereotaxy using bone fiducial markers for deep brain stimulation. J Neurosurg 2005;103(3):404-13. Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation for the treatment of central pain. Acta Neurochir 1991;52: 137-9. Arle JE, Apetauerova D, Zani J, Deletis DV, Penney DL, Hoit D, Gould C, Shils JL. Motor cortex stimulation in patients with Parkinson disease: 12-month follow-up in 4 patients. J Neurosurg 2008;109(1):133-9. Mogilner AY, Rezai AR. Epidural motor cortex stimulation with functional imaging guidance. Neurosurg Focus 2001;11(3):E4. Pillay PK, Barnett G, Awad I. MRI-guided stereotactic placement of depth electrodes in temporal lobe epilepsy. Br J Neurosurg 1992;6(1):47-53. Parrent AG, Blume WT. Stereotactic amygdalohippocampotomy for the treatment of medial temporal lobe epilepsy. Epilepsia 1999;40(10):1408-16. Berger MS, Deliganis AV, Dobbins J, Keles GE. The effect of extent of resection on recurrence in patients with low grade cerebral hemisphere gliomas. Cancer 1994; 74(6):1784-91. Laws ER, Parney IF, Huang W, Anderson F, Morris AM, Asher A, Lillehei KO, Bernstein M, Brem H, Sloan A, Berger MS, Chang S. Survival following surgery and prognostic factors for recently diagnosed malignant glioma: data from the Glioma Outcomes Project. J Neurosurg 2003;99(3):467-73. Hill D, Maurer C, Jr., Maciunas R, Barwise J, Fitzpatrick J, Wang M. Measurement of intraoperative brain surface deformation under a craniotomy. Neurosurgery 1998; 43(3):514-26; discussion 27-8. Black P, Moriarty T, Alexander E, 3rd, Stieg P, Woodard E, Gleason P, Martin C, Kikinis R, Schwartz R, Jolesz F. Development and implementation of intraoperative magnetic resonance imaging and its neurosurgical applications. Neurosurgery 1997;41(4): 831-42; discussion 42-5. Black PM, Alexander E, 3rd, Martin C, Moriarty T, Nabavi A, Wong TZ, Schwartz RB, Jolesz F. Craniotomy for tumor treatment in an intraoperative magnetic resonance imaging unit. Neurosurgery 1999;45(3): 423-31; discussion 31-3. Kanner AA, Vogelbaum MA, Mayberg MR, Weisenberger JP, Barnett GH. Intracranial navigation by using low-field intraoperative magnetic resonance imaging: preliminary experience. J Neurosurg 2002;97 (5): 1115-24.


79. Schulder M, Liang D, Carmel PW. Cranial surgery navigation aided by a compact intraoperative magnetic resonance imager. J Neurosurg 2001;94(6):936-45. 80. Fahlbusch R, Ganslandt O, Nimsky C. Intraoperative imaging with open magnetic resonance imaging and neuronavigation. Childs Nerv Syst 2000;16(10–11): 829-31. 81. Bohinski RJ, Kokkino AK, Warnick RE, Gaskill-Shipley MF, Kormos DW, Lukin RR, Tew JM, Jr. Glioma resection in a shared-resource magnetic resonance operating room after optimal image-guided frameless stereotactic resection. Neurosurgery 2001;48(4):731-42; discussion 42-4. 82. Wirtz CR, Bonsanto MM, Knauth M, Tronnier VM, Albert FK, Staubert A, Kunze S. Intraoperative magnetic resonance imaging to update interactive navigation in neurosurgery: method and preliminary experience. Comput Aided Surg 1997;2(3–4):172-9. 83. Rubino G, Farahani K, McGill D, Van De Wiele B, Villablanca J, Wang-Mathieson A. Magnetic resonance imaging-guided neurosurgery in the magnetic fringe fields: the next step in neuronavigation. Neurosurgery 2000;46(3):643-53; discussion 53-4. 84. Nimsky C, Ganslandt O, Fahlbusch R. Functional neuronavigation and intraoperative MRI. Adv Tech Stand Neurosurg 2004;29:229-63. 85. Nimsky C, Ganslandt O, Fahlbusch R. Implementation of fiber tract navigation. Neurosurgery 2006;58(4 Suppl 2): ONS-292-303; discussion ONS-4. 86. Sutherland G, Kaibara T, Louw D, Hoult D, Tomanek B, Saunders J. A mobile high-field magnetic resonance system for neurosurgery. J Neurosurg 1999;91(5):804-13. 87. Hall W, Liu H, Martin A, Pozza C, Maxwell R, Truwit C. Safety, efficacy, and functionality of high-field strength interventional magnetic resonance imaging for neurosurgery. Neurosurgery 2000;46(3):632-41; discussion 41-2. 88. Hall WA, Galicich W, Bergman T, Truwit CL. 3-Tesla intraoperative MR imaging for neurosurgery. J Neurooncol 2006;77(3):297-303. 89. Hall WA, Liu H, Truwit CL. Navigus trajectory guide. Neurosurgery 2000;46(2):502-4. 90. Hadani M, Katznelson E, Zuk Y, Sinai R. A novel, realtime intraoperative MR imaging-guided system: a wideopen, compact, and displaceable IMR imaging system for conventional neurosurgical operating rooms. J Neurosurg 2000;92:560. 91. Levivier M, Wikler D, De Witte O, Van de Steene A, Baleriaux D, Brotchi J. PoleStar N-10 low-field compact intraoperative magnetic resonance imaging system with mobile radiofrequency shielding. Neurosurgery 2003;53 (4):1001-6; discussion 7. 92. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J, Maniker AH, Carmel PW. Cranial surgery with an expanded compact intraoperative magnetic resonance imager. Technical note. J Neurosurg 2006;104 (4):611-7. 93. Azmi H, Biswal B, Salas S, Schulder M. Functional imaging in a low-field, mobile intraoperative magnetic

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

39

resonance scanner: expanded paradigms. Neurosurgery 2007;60(1):143-8; discussion 8-9. Nimsky C, Ganslandt O, Buchfelder M, Fahlbusch R. Glioma surgery evaluated by intraoperative low-field magnetic resonance imaging. Acta Neurochir Suppl 2003;85:55-63. Wirtz C, Knauth M, Staubert A, Bonsanto M, Sartor K, Kunze S, Tronnier V. Clinical evaluation and follow-up results for intraoperative magnetic resonance imaging in neurosurgery [In Process Citation]. Neurosurgery 2000;46(5):1112-20; discussion 20-2. Bohinski RJ, Warnick RE, Gaskill-Shipley MF, Zuccarello M, van Loveren HR, Kormos DW, Tew JM, Jr. Intraoperative magnetic resonance imaging to determine the extent of resection of pituitary macroadenomas during transsphenoidal microsurgery. Neurosurgery 2001;49(5): 1133-43; discussion 43-4. Fahlbusch R, Ganslandt O, Buchfelder M, Schott W, Nimsky C. Intraoperative magnetic resonance imaging during transsphenoidal surgery. J Neurosurg 2001; 95(3):381-90. Nabavi A, Black PM, Gering DT, Westin CF, Mehta V, Pergolizzi RS, Jr., Ferrant M, Warfield SK, Hata N, Schwartz RB, Wells WM, 3rd, Kikinis R, Jolesz FA. Serial intraoperative magnetic resonance imaging of brain shift. Neurosurgery 2001;48(4):787-97; discussion 97-8. Bernays RL, Kollias SS, Khan N, Brandner S, Meier S, Yonekawa Y. Histological yield, complications, and technological considerations in 114 consecutive frameless stereotactic biopsy procedures aided by open intraoperative magnetic resonance imaging. J Neurosurg 2002.97: 354-62. Schwartz TH, Marks D, Pak J, Hill J, Mandelbaum DE, Holodny AI, Schulder M. Standardization of amygdalohippocampectomy with intraoperative magnetic resonance imaging: preliminary experience. Epilepsia 2002;43(4):430-6. Buchfelder M, Ganslandt O, Fahlbusch R, Nimsky C. Intraoperative magnetic resonance imaging in epilepsy surgery. J Magn Reson Imaging 2000;12(4):547-55. Samdani AF, Schulder M, Catrambone JE, Carmel PW. Use of a compact intraoperative low-field magnetic imager in pediatric neurosurgery. Childs Nerv Syst 2005; 21(2):108-13; discussion 14. Martin AJ, Larson PS, Ostrem JL, Keith Sootsman W, Talke P, Weber OM, Levesque N, Myers J, Starr PA. Placement of deep brain stimulator electrodes using realtime high-field interventional magnetic resonance imaging. Magn Reson Med 2005;54(5):1107-14. Jankovski A, Raftopoulos C, Vaz G, Hermoye L, Cosnard G, Francotte F, Duprez T. Intra-operative MRI at 3T: short report. Jbr-Btr 2007;90(4):249-51. Matsumae M, Koizumi J, Fukuyama H, Ishizaka H, Mizokami Y, Baba T, Atsumi H, Tsugu A, Oda S, Tanaka Y, Osada T, Imai M, Ishiguro T, Yamamoto M, Tominaga J, Shimoda M, Imai Y. World’s first

615

616

39 106. 107. 108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.


magnetic resonance imaging/x-ray/operating room suite: a significant milestone in the improvement of neurosurgical diagnosis and treatment. J Neurosurg 2007; 107(2):266-73. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-9. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psyc 1983;46:797-803. Betti O, Derechinsky V. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurochir Suppl 1984;33:385-90. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988;22:454-64. Lunsford LD, Flickinger JC, Linder G, Maitz A. Stereotactic radiosurgery of the brain using the first United States 210 cobalt-60 source gamma knife. Neurosurg 1989;24:151-9. Hoh DJ, Liu CY, Pagnini PG, Yu C, Wang MY, Apuzzo ML. Chained lightning, part I: Exploitation of energy and radiobiological principles for therapeutic purposes. Neurosurgery 2007;61(1):14-27; discussion-8. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25:557-61. Kondziolka D, Flickinger JC, Perez B. Judicious resection and/or radiosurgery for parasagittal meningiomas: outcomes from a multicenter review. Gamma Knife Meningioma Study Group. Neurosurgery 1998;43(3): 405-13; discussion 13-4. Lunsford LD, Witt TC, Kondziolka D, Flickinger JC. Stereotactic radiosurgery of anterior skull base tumors. Clin Neurosurg 1995;42:99-118. Wowra B, Muacevic A, Jess-Hempen A, Tonn JC. Safety and efficacy of outpatient gamma knife radiosurgery for multiple cerebral metastases. Expert Rev Neurother 2004;4(4):673-9. Mehta VK, Lee QT, Chang SD, Cherney S, Adler JR, Jr. Image guided stereotactic radiosurgery for lesions in proximity to the anterior visual pathways: a preliminary report. Technol Cancer Res Treat 2002;1(3):173-80. Romanelli P, Heit G, Chang SD, Martin D, Pham C, Adler J. Cyberknife radiosurgery for trigeminal neuralgia. Stereotact funct Neurosurg 2003;81(1–4): 105-9. Villavicencio AT, Lim M, Burneikiene S, Romanelli P, Adler JR, McNeely L, Chang SD, Fariselli L, McIntyre M, Bower R, Broggi G, Thramann JJ. Cyberknife radiosurgery for trigeminal neuralgia treatment: a preliminary multicenter experience. Neurosurgery 2008;62(3): 647-55; discussion-55. Chen JC, Greathouse HE, Girvigian MR, Miller MJ, Liu A, Rahimian J. Prognostic factors for radiosurgery treatment of trigeminal neuralgia. Neurosurgery 2008;62(5 Suppl):A53-60; discussion A-1.

120. Jawahar A, Kondziolka D, Kanal E, Bissonette DJ, Lunsford LD. Imaging the trigeminal nerve and pons before and after surgical intervention for trigeminal neuralgia. Neurosurgery 2001;48(1):101-6; discussion 6-7. 121. Tishler RB, Loeffler JS, Lunsford LD, Duma C, Alexander E, Kooy H, Flickinger J. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993;27:215-21. 122. Sims E, Doughty D, Macaulay E, Royle N, Wraith C, Darlison R, Plowman PN. Stereotactically delivered cranial radiation therapy: a ten-year experience of linac-based radiosurgery in the UK. Clin Oncol (R Coll Radiol) 1999;11(5):303-20. 123. Witt TC, Kondziolka D, Baumann SB, Noll DC, Small SL, Lunsford LD. Preoperative cortical localization with functional MRI for use in stereotactic radiosurgery. Stereotact funct Neurosurg 1996; 66(1–3):24-9. 124. Liu WC, Schulder M, Narra V, Kalnin AJ, Cathcart C, Jacobs A, Lange G, Holodny AI. Functional magnetic resonance imaging aided radiation treatment planning. Med Phys 2000;27(7):1563-72. 125. Maruyama K, Kamada K, Shin M, Itoh D, Masutani Y, Ino K, Tago M, Saito N. Optic radiation tractography integrated into simulated treatment planning for Gamma Knife surgery. J Neurosurg 2007;107(4): 721-6. 126. Maruyama K, Kamada K, Ota T, Koga T, Itoh D, Ino K, Aoki S, Tago M, Masutani Y, Shin M, Saito N. Tolerance of pyramidal tract to gamma knife radiosurgery based on diffusion-tensor tractography. Int J Radiat Oncol Biol Phys 2008;70(5):1330-5. 127. Deogaonkar M, Walter BL, Boulis N, Starr P. Clinical problem solving: finding the target. Neurosurgery 2007;61(4):815-24; discussion 24-5. 128. Okun MS, Vitek JL, DeLong MR. Not our knife: gamma knife surgery for Parkinson disease. Arch Neurol 2002;59(8):1334-5; author reply 5. 129. De Salles AA, Frighetto L, Lacan G, Melega W. Radiosurgery can achieve precision needed for functional neurosurgery. Arch Neurol 2003;60(10):1494-6; author reply 6. 130. Sato S, Ohye C, Shibazaki T, Zama A, Cai X. Neurophysiological evaluation of the optimum target in gamma thalamotomy: indirect evidence. Stereotact Funct Neurosurg 2005;83(2–3):108-12; discussion 13-4. 131. Young RF, Shumway-Cook A, Vermeulen SS, Grimm P, Blasko J, Posewitz A, Burkhart WA, Goiney RC. Gamma knife radiosurgery as a lesioning technique in movement disorder surgery. J Neurosurg 1998; 89(2):183-93. 132. Kondziolka D, Ong JG, Lee JY, Moore RY, Flickinger JC, Lunsford LD. Gamma Knife thalamotomy for essential tremor. J Neurosurg 2008;108(1):111-7.


133. Mathieu D, Kondziolka D, Niranjan A, Flickinger J, Lunsford LD. Gamma knife thalamotomy for multiple sclerosis tremor. Surg Neurol 2007;68(4):394-9. 134. Keep MF, Mastrofrancesco L, Erdman D, Murphy B, Ashby LS. Gamma knife subthalamotomy for Parkinson disease: the subthalamic nucleus as a new

39

radiosurgical target. Case report. J Neurosurg 2002; 97(5 Suppl):592-9. 135. Anschel DJ, Romanelli P, Benveniste H, Foerster B, Kalef-Ezra J, Zhong Z, Dilmanian FA. Evolution of a focal brain lesion produced by interlaced microplanar X-rays. Minim Invasive Neurosurg 2007;50(1):43-6.

617

48 Novel Therapies for Brain Tumors G. Al-Shamy . R. Sawaya

Introduction Approximately 18,000 new cases of primary brain tumors are diagnosed annually in the United States [1]. Despite significant advancements in neurosurgical techniques, radiotherapy, imaging modalities, and molecular neuro-oncology, there remains little improvement in clinical outcome for most patients. In fact, the median survival time of patients with malignant gliomas, particularly glioblastoma multiforme (GBM), the most common primary brain tumor, remains less than 1 year, with more than 90% of patients succumbing within 5 years of diagnosis [2]. Several factors are thought to underlie the lack of progress in developing effective treatments for malignant brain tumors. First, the central nervous system (CNS) presents a unique environment with limited capacity for self repair. The presence of the blood-brain barrier (BBB) further complicates systemic delivery of chemotherapeutic agents to CNS lesions. In addition, malignant brain tumor cells have unique innate properties that pose additional problems. These tumors are inherently aggressive, as highlighted by their remarkable degree of resistance to conventional therapies. Their widely infiltrating nature also makes them suboptimal candidates for surgical intervention. Finally, the lack of predictive preclinical models, coupled with our relatively poor understanding of glioma pathogenesis, has impeded the development of novel treatment options. The current ‘‘gold standard’’ for the management of malignant gliomas involves an attempt at complete or maximal safe surgical resection #


(as specified in the National Comprehensive Cancer Network CNS guidelines [3]) in conjunction with radiotherapy and temozolomide (TMZ) chemotherapy, followed by six monthly cycles of TMZ. Although the above multimodality treatment regimen provides an additional improvement above the median survival time seen in patients treated without TMZ (14.6 and 12 months, respectively), clinical recurrence or progression is nearly universal [4,5]. Conventional treatment modalities have largely failed to yield significant progress in the treatment of malignant gliomas, despite extensive research efforts over the past two decades. This has led to increasing interest in exploring alternative treatment strategies, including tumor-specific immunotherapy, gene therapy, molecularly targeted therapeutics and tumor oncolysis. Additionally, recent advances in the knowledge of basic tumor biology have led to a paradigm shift in the way brain tumors are being studied, allowing for novel approaches to understanding gliomagenesis and glioma therapeutics. Strategies to overcome the therapy-limiting properties of the CNS, including attempts at improving delivery of therapeutic agents via targeted delivery systems, have now become a major focus of brain tumor research. The aim of this chapter is to consolidate the various recently developed novel glioma therapies highlighted above, along with several others. We will not only examine our current understanding of the molecular events underlying malignant glioma pathogenesis but also focus on the history, evolution, and clinical implementation of new treatment modalities. The purpose of this review

750

48

Novel therapies for brain tumors

is to provide the reader with a broad perspective on the various applications of novel brain tumor therapeutics.

Direct Delivery of Chemotherapeutic Agents In light of the limitations posed by the BBB in systemic drug delivery, direct intracranial delivery of chemotherapeutic agents presents a unique approach for treating malignant gliomas by obviating the need for a drug to cross the BBB and limiting the toxicity of systemic anticancer agents. This technology makes it possible to achieve very high local concentrations of chemotherapeutic agents at the tumor site, while avoiding the side effects associated with systemic administration of extremely high doses. The two main approaches that have yielded the most promising results are polymerically-controlled release and convectionenhanced delivery. Each system has its own advantages and disadvantages, as detailed below.

Polymerically-Controlled Release In an effort to improve local control after brain tumor resection, researchers have developed controlled-release polymers that are implanted directly at the resection site to allow for restricted slow release of chemotherapeutic agents into the tumor bed. This system is particularly useful because it can provide reliable sustained drug release for periods of days, or even years [6]. Constructed from either biodegradable or nonbiodegradable polymers, the polymerically-controlled release system depends on diffusion of the drug through the polymer matrix. Biodegradable polymer systems have been considered more attractive for clinical application because they do not require removal at a later date. Importantly, the diffusion kinetics of these systems have been well characterized [7], and the use of these systems has allowed the

emergence of a number of clinical studies exploring alternative treatment strategies, including gene therapy and tumor-specific immunotherapy, with further applications under investigation. The most common type of polymer currently used intracranially is poly[bis(p-carboxyphenoxy) propane-sebacic acid]. This polymer has achieved the best success when loaded with nitrosourea (BCNU) in the form of Gliadel®. A randomized, placebo-controlled, double blind, prospective, phase III clinical trial demonstrated an additional median survival benefit of 8 weeks with the use of Gliadel® polymers in patients with recurrent GBM [8]. This study failed to identify any local or systemic adverse effects attributed directly to Gliadel®. Several additional studies have since evaluated the effectiveness and safety of Gliadel® in the initial treatment of malignant gliomas. A prospective, randomized double blind clinical trial in Europe [9] reported a statistically significant benefit in survival with Gliadel® relative to placebo in patients with GBM (53 and 40 weeks, respectively; p = 0.0083). Therapy with Gliadel®, even when combined with radiotherapy, has been shown to be well tolerated, with no significant increase in toxicity, infection, or inflammation [10]. This method of delivery has proved to be particularly attractive clinically in that it adds minimal complexity, requires no additional surgical procedures, and minimizes solubility limitations. A main limitation of polymerically-controlled release, however, is that local penetration of the drug is frequently hindered by diffusion [11]. In addition to the local environment of a tumor, the extracellular matrix of the brain parenchyma also imposes its own limits on diffusive transport. Together these factors confine high drug concentrations to within approximately 3 mm of the delivery site [12]. The poor drug penetration and the consequent dependence of the drug dosage on implant size has therefore hindered the therapeutic benefits of polymerically-controlled release, despite all of the advantages of the system.


Convection-Enhanced Delivery Convection-Enhanced Delivery (CED) is an alternative method of controlled local drug release that has been developed to deliver compounds throughout the brain so as to overcome the diffusion barrier seen with polymerically-controlled release systems [13]. CED utilizes an applied external positive pressure infusion to induce fluid convection in the brain and thus force chemotherapeutic drugs throughout the parenchyma via interstitial spaces. Fluid is typically administered via a small catheter using a pump [14]. The main benefit of administering drugs by CED is a greater distribution volume and a longer infusion time, allowing for continued drug exposure. An additional benefit arises from the flexibility of the technique and its ability to be used with gene therapy as well as with immunotherapeutics. Multiple therapeutic agents have been evaluated in clinical trials for glioma treatment using this method [15]. Perhaps the most promising agent investigated thus far is IL13-PE38QQR (cintredekin besudotox). This synthetic drug is derived from a human protein, interleukin 13 (IL13), linked to a bacterial toxin, Pseudomonas exotoxin (PE). The IL13 portion of the drug is able to bind to specific receptors on the tumor cells, allowing for a form of targeted therapy. A recent phase III double blind, randomized controlled trial compared the clinical outcome between implantation of Gliadel® wafers containing BCNU (carmustine) and CED using IL13-PE38QQR in the treatment of patients with recurrent GBMs. The median survival times were comparable between the two modalities, though there was some suggestion of additional benefit with CED using IL13-PE38QQR. An additional advantage of the use of CED is its ability to permit the monitoring of the distribution of chemotherapeutic agents and to thus provide some measure of controlled delivery. Furthermore, through the use of gadolinium liposomes, the infusion of particles can be monitored

48

directly using magnetic resonance imaging (MRI). More recently, this imaging method has been validated in a primate model [16]. Just as with polymerically-controlled release systems, CED has its own shortcomings and complications. Perhaps most importantly, it is an invasive procedure that has the potential for inducing high intracranial pressures. It also has the disadvantage of having an unpredictable drug distribution [17]. Current research is aimed at developing better methods to track infused agents and to optimize the efficacy of CED for brain tumor therapeutics. Interstitial drug delivery of chemotherapeutic agents does provide an effective means of bypassing the BBB, minimizing systemic toxicity, and producing high concentrations of the drug at the site of interest. Nontheless, clinical trials have shown only a modest improvement in survival time with this method when compared with conventional therapy, owing to inherent limitations of the technology. Further development may allow this unique approach to increase the efficacy of novel antitumor therapeutics and conventional multimodality therapies and to thereby potentially improve the survival benefits in the clinic.

Molecularly Targeted Therapy The recent advances in our understanding of the molecular biology of brain tumors and the discovery of various dysregulated cell signaling pathways found in glioma cells (> Figure 48-1) have set forward a number of unconventional targets that could be used in the treatment of malignant gliomas. A number of approaches aimed at inhibiting these signaling pathways, ranging from upstream growth factor ligands and their receptors to downstream intracellular effectors, have been explored. To date, the targets identified include vascular endothelial growth factor (vEGF), platelet-derived growth factor (PDGF), agents targeting components of the Ras- and AKT (proto-oncogene

751

752

48


. Figure 48-1 A schematic overview of the major signaling pathways involved in glioma progression, including those involving the epidermal growth factor receptor/phosphatidylinositol-3’ kinase (EGFR/PI3K), platelet-derived growth factor (PDGFR), protein kinase C (PKC), and vascular endothelial growth factor receptor (vEGFR). Inhibitors currently in clinical trials are highlighted in red at the various stages of the pathways on which they act. (Abbreviations: mTOR, mammalian target of rapamycin; PTEN, phosphatase and tensin homolog)

protein)-mediated pathways, and the human epidermal growth factor receptor (HER2/neu). Each of these molecules has been studied extensively and found to play a critical role in tumorigenesis and disease progression [18].

Human Epidermal Growth Factor Receptor The epidermal growth factor receptor (EGFR), a member of the tyrosine kinase receptor family, is


a cell-surface transmembrane protein that has been implicated in cell growth and proliferation, as well as in cell survival, motility, and resistance to chemotherapy and radiation therapy [19]. The EGFR gene is located on chromosome seven and encodes a 170 kDa transmembrane glycoprotein with intrinsic tyrosine kinase activity [20]. In GBMs, the EGFR is both overexpressed (approximately 50% of tumors have chromosome seven EGFR locus amplification) and mutated (by deletion of exons 2–7), giving rise to a ligand-independent, constitutively activated form called EGFRvIII [20,21]. EGFRvIII strongly and persistently activates the phosphotidylinositol-3’ kinase (PI3K) survival and antiapoptotic pathway, a signaling cascade that is essential in providing radioresistance to glioma cells [22,23]. More recently, a study by Pelloski and colleagues [24] has also suggested that having an EGFRvIII-positive GBM is an independent prognostic factor for poor survival outcome. Recently, three EGFR inhibitors (> Figure 48-1) have been approved by the FDA for use in clinical trials. The results with gefitinib (Iressa), the first of these inhibitors, were largely disappointing. A phase II study using gefitinib as the sole agent for treating recurrent tumors showed no radiographic tumor regression [25]. In a second study, 98 newly diagnosed GBM patients receiving gefitinib in conjunction with radiation therapy failed to show any clinical improvement or survival benefit [26]. Importantly, in both of these studies, there was no relation between clinical outcome and degree of EGFR expression. The results with erlotinib (Tarceva), a related drug, have been more variable. A number of phase II trials showed tumor regression, with 6-month progression-free survival (PFS) rates ranging from 17 to 27% [27,28] in patients treated with Tarceva. Promising results have also come from a phase I trial combining Tarceva with TMZ [29]. However, these results failed to be duplicated when Tarceva was added to the standard protocols for patients newly diagnosed with GBM [30]. Cetuximab, the

48

third EGFR inhibitor agent (a monoclonal antibody), is currently being studied as a possible treatment in patients with recurrent GBM [31]. Two possibilities have been suggested for the relative ineffectiveness of EGFR inhibitors. One of these has to do with the complexity of cellular signaling and the interactions of various downstream molecules and pathways. That is, inhibition of the receptor alone may not be sufficient to mitigate downstream signaling, and additional inhibition of downstream effectors may be necessary. In fact, the protein kinase B (PKB)/AKT signaling pathway has been of interest recently, as its inhibition significantly increases the effectiveness of EGFR inhibitors. Studies combining EGFR inhibitors with rapamycin, an inhibitor of the PKB/AKT signaling pathway, have yielded promising results [32,33]. The second potential reason for the lack of efficacy of EGFR inhibitors stems from the fact that 40% of the patients with EGFR overexpression have the mutated EGFRvIII. This mutation bypasses the need for ligand binding and yields a constitutively active receptor even in the absence of growth promoting signals and factors. Consequently, recent work has focused on the development of a vaccine specific to EGFRvIII.

Vascular Endothelial Growth Factor (vEGF) and Receptor (vEGFR) A characteristic feature of GBM is exuberant angiogenesis, a key event in tumor growth and progression. One possible mechanism underlying tumor neovascularization may involve the recruitment of native blood vessels to the tumor through expression of hypoxia-inducible factor (HIF)-1 alpha and vascular endothelial growth factor (vEGF) in perinecrotic pseudopalisading glioma cells. VEGF is a secreted peptide that acts through its receptors, FLT1 and FLK-1-KDR (vEGFR2), to stimulate endothelial cell division and the formation of new blood vessels [34].

753

754

48


Recently, Avastin (bevacizumab), a recombinant human antibody specifically designed to block vEGF, was first used for patients with recurrent brain tumors in combination with CPT 11 (irinotecan), a camptothecin-derived anticancer agent with DNA topoisomerase one inhibitory activity [35]. Although the initial results were very encouraging (after the initial course of treatment, significant tumor regression was observed), long-term survival data proved to be disappointing. A number of recent trials studying the effect of Avastin and CPT 11 on malignant gliomas suggest that patients treated with this combination show a rapid initial response followed by a rapid regrowth of the tumor thereafter [36]. Taken together, the results of these studies demonstrated a median survival interval similar to that seen with conventional therapy, but with the added complications of increased risk of internal bleeding and blood clots. Vatalanib, a vEGFR inhibitor, has also been evaluated recently for treatment of patients with gliomas in a multicenter phase I/II trial, both alone and in combination with chemotherapy. The results in both cases were equivalent, with 66% of patients showing stable disease after treatment [37]. More recently, a phase II trial of AZD2171 (Astrazeneca Pharmaceuticals), a vEGFR2 inhibitor, has been initiated in the treatment of recurrent GBM [38,39].

tumor growth by indirect mechanisms pertaining to the tumor stroma or vasculature [45–47]. Gleevec (STI-571, imatinib mesylate), an inhibitor of the bcr-abl tyrosine kinase involved in the growth of chronic myelogenous leukemia, has recently been shown to inhibit glioma growth in animal studies [48]. The reason lies in the large sequence homology and biochemical similarity between the tyrosine kinase domain of the bcrabl protein and the PDGFR. Yet, these results could not be repeated in clinical studies. As a monotherapy, Gleevec had minimal success, failed to show any significant clinical benefit, and was associated with increased risk of intracranial hemorrhage [49]. Additional preclinical studies have suggested that combining Gleevec with hydroxyurea may provide additional cytotoxic effects. In vitro studies have demonstrated that Gleevec increases the chemo- or radiosensitivity of GBM cells in culture [50–52], suggesting that it may act to enhance the activity of chemotherapeutic agents currently used to treat GBM. Hydroxyurea, a cytotoxic agent that inhibits DNA synthesis, is widely used in cancer therapy and penetrates the BBB [53,54]. More recently, a study combining Gleevec and hydroxyurea demonstrated encouraging antitumor efficacy, with a reported 6-month PFS rate of 32% [55]. This result was later confirmed in a phase II study [56].

Protein Kinase C Pathway Platelet-Derived Growth Factor Receptor A number of preclinical studies have provided considerable support for the hypothesis that autocrine signaling by platelet-derived growth factor (PDGF) plays a key role in gliomagenesis [40–42]. These findings suggest that inhibition of PDGF receptors (PDGFRs) may be able to arrest GBM progression by disrupting autocrine signaling and, consequently, the glioma cell cycle [43,44]. Alternatively, inhibition of PDGFRs may affect GBM

Protein kinase C (PKC) is a serine/threonine kinase that has been shown to regulate tumor cell proliferation, migration/invasion, and angiogenesis [39]. Tamoxifen, an antiestrogen medication historically used in the treatment of breast cancer, plays a role in inhibiting the enzymatic reaction of PKC. One stage II clinical trial reported a considerable survival benefit associated with the administration of oral tamoxifen to patients with recurrent gliomas [57]. Tumor regression was reported in 25% of patients and


stabilization of tumor growth for an additional 20%. Despite the reported benefits, however, this modality of therapy did pose some potential problems. The dosage of tamoxifen used in this study was significantly higher than the dose typically used in breast cancer patients, which poses an increased risk of blood clots, weight gain, and uterine cancer in women and decreased libido in men as common side effects. Nevertheless, the capacity of tamoxifen to inhibit the chemoresistance of brain tumor cells made it a suitable agent to use in combination with traditional chemotherapeutic approaches. Tamoxifen has thus been studied in combination with carboplatin as well as BCNU for the treatment of gliomas. Mastronardi and colleagues studied the role of tamoxifen and carboplatin for patients with newly diagnosed brain tumors. This group reported a 1- and 2-year survival rate of 52 and 32%, respectively [58]. Unfortunately, this result could not be replicated in later studies. Tamoxifen has also been evaluated in combination with BCNU as the initial treatment after radiation in a number of phase II and phase III clinical studies [59,60], with variable results. Some studies reported long-term survival effects with tamoxifen, and others reported no significant increase in overall median survival. This controversy was recently explained by a Canadian study using magnetic resonance spectroscopy. The authors demonstrated that tamoxifen appears to work on only a minority of patients and that one can predict which patients will respond to it based on the presence of different metabolites [61].

48

of the second messenger molecule PIP3. Consequently, PTEN inactivation results in accumulation of PIP3 in cells and renders the PI3K pathway constitutively active via activity of the serine/threonine kinase AKT/PKB cascade downstream of PI3K. Interestingly, PTEN loss has also been shown to promote resistance to EGFR kinase inhibitors by dissociating EGFR/ EGFRvIII inhibition from downstream inhibition of PI3K signaling cascades. Overall, loss of PTEN and constitutive activation of PI3K signaling results in promotion of antiapoptotic effects, cell growth and proliferation. In glioma patients, activation of PI3K is associated with poor outcome [63]. Consistent with this, increased levels of AKT and PKB phosphorylation downstream of PI3K have been clinically associated with resistance to Tarceva in malignant glioma patients. None of the tumors expressing high levels of PKB/AKT responded to Tarceva administration, whereas 8 of 18 tumors with low PKB/AKT levels responded to treatment. The mammalian target of rapamycin (mTOR) is another serine/threonine kinase found downstream of PI3K. Rapamycin and its synthesized analog, temsirolimus, inhibit mTOR signaling and have been evaluated for the treatment of recurrent gliomas. Phase II studies with temsirolimus have failed to show any survival benefit [64,65]. Perifosine, an oral AKT inhibitor, is currently undergoing clinical evaluation in patients with malignant gliomas. Despite all of these negative results, molecularly targeted therapy offers great promise.

The Role of Immunotherapy PTEN Signaling Pathway The phosphatase and tensin homolog (PTEN) is a phosphatase that inhibits PI3K and is lost in 50% of patients with GBM [62]. PTEN has been shown to negatively regulate the phosphatidylinositol-3-kinase (PI3K) pathway by removing a phosphate from the inositol ring

Brain tumors have long been considered to arise at a site that is privileged with respect to immune surveillance. A number of studies have demonstrated that gliomas express immunosuppressive characteristics (> Figure 48-2), both locally [66] and systemically [67]. Gliomas have been shown to be associated with overall lymphopenia,

755

756

48


. Figure 48-2 A schematic representation of the interaction of glioma cells with the native immune system. By employing various signaling mechanisms, glioma cells are able to evade the immune response and proliferate within the brain. A number of strategies currently being investigated are aimed at enhancing the immune response as an effective approach to cancer treatment. (Abbreviations: GM-CSF, granulocyte/macrophage-colony stimulating factor; IL-2, interleukin-2; TGF-ß, transforming growth factor beta)

depressed antibody production, and impaired antigen-presenting cell (APC) function [68–70]. Additionally, patients with gliomas present with depressed peripheral T-cell responsiveness as well as depressed T-cell receptor-mediated signaling [71]. Taken together, these studies strongly suggest that weakening the immune system is a fundamental characteristic of the malignancy of GBM. A natural corollary to this idea would be that strengthening one’s immune system may be an effective approach to cancer treatment. This concept has been further validated in experimental rodent model systems of intracranial gliomas, where boosting otherwise impaired tumor-specific immune responses can successfully eradicate wellformed neoplasms [72,73]. Overexpression of transforming growth factor (TGF)-ß2 is thought to be a primary contributing factor to the immunosuppressed state seen in patients with

gliomas and, consequently, a significant cause of the failure of current immunotherapeutic strategies [74]. To combat this, Gorelik and colleagues [75] recently demonstrated that with T-cell-specific blockade of TGF signaling, an effective immune response could be generated that was capable of eradicating tumors in mice. A similar strategy was employed by Ruffini and colleagues, who demonstrated that by effectively antagonizing TGF-ß2 secretion with a neutralizing antibody, the in vitro proliferative capacity and antitumor cytotoxicity of adherent lymphokine-activated killer (LAK) cells could be enhanced [76]. More recently, treatments that involve tumorspecific immune reactions have received great attention because of their high benefit-to-risk potential. Both passive and active immunotherapeutic approaches have been attempted, both alone and in combination, in patients with gliomas.


Passive Immunotherapeutic Approach Receptor/Antigen Targeting Recent advances in the understanding of the underlying molecular biology of brain tumors have led to the discovery of cell-surface epitopes specific to cancer cells. Developing antibodies against these epitopes offers a potential way to selectively target and eradicate neoplastic cells dispersed within normal brain tissue and is a very attractive means of eliminating all intracranial neoplastic foci left behind after surgical resection of the primary tumor mass. Tenascin, an extracellular matrix protein with unknown function, is one of the recently discovered targets that has been studied in clinical trials. It has been detected in almost all highgrade gliomas [77], and monoclonal antibodies specific to it have been generated successfully. The BC2 form of the antibody has been conjugated to radioactive iodine-131 (I-131) and infused directly into the tumor cavity after resection of both newly diagnosed and recurrent high-grade gliomas [78–80]. Radioimmunotherapy was used in the above trials in order to provide an additional radiation boost to the tumor bed without incurring the known toxicity effects of external beam irradiation. More recently, a phase II study using an I-131-labeled antitenascin monoclonal antibody in patients with recurrent brain tumors reported a significantly better median survival time with this method compared with conventional therapy [81]. This encouraging result has been reproduced in a different study but with a higher incidence of hematological and neurological toxicity [82]. A variation on this approach involves the use of interleukin-13, which is conjugated to Pseudomonas endotoxin, a bacterial toxin that has been shown to be lethal to glioma cells in previous studies. A phase I/II study employing this agent

48

to treat recurrent brain tumors recorded a median survival interval of 46 weeks [83]. Another variation on this approach involves transferrin-CRM, which is conjugated to a modified diphtheria toxin. Using this conjugate, a phase I study reported an encouraging result, with significant tumor regression in more than 50% of patients treated [84]. A subsequent phase II study was able to reproduce this positive result, with 35% of patients reported to have tumor regression and a median survival time of 37 weeks [85].

Cytokine/Modulator Therapy Cytokines play a critical role in the induction, stimulation, and promotion of the immune response. Consequently, they have received significant attention in research as a possible tool to potentiate the immune response and induce cell-mediated antitumor immunity. A number of different studies have used different approaches for the treatment of gliomas, including systemic, intrathecal, and intratumoral administration of cytokines. One of the first cytokines to be studied in gliomas was interleukin-2 (IL-2), an important T-cell growth factor that has been implicated in the proliferation of CD8 + T cells and the modulation of their cytotoxic activity. A number of early in vitro studies supported a rationale for the use of IL-2 in patients with malignant brain tumors. This idea was largely based on the ability of IL-2 to block the T-cell depressing activity of TGF-ß, a key mediator of the suppressed immune response observed in glioma patients [86,87]. However, the use of IL-2 proved to have several important limitations, including not only being poor at blocking the effects of TGF-ß but also producing increased vascular permeability leading to brain edema. Several clinical trials have also studied the role of interferons alpha and beta for treating patients with recurrent gliomas. The primary

757

758

48


constraints in these studies were the lack of reproducibility as well as poor study design [88,89]. Nevertheless, more recent studies have described novel approaches for the delivery of these cytokines. These include techniques that are cell based, gene therapy based, and neural stem cell based. These methods allow for the robust delivery of tumoricidal cytokines directly to neoplastic regions, thereby circumventing the limitations of systemic therapy.

Immunological Cell Transfer In the normal immune system, immune surveillance is regulated by natural killer (NK) cells and to a lesser degree by cytotoxic T cells. In a neoplastic setting, this same mechanism is employed to eliminate newly formed neoplastic cells, with the goal of preventing their propagation and consequently hindering the formation of clinically detectable masses. Recently, a new approach called adoptive cell transfer has been developed to try to harness the cytotoxic abilities of T-cell populations by transplanting potentially tumoricidal T cells into patients with malignant brain tumors. Delivery methods can vary from systemic infusion to local intracerebral inoculation. To date, use of LAK cells has produced the most promising results. LAK cells are peripherally circulating populations of lymphocytes that have been shown to be capable of lysing NK cellresistant tumor cells in vitro after exogenous stimulation with IL-2 [90]. For adult patients with recurrent malignant gliomas, Hayes and colleagues have reported an improved long-term survival time following IL-2 and LAK cell infusion into the tumor cavity after resection [91]. Unfortunately, as with many of the other methods detailed thus far, these results have not been reproducible and are complicated further by side effects resulting from administration of IL-2. Additional significant limitations to treatment

using T-cell populations that have been expanded and sensitized in vitro are their lack of specificity and their inability to retain memory (unlike normal immune cells in the body).

The Role of APCs/Dendritic Cells Perhaps the most critical step in inducing tumor antigen-specific immunity involves devising ways to effectively deliver tumor antigens to T cells and developing methods to enhance detection of tumor antigens. Currently, these are both areas of intense research in tumor biology. The basic understanding of normal immunological principles has greatly underscored the function of these cells in a neoplastic setting. The presentation of antigens to naı¨ve T cells triggers a cascade of events comprising an active immune response that includes clonal expansion, the formation of memory cells, and finally, cytolytic action. To date, the most potent APC has been recognized to be the dendritic cell. Dendritic cells (DCs) are bone marrow-derived cells that can be isolated in cultures of peripheral blood mononuclear cells. Briefly, large numbers of DCs can be obtained by stimulating bone marrow cultures with granulocyte/macrophage-colony stimulating factor (GM-CSF). These DCs are capable of presenting both endogenous and exogenous antigens to naı¨ve T cells in an HLA-restricted manner [92]. Findings from a number of preclinical animal model systems have suggested that immunizing mice or rats with DCs pulsed with tumor antigens is sufficient to prime a cytotoxic lymphocytic response that is specific to tumor antigens and consequently can provide protective immunity against CNS tumors [93–95]. The first study to use a DC-based vaccine to target malignant gliomas was performed by Heimberger and colleagues [93]. In this work, bone marrow-derived DCs were pulsed with a specific subunit of the constitutively active EGFRvIII mutation. Immunized DCs were then administered to


mice intraperitoneally once a week for 3 weeks. Importantly, even after being challenged with tumor antigens more than once, the group receiving DC injections showed increased survival, demonstrating the development and induction of immunologic memory [93]. A follow-up study employed a tumor model system that mirrored the antigenic properties of spontaneous human gliomas. After being pulsed with tumor-specific antigens, DCs were administered to mice intraperitoneally each week, as in the above experiment, for a 4-week series. Animals were later challenged with intracranial injections of tumor cells. According to this study, mice immunized with DC injections exhibited a 160% increase in survival time; moreover, more than 50% of the animals showed long-term survival. Finally, mice rechallenged with tumor also exhibited enhanced survival, once again demonstrating the induction of immunological memory [94]. Based on the apparent successes observed in the animal studies above, Yu and colleagues [96] embarked on a phase I clinical trial administering a DC-based vaccine to glioma patients. In this trial, patients received autologous peripheral DCs prepared with IL-4 and GM-CSF stimulation after a brief exposure to peptides eluted from the surface of autologous glioma cells. Despite the small sample size, this study reported that patients receiving DC vaccinations showed a prolonged survival time compared with age- and sex-matched controls, indicating that DC vaccination may confer some survival benefit. This study conclusively demonstrated the safety, feasibility, and biological activity of the procedure, with evidence of an increased intratumoral cytotoxic CD8 + and memory T-cell (CD45RO+) infiltration. Another phase I trial investigating the feasibility of this procedure was conducted by Liau and colleagues at UCLA [97]. Twelve patients with histologically confirmed GBM received three biweekly intradermal injections of 1, 5, or 10 million autologous DCs, pulsed with the same amount (100 mg per injection) of acid-eluted

48

autologous tumor peptides. In this study, DC vaccinations were not associated with any evidence of dose-limiting toxicity or serious adverse effects. Additionally, one patient exhibited an objective clinical response documented by magnetic resonance imaging (MRI), and six patients developed significant systemic antitumoral cytotoxic T-lymphocyte responses, but these did not consistently translate into objective clinical responses or correlate with increased survival times. Based on the results of these studies, a phase II clinical trial is currently under way.

Gene Technique-based Immunotherapy More recently, advances in recombinant gene technology have opened the doors for the development of novel tumor-specific therapies. Genetic constructs can routinely be modified by viral vectors (or, alternatively, plasmid DNA) to express a variety of genes encoding tumor antigens, cytokines, or accessory molecules. In this context, genetically modifying tumor cells could increase their immunogenicity and potentially enhance the systemic immune response generated against an intracranial tumor. In principle, genetic modification can also lead to tumor cell death via induction of immunological cascades. Viruses, especially adenoviruses, have been pivotal in the application and development of these recombinant approaches, as these organisms can efficiently deliver their genome into eukaryotic cells and then transcribe and translate it within them. By replacing a part of their genome, scientists can readily convert adenoviruses from lytic pathogens to lysogenic vectors for gene delivery. Among the different genes whose delivery has been attempted by this approach, only a handful have been studied in clinical trials. These include p53, the herpes simplex virus-thymidine kinase (HSV-tk) gene, and the genes for interferons alpha and beta.

759

760

48


Adenovirus-Mediated p53 Gene Therapy The gene for the p53 tumor suppressor is located on the short arm of chromosome 17 and is one of the most frequently mutated genes in human gliomas [98]. p53 plays a critical role in cell cycle arrest and apoptosis, and its loss is a key factor in the development of glial neoplasms [99]. The North American Brain Tumor Consortium (NATBC) recently completed a phase I trial of adenovirus-mediated p53 gene therapy in 12 patients with recurrent malignant gliomas [100]. Patients first underwent stereotactic surgical implantation of a catheter into the center of tumor mass. Ad-p53, an adenovirus (type 5) in which the E1 coding region is replaced with p53 cDNA, whose expression is driven by a cytomegalovirus promoter, was delivered via this catheter. Three days after infusion of the virus, patients underwent open craniotomy, and the tumor and catheter were resected en bloc. After tumor resection, Ad-p53 was directly injected into the post-resection tumor bed, and the craniotomy was subsequently closed. Despite the serious limitation of having a narrow distribution of the therapeutic gene in the brain parenchyma, the results of this study demonstrated the feasibility and safety of such a procedure. Clinical toxicity was minimal, and time to recurrence within the treated group was 7 months. The problem of poor tissue penetration could potentially be circumvented, while also increasing the efficacy of adenoviral vectors, by using replication-competent vectors. This technique can not only exert a cytolytic effect on infected cells but also allow for subsequent propagation of the virus to neighboring cells, resulting in additional tumor/ tissue penetration. The main limitation of these vectors, however, is that the host’s normal brain cells must be spared. One potential strategy to avoid this problem is to produce mutant E1 genes, leading to the development of conditionally replicative adenoviruses [101]. An example of

this is ONYX-015, a conditionally replicative adenovirus capable of specifically affecting tumor cells. ONYX-015 contains a mutation in the p53binding protein E1B-55K, thus rendering it unable to transfect normal cells, which have normal p53 proteins. A recent phase I study of 24 glioma patients demonstrated the safety of this technique but failed to conclusively prove the specificity of the drug for glioma cells [102]. Recently, a new oncolytic adenovirus called delta 24 has been developed in an attempt to increase the specificity of these vectors for tumor cells. Delta 24, an oncolytic virus with a 24-base-pair deletion in the viral E1A gene, results in selective replication in cells harboring a mutation in the retinoblastoma (Rb) protein or its regulatory pathway [103]. In animal studies, delta 24-RGD has been shown to be effective against gliomas [104]. Importantly, the pathways involved in adenovirus-mediated cell death remain unclear. Delta 24 appears to induce the formation of acidic vesicular organelles, a process termed autophagy, in both in vitro and in vivo models [105]. More recently, two independent studies reported that the combined use of oncolytic adenovirus delta 24-RGD and chemotherapeutic agents results in an enhanced antiglioma effect in vivo [106,107].

Herpes Simplex Virus-based Gene Therapy A recently developed gene therapy model that has received much attention over the past decade is so-called suicide gene therapy. This technique is based on the transduction of tumor cells with herpes simplex virus-thymidine kinase (HSV-tk) in the presence of gancyclovir as a prodrug. Over the past few years, the method of specific gene delivery as well as the understanding of the biology of the suicide effect has evolved tremendously. Initial studies on the transfer of HSV-tk depended largely on the use


of recombinant-deficient retroviruses (RVs). RVs were generated and surgically released within the tumor bed by injection of genetically modified mouse fibroblasts (virus producer cells, or VPCs) [108,109]. In vitro experiments and in vivo animal tumor model systems of glioma have demonstrated the feasibility of RV-mediated gene transduction and the ability of this method to kill glioma cells by insertion of toxicity-generating transgenes. In spite of these encouraging reports of effective RV-mediated gene therapy in experimental animal model systems, clinical trials in glioma patients have been largely disappointing. Phase I and II clinical studies in patients with recurrent malignant gliomas have demonstrated some efficacy of RVmediated gene therapy in addition to a favorable safety profile [110,111]. However, a phase III study of adjuvant gene therapy in 248 GBM patients failed to report any survival benefit with administration of HSV-tk retrovirus VPCs [112]. The failure of RV gene therapy systems prompted researchers to investigate the development of alternative therapeutic viral vector systems to increase efficacy and yield better transduction efficiency. One of these methods utilizes replication-competent vectors [113]. Such vectors permit higher transduction rates in addition to enhanced cytolytic effects on transduced cells. Animal studies using replication-competent vectors have yielded promising results [114]. Additionally, our better understanding of the bystander cell death theory has begun to shed further light on the limitations of the RV gene therapy system. After treatment with the antiviral drug gancyclovir, untransfected tumor cells adjacent to transfected cells are killed as well. This bystander effect is thought to be mediated by gap junction intercellular transfer of toxic phosphorylated gancyclovir molecules by connexin protein family members [115,116]. In fact, several groups have suggested that the combined activity of HSV-tk and connexin 43 could enhance gene therapy [117].

48

The Role of Antisense Therapy The last decade has fostered the use of a novel and extremely potent method of experimentally manipulating gene expression known as RNA interference (RNAi and shRNA) [118,119]. RNAi is a process whereby double-stranded RNA (dsRNA) sequences target messenger RNA (mRNA) for destruction in a sequence-dependent manner. These small dsRNAs are delivered into target cells as short hairpin precursors (shRNA) by expression plasmids. DNA expression plasmids are encapsulated inside liposomes, which serve as a safeguard to protect DNA from ubiquitous degradation in vivo. The surface of the liposome has specific molecules conjugated to it that will bind endogenous receptors, inducing receptor-mediated transcytosis through the BBB, as well as transport by endocytosis to the nuclear compartment of brain cells [120,121]. Consequently, this method enables the efficient expression of plasmid DNA in brain tumor cells after intravenous administration of the gene. Once inside the cell, shRNA is processed by an enzyme called Dicer into active 21-nucleotide RNA fragments that can recognize target mRNA sequences via base-pairing interactions. The end result of this process is the suppression of specific genes at the posttranslational level. Several genes implicated in the malignant growth of gliomas have been studied as potential targets for gene therapy using RNA interference. One of these genes is TGF-b2, known to play a pivotal role in tumor progression by regulating key mechanisms such as proliferation, metastasis, and angiogenesis. One group in Germany recently developed a specific antisense oligonucleotide against TGF-b2 (AP12009). After demonstrating the safety and efficacy of this fragment in vitro, the researchers set out to compare the effect of AP12009, as administered by CED, with that of chemotherapeutic agents (TMZ or PCV, combined procarbazine, lomustine, and vincristine) in a phase II trial of patients with malignant

761

762

48


recurrent gliomas. They reported improved survival times in the former group, with some patients exhibiting long-term total remission. Thus, both the preclinical and the clinical results implicate targeted TGF-b2 suppression as a promising therapeutic approach for malignant tumor therapy [122,123]. The EGFR, another key regulator of the pathogenesis of brain tumors, has also been studied extensively as a possible target for silencing using RNA interference. In an experimental human brain tumor model system in mice with severe combined immunodeficiency, weekly intravenous injections of RNAi directed against the EGFR led to reduced expression of EGFRs within the tumor and an 88% increase in survival time of these mice [124]. The effectiveness and safety of this technology demonstrated in preclinical and clinical trials suggests it to be a promising method for further study in human clinical trials. One particular avenue of interest might involve designing an RNAi-based method to suppress both wild type and mutant EGFRvIII mRNAs.

Summary The past two decades have seen tremendous strides in cancer biology research. The discovery of signaling cascades mediating tumor cell growth and propagation has allowed for a deeper understanding of the basic biology of tumors and revealed a series of mutations that may directly or indirectly be involved in the pathogenesis of brain tumors. New studies aimed at dissecting the complex delicate interactions across a number of signaling pathways in the pathogenesis of gliomas has led to the rational development of different targeted therapies against one or a number of these pathways. Some of these therapies include the use of novel chemotherapeutic agents, immunotherapy, or viral-mediated gene therapy.

In general, preclinical studies have shown promising results with a number of these modalities. The great challenge, however, has been to translate these results into significant improvement of the clinical outcome in patients. A major problem lies in the fact that most clinical trials use these novel agents as monotherapies as opposed to adding them to existing protocols as adjuvant therapies. Consequently, most of these agents have failed to demonstrate survival benefit in unselected patient populations. The greatest limitation, however, remains the striking hetereogeneity of human gliomas, both in terms of cellular phenotype and behavior and with respect to the vast number of genetic aberrations seen, which can vary widely among individuals and sometimes even within the same tumor. Coupled with the confined delivery of drugs to the tumor bed, this is the single most significant problem underlying therapeutic failure. Despite all the advances highlighted in this chapter, prognosis for GBM remains poor and survival outcomes dismal. Future success in achieving improved targeted therapies depends on the proper segregation of patients based on the major mutations and molecular profile of their respective tumors. If scientists can better identify correlative biomarkers, one can envision ‘‘individualized’’ targeted therapy based on predictive molecular or genetic signatures of individual tumors from each patient. Consequently, this will allow therapy to be targeted against an individual mutation or, more broadly, tailored to individual patients. Strategies that might be employed in order to achieve this include genomic analysis to identify new targets and promising treatment combinations, in addition to methods to improve therapeutic delivery systems. The challenge lies in being able to unify our understanding of tumor biology with a keen grasp of the interactions between numerous therapeutic modalities and signaling pathways. Our success in combating these tumors will rely on our ability to effectively implement a multimodality paradigm including surgery, radiotherapy, and a form of targeted therapy.


Acknowledgments We thank David M. Wildrick, Ph.D., for editorial assistance with the manuscript.

References 1. Ries LAG, Eisner MP, Kosary CL, et al. (eds.). SEER cancer statistics review, 1975–2002 (http://seer.cancer. gov/csr/1975_2002/). Bethesda, MD: National Cancer Institute; 2005. 2. Deorah S, Lynch CF, Sibenaller ZA, Ryken TC. Trends in brain cancer incidence and survival in the United States: surveillance, epidemiology, and end results program, 1973 to 2001. Neurosurg Focus 2006;20:E1. 3. Central nervous system cancers. National comprehensive cancer network clinical practice guidelines in oncology, vol. 1. http://www.nccn.org/professionals/ physician_gls/PDF/cns.pdf, 2008 (accessed on March 27, 2008). 4. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352:987-96. 5. Stupp R, Dietrich PY, Ostermann Kraljevic S, et al. Promising survival for patients with newly diagnosed glioblastoma multiforme treated with concomitant radiation plus temozolomide followed by adjuvant temozolomide. J Clin Oncol 2002;20:1375-82. 6. Sampath P, Brem H. Implantable slow-release chemotherapeutic polymers for the treatment of malignant brain tumors. Cancer Control 1998;5:130-7. 7. Mak M, Fung L, Strasser JF, Saltzman WM. Distribution of drugs following controlled delivery to the brain interstitium. J Neurooncol 1995;26:91-102. 8. Brem H, Piantadosi S, Burger PC, et al. Placebocontrolled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. The polymer-brain tumor treatment group. Lancet 1995;345:1008-12. 9. Valtonen S, Timonen U, Toivanen P, et al. Interstitial chemotherapy with carmustine-loaded polymers for highgrade gliomas: a randomized double-blind study. Neurosurgery 1997; 41:44-8; discussion 48-9. 10. Brem H, Ewend MG, Piantadosi S, et al. The safety of interstitial chemotherapy with BCNU-loaded polymer followed by radiation therapy in the treatment of newly diagnosed malignant gliomas: phase I trial. J Neurooncol 1995;26:111-23. 11. Zamecnik J. The extracellular space and matrix of gliomas. Acta Neuropathol 2005;110:435-42. 12. Fleming AB, Saltzman WM. Pharmacokinetics of the carmustine implant. Clin Pharmacokinet 2002;41:403-19.

48

13. Bobo RH, Laske DW, Akbasak A, et al. Convectionenhanced delivery of macromolecules in the brain. Proc Natl Acad Sci USA 1994;91:2076-80. 14. Morrison PF, Chen MY, Chadwick RS, et al. Focal delivery during direct infusion to brain: role of flow rate, catheter diameter, and tissue mechanics. Am J Physiol 1999;277:R1218-29. 15. Lopez KA, Waziri AE, Canoll PD, Bruce JN. Convectionenhanced delivery in the treatment of malignant glioma. Neurol Res 2006;28:542-8. 16. Saito R, Krauze MT, Bringas JR, et al. Gadoliniumloaded liposomes allow for real-time magnetic resonance imaging of convection-enhanced delivery in the primate brain. Exp Neurol 2005;196:381-9. 17. Sawyer AJ, Piepmeier JM, Saltzman WM. New methods for direct delivery of chemotherapy for treating brain tumors. Yale J Biol Med 2006;79:141-52. 18. Arteaga CL. The epidermal growth factor receptor: from mutant oncogene in nonhuman cancers to therapeutic target in human neoplasia. J Clin Oncol 2001;19:32S-40S. 19. Chamberlain MC. Treatment options for glioblastoma. Neurosurg Focus 2006;20:E2. 20. Kuan CT, Wikstrand CJ, Bigner DD. EGF mutant receptor vIII as a molecular target in cancer therapy. Endocr Relat Cancer 2001;8:83-96. 21. Mellinghoff IK, Wang MY, Vivanco I, et al. Molecular determinants of the response of glioblastomas to EGFR kinase inhibitors. N Engl J Med 2005;353:2012-24. 22. Li B, Yuan M, Kim IA, et al. Mutant epidermal growth factor receptor displays increased signaling through the phosphatidylinositol-3 kinase/AKT pathway and promotes radioresistance in cells of astrocytic origin. Oncogene 2004;23:4594-602. 23. Choe G, Horvath S, Cloughesy TF, et al. Analysis of the phosphatidylinositol 3 0 -kinase signaling pathway in glioblastoma patients in vivo. Cancer Res 2003;63:2742-6. 24. Pelloski CE, Ballman KV, Furth AF, et al. Epidermal growth factor receptor variant III status defines clinically distinct subtypes of glioblastoma. J Clin Oncol 2007;25:2288-94. 25. Rich JN, Reardon DA, Peery T, et al. Phase II trial of gefitinib in recurrent glioblastoma. J Clin Oncol 2004;22:133-42. 26. Uhm JH, Ballman KV, Giannini C, et al. Phase II study of ZD1839 in patients with newly diagnosed grade 4 astrocytoma [abstract]. J Clin Oncol 2004;22:1505. 27. Cloughesy T, Yung A, Vrendenberg J, et al. Phase II study of erlotinib in recurrent GBM: molecular predictors of outcome [abstract]. J Clin Oncol 2005;23:1507. 28. Vogelbaum MA, Peereboom D, Stevens G, et al. Phase II trial of the EGFR tyrosine kinase inhibitor erlotinib for single agent therapy of recurrent glioblastoma multiforme: interim results [abstract]. J Clin Oncol 2004;22:1558. 29. Prados M, Chang S, Burton E, et al. Phase I study of OSI774 alone or with temozolomide in patients with malignant glioma [abstract]. J Clin Oncol 2003;22:394.

763

764

48


30. Peereboom DM, Brewer CJ, Suh JH, et al. Phase II trial of erlotinib with temozolomide and concurrent radiation therapy in patients with newly diagnosed glioblastoma multiforme: final results [abstract]. Neuro-oncol 2006; 8:448. 31. Sadones J, Chaskis C, Joosens EJ, et al. A stratified phase II study of cetuximab for the treatment of recurrent glioblastoma multiforme: preliminary results [abstract]. J Clin Oncol 2006;24:1558. 32. Reardon DA, Quinn JA, Vredenburgh JJ, et al. Phase 1 trial of gefitinib plus sirolimus in adults with recurrent malignant glioma. Clin Cancer Res 2006;12:860-8. 33. Doherty L, Gigas DC, Kesari S, et al. Pilot study of the combination of EGFR and mTOR inhibitors in recurrent malignant gliomas. Neurology 2006;67:156-8. 34. Fournier E, Birnbaum D, Borg JP. [Receptors for factors of the VEGF (vascular endothelial growth family)]. Bull Cancer 1997;84:397-405. 35. Stark-Vance V. Bevacizumab and CPT-11 in the treatment of relapsed malignant glioma [abstract]. Neurooncol 2005;7:369. 36. Goli KJ, Desjardins A, Herndon JE, et al. Phase II trial of bevacizumab and irinotecan in the treatment of malignant gliomas [abstract]. J Clin Oncol 2007; 25:2003. 37. Conrad C, Friedman H, Reardon D, et al. A phase I/II trial of single-agent PTK787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in patients with recurrent glioblastoma multiforme (GBM) [abstract]. J Clin Oncol 2004;22:1512. 38. Wong ML, Kaye AH, Hovens CM. Targeting malignant glioma survival signalling to improve clinical outcomes. J Clin Neurosci 2007;14:301-8. 39. Sathornsumetee S, Reardon DA, Desjardins A, et al. Molecularly targeted therapy for malignant glioma. Cancer 2007;110:13-24. 40. Newton HB. Molecular neuro-oncology and development of targeted therapeutic strategies for brain tumors. Part 1: Growth factor and Ras signaling pathways. Expert Rev Anticancer Ther 2003;3:595-614. 41. Fleming TP, Saxena A, Clark WC, et al. Amplification and/or overexpression of platelet-derived growth factor receptors and epidermal growth factor receptor in human glial tumors. Cancer Res 1992;52:4550-3. 42. Lokker NA, Sullivan CM, Hollenbach SJ, et al. Plateletderived growth factor (PDGF) autocrine signaling regulates survival and mitogenic pathways in glioblastoma cells: evidence that the novel PDGF-C and PDGF-D ligands may play a role in the development of brain tumors. Cancer Res 2002;62:3729-35. 43. George D. Platelet-derived growth factor receptors: a therapeutic target in solid tumors. Semin Oncol 2001; 28:27-33. 44. Strawn LM, Mann E, Elliger SS, et al. Inhibition of glioma cell growth by a truncated platelet-derived growth factor-beta receptor. J Biol Chem 1994;269:21215-22.

45. Pietras K, Rubin K, Sjoblom T, et al. Inhibition of PDGF receptor signaling in tumor stroma enhances antitumor effect of chemotherapy. Cancer Res 2002;62:5476-84. 46. Plate KH, Breier G, Farrell CL, Risau W. Platelet-derived growth factor receptor-beta is induced during tumor development and upregulated during tumor progression in endothelial cells in human gliomas. Lab Invest 1992;67:529-34. 47. Pietras K. Increasing tumor uptake of anticancer drugs with imatinib. Semin Oncol 2004;31:18-23. 48. Geng L, Shinohara ET, Kim D, et al. STI571 (Gleevec) improves tumor growth delay and survival in irradiated mouse models of glioblastoma. Int J Radiat Oncol Biol Phys 2006;64:263-71. 49. Wen PY, Yung WK, Hess K, et al. Phase I study of STI571 (Gleevec) for patients with recurrent malignant gliomas and meningiomas (NABTC 99–08) [abstract]. J Clin Oncol 2002;21:73. 50. Holdhoff M, Kreuzer KA, Appelt C, et al. Imatinib mesylate radiosensitizes human glioblastoma cells through inhibition of platelet-derived growth factor receptor. Blood Cells Mol Dis 2005;34:181-5. 51. O’Reilly T, Wartmann M, Maira SM, et al. Patupilone (epothilone B, EPO906) and imatinib (STI571, Glivec) in combination display enhanced antitumour activity in vivo against experimental rat C6 glioma. Cancer Chemother Pharmacol 2005;55:307-17. 52. Katayama R, Huelsmeyer MK, Marr AK, et al. Imatinib mesylate inhibits platelet-derived growth factor activity and increases chemosensitivity in feline vaccineassociated sarcoma. Cancer Chemother Pharmacol 2004;54:25-33. 53. Dogruel M, Gibbs JE, Thomas SA. Hydroxyurea transport across the blood-brain and blood-cerebrospinal fluid barriers of the guinea-pig. J Neurochem 2003;87:76-84. 54. Gwilt PR, Manouilov KK, McNabb J, Swindells SS. Pharmacokinetics of hydroxyurea in plasma and cerebrospinal fluid of HIV-1-infected patients. J Clin Pharmacol 2003;43:1003-7. 55. Dresemann G. Imatinib and hydroxyurea in pretreated progressive glioblastoma multiforme: a patient series. Ann Oncol 2005;16:1702-8. 56. Reardon DA, Egorin MJ, Quinn JA, et al. Phase II study of imatinib mesylate plus hydroxyurea in adults with recurrent glioblastoma multiforme. J Clin Oncol 2005; 23:9359-68. 57. Couldwell WT, Hinton DR, Surnock AA, et al. Treatment of recurrent malignant gliomas with chronic oral high-dose tamoxifen. Clin Cancer Res 1996;2:619-22. 58. Mastronardi L, Puzzilli F, Couldwell WT, et al. Tamoxifen and carboplatin combinational treatment of high-grade gliomas. Results of a clinical trial on newly diagnosed patients. J Neurooncol 1998;38:59-68. 59. Napolitano M, Keime-Guibert F, Monjour A, et al. Treatment of supratentorial glioblastoma multiforme with radiotherapy and a combination of BCNU and


60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

tamoxifen: a phase II study. J Neurooncol 1999; 45:229-35. Vertosick FT, Selker RG. The treatment of newly diagnosed glioblastoma multiforme using high dose tamoxifen (TMX), radiotherapy and conventional chemotherapy [abstract # 2887]. Proc Am Assoc Cancer Res 1997; 2887. Preul MC, Caramanos Z, Villemure JG, et al. Using proton magnetic resonance spectroscopic imaging to predict in vivo the response of recurrent malignant gliomas to tamoxifen chemotherapy. Neurosurgery 2000;46:306-18. Bianco R, Shin I, Ritter CA, et al. Loss of PTEN/ MMAC1/TEP in EGF receptor-expressing tumor cells counteracts the antitumor action of EGFR tyrosine kinase inhibitors. Oncogene 2003;22:2812-22. Chakravarti A, Zhai G, Suzuki Y, et al. The prognostic significance of phosphatidylinositol 3-kinase pathway activation in human gliomas. J Clin Oncol 2004;22:1926-33. Galanis E, Buckner JC, Maurer MJ, et al. Phase II trial of temsirolimus (CCI-779) in recurrent glioblastoma multiforme: a North Central Cancer Treatment Group Study. J Clin Oncol 2005;23:5294-304. Chang SM, Wen P, Cloughesy T, et al. Phase II study of CCI-779 in patients with recurrent glioblastoma multiforme. Invest New Drugs 2005;23:357-61. Black KL, Chen K, Becker DP, Merrill JE. Inflammatory leukocytes associated with increased immunosuppression by glioblastoma. J Neurosurg 1992;77:120-6. Elliott LH, Brooks WH, Roszman TL. Inability of mitogen-activated lymphocytes obtained from patients with malignant primary intracranial tumors to express high affinity interleukin 2 receptors. J Clin Invest 1990;86:80-6. Bhondeley MK, Mehra RD, Mehra NK, et al. Imbalances in T cell subpopulations in human gliomas. J Neurosurg 1988;68:589-93. Kikuchi T, Abe T, Ohno T. Effects of glioma cells on maturation of dendritic cells. J Neurooncol 2002;58:125-30. Yang T, Witham TF, Villa L, et al. Glioma-associated hyaluronan induces apoptosis in dendritic cells via inducible nitric oxide synthase: implications for the use of dendritic cells for therapy of gliomas. Cancer Res 2002;62:2583-91. Morford LA, Elliott LH, Carlson SL, et al. T cell receptormediated signaling is defective in T cells obtained from patients with primary intracranial tumors. J Immunol 1997;159:4415-25. Ehtesham M, Samoto K, Kabos P, et al. Treatment of intracranial glioma with in situ interferon-gamma and tumor necrosis factor-alpha gene transfer. Cancer Gene Ther 2002;9:925-34. Plautz GE, Touhalisky JE, Shu S. Treatment of murine gliomas by adoptive transfer of ex vivo activated tumordraining lymph node cells. Cell Immunol 1997;178:101-7. Bodmer S, Strommer K, Frei K, et al. Immunosuppression and transforming growth factor-beta in

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

48

glioblastoma. Preferential production of transforming growth factor-beta 2. J Immunol 1989;143:3222-9. Gorelik L, Flavell RA. Immune-mediated eradication of tumors through the blockade of transforming growth factor-beta signaling in T cells. Nat Med 2001;7:1118-22. Ruffini PA, Rivoltini L, Silvani A, et al. Factors, including transforming growth factor beta, released in the glioblastoma residual cavity, impair activity of adherent lymphokine-activated killer cells. Cancer Immunol Immunother 1993;36:409-16. Wikstrand CJ, Fung KM, Trojanowski JQ, et al. Immunohistochemistry and antigens of diagnostic significance. In: Bigner DD, McLendon RE, Bruner JM, editors. Russell and Rubinstein’s pathology of the nervous system. New York, NY: Oxford University Press; 1998. p. 251-304. Riva P, Arista A, Tison V, et al. Intralesional radioimmunotherapy of malignant gliomas. An effective treatment in recurrent tumors. Cancer 1994;73:1076-82. Riva P, Arista A, Sturiale C, et al. Glioblastoma therapy by direct intralesional administration of I-131 radioiodine labeled antitenascin antibodies. Cell Biophys 1997; 24-25:37-43. Riva P, Franceschi G, Arista A, et al. Local application of radiolabeled monoclonal antibodies in the treatment of high grade malignant gliomas: a six-year clinical experience. Cancer 1997;80:2733-42. Cokgor I, Akabani G, Kuan CT, et al. Phase I trial results of iodine-131-labeled antitenascin monoclonal antibody 81C6 treatment of patients with newly diagnosed malignant gliomas. J Clin Oncol 2000;18:3862-72. Reardon DA, Akabani G, Coleman RE, et al. Phase II trial of murine (131)I-labeled antitenascin monoclonal antibody 81C6 administered into surgically created resection cavities of patients with newly diagnosed malignant gliomas. J Clin Oncol 2002;20:1389-97. Prados M, Kunwar S, Lang FF, et al. Final results of phase I/II studies of IL13-pE38QQR administered intratumorally (IT) and/or peritumorally (PT) via convectionenhanced delivery (CED) in patients undergoing tumor resection for recurrent malignant glioma [abstract]. J Clin Oncol 2005;23:1506. Laske DW, Youle RJ, Oldfield EH. Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med 1997;3:1362-8. Weaver M, Laske DW. Transferrin receptor ligandtargeted toxin conjugate (Tf-CRM107) for therapy of malignant gliomas. J Neurooncol 2003;65:3-13. Kuppner MC, Hamou MF, Sawamura Y, et al. Inhibition of lymphocyte function by glioblastoma-derived transforming growth factor beta 2. J Neurosurg 1989;71:211-7. Siepl C, Bodmer S, Frei K, et al. The glioblastomaderived T cell suppressor factor/transforming growth factor-beta 2 inhibits T cell growth without affecting the interaction of interleukin 2 with its receptor. Eur J Immunol 1988;18:593-600.

765

766

48


88. Brandes AA, Scelzi E, Zampieri P, et al. Phase II trial with BCNU plus alpha-interferon in patients with recurrent high-grade gliomas. Am J Clin Oncol 1997;20:364-7. 89. Yung WK, Prados M, Levin VA, et al. Intravenous recombinant interferon beta in patients with recurrent malignant gliomas: a phase I/II study. J Clin Oncol 1991;9:1945-9. 90. Hiserodt JC, Vujanovic NL, Reynolds CW, et al. Studies on lymphokine activated killer cells in the rat: analysis of precursor and effector cell phenotype and relationship to natural killer cells. Prog Clin Biol Res 1987; 244:137-46. 91. Hayes RL, Koslow M, Hiesiger EM, et al. Improved long term survival after intracavitary interleukin-2 and lymphokine-activated killer cells for adults with recurrent malignant glioma. Cancer 1995;76:840-52. 92. Porgador A, Gilboa E. Bone marrow-generated dendritic cells pulsed with a class I-restricted peptide are potent inducers of cytotoxic T lymphocytes. J Exp Med 1995;182:255-60. 93. Heimberger AB, Archer GE, Crotty LE, et al. Dendritic cells pulsed with a tumor-specific peptide induce longlasting immunity and are effective against murine intracerebral melanoma. Neurosurgery 2002; 50:158-64; discussion 164-6. 94. Heimberger AB, Crotty LE, Archer GE, et al. Bone marrow-derived dendritic cells pulsed with tumor homogenate induce immunity against syngeneic intracerebral glioma. J Neuroimmunol 2000;103:16-25. 95. Ni HT, Spellman SR, Jean WC, et al. Immunization with dendritic cells pulsed with tumor extract increases survival of mice bearing intracranial gliomas. J Neurooncol 2001;51:1-9. 96. Yu JS, Wheeler CJ, Zeltzer PM, et al. Vaccination of malignant glioma patients with peptide-pulsed dendritic cells elicits systemic cytotoxicity and intracranial T-cell infiltration. Cancer Res 2001;61:842-7. 97. Liau LM, Prins RM, Kiertscher SM, et al. Dendritic cell vaccination in glioblastoma patients induces systemic and intracranial T-cell responses modulated by the local central nervous system tumor microenvironment. Clin Cancer Res 2005;11:5515-25. 98. Bogler O, Huang HJ, Kleihues P, Cavenee WK. The p53 gene and its role in human brain tumors. Glia 1995; 15:308-27. 99. Prives C, Hall PA. The p53 pathway. J Pathol 1999; 187:112-26. 100. Lang FF, Bruner JM, Fuller GN, et al. Phase I trial of adenovirus-mediated p53 gene therapy for recurrent glioma: biological and clinical results. J Clin Oncol 2003;21:2508-18. 101. Curiel DT. The development of conditionally replicative adenoviruses for cancer therapy. Clin Cancer Res 2000;6:3395-9.

102. Chiocca EA, Abbed KM, Tatter S, et al. A phase I open-label, dose-escalation, multi-institutional trial of injection with an E1B-attenuated adenovirus, ONYX-015, into the peritumoral region of recurrent malignant gliomas, in the adjuvant setting. Mol Ther 2004; 10:958-66. 103. Fueyo J, Gomez-Manzano C, Alemany R, et al. A mutant oncolytic adenovirus targeting the Rb pathway produces anti-glioma effect in vivo. Oncogene 2000; 19:2-12. 104. Jiang H, Gomez-Manzano C, Alemany R, et al. Comparative effect of oncolytic adenoviruses with E1A55 kDa or E1B-55 kDa deletions in malignant gliomas. Neoplasia 2005;7:48-56. 105. Jiang H, Gomez-Manzano C, Aoki H, et al. Examination of the therapeutic potential of Delta-24-RGD in brain tumor stem cells: role of autophagic cell death. J Natl Cancer Inst 2007;99:1410-4. 106. Alonso MM, Jiang H, Yokoyama T, et al. Delta-24RGD in combination with RAD001 induces enhanced anti-glioma effect via autophagic cell death. Mol Ther 2008;16:487-93. 107. Alonso MM, Gomez-Manzano C, Bekele BN, et al. Adenovirus-based strategies overcome temozolomide resistance by silencing the O6-methylguanine-DNA methyltransferase promoter. Cancer Res 2007;67:11499-504. 108. Culver KW, Ram Z, Wallbridge S, et al. In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors. Science 1992; 256:1550-2. 109. Oldfield EH, Ram Z, Culver KW, et al. Gene therapy for the treatment of brain tumors using intra-tumoral transduction with the thymidine kinase gene and intravenous ganciclovir. Hum Gene Ther 1993;4:39-69. 110. Trask TW, Trask RP, Aguilar-Cordova E, et al. Phase I study of adenoviral delivery of the HSV-tk gene and ganciclovir administration in patients with current malignant brain tumors. Mol Ther 2000;1:195-203. 111. Prados MD, McDermott M, Chang SM, et al. Treatment of progressive or recurrent glioblastoma multiforme in adults with herpes simplex virus thymidine kinase gene vector-producer cells followed by intravenous ganciclovir administration: a phase I/II multi-institutional trial. J Neurooncol 2003;65:269-78. 112. Rainov NG. A phase III clinical evaluation of herpes simplex virus type 1 thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical resection and radiation in adults with previously untreated glioblastoma multiforme. Hum Gene Ther 2000; 11:2389-401. 113. Wildner O, Morris JC, Vahanian NN, et al. Adenoviral vectors capable of replication improve the efficacy of HSVtk/GCV suicide gene therapy of cancer. Gene Ther 1999;6:57-62. 114. Nanda D, Vogels R, Havenga M, et al. Treatment of malignant gliomas with a replicating adenoviral vector


115.

116.

117.

118. 119.

120.

expressing herpes simplex virus-thymidine kinase. Cancer Res 2001;61:8743-50. Mesnil M, Piccoli C, Tiraby G, et al. Bystander killing of cancer cells by herpes simplex virus thymidine kinase gene is mediated by connexins. Proc Natl Acad Sci USA 1996;93:1831-5. Warn-Cramer BJ, Cottrell GT, Burt JM, Lau AF. Regulation of connexin-43 gap junctional intercellular communication by mitogen-activated protein kinase. J Biol Chem 1998;273:9188-96. Marconi P, Tamura M, Moriuchi S, et al. Connexin 43enhanced suicide gene therapy using herpesviral vectors. Mol Ther 2000;1:71-81. Hannon GJ. RNA interference. Nature 2002;418:244-51. Martinez J, Patkaniowska A, Urlaub H, et al. Singlestranded antisense siRNAs guide target RNA cleavage in RNAi. Cell 2002;110:563-74. Shi N, Zhang Y, Zhu C, et al. Brain-specific expression of an exogenous gene after i.v. administration. Proc Natl Acad Sci USA 2001;98:12754-9.

48

121. Shi N, Boado RJ, Pardridge WM. Receptor-mediated gene targeting to tissues in vivo following intravenous administration of pegylated immunoliposomes. Pharm Res 2001;18:1091-5. 122. Schlingensiepen KH, Fischer-Blass B, Schmaus S, Ludwig S. Antisense therapeutics for tumor treatment: the TGF-beta2 inhibitor AP 12009 in clinical development against malignant tumors. Recent Results Cancer Res 2008;177:137-50. 123. Hau P, Jachimczak P, Schlingensiepen R, et al. Inhibition of TGF-beta2 with AP 12009 in recurrent malignant gliomas: from preclinical to phase I/II studies. Oligonucleotides 2007;17:201-12. 124. Zhang Y, Zhang YF, Bryant J, et al. Intravenous RNA interference gene therapy targeting the human epidermal growth factor receptor prolongs survival in intracranial brain cancer. Clin Cancer Res 2004;10:3667-77.

767

43 Pathology Techniques in Stereotactic and Image Guided Biopsy P. T. Chandrasoma . N. E. Klipfel

Stereotactic brain biopsy has become a standard and widely available technique in the past two decades for obtaining tissue from intracranial lesions. The success of the procedure is dependent on (1) the neurosurgeon obtaining a representative sample from the lesion and (2) the ability of the pathologist to make an accurate diagnosis. It has been well documented that stereotactic brain biopsy is a highly effective diagnostic procedure in major academic centers where neurosurgeons specialize in stereotactic procedures and special expertise in neuropathology is readily available. The reported non-diagnostic rate in major centers varies between 4 and 7.2% [1–6]. The incidence of diagnostic failure does not appear to depend on the computed tomographic (CT) morphology of the lesion or the experience of the surgeon [7]. The high level of accuracy is maintained even when results of biopsy are compared with subsequent resection [8]. With the increasing use of the procedure in community hospitals, the success of stereotactic biopsy at the present time depends, among other factors, on the ability of the community hospital’s pathologist to make a diagnosis [9]. The pathologist’s expertise in many such settings is limited by lack of specific training in stereotactic biopsies and the lack of experience working with the very small volumes of tissue available from stereotactic biopsies. It is critical to develop a technique for pathological processing of stereotactic biopsies that will utilize the strengths that community pathologists already have, so as to optimize diagnosis [4]. #


When the first stereotactic biopsies were received in our laboratory in the early 1980s, it was clear that they represented a new problem different from that of handling biopsies taken at open craniotomy. First, the specimens were far more representative of the lesion because the location of the biopsy was within a selected target in the CT lesion; in an open biopsy, on the other hand, the biopsy was commonly taken from the periphery of a grossly abnormal area. Second, the specimen was of extremely small size, and our first attempts at using frozen section were not very successful. Third, the ability to obtain additional material from the lesion was more limited than in an open craniotomy, so that requests for additional tissue were less acceptable. This led, in our laboratory, to a trial with the smear technique. Our experience over the past decades has led to the conviction that the best way to handle a stereotactic biopsy is with the initial performance of a smear [10]. This technique provides excellent cytological detail without the artifact caused by freezing the tissue. This utilizes the application of cytopathological expertise, which is usually developed far more in the community pathologist than is neuropathology, making this technique especially appropriate for use in the community setting [4,11].

Types of Biopsies and Smear Technique The type of biopsy obtained varies with different neurosurgeons. At the University of Southern

664

43

Pathology techniques in stereotactic and image guided biopsy

California, the biopsy is taken from a single point target within the lesion with the equivalent of a pediatric bronchoscopic forceps. The specimen consists of one to three samples of tissue that are 1–2 mm in greatest dimension. These are extremely small samples, but the diagnostic success achieved from this sample, coupled with the exceedingly low morbidity and mortality reported with this type of biopsy from our institution, has confirmed the effectiveness of this method [1]. In other centers, larger numbers of specimens might be obtained, often from the entire transit zone within the lesion or from multiple targets. Alternatively, the use of sidecutting biopsy needles provides a core of tissue from the lesion 1+ cm in length. Some authorities report increased diagnostic success with larger specimens [12]. The technique for making a smear is as follows [10]: A small piece from each of the biopsy samples is cut with a scalpel blade and placed on a glass slide. Any necessary orientation can be maintained by labeling the different tissue pieces on the slide. The tissue is smeared on the slide with a second glass slide, using pressure while drawing the slides apart. The amount of pressure required for smearing the tissue varies with the type of specimen. Normal brain and most glial neoplasms smear easily. Reactive gliosis, some fibrillary astrocytomas, and schwannomas smear with difficulty. In specimens that resist smearing, the tissue should be broken up by to-and-fro movements of the slide to effect maximum smearing. While this produces considerable crush artifact, it is preferable to having a thick uncrushed tissue fragment that is impossible to interpret. The way in which tissue smears on the slide is unpredictable, and it is important that both slides used for smearing are stained; they can vary considerably in appearance and cellularity. The smears are immediately fixed in methyl alcohol and stained by the rapid hematoxylin and eosin stain. The method used

for staining is identical to that which is used for frozen sections and requires no special materials. Selection of the tissue used for smearing is simple with the specimens usually received at our institution. When multiple samples are received from different target points or when the tissue specimen is a long core sample, selection of which part of the specimen to use for the smear requires clear understanding of the relationship of the different specimens or areas to the target point. In these cases, it is frequently necessary to process several different pieces of tissue. In evaluating the smear, the ease or difficulty of smearing must be taken into consideration. Tissue that spreads out easily will appear less cellular on the smear than that which is cohesive. In cohesive areas of the smear, it is more difficult to evaluate cytological features because of stratification of cells and tissue opacity. Tissue that has required crushing by repeated movements of the slides will be affected by drying artifact, which can produce considerable nuclear enlargement. The pathologist intending to use the smear technique should be familiar with the normal appearance of different areas of the brain on smear. This can be done by (1) reference to a variety of cytopathological texts and (2) keeping a reference set of smears made from autopsy brain tissue. It is of particular importance to recognize the normally cellular layers of the cerebellum. Also, the temporal lobe is normally moderately hypercellular and contains numerous neurons which can be confused with neoplastic astrocytes. If a diagnosis cannot be rendered through evaluation of the smear, whether from a thick preparation or insufficient diagnostic features (e.g., architectural features), the remaining tissue should be processed for frozen section. If the tissue is nondiagnostic (due to sampling or necrosis), additional tissue should be requested.


Need for Correlation with Computed Tomography and Clinical Features The two most important factors determining the success of stereotactic biopsy is the communication between the pathologist and the neurosurgeon and collection of the ideal tissue sample. The pathologist must be aware of the patient’s clinical history, the radiological appearance of the lesion, the differential diagnosis, and the indication for the biopsy. The pathologist must develop basic skills in CT radiology and clearly understand how the target point selected for the biopsy relates to the lesion as a whole. The pre-biopsy clinical discussion can be done in one of two ways: (1) the pathologist being present in the operating room as the biopsy is done and examining the CT scan there or (2) the neurosurgeon coming to the frozen section room with the specimen and CT scan and discussing the case during the processing of the specimen. It is highly recommended that one of these schemes is followed.

43

and extraparenchymal. These distinctions are not always precise. We have encountered cases of secretory meningioma associated with severe cerebral edema that were mistakenly thought to be intraparenchymal by radiological features. Modern radiological imaging modalities, such as magnetic resonance imaging and positron emission tomography, can indicate the histological type and grade of cerebral lesions. While these indications are not always accurate, the pathologist should be aware of these radiological opinions. In general, radiological appearances will divide cerebral hemispheric lesions into (1) suspected high grade neoplasms, which are characterized by irregular lesions with irregular contrast enhancement, surrounding edema and mass effect; (2) suspected low grade gliomas, which are typically large, irregular, nonenhancing lesions; and (3) lesions with features that point to specific entities such as pilocytic astrocytoma, pleomorphic xanthoastrocytoma, and ganglioglioma. In all these categories, the possibility of a nonneoplastic condition exists. Selection of the target point within the lesion is of critical importance in obtaining a diagnostic sample and is the most important part of the procedure [17].

Biopsy of Mass Lesions in the Cerebral Hemisphere Mass lesions in the cerebral hemispheres represent the commonest indication for stereotactic brain biopsy. Within this category, it is helpful to separate pediatric patients and those who are positive for human immunodeficiency virus (HIV), in whom diagnostic considerations are frequently different (vide infra) [13–16]. In populations with a high prevalence of HIV positive patients, malignant lymphoma and toxoplasmosis should be kept in the differential even in those who are not known to be at risk. Also important to distinguish from lesions of the cerebral hemispheres are mass lesions that are recognizable as intraventricular, suprasellar, pineal,

Biopsy in Clinically Suspected High Grade Neoplasms Biopsies in suspected high grade lesions are usually obtained from the enhancing edge of the lesion because of the possibility that the central area is necrotic. While this is the best approach, this principle commonly results in the undergrading of high grade lesions due to sampling and might cause diagnostic difficulty if the biopsy is from the infiltrative edge of the neoplasm. In most high grade neoplasms, there are (1) areas composed entirely of neoplasm where the brain has been completely replaced, (2) areas of necrosis without viable tumor cells, and

665

666

43


(3) areas of neoplastic cell infiltration of the adjacent brain. On radiological imaging, these areas are not precisely distinguishable. Also, adjacent brain showing secondary changes, such as reactive gliosis and edema, might be perceived as neoplasm radiographically. It must be accepted that target selection in these lesions is difficult and does not always produce an ideal specimen. An ideal specimen from a pathologist’s standpoint is one composed entirely of viable neoplasm where it has completely replaced normal brain. With such a specimen, there is no question as to the diagnosis of a malignant neoplasm, the pathologist must then identify the histological type. The vast majority of biopsied high grade lesions are high grade astrocytomas, either anaplastic astrocytoma or glioblastoma multiforme. These are characterized by increased cellularity, pleomorphism, mitoses and cytological atypia in cells recognizable as astrocytes (> Figure 43-1). Glioblastoma multiforme differs from anaplastic astrocytoma by the presence of necrosis (> Figure 43-2) or microvascular proliferation (> Figure 43-3). In stereotactic biopsies, because selection of the target point attempts to avoid necrotic areas, there is a tendency to under-grade glioblastoma multiforme. If there is a clinical

need to differentiate anaplastic astrocytoma from glioblastoma multiforme, biopsies from multiple target points within the lesion should be done. Accurate grading of astrocytomas is based on the presence of atypia, mitotic activity, vascular proliferation and necrosis. While the small size of the sample may lead to under-grading, the accuracy is generally adequate for clinical management. Astrocytes are recognized by the presence of thin cytoplasmic fibrils that form a background network between the neoplastic cells (> Figure 43-1). The presence of gemistocytes, which are large cells with abundant eosinophilic cytoplasm, are seen in some astrocytic neoplasms. In rare astrocytomas where the cells are so poorly differentiated as to lack prominent fibrils or show astrocytic features, immunoperoxidase staining of formalin fixed, paraffin embedded tissue for gliofibrillary acidic protein might be helpful. In cases where a diagnosis of high grade astrocytoma is uncertain, immunoperoxidase staining with the MIB1 epitope (Ki-67 antigen) will show the proliferative neoplastic cells staining >5% in anaplastic astrocytomas and provides a good correlation with prognosis in this group of neoplasms [18]. Other neoplasms that might be encountered in a suspected high-grade hemispheric lesion are

. Figure 43-1 Smear of anaplastic astrocytoma showing cellular proliferation of astrocytes with enlarged atypical nuclei (arrow) and cytoplasmic fibrils (H&E, 400)

. Figure 43-2 Smear of glioblastoma showing necrotic debris (H&E, 200)


43

. Figure 43-3 Smear of glioblastoma showing microvascular proliferation (arrow) (H&E, 400)

. Figure 43-4 Smear of metastatic adenocarcinoma showing cohesive groups of epithelial cells with intracytoplasmic vacuoles (long arrow) and glandular lumens (short arrow) (H&E, 400)

anaplastic oligodendrogliomas, mixed gliomas, metastatic carcinoma (> Figure 43-4), metastatic melanoma (> Figure 43-5), high-grade malignant lymphoma (> Figure 43-6), and a variety of very rare lesions. These neoplasms have

distinctive cytological features that permit diagnosis on smear. In some cases, recognition of cells on the smear is not absolute but permits processing of the tissue in such a manner as to perform special studies to characterize the cells.

667

668

43


. Figure 43-5 Smear of metastatic melanoma showing loosely cohesive cells with cytoplasmic pigment (arrow) (H&E, 400)

. Figure 43-6 Smear of malignant lymphoma showing discohesive round cells with typical chromatin pattern of transformed lymphocytes and lymphoglandular bodies (arrow) (H&E, 400)

The two special procedures commonly utilized for stereotactic biopsies are (1) immunoperoxidase staining for antigens that characterize the derivation of neoplastic cells such as CD45 (lymphoid), GFAP (astrocytic); keratin (epithelial), and HMB45 (melanocytic) and (2) electron microscopy, which is usually done only in academic institutions. The combination of smear, permanent

section, immunoperoxidase staining and rarely EM will provide an accurate diagnosis in all cases of high grade lesions in which a representative sample is provided to the pathologist. At this time, molecular genetics is not used for diagnosis, but testing can provide treatment indications, such as fluorescent in situ hybridization (paraffin tissue) for 1p and 19q losses in oligodendroglioma [19]. In cases where the sample is less than ideal, the diagnostic accuracy is less, and the likelihood of diagnostic error increases as the sample quality decreases. A less than ideal sample is one taken from an area where a high-grade astrocytoma infiltrates normal brain, containing elements of both neoplasm and brain. In such cases, a diagnosis of high grade astrocytoma can be made when the neoplastic cells are recognized as such by virtue of their anaplasia. In such biopsies, however, there are two possible sources of error: (1) under-grading of the astrocyctoma, which is not a serious error because the diagnosis that will be given will be an anaplastic astrocytoma [20], and (2) misdiagnosis of reactive gliosis as astrocytoma. Reactive gliosis is usually characterized by moderate cellularity, cellular polymorphism and minimal cytological atypia without the anaplasia that characterizes anaplastic astrocytoma. However, experience is needed in these cases, particularly when the neurosurgeon reports that there is a high clinical likelihood of a high grade glioma and insists that the sample is representative. In clinically suspected high-grade lesions, a pathological diagnosis of low grade (welldifferentiated) astrocytoma should never be rendered by the pathologist or accepted by the neurosurgeon. In such cases, biopsies showing increased cellularity without anaplasia represent areas of reaction around the lesion, and therefore samples that are not representative of the pathological lesion. When the lesion is a high grade astrocytoma, the pathological diagnosis of low grade astrocytoma represents clinically significant under-grading. When the lesion turns


out to be anything other than an astrocytoma, the incorrect diagnosis of low grade astrocytoma has the potential to be devastating. Cases of cerebral abscesses rarely have had an initial diagnosis rendered on stereotactic biopsy as low grade astrocytoma.

Biopsy in Clinically Suspected Low Grade Glioma Low grade astrocytomas of the cerebral hemispheres tend to be large, ill-defined, non-enhancing lesions on CT scan without much associated edema or mass effect. Normal brain tissue is replaced by the neoplasm in the central area, with extensive infiltration at the periphery. Low grade astrocytoma is characterized by low to moderate cellularity and slight cytological atypia (> Figure 43-7). In comparison, normal brain shows low cellularity, capillaries, a polymorphic population of astrocytic, microglial, oligodendroglial and neural cells and granular background with few fibrils (> Figure 43-8). Anaplasia, mitotic figures, microvascular proliferation, and necrosis are absent. Rarely, these neoplasms have a cystic component. . Figure 43-7 Smear of low grade astrocytoma showing low cellularity, minimal cytological atypia (arrow) and a fibrillary background. The astrocytic nuclei are enlarged (H&E, 400)

43

In solid lesions, the target point selected is usually in the central part of the lesion, where the tissue consists entirely of neoplastic cells. In such samples, the smear shows slightly increased cellularity, a monomorphous population of fibrillary astrocytes with prominent cytoplasmic fibrillary processes forming the background of the smear, and mild cytological atypia (> Figure 43-7). The pathological diagnosis rests on the differentiation of the neoplastic astrocytic proliferation from normal brain. This is easier in the smear preparation, where the abnormal fibrillary background of the neoplasm varies greatly from the finely granular eosinophilic background of normal brain. A diagnosis of low grade astrocytoma can be made with much greater confidence on a smear preparation than in sections (either frozen or paraffin) for this reason. In sections, the background is less distinctive and the diagnosis rests on identification of cellularity and cytological atypia, which might deviate only slightly from normal or reactive. MIB-l immunoperoxidase staining in low grade astrocytomas in most cases will show a very low proliferative index with Figure 43-9). These cases are impossible to distinguish from fibrillary astrocytomas by the smear alone. This type of gliosis occurs adjacent to craniopharyngioma and syringomyelia; clinical and CT data are essential for diagnosis in such cases. Cystic low grade astrocytomas represent one of the most difficult diagnostic problems on stereotactic biopsy, because nonneoplastic cystic lesions are surrounded by reactive gliosis. Cystic astrocytomas usually contain protein-rich clear yellow fluid, which commonly represents the

first sample. Biopsy of either the wall or a mural nodule shows cellular material that can make it very difficult to distinguish between low grade astrocytoma from reactive gliosis. In such cases, a good plan is to follow the patient with repeat biopsy if the cyst refills or the lesion progresses, clinical events that make astrocytoma more likely. Some low grade glial neoplasms are composed either predominantly of oligodendroglial cells (> Figure 43-10) or of mixtures of oligodendroglia and astrocytes. These cases are commonly associated with microcalcification. In such cases, the smear shows the presence of the typical oligodendroglial cells in varying numbers. These are round cells with uniform round nuclei which have a finely granular chromatin pattern. They appear either as naked nuclei or have abundant and faintly eosinophilic cytoplasm. When oligodendroglial cells are few in number, it is usually not necessary to change the diagnosis of low grade astrocytoma. In cases where oligodendroglial cells are present in substantial numbers, a diagnosis of low grade mixed glioma should be made. In cases where oligodendroglial cells dominate the smear and sections and astrocytes are few in number, the differential diagnosis is a mixed oligoastrocytoma or the infiltrating edge of a pure oligodendroglioma with associated

. Figure 43-9 Smear of tissue adjacent to a craniopharyngioma showing reactive gliosis with Rosenthal fibers (arrow) (H&E, 400)

. Figure 43-10 Smear of oligodendroglioma showing a cellular smear with uniform round cells and calcification (arrow) (H&E, 400)


reactive gliosis. When the astrocytic cells are mainly gemistocytic, reactive gliosis is favored. MIB1 immunoperoxidase staining is useful in these well-differentiated mixed gliomas to detect the possibility of a more aggressive biological behavior.

43

. Figure 43-11 Smear of subependymal giant-cell astrocytoma showing enlarged cells with atypical nuclei and abundant eosinophilic cytoplasm (H&E, 400)

Biopsy in Clinically Distinctive Neoplasms Pediatric or young adult neoplastic lesions in the cerebral hemisphere can have clinically distinctive features. These include pilocytic astrocytoma (young age, circumscription, cystic component), pleomorphic xanthoastrocytoma (young age, surface location) and ganglioglioma (young age, temporal lobe location, calcification). Smears and sections from these neoplasms frequently have higher cellularity and pleomorphism than typical low grade astrocytomas, and these features might lead to an erroneous diagnosis of high grade astrocytoma. Good communication between the pathologist and the neurosurgeon is essential for accurate diagnosis.

Biopsy of Mass Lesions in Distinctive Locations The pathologist must be aware of the exact location of the lesion being biopsied. This will permit an appropriate differential diagnosis to be developed. Mass lesions related to the ventricles include ependymoma, choroid plexus neoplasms, colloid cyst of the third ventricle, craniopharyngioma, central neurocytoma, intraventricular meningioma, and subependymal giant cell astrocytoma (> Figure 43-11). In some of these lesions, the pathological features on smear and sections might resemble other lesions, leading to potential misdiagnosis in the absence of adequate clinical and radiographic information. Examples

include subependymal giant-cell astrocytoma, which might mimic an anaplastic gemistocytic astrocytoma, and a central neurocytoma, which can closely resemble oligodendroglioma. The differential diagnosis of mass lesions in the pineal region includes pinealocyte neoplasms and germ cell neoplasms. Germinoma is the commonest of these and is characterized on smear by the presence of a triple cell population consisting of large malignant germ cells which have large nuclei with prominent nucleoli, small lymphocytes, and epithelioid histiocytes. Germinoma can present difficulties in diagnosis by stereotactic biopsy when there is extensive granulomatous inflammation within the neoplasm. Nongerminomatous germ cell neoplasms include teratoma, choriocarcinoma, embryonal carcinoma and yolk sac carcinoma. The presence of elevated levels in serum and cerebrospinal fluid of beta-hCG (choriocarcinoma) and alpha-fetoprotein (yolk sac carcinoma and embryonal carcinoma) is useful from a diagnostic standpoint. It should be noted that these neoplasms, particularly choriocarcinoma, tend to bleed profusely at biopsy, and therefore, caution is necessary.

671

672

43


The differential diagnosis of suprasellar lesions includes craniopharyngioma, pituitary adenoma, hypothalamic glioma, meningioma, germinoma and chordoma. Many of these lesions have distinctive radiological features such as the presence of calcification in craniopharyngioma and the origin from the skull base in chordoma. One major difficulty in diagnosis is the biopsy from the wall of a suprasellar cystic lesion of a young patient that shows a population of fibrillary astrocytes, sometimes with Rosenthal fibers. Care must be taken before this is mistakenly interpreted as a hypothalamic glioma. In such cases, obtaining a sample of cyst fluid is very helpful; craniopharyngioma has the typical brown, oily fluid, which can be shown to contain cholesterol crystals on a direct smear, whereas hypothalamic astrocytoma has a thick, yellow, protein-rich fluid devoid of cholesterol crystals. Lesions of the cerebellum are extremely varied. In the elderly, the commonest cerebellar mass lesion by far is metastatic carcinoma. In the young, juvenile pilocytic astrocytoma, hemangioblastoma and dysplastic gangliocytoma are the main considerations in hemispheric lesions and medulloblastoma (> Figure 43-12) and ependymoma (> Figure 43-13) in midline lesions. The diagnosis of hemangioblastoma is rarely made by stereotactic biopsy which usually produces a sample with little recognizable cellular material. Dysplastic gangliocytoma has a distinctive radiographic appearance which should be recognized. Extra-axial lesions are rarely biopsied stereotactically because they are amenable to resection. However, meningiomas (> Figure 43-14 and > 43-15) might sometimes be biopsied stereotactically when there are unusual features. Secretory meningiomas tend to produce marked cerebral edema, which might lead to misinterpretation of radiographic images as intra-axial neoplasms. Secretory meningiomas show typical meningothelial cells, which are spindle cells with ovoid nuclei, but they contain prominent cytoplasmic inclusions, which stain immunohistochemically

. Figure 43-12 Smear of medulloblastoma showing hypercellularity, uniform oval nuclei, indistinct cytoplasm and rosettes (arrow) (H&E, 400)

. Figure 43-13 Smear of ependymoma showing columnar cells with oval nuclei arranged around small vessels (perivascular pseudorosettes, arrow) (H&E, 400)

with carcinoembryonic antigen and might be misdiagnosed as adenocarcinoma.

Biopsies of Mass Lesions in HIV-Positive Patients The main indications for stereotatic biopsy in mass lesions of the brain in HIV-positive patients in our institution are (1) to make a diagnosis of


. Figure 43-14 Smear of meningioma showing whorl of cohesive cells with oval nuclei, one of which contains an intranuclear cytoplasmic pseudoinclusion (arrow) (H&E, 400)

. Figure 43-15 Smear showing psammoma bodies (laminated calcifications) in a meningioma (H&E, 400)

malignant lymphoma in a patient whose mass lesion has not responded to empiric treatment against Toxoplasma and (2) to confirm the diagnosis of progressive multifocal leukoencephalopathy [13–16]. In this setting, cerebral toxoplasmosis is rarely encountered, unlike previously, when this was the commonest diagnosis

43

. Figure 43-16 Section of cerebral toxoplasmosis showing a pseudocyst, scattered tachyzoites (short arrow) and (long arrow) chronic inflammation (H&E, 400)

in HIV-positive patients undergoing stereotactic brain biopsy. We still encounter toxoplasmosis in patients who are either not known to be at risk for HIV or in those presenting emergently in whom stereotactic biopsy is performed before HIV status is known. Toxoplasmosis is characterized by the presence of necrosis, chronic inflammation, reactive gliosis, and Toxoplasma organisms, either as pseudocysts or small crescentic tachyzoites (> Figure 43-16). The diagnosis can be confirmed by the use of immunoperoxidase staining for Toxoplasma antigens. Malignant lymphoma of the brain has become a common diagnosis at stereotactic biopsy in our institution. In addition to the epidemic of this neoplasm in HIV-positive patients, we have encountered an increase in the frequency of central nervous system (CNS) lymphoma in the HIV-negative population, both over and under 65 years of age [21–22]. Malignant lymphomas of the CNS tend to be high grade lymphomas and are commonly associated with extensive necrosis. Necrosis presents a problem in diagnosis when the biopsies show only nonviable tissue. Repeat biopsies are frequently needed to demonstrate the malignant lymphoid cells,

673

674

43


which are distinctive on smear as large, round cells with large nuclei, prominent nucleoli and lymphoglandular bodies (detached spherical fragments of cytoplasm) (> Figure 43-6). These cells are present as diffusely infiltrating cells but might have a perivascular distribution, especially in the periphery of the lesion. The diagnosis of lymphoma can be confirmed with immunoperoxidase staining for CD45 (common leukocyte antigen). Most CNS lymphomas are B-cell lymphomas. Progressive multifocal leukoencephalopathy (PML) usually produces a distinctive radiological appearance with multifocal white matter lesions. Histological diagnosis might be necessary in cases that are being considered for high dosage antiviral drug therapy. In these cases, stereotactic biopsy represents the optimal method of obtaining tissue. This condition (PML) is characterized by demyelination (associated with numerous foamy macrophages), necrosis, and a reactive glial proliferation, which shows the presence of infected cells. The gliosis is commonly very cellular with numerous highly atypical, giant astrocytes, and occasional mitotic figures. In the absence of the appropriate clinical background, the smear features can mimic an anaplastic astrocytoma. The diagnostic feature is the presence of enlarged oligodendroglial cells with large round nuclei that contain viral inclusions. In smears and sections, these distinctive nuclei have ground-glass basophilia and clumped chromatin is frequently visible (> Figure 43-17). When inclusions are rare, as in a biopsy from the edge of a lesion, immunoperoxidase staining for the JC papovavirus is helpful in confirming the diagnosis of PML (> Figure 43-18). In many HIV-positive patients, the brain biopsy tissue appears abnormal, even when no specific lesion is identified. These abnormalities include edema, chronic inflammatory cell infiltration and reactive gliosis with astrocytic atypia. The relationship of such abnormalities, which can broadly be classified as ‘‘subacute HIV encephalitis,’’ to AIDS dementia and microglial nodules is uncertain. However, it almost certainly

. Figure 43-17 Section of progressive multifocal leukoencephalopathy showing infected enlarged nuclei with a groundglass appearance (arrow) (H&E, 400)

. Figure 43-18 Immunoperoxidase stain for JC virus antigen showing positive (dark-stained) nuclei in PML (400)

reflects infection of neural elements with HIV. HIV infection of the tissue must be assumed in all biopsies of HIV-positive patients, and stringent precautions must be taken to ensure that those handling the tissue are protected during all stages of the procedure. In our institution, the incidence of HIV positivity is high enough to recommend that all tissue at stereotactic brain biopsy be handled with extreme precautions. It is of great importance that the pathologist be informed as to the HIV-positive status of a patient when this information is available to the neurosurgeon.


Biopsy of a Recurrent Lesion after Radiation for Glioma A common and difficult problem encountered at stereotactic biopsy is in a patient who develops a mass lesion after radiation therapy for a glioma. In these cases, the differential diagnosis is between recurrent astrocytoma and changes induced by radiation. Radiation changes are characterized by necrosis; vascular changes, including neovascularization, hyalinization and endothelial atypia; chronic inflammation; and reactive gliosis. The reactive glial cells often show cytological atypia, which can sometimes be severe. The reactive cells that show radiation-induced cytologic atypia can be impossible to differentiate from neoplastic residual astrocytes. Examination of radiated brain for lesions other than astrocytomas (e.g., metastatic carcinoma) demonstrates that radiation-induced cytological atypia in astrocytes can be very severe. The diagnosis of recurrent astrocytoma should be made in these cases only when there is evidence of neoplastic astrocytic proliferation. This includes hypercellularity of the astrocytic proliferation and areas where the tissue is overrun by the proliferating neoplastic astrocytes (> Figure 43-19). In biopsies that show necrotic

. Figure 43-19 Smear of recurrent astrocytoma after radiation showing high cellularity and marked cytological atypia (arrow) (H&E, 400)

43

and gliotic tissue with scattered atypical astrocytes, the diagnosis is radiation change with a descriptive comment about the atypical astrocytes. This comment may include a statement that it is not possible to distinguish reactive astrocytes with radiation change from residual single neoplastic astrocytes. The critical information provided in such a diagnosis should be the fact that the biopsy contains no evidence that the clinical lesion is the result of recurrent neoplastic astrocytic proliferation. The pathologist cannot reliably confirm the sterilization of a glial neoplasm by radiation.

Handling of a Biopsy that Shows Inflammation Stereotactic biopsies from inflammatory lesions of the brain present the greatest amount of difficulty [7]. The only time a definitive diagnosis can be made in such cases is when the etiology of the inflammatory lesion is identified, usually by recognizing an infectious agent. Diagnosis of PML, herpes and cytomegalovirus encephalitis, tuberculoma (where acid-fast stains show mycobacteria), fungal granulomas such as those caused by Cryptococcus and Coccidioides, cysticercosis, toxoplasmosis, and cerebritis caused by Mucor are examples of specific inflammatory lesions. Tissue for infectious work-up (culture and/or PCR) should be submitted from the sterile operating room environment. When the biopsy shows inflammation without a recognizable etiological agent (> Figure 43-20), diagnosis becomes very uncertain. With a small sample being obtained from a lesion, it is difficult to know whether the observed findings are representative of the whole lesion or whether the changes represent inflammation and reactive gliosis at the edge of a neoplastic lesion that has not been sampled. If the target selected was in the peripheral part of the lesion, the finding of inflammation and reactive gliosis is an indication

675

676

43


. Figure 43-20 Section showing nonspecific inflammation composed of perivascular and scattered infiltrate of reactive lymphoid cells (H&E, 400)

for taking a biopsy from a second, more central intralesional target point at some distance from the first. If this also shows similar features, a diagnosis of nonspecific chronic inflammation and reactive gliosis is made. This assumes that there is communication between neurosurgeon and pathologist and a high level of trust in the ability of the procedure to procure tissue accurately from the defined target. In cases where the final diagnosis is nonspecific inflammation and fibrosis, patient follow-up must be ensured. In those cases where the lesion progresses on followup imaging, repeat biopsy is necessary. We have encountered a few cases where a neoplasm was diagnosed on a follow-up biopsy in a lesion that progressed. When the smear shows nonspecific inflammation and reactive gliosis, the tissue must be processed to optimize diagnosis. The pathologist must ask the neurosurgeon to provide the maximum amount of tissue possible. The available tissue must be triaged depending on the features observed in the smear. In rare cases, the dominant inflammatory cell is a neutrophil; in these cases, a bacterial etiology is likely and routine and anaerobic bacterial cultures are necessary. In one case,

we isolated Nocardia species from such a specimen. Usually, however, the inflammatory cells consist of a mixture of lymphocytes and plasma cells. In such cases, a small sample of tissue can be taken for electron microscopy and the rest processed for permanent sections. Immunoperoxidase staining and, if necessary, polymerase chain reaction studies can be performed on paraffin sections for herpes simplex virus, cytomegalovirus, JC virus (PML), and Toxoplasma. In a few cases, granulomatous inflammation is identified in smears and sections. This is characterized by the presence of epithelioid histiocytes in aggregates, a feature that is more easily recognizable in sections than smears. When nonnecrotizing granulomatous inflammation is present in a biopsy of a mass in the pineal or suprasellar region, the possibility of germinoma must be considered and additional tissue requested. In other locations, and where necrosis is present, tissue must be sent for mycobacterial and fungal cultures and sections with granulomas stained with acid-fast and fungal stains in an attempt to establish a diagnosis. Cerebral sarcoidosis is also a rare possibility in cases with nonnecrotizing granulomas.

References 1. Apuzzo MLJ, Chandrasoma PT, Cohen D, et al. Computed imaging stereotaxy: experience and perspective related to 500 procedures applied to brain masses. Neurosurgery 1987;20:930-7. 2. Bleggi-Torres LF, de Noronha L, Gugelmin ES, et al. Accuracy of the smear technique in the cytological diagnosis of 650 lesions of the central nervous system. Diagn Cytol 2000;24:293-5. 3. Firlik KS, Martinez AJ, Lunsford LD. Use of cytological preparations for the intraoperative diagnosis of stereotactically obtained brain biopsies: a 19-year experience and survey of neuropathologists. J Neursurg 1999;91:454-8. 4. O’Neill KS, Dyer PV, Bell BA, et al. Is preoperative smear cytology necessary for CT-guided stereotactic biopsy? Br J Neurosurg 1992;6: 421-7. 5. Yu X, Liu Z, Tian Z, et al. Stereotactic biopsy for intracranial space-occupying lesions: clinical analysis of 550 cases. Stereotact funct neurosurg 2000;75:103-8.


6. Ferreira MP, Ferreira NP, Pereira AA, et al. Stereotactic computed tomography-guided brain biopsy: diagnostic yield based on a series of 170 patients. Surg Neurol 2006;65:27-32. 7. Ranjan A, Rajshekhar V, Joseph T, et al. Nondiagnostic CT-guided stereotactic biopsies in a series of 407 cases: influence of CT morphology and operator experience. J Neurosurg 1993;79:839-44. 8. Chandrasoma PT, Smith MM, Apuzzo MLJ. Stereotactic brain biopsy in brain masses: comparison of results at biopsy versus resected surgical specimen. Neurosurgery 1989;24:160-5. 9. Abernathey CD, Ramsey M, Knight K. Utilization of image-derived computer-assisted stereotaxis in a community-based practice setting. Stereotact Funct Neurosurg 1992;58:99-102. 10. Chandrasoma PT, Apuzzo MLJ. Stereotactic Brain Biopsy. New York: Igaku-Shoin, 1989. 11. Burger PC. Use of cytological preparations in the frozen section diagnosis of central nervous system neoplasms. Am J Surg Pathol 1985;9:344-9. 12. Kleihues P, Volk B, Anagnostopoulos J, et al. Morphologic evaluation of stereotactic brain tumour biopsies. Acta Neurochir Suppl 1984;33:171-81. 13. Chappell ET, Guthrie BL, Orenstein J. The role of stereotactic biopsy in the management of HIV related focal brain lesions. Neurosurgery 1992;30:825-9. 14. Cajulis, RS, Hayden R, Frias-Hidveqi D, et al. Role of cytology in the intraoperative diagnosis of HIV-positive patients undergoing stereotactic brain biopsy. Acta Cytol 1997;41:481-6.

43

15. Skolasky RL, Dal Pan GJ, Olivi A, et al. HIV-associated primary CNS morbidity and utility of brain biopsy. J Neurol Sci 1999;163:32-8. 16. Vallat-Decouvelaere AV, Chre´tien F, et al. The neuropathology of HIV infection in the era of highly active antiretroviral therapy. Ann Pathol 2003;23:408-23. 17. Chandrasoma PT: Problems relating to pathological interpretation in stereotactic biopsy procedures. In: Apuzzo MLJ, editor. Brain surgery: complication avoidance and management. New York: Churchill Livingstone, 1993. p. 425-31. 18. Wakimoto H, Aoyagi M, Nakayama T, et al. Prognostic significance of Ki-67 labeling indices obtained using MIB-1 monoclonal antibody in patients with supratentorial astrocytomas. Cancer 1996;77:373-80. 19. Kouwenhoven MC, Kros JM, French PF, et al. 1p/19q loss within oligodendroglioma is predictive for response to first line temozolomide but not to salvage treatment. Eur J Cancer 2006;42:2499-503. 20. Apuzzo MLJ, Hinton DR. Clinically relevant issues attendant to pathology. In: Apuzzo MLJ, editor. Malignant cerebral glioma. Park Ridge, IL: American Association of Neurological Surgeons, 1990. p. 19-21. 21. Namiki TS, Nichols P, Young T, et al. Stereotactic biopsy diagnosis of central nervous system lymphoma. Am J Clin Pathol 1988;90:40-5. 22. Haldorsen IS, Krossnes BK, Aarseth JH, et al. Increasing incidence and continued dismal outcome of primary central nervous system lymphoma in Norway 1989–2003. Cancer 2007;110:1803-14.

677

38 Robotic Neurosurgery P. L. Gildenberg

What is a Robot? In order to discuss robotic surgery, it is first necessary to ask, ‘‘What is a robot?’’ The term robot is assigned to many devices that have little in common. As a minimum, to be called a robot, the device should have a program or series of steps that are initiated by a command. Some items do not meet even that simple definition, but are called robotic as a marketing tool, rather than a technological descriptor. To begin our discussion, let us look at some of the earliest examples. The earliest use of the term robot referred to the mechanical men and women in Karel Capek 1923 play ‘‘R.U.R.’’ or ‘‘Rossum’s Universal Robots.’’ The term stems from the Czech word robota, meaning to work or compulsory labor, and is related to the Greek arbeit, also meaning to work. The ‘‘robotniks’’ were peasants who owed such labor. The primary dictionary definition is ‘‘a machine that looks like a human and performs various complex human tasks, such as walking or talking,’’ although this describes an animatron, which may or may not be a robot. Subsequent definitions include ‘‘any machine or mechanical device that operates automatically with humanlike skill’’ (Random House) or ‘‘an automatic apparatus or device that performs functions ordinarily ascribed to human beings or operates with what appears to be almost human intelligence’’ (Webster). Surprisingly, NASA defines a robot as simply ‘‘a mechanical device that operates automatically.’’ The first ‘‘robots’’ were mechanical devices that obtained power from such things as falling liquid, such as water. In 1,000 BCE, an ancient #


Egyptian water clock depended on the rate water drained through holes in a container. As early as 300 BCE, Heron of Alexandria described, in a series of five volumes that still survive, a group of hydraulic devices that were robotic in nature [1]. Placing a cup on a platform (the command) would fill the cup with wine and then stop (the program). Overfilling a wine ‘‘greedy cup’’ would cause the wine to drain from an opening on the bottom of the cup to deposit the wine on the lap of the greedy guest. Heron also used pneumatic pressure by running water through a pipe into an air filled container so the escaping air made several silver birds chirp as they moved their wings. He also invented the cam, which he used to program a robotic puppet theater powered by ropes, wires and levers to cause several figures to move in such a way that they seemed to interact within a group. The program could be changed by shortening or lengthening the ropes or levers, which may constitute the earliest computer programming. About that same time, a coin operated (the command – insertion of a coin) dispenser (the program) of a measured amount of water and a piece of soap (actually a ball of pumice) was invented in Byzantium to be used for worshippers to wash their hands as they entered a temple. None of those robots resembled a person. A robot that is made to look like a person or animal is called an animatron. If speech or sounds supposedly emanating from the animatron are added, the result is a resulting robot is an audio-animatron. Leonardo da Vinci is considered by many to be the inventor of the first animatronic robot. He made a robot that looked like a knight in armor

584

38

Robotic neurosurgery

who extended his arms or returned them to his sides (the program) or sat down (another independent program) or arose (another program) on command, a working replica of which is in the da Vinci Museum in Florence [1]. Among a multitude of his inventions, he also made a walking lion that could be programmed to perform acts which could be varied through a series of gear ratios, which could be selected. Some time later, he also used such gears in machines designed to move men and weapons, a precursor to the tank. The ancient Greek water clock, which relied on power from falling water to implement the routine or program, would qualify, as would a present day alarm clock ‘‘programmed’’ to sound a bell at a predetermined time. Other early examples include an animatron whose voice was supplied by a recording of a human voice, long before computerized speech simulation was developed. Elektro, was designed to look somewhat like a person. He (or it) was introduced at the 1939 World’s Fair in New York, and probably represents the first robot built in the US. As people approached, he would greet them or offer other pre-recorded messages. A person commanded each program to run, which was necessary in that pre-computer stage. As direct descendants, the best examples of present day audio-anamatronics may be seen at Disneyland. Human and animal figures go through a series of actions synchronized with pre-recorded speech or animal sounds. There may be a multitude of programmed gestures and facial expressions. The computer operated program may sometimes be extremely complicated, so that one audio-animatron seems to interact with another. You can see somewhat less sophisticated animatrons at your local Chuck E. Cheese. In its simplest form, the program is predetermined by the operator, and the robot has no decision-making capabilities, but the command may initiate a series of steps. A simple robot may be under direct moment-to-moment

control of the operator, for instance a wireless remote-controlled toy car. A program might initiate a series of operations, as in a robotic coffee maker that may at the programmed time grind the beans, put a measured amount into the basket, and let a measured amount of water drip through the coffee grounds to fill the cup below. Robots may be electronic, electrical, or mechanical programmable self-controlled devices. Typically, a modern robot performs a task by following a set of instructions stored in an onboard computer or robot controller that specifies exactly what it must do to complete the job. The computer sends commands to each of the robot’s motorized joints, which function much like human joints to move various parts of the robot. The operator input is channeled through a computer interface that is programmed by the operator and then instructs the robot controller to accomplish the programmed task. The commands that constitute the program may be a series of steps in a routine or subroutine that is initiated by the command. Alternatively, the robot can also be used to denote a computer interfaced technique wherein the operator controls each step of the program by a series of stepby-step manual inputs, which constitutes a master-slave electronic manipulator, as in the da Vinci surgical robot. The use of feed-back that is recognized by the computer and influences subsequent steps requires sensory input, which may mimic the human in that it can be visual (video), tactile (a sensory device such as a pressure transducer), thermal (a thermostat), or proprioceptive (to let the robot know the position of its arm or arms, from which position further actions may ensue). Such systems may also be used to improve spacial accuracy of a robot, for instance with continuous video or other visual registration feedback to correct the trajectory or the tool held at the end of the robot arm and make it increasingly accurate as the target is approached.


Additional feed-back controls and multistep programs may be superimposed on a simple design. At the given command, a series of activities can occur, with the complexity increasing significantly, especially if the feedback from the environment is used to determine which steps of the program the robot will follow. There may be a number of environmental variables, with significant increase in the complexity of the robot program. When this interaction with the environment becomes so complex that it suggests the robot is thinking for itself, such artificial intelligence (AI) may provide the robot with a complex multitude of options for responding to the environment. An annual competition involves designing a vehicle that can navigate on the open road, following a specific route, with no input except the sensors in the robotically controlled vehicle. Another example of complex feedback being integrated into the robotic program is a robotic floor sweeper, which may have a feed-back system so it can decide on a path around the room, while avoiding objects on the floor or change directions when a wall is contacted. If a piece of furniture is moved, the robot can correct its route and remember it during future excursions. An electro-mechanical suturing robot is presently being developed in Houston. It is essentially a hand-held device that passes a suture through tissue on command of the surgeon. Because the feed-back involves the surgeon who positions the robot while looking at the surgical field, initiating a series of steps by pressing a control button and moving on to the next step, the surgeon is part of the feed-back loop, so we refer to the device as a ‘‘semi-robotic’’ suturing device. The robot must have a customized tool, an end effector, mounted at the end of the arm to perform the specific tasks for which the robot has been programmed. It may be welding tips for spot welding on an auto assembly line, a paint sprayer to paint a vehicle part or a wall,

38

or suction cups to pick up a part from an assembly line. A robot designed for surgery may have an end effector designed to hold an endoscope or retractor, an electrode or biopsy needle holder, a small forceps or needle holder, an electrocautery, or a dissecting tool. A robot for use in orthopedic surgery may have a drill or machining assembly mounted at the end of the arm.

When does a Device Become a Robot? There is no exact line separating non-robotic mechanical devices from a robot, and the opinion whether some mechanical device is a robot depends on whether you ask the inventor, the engineer, or the investor. (In that regards, the word ‘‘robot’’ is in the same category as the word ‘‘nano,’’ which Richard Smalley, who won the Nobel Prize for nanotechnology, said only half joking, ‘‘Nanotechnology is anything that bears the name ‘nano’ and makes a lot of money.’’) A surgical or industrial robot usually does not have arms and legs like a person, or even two arms. In fact the most commonly used industrial robot consists mainly of only one arm with a tool at the end (> Figure 38‐1). The tool mounted at the end of the arm is selected depending on the robot’s task, such as welding, painting, picking up or moving a part on an assembly line, or fitting together components of the article being manufactured. The arm of a robot may move by hydraulic or pneumatic pressure that can be finely regulated. Alternatively, a robot may move by electromechanical motors or actuators, which have become more available and precise and controllable as computers have become more sophisticated. The robot joint may move by a fly-by-wire system that that has either a computer control system or the mechanical drive near the part that is to be moved, and such systems operate flight control surfaces in airplanes.

585

586

38


. Figure 38‐1 (a) and (b): a model of a Kuka industrial robot, typical of those used in automobile manufacturing and the type used in the CyberKnife

The industrial robot arm (> Figure 38‐1) has multiple joints, usually six or seven, corresponding to degrees of freedom. Often, the joints can be compared with the shoulder, elbow, wrist, or fingers, which can extend, flex, or even rotate. In order for such a robot to move the ‘‘hand’’ smoothly from point A to point B, it may be necessary for all the joints to move sequentially or simultaneously. The path between the points can be defined, and the computer that serves as a robot controller then calculates how much to move each joint to accomplish the action. In order for the robot to interact with its environment, the robot must know precisely where a target or task is in relation to the robot. In surgery, this may take the same form as registering the patient’s head and target to any image guidance device. Registration may be accomplished by the use of surface or inserted fiducials, again like image guided surgery. On a production line, registration may involve determination of the position of the target in relation to the robot body by accurately placing the target at a specific spot relative to the robot, or by providing the robot accurate feedback from the tool at the end of the robot arm so the robot can adjust the localization of the tool accordingly. For instance, in the hair transplantation robot, described

below, both registration of the target on the scalp to the location of the robot arm and identification of the target are done by stereoscopic video cameras mounted on the tool at the end of the robot arm. In neurosurgery, registration may be done by any technique that is used for image-guided or stereotactic surgery to determine the precise position of the patient or organ relative to the robot. What jobs are best done by a robot? Those that are predictable, repetitive, stressful, intense, tedious, and/or spatially oriented. Those that require a repetitive series of steps, even though complex, may qualify. Is short, those characteristics that make a task boring or tedious for a human may make it ideal for a robot. A passive robot is one that attains a position that is to be maintained for a longer period of time than would be comfortable for a surgical assistant. Examples are robots that serve not as the surgeon, but as the surgical assistant, holding retractors or positioning endoscopes. Such robots are easier to develop, and have been the first robots developed for clinical use. An active robot, on the other hand, is programmed to do things (the program) and may have a number of tasks to do during surgery. One approach in the development of robotic surgery is to take stock of how


many individual steps are done in a particular type of surgery. Then concentrate on those individual tasks that are repetitive, spatially oriented and tedious and consider whether a robot could do the job better than the surgeon or assistant. When is a robot (or other technologic advance) of use to the neurosurgeon, or to anyone else, for that matter? A scale that has recently been called ‘‘the Gildenberg technology scale’’ has four phases.

Phase 1 – The device can do what the surgeon can do manually, but not as efficiently or not as fast. This is of interest only to the inventor. Phase 2 – The device can do what the surgeon does, just as fast and just as efficiently. This is of interest to the developer and possibly to potential or actual investors. It may be more useful for marketing than for surgery. Phase 3 – The device can do what the surgeon does, but faster or better. This is of interest to other surgeons and to investors, may be cost effective, and may be used at many institutions. Phase 4 – The device can do something the surgeon cannot do without the robot. This is disruptive or revolutionary technology, of interest to all.

There are very few techniques that reach phase 4 (See chapters E-5 and E-14 on the CyberKnife). At present, the CyberKnife may be the only neurosurgical device that reaches that level. It consists of an x-ray tube mounted at the end of an arm of an industrial robot. The targeting and distribution of the radiation cannot be achieved without the robot and its programming, since the non-isocentric program is too complex to be calculated manually. Because of the flexibility of the programming, the same device can be used for stereotactic radiosurgery, hypofractionated radiotherapy as well as conventional radiation therapy throughout the body. One particularly

38

impressive feature is the ability to radiate pulmonary lesions during respiration, since the robot can follow the movement of the tumor on a moment-to-moment basis. The other outstanding example of a phase 4 robotic device is Lasik eye surgery that reconfigures the shape of the cornea in order to correct vision. The device determines the pre-operative error in refraction, calculates the way the cornea should be recontoured, and contours the cornea. Although the surgeon has ultimate control, the procedure cannot be done manually. On the bottom line, however, is that most surgical robots are planned to permit the surgeon to do a better job, and not to replace the surgeon.

Surgical Robots Although there has been great optimism since the early 1990s about the incorporation of robotics into surgery, including neurosurgery [2], progress has been very slow, despite major advances in computer science. Such development is expensive and consequently requires either an investor not afraid of significant risk or an industrial collaboration. It requires a developmental team with a variety of interdigitating specializations, takes a great deal of time to evolve, and confronts many regulatory issues. In addition, there are concerns about intellectual property and patents that make development difficult in the need to avoid infringement on existing patents or to know when infringement may occur [3]. Regulatory concerns are many, since the safety of the robot performing all or part of a procedure is paramount. Advances in other less expensive image guidance technology may supply many of the benefits of robotics at less expense and be sooner to market. Ideally, a surgical robot should be proven to be cost effective before being introduced into the marketplace, by making the surgery more efficient. Even if not cost justified, the term ‘‘robotic’’ is attractive and an excellent

587

588

38


marketing tool, so it may find buyers even in Phase 2. Some robots designed for surgery are adaptations of existing industrial robots, which may restrict their capabilities, but make it less expensive to develop. Other surgical robots are designed ‘‘from the ground up,’’ which involves much more laborious and expensive development, but may result in a more functional system. The task of some robots is to attain a certain position and then remain at that position, such as a robot designed as a retractor holder. The position may be modified to another position, which is then held until the robot is commanded to change positions again. Other robots are programmed to perform tasks that require frequent or continuous movements, such as auto industry robots that may apply paint to the outside surface of an automobile. Most robots of interest to neurosurgeons are not configured at all like a person. Most have but a single arm, but a few may have two. The tool at the end of the arm may be a retractor holder, a holder of an endoscope, a holder for insertion of a biopsy cannula or electrode, or even a surgical microscope. Thus, the combination of a robot and an image guidance system may accurately place an electrode or aim an operating microscope at predetermined stereotactic coordinates or at a point-in-space determined from a preoperative CT scan or MRI using targeting technology similar to image guided surgery. The use of a robot to insert an electrode stereotactically may not provide a clinical advantage, even if less than a millimeter improvement of the accuracy of a stereotactic frame is achieved. The robotic procedure may take longer than using a conventional stereotactic frame or image guidance, most of which have a computer program with appropriate planning software. In competition with such computerized but manually operated guidance systems, the surgical robot may not reach a level of cost effectiveness. In evaluating cost-effectiveness, many interacting processes must be considered. Operating room

workflow may be optimized by use of a robot which may decrease operating room time by taking advantage of potentially more efficient technology [4]. Conversely, the set-up time for a robot may disrupt operating room workflow. One of the first surgical uses of a robot was by Sakaguchi [5] in 1985, whose robot performed percutaneous nephrostomy. The localization was performed by establishing an entry point on the skin and a target point in the dilated renal pelvis. Then the trajectory was calculated and the robot programmed to follow that course. A more recent contemporary technique to aspirate or biopsy with a needle involves registration of the target by three-dimensional ultrasound [6,7], with robotic placement of the needle. The surgical robot that presently has dominated the field is the da Vinci robot, made by Intuitive Surgical. A perusal of its design and capabilities may provide a basis for discussing other robots that are being developed for neurosurgical use. Each motion is controlled directly by the surgeon, so one may argue that it is an electro-mechanical or master-slave device rather than a robot per se. It has very few if any internal program sequences but the controls are sophisticated but yet intuitive. The main claim to being a robot is that it allows scaling of movements, so a surgeon may, for instance, move his or her finger one centimeter, but the robot only moves the instrument 1 mm. It may also dampen physiologic tremor to hold the instruments with a steady hand. Surgeons generally feel that the ability to scale movements provides more of a benefit than of dampening physiologic tremor. The presently available commercial version does not have haptic feed-back, although attempts to provide that have been reported [8] and recently have had considerable success [9]. It has, however, what is arguably one of the best stereoscopic endoscopic views of the operative field available. A study demonstrated that suturing with the three-dimensional vision of the da Vinci robot was 65% faster


than suturing when the surgeon had a twodimensional view [10]. By using a master-slave control, the surgeon may command the da Vinci system at each movement, which does not require preoperative planning of steps prior to surgery, so the surgeon has the ability to respond to contingencies as they arise. The surgeon sits across the room at a console and looks through two openings at two monitors that together provide exquisite stereoscopic vision. Two or three working channel arms plus the endoscope are manually inserted through access ports. The tools for each working channel are selected and secured to the intended arm. A variety of tools is available and can be changed during the surgery without withdrawing the working channel, ordinarily by a surgeon who is stationed next to the patient. The da Vinci system includes tools for various functions, such as cutting, coagulating, and suturing. That, in combination with direct real-time visualization of the operating field at the business end of the endoscopic tools provides continuous feedback to the surgeon, who exerts moment-to-moment control. Despite minimal if any haptic sensory feed-back, the combination of micromanipulator control over the tools and excellent stereoscopic visualization provides the surgeon with improved capabilities for a number of types of surgery [11]. The da Vinci Surgical Robot was invented by Philip S. Green, of the Stanford Research Institute, who also invented clinical useful ultrasound in the late 1960s [12]. The idea of having a surgeon sit at a console to perform the surgery, which could be miles away from the operating room was named the Green Telepresence System. It was initially called Mona, after da Vinci’s Mona Lisa, but the name was changed to da Vinci in 1999. In 2000 it became the first robotic surgical system to be approved by the FDA. The da Vinci robot is presently in common use in pelvic surgery, especially such procedures as radical prostatectomy [13–15]. The view of that surgical field in that procedure has always

38

been problematic because of the prostate’s position behind the pubic rim, and yet great care must be taken to avoid damage to small nerves and blood vessels overlying it. The Da Vinci provides excellent visualization of the tissue and tools throughout the procedure. Gynecologic surgery has been the next most frequent indication for the use of the Da Vinci [16]. Certain abdominal surgery procedures are now being done robotically [17–19]. Although it was originally intended primarily for thoracic surgery, only recently have techniques been developed for the da Vinci system to allow mitral valve replacement and coronary artery bypass on the beating heart [20,21]. The da Vinci has not to date made significant roads into neurosurgery. There seem to be several reasons for that. First, those surgical exposures which require excellent visualization of critical structures are already done in conventional minimally invasive surgery, often with an operating microscope. Second, the da Vinci arm bearing the endoscopes prevents direct access of the surgeon to the surgical field, which may be critical if an emergency should arise. Third and perhaps most important, neurosurgeons already have a superb stereoscopic view with depth perception deep within the surgical field by means of the stereoscopic operating microscope, which can be aimed with image guidance if necessary, and the target outlined by a heads-up display. Since the surgeon uses the tools held in his or her hand, there is haptic feedback that is not available through the robot. On the other hand, microscopic neurosurgery may be reaching the limits of dexterity of the surgeon’s hand [20], so having the option of scaling the movements may be significant in future procedures. Consequently, neurosurgeons already have excellent stereoscopic microscopic visualization, although the availability of scaling (movement reduction) with instruments free from physiologic tremor may entice neurosurgeons to begin to use the da Vinci system.

589

590

38


Robotics is beginning to make inroads into orthopedic surgery. There are three types of orthopedic related robots, one directing a cutting guide block or drilling sleeve, another to constrain the range of movement of a surgical instrument, and one which directs a milling device automatically in order to secure a custom made prosthesis according to the preoperative plan [22]. One of the earliest surgical robotic systems is the RoboDoc, that fits into the third category, When it was introduced in 1992, it was the first robot to assist the orthopedic surgeon in a total hip arthroplasty [23]. It integrates planning and production of both the acetabular cup and the femoral stem, providing a significant advantage over conventional artificial hip implantation. It is used in conjunction with the OrthoDoc Preoperative Planning Workstation to determine the optimal shape of each component of the implantable hip and femur stem to achieve maximal 95% contact with the bone for optimal stability of the prosthesis. A similar system incorporating tremor filter and motion scaling was reported that same year from Johns Hopkins and was called the Steady Hand System. In many respects, this system resembled the da Vinci system [24], which became their model, in that it allowed scaling of movements and filtering of tremor. According to a 2002 report from the company, the intellectual property was assigned to ImageGuide, that began collaboration with GE at that time. There is one robotic surgical technique that I introduced that is still under development that used a smaller version of the kuka industrial robot as the CyberKnife – robotic hair transplantation [25] (> Figure 38‐2). It is based on experience with image guided and stereotactic surgery. It is equivalent to robotically inserting an electrode into a brain target, but you stop at the skull and may do it 2,000 times. A normal non-robotic hair transplantation procedure involves removing a strip of hair-bearing scalp from the occiput. Perhaps four to six technicians with desktop

. Figure 38‐2 Robotic hair transplantation involves planning the procedure in advance, retrieval of individual hair follicles or follicular units, and inserting them into preprogrammed retrieval sites in the bald area

operating microscopes and razor blades trim the tissue until just one follicle or follicular unit is in each piece, which may number 1,000–2,000. The surgeon makes that same number of punctate incisions in the bald part of the scalp, and then inserts each follicular unit graft bearing one or several hair shafts into each incision. The process ordinarily takes 6 h or more. It is tedious, time consuming, stressful, repetitive, and spatially oriented, which makes it a perfect job for a robot. The robot control and programming have several similarities to the CyberKnife program (see chapter E-5), in that it uses an industrial robot with a specifically designed tool at the end of the arm. However, since all the donor follicular units and scalp recipient targets are on or near the surface, registration of the target can be done with stereoscopic video cameras mounted on the tool at the end of the robot arm. The video can provide both the scanning of the head for preoperative planning, registration of the head to the robot, and spacial accuracy of less than one millimeter for follicular unit retrieval. Both the retrieval of grafts and the insertion sites are planned on the same computer. Although


the computer holds all the information for the entire procedure, the surgeon still has the ultimate control [26].

Neurosurgical Robots There have been many preliminary reports of the use of robotics in neurosurgery. It is often difficult to know whether those devices were developed into a commercially available system under a different name. Existing companies may acquire a robotic system during pre-clinical development in order to merge some of its technology into their own technology. Further it is somewhat frustrating to review the development of new technology, such as the use of robotics in surgery, since much of the information is proprietary and involves industrial development, so never appears in the surgical literature. These considerations make it particularly difficult to follow each robotic system throughout its development. I have tried to review those robotic systems that were demonstrated to be clinically useful, but some systems merit mention may not be included, and some that are discussed may not attain commercial success. In order for a robotic system to be attractive to neurosurgery and to achieve phase 4, several criteria would have to be met. Foremost, the surgery would have to be both more accurate and faster than a neurosurgeon can do with an operating microscope. The system would be image guided, since it would be in direct competition with manual image guided techniques. It would have to be under the direct control of the surgeon, but add programming of some or all the steps of the surgery so that it would enhance image guidance, rather than interfere with it. It should be adaptable to telerobotic surgery, especially since there are an increasing number of places where there are few neurosurgeons available. It must be cost-effective, since we are headed toward a future of tightening of health

38

care budgets. Several neurosurgical robots under development are still in Phase 1 or Phase 2 of development. Reviews of existing neurosurgical robots have recently been presented by Louw [27] and McBeth [28]. Benabid [29,30] was the first neurosurgeon to report the use of a robot to carry out stereotactic targeting in 1987, when he reported preliminary results with a six-axis robot linked to a stereotactic frame. The probe holder was positioned robotically to reach a target that had been calculated from x-rays and angiograms. By 1992, he had experience in 140 cases, and hailed its use particularly in image guided robotic endoscopy [30]. The NeuroMate (Integrated Surgical Systems, ISS) was the first neurosurgical robot approved by the FDA and consequently the first commercially available. The navigation is based on pre-operative 3-D imaging, and it can be used in either a frame based or frameless configuration. The system can be either ceiling or floor mounted, with accuracy comparable to other image guided or stereotactic techniques [31]. It has been used most frequently to guide an operating microscope [32] or to implant DBS electrodes [33,34]. Its reported use for electrode insertion is similar to using a frame in the manual mode, which would probably put it into phase 2 of the Gildenberg technology scale. The Zeiss MKM stereotactically guided microscope, the Mehrkoordinaten Manipulator (MKM) robotic navigation system for frameless stereotactic procedures, which was introduced in the late 1990s sand was considered by them to be a robotic device in that it is registered to the patient as in other image guided surgery, after which computerized techniques maintain its optical orientation to the surgical field. The accuracy compared favorably to the BRW stereotactic frame [35]. It was one of the first image guided surgery devices and generated a great deal of interest [36]. It provides not only stereotactic optics of a neurosurgical surgical microscope, but superimposes on the view of the surgical field a definition of the target that is not dissimilar

591

592

38


to the system that Kelly [37] had introduced as early as 1980. Kelly’s apparatus was a frame-based stereotactically surgical microscope that was manually positioned, and provided the target visible in a heads-up display through the microscope. Kassell [38] linked it to a one of the first remote telepresence system, which he named SuMIT, for Surgical Manipulator Interface Technology. In 1987, Young [39] introduced the Unimation PUMA (Programmable Universal Machine for Assembly) industrial robot modified for neurosurgery, which is presumably the same PUMA 560 that Erich Mu¨he used in 1985 to perform the first robotic-assisted laparoscopic cholecystectomy [40]. A urologic version of that robot was successfully programmed to perform transurethral resection of the prostate (TURP), which, in turn, led to a version specifically related to urologic procedures such as prostatectomies [41]. In neurosurgical use of the PUMA robot, frame-based coordinates were calculated from pre-operative imaging with a dedicated CTscanner. A biopsy was the first robotic image guided procedure using this system [42]. Target coordinates and trajectory were calculated and fed into the robot

control computer, similar to non-robotic image guidance programs. Experience with its use in a series of children for brain tumor resection was reported by Drake (> Figure 38‐3) [43]. The Minerva robot, developed at the University of Lausanne, Switzerland, was described in 1990, and offered the neurosurgeon frameless guidance. The robotic configuration included an intraoperative CTscanner [44]. It has been used clinically for both biopsy and electrode insertion [45–47]. A Scandanavian neurosurgical robot from Finland was introduced in 1986. the Oulu Neuronavigator System/Leksell Index System. A clinical trial was done from 1980–1984 involving 93 successful operations in 77 patients [48]. A version that involved ultrasonic localization of the target was also introduced about the same time [49]. Louw [27] and his group from Canada, which also included Rizun and Sutherland, who is the project director, presented their new neurosurgical robot, the NeuroArm, in 2004 (> Figure 38‐4). It is one of the most sophisticated robots to be used primarily for neurosurgery. The system was designed from the ground up, rather than attempting to convert

. Figure 38‐3 The use of a PUMA robot in pediatric brain tumor resection [43]


. Figure 38‐4 The NeuroArm robot neurosurgery [27]

designed

specifically

for

a commercial system to the operating room, which has allowed the developers much more flexibility. Their system would include haptic feedback, the sensation that is conveyed when the surgeon grasps tissue or an instrument [50], in addition to motion scaling and physiologic tremor suppression [51]. It can be used either with one arm in an MRI compatible mode, or as an ambidextrous or two-arm configuration. The surgeon sits at a workstation in an adjacent control room, where he or she has access to several monitors, including a stereoscopic viewer. They use servo control of individual joints of the robot arm, in a master-slave system, similar to the da Vinci. Each arm has eight degrees of freedom, which provides particularly good dexterity. The two-arm configuration is designed for fine

38

microsurgical movements, depending on the tools that are mounted on each arm. The MRI configuration permits only the use of one arm. An area where robotics is of considerable interest at present is telesurgery. If a surgeon can perform surgery when sitting across the operating room, why not provide that capability to a hospital miles away where they have no neurosurgeon? There has been considerable interest in such a system by the military and space programs. As computer communication becomes commonplace, the restrictions of offering such a services to remote locations are becoming less daunting. The NeuRobot was introduced in 2002 by Hongo and Goto in Shinshu University School of Medicine, Matsumoto, Japan [52]. It consisted of a master and a slave micromanipulator and three-dimensional display. By 2003, a telecontrolled manipulator was incorporated [53,54]. In 2003, the use of an AESOP 3000 robot was used for telesurgery between Baltimore and Sao˜ Paulo, Brazil, in one case to hold and direct a laparoscope, and in another case to use a robot specifically designed for access to the kidney [55]. A recent report from China by Tian and his group [56], using their CAS-BH5 robotic system, demonstrated the usefulness of telemanipulation in 10 patients via a digital data network between Beijing and Yan’an, 1,300 km away. Lum, from a group from the University of Washington in Seattle led by Hannaford [57], have recently introduced a light weight robot designed for military mobility, sponsored by the US Defense Advanced Research Projects Agency (DARPA). Its mobility was of primary importance, so it could be transported to the field of battle, for which it underwent a simulated trial in the California desert [58]. Computer Motion, Inc. has developed several surgical robotic systems. One of its premier devices was the voice-controlled Automatic Endoscopic System for Optimal Positioning (AESOP),

593

594

38


which allowed the surgeon to direct or move an endoscope or working channel with either a manual mode or with voice commands. A study in 1997 with medical students as subjects showed that there was a steep learning curve at the beginning, but once facility with vocal commanding the robot was achieved, the stability and movement of the endoscope image provided a significant benefit [59]. A report 2 years later concluded that the use of the voice-controlled AESOP robot for placing the endoscope allowed a single surgeon to perform endoscopic mitral valve surgery, a distinct cost advantage [60]. The AESOP was also found to be an advantage in endoscopic pituitary surgery [61]. A recent report added a capability for eye gaze tracking to aim the AESOP robotically positioned endoscope [62]. The AESOP has been used for urology [63], general surgery [64], laparoscopy [65], laryngoscopy [66], but the main use that may affect neurosurgery is the use in the endoscopic approach to the sella [61]. Computer Motion also produced the Zeus system which provided voice control of the endoscope and manipulation through the working channels, which was introduced into the US through Medtronic. Both of these systems were successfully used for endoscopic cholecystectomy [67], coronary artery bypass surgery [26] and gynecologic surgery [68], but it is uncertain whether they were used in neurosurgery. A direct comparison between the AESOP, Zeus, and da Vinci demonstrated that the da Vinci required less operating room time than the other two systems [69]. In 1995, Giorgi [70] presented a preliminary report of a robotically directed operating microscope, developed while he was at the University of Maryland. After he had returned to Italy, in 2000, he presented clinical experience in 14 children with a more sophisticated robotically driven operating microscope. The microscope was registered to the patient’s head with three progressive scan-synchronized infrared cameras mounted around the lens of the microscope,

which could be directed with a six-axis joystick that was used as the microscope handle [71]. Also in 1995, a Robot-Assisted Microsurgery System (RAMS) was developed with NASA collaboration. MRI was integrated into the system. There was a master-slave or micromanipulator control system with six degrees of freedom for three-dimensional manipulation. Outstanding features were an adjustable filter of physiological tremor and adjustable motion scaling which enhanced surgical precision [72]. A hexapod-based robot system was introduced in 2004, the Evolution 1, (Universal Robot Systems, Schwerin, Germany) [73]. It was tracked by the Stealth Neuronavigation System (Medtronic), and was used for endoscopic transsphenoidal surgery with good accuracy and control. It had the advantage of being able to track two instruments simultaneously. An advanced design was introduced in 2007, which also combined frameless stereotaxy, endoscopy, and robotics. The user interface was simplified and additional safety features were introduced [74]. Several robotic systems involving spine surgery are of interest to neurosurgeons. Perhaps the first such system, an ingenious small robot that is attached directly to the spine was introduced for the accurate placement of pedicle and translaminar screws [75]. One of the difficulties in the advance of robotics in endoscopic surgery is the need to miniaturize instruments that can be inserted through working channels [76]. Such devices are becoming more realistic with improvements in miniaturization and computer design. One device that may be useful in neurosurgery, as well as other vascular procedures, is a hand-held robot that streamlines the task of suturing. It is called a semi-robotic suturing device, since the surgeon remains part of the feed-back loop. It is anticipated that the microversion will make vascular suturing more efficient and less stressful to tissues for the expert surgeon. A larger design will be used to provide


expertise to those less skilled in surgical techniques who are occasionally called upon in the early management of wounds, such as in the field or on-board ship for the military, where a skilled surgeon may not be immediately available. It is still under development at Suture Robotics in Houston. Where are we in regard to neurosurgical robotics becoming a reality? Although robotics is making inroads into certain surgical procedures, most progress is in visualization of minimally invasive or microsurgical targets in regions that are difficult to visualize or gain access to. The site where this is a particular concern is in the pelvis, and the Da Vinci robot has become a cost effective tool for surgery in that region. It has more recently become useful in thoracic surgery, both because it minimizes the exposure and because it has been successfully adapted to coronary artery and mitral valve procedures. In all these cases, surgery can be done more effectively and efficiently with the Da Vinci or similar system than without. Attempts to introduce robotics into neurosurgery have been less successful than in other specialties. The superb stereoscopic visualization of the surgical field through a small incision is already available with image guided neurosurgery and an operating microscope. To date, there has arguably not been a neurosurgical procedure that can be done more efficiently or more effectively with a robot than without (Phase 3), except the CyberKnife in stereotactic radiosurgery, although non-robotic image guidance provides much of the function usually assigned to a robot. As robotic surgery becomes more sophisticated, as computer programming introduces subroutines that can be done efficiently by robotics, as we reach the limits of manual microsurgical techniques, the ability to scale movements and eliminate physiological tremor will invite robotic participation. The problem of access to limited surgical exposure because of the presence of a robot will recede as smaller

38

and smaller robots with more capability are developed, perhaps with flexible optics that would permit free access to the surgical field and still remain out of the surgeon’s way.

References 1. History of Robots. The History Channel. 4–17–2007. 2. Preising B, Hsia TC, Mittelstadt B. A literature review: robots in medicine. IEEE Eng Med Biol Mag 1991;10:13-22. 3. McLean TR, Torrance AW. Are the Brookhill-Wilk patents impediments to market growth in cybersurgery? Int J Med Robot 2008;4:3-9. 4. Sutherland JV, Van den Heuvel WJ, Ganous T, Burton MM, Kumar A. Towards an intelligent hospital environment: OR of the future. Stud Health Technol Inform 2005;118:278-312. 5. Sakaguchi T. Percutaneous puncture with a robot. Hinyokika Kiyo 1985;31:1265-8. 6. Boctor EM, Choti MA, Burdette EC, Webster Iii RJ. Three-dimensional ultrasound-guided robotic needle placement: an experimental evaluation. Int J Med Robot 2008;4:180-91. 7. Boctor EM, Webster RJ III, Mathieu H, Okamura AM, Fichtinger G. Virtual remote center of motion control for needle placement robots. Comput Aided Surg 2004;9:175-83. 8. Shimachi S, Kameyama F, Hakozaki Y, Fujiwara Y. Contact force measurement of instruments for forcefeedback on a surgical robot: acceleration force cancellations based on acceleration sensor readings. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8: 97-104. 9. Shimachi S, Hirunyanitiwatna S, Fujiwara Y, Hashimoto A, Hakozaki Y. Adapter for contact force sensing of the da Vinci robot. Int J Med Robot 2008;4:121-30. 10. Badani KK, Bhandari A, Tewari A, Menon M. Comparison of two-dimensional and three-dimensional suturing: is there a difference in a robotic surgery setting? J Endourol 2005;19:1212-5. 11. Bhayani SB, Link RE, Varkarakis JM, Kavoussi LR. Complete davinci versus laparoscopic pyeloplasty: cost analysis. J Endourol 2005;19:327-32. 12. Da Vinci’s Hands. European Patent Office, 2008. 13. Rassweiler J, Hruza M, Teber D, Su LM. Laparoscopic and robotic assisted radical prostatectomy – critical analysis of the results. Eur Urol 2006;49:612-24. 14. Takenaka A, Leung RA, Fujisawa M, Tewari AK. Anatomy of autonomic nerve component in the male pelvis: the new concept from a perspective for robotic

595

596

38 15.

16.

17.

18.

19.

20.

21.

22. 23.

24.

25.

26.

27.

28. 29.


nerve sparing radical prostatectomy. World J Urol 2006;24:136-43. Walsh PC. A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. J Urol 2003;169:1588-9. Elliott DS, Chow GK, Gettman M. Current status of robotics in female urology and gynecology. World J Urol 2006;24:188-92. Di Marco DS, Chow GK, Gettman MT, Elliott DS. Robotic-assisted laparoscopic sacrocolpopexy for treatment of vaginal vault prolapse. Urology 2004;63:373-6. Heemskerk J, Van Gemert WG, Greve JW, Bouvy ND. Robot-assisted versus conventional laparoscopic Nissen fundoplication: a comparative retrospective study on costs and time consumption. Surg Laparosc Endosc Percutan Tech 2007;17:1-4. Sanchez BR, Mohr CJ, Morton JM, Safadi BY, Alami RS, Curet MJ. Comparison of totally robotic laparoscopic Roux-en-Y gastric bypass and traditional laparoscopic Roux-en-Y gastric bypass. Surg Obes Relat Dis 2005;1:549-54. Chitwood WR Jr, Nifong LW. Minimally invasive videoscopic mitral valve surgery: the current role of surgical robotics. J Card Surg 2000;15:61-75. Dogan S, Aybek T, Westphal K, Mierdl S, Moritz A, Wimmer-Greinecker G. Computer-enhanced totally endoscopic sequential arterial coronary artery bypass. Ann Thorac Surg 2001;72:610-1. Sugano N. Computer-assisted orthopedic surgery. J Orthop Sci 2003;8:442-8. Cain P, Kazanzides P, Zuhars J, Mittelstadt B, Paul H. Safety considerations in a surgical robot. Biomed Sci Instrum 1993;29:291-94. Leven J, Burschka D, Kumar R, Zhang G, Blumenkranz S, Dai XD, Awad M, Hager GD, Marohn M, Choti M, Hasser C, Taylor RH. DaVinci canvas: a telerobotic surgical system with integrated, robot-assisted, laparoscopic ultrasound capability. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8:811-8. Riordan T. Patents: implanting hair is tedious, exacting work – the perfect work for a robot. New York Times, C4. 9–15–2003. New York Times, NY, USA. Gildenberg PL. Method and apparatus for stereotactic robotic hair transplantation. USPTO Patent 6, 585,746, 1999. Louw DF, Fielding T, McBeth PB, Gregoris D, Newhook P, Sutherland GR. Surgical robotics: a review and neurosurgical prototype development. Neurosurgery 2004;54:525-36. McBeth PB, Louw DF, Rizun PR, Sutherland GR. Robotics in neurosurgery. Am J Surg 2004;188:68S-75S. Benabid AL, Cinquin P, Lavalle S, Le Bas JF, Demongeot J, De Rougemont J. Computer-driven robot for stereotactic surgery connected to CT scan and magnetic

30.

31.

32. 33.

34.

35.

36. 37.

38.

39. 40. 41.

42.

43.

44.

45.

resonance imaging. Technological design and preliminary results. Appl Neurophysiol 1987;50:153-4. Benabid AL, Lavallee S, Hoffmann D, Cinquin P, Demongeot J, Danel F. Potential use of robots in endoscopic neurosurgery. Acta Neurochir Suppl (Wien) 1992;54:93-7. Li QH, Zamorano L, Pandya A, Perez R, Gong J, Diaz F. The application accuracy of the Neuromate robot–a quantitative comparison with frameless and frame-based surgical localization systems. Comput Aided Surg 2002;7:90-8. Heilbrun MP, McDonald JD. The future of image guided surgery. Clin Neurosurg 2000;46:89-101. Varma TR, Eldridge P. Use of the Neuromate stereotactic robot in a frameless mode for functional neurosurgery. Int J Med Robot 2006;2:107-13. Varma TR, Eldridge PR, Forster A, Fox S, Fletcher N, Steiger M, Littlechild P, Byrne P, Sinnott A, Tyler K, Flintham S. Use of the Neuromate stereotactic robot in a frameless mode for movement disorder surgery. Stereotact Funct Neurosurg 2003;80:132-5. Willems PW, Noordmans HJ, Berkelbach van der Sprenkel JW, Viergever MA, Tulleken CA. An MKMmounted instrument holder for frameless pointstereotactic procedures: a phantom-based accuracy evaluation. J Neurosurg 2001;95:1067-74. Jordan A. Miracle microscope offers hope. Mich Health Hosp 1997;33:30. Kelly PJ, Alker GJ Jr. A method for stereotactic laser microsurgery in the treatment of deep seated CNS neoplasms. Appl Neurophysiol 1980;43:210-5. Kassell NF, Downs JH III, Graves BS. Telepresence in neurosurgery: the integrated remote neurosurgical system. Stud Health Technol Inform 1997;39:411-9. Young RF. Application of robotics to stereotactic neurosurgery. Neurol Res 1987;9:123-8. The story behind robotic-assisted endoscopic surgery. European Patent Office, 2008 www.epo.prg. Davies BL, Hibberd RD, Ng WS, Timoney AG, Wickham JE. The development of a surgeon robot for prostatectomies. Proc Inst Mech Eng 1991;205:35-8. Kwoh YS, Hou J, Jonckheere EA, Hayati S. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng 1988;35:153-60. Drake JM, Joy M, Goldenberg A, Kreindler D. Computer- and robot-assisted resection of thalamic astrocytomas in children. Neurosurgery 1991;29:27-33. Glauser D, Flury P, Durr P, Funakubo H, Burckhardt CW, Favre J, Schnyder P, Fankhauser H. Configuration of a robot dedicated to stereotactic surgery. Stereotact Funct Neurosurg 1990;54–55:468-70. Fankhauser H, Glauser D, Flury P, Piguet Y, Epitaux M, Favre J, Meuli RA. Robot for CT-guided stereotactic neurosurgery. Stereotact Funct Neurosurg 1994;63:93-8.


46. Glauser D, Fankhauser H, Epitaux M, Hefti JL, Jaccottet A. Neurosurgical robot Minerva: first results and current developments. J Image Guid Surg 1995;1:266-72. 47. Hefti JL, Epitaux M, Glauser D, Fankhauser H. Robotic three-dimensional positioning of a stimulation electrode in the brain. Comput Aided Surg 1998;3:1-10. 48. Heikkinen ER. Stereotactic neurosurgery: new aspects of an old method. Ann Clin Res 1986;18 Suppl 47:73-83. 49. Koivukangas J, Ylitalo J, Alasaarela E, Tauriainen A. Three-dimensional ultrasound imaging of brain for neurosurgery. Ann Clin Res 1986;18 Suppl 47:65-72. 50. Rizun P, Gunn D, Cox B, Sutherland G. Mechatronic design of haptic forceps for robotic surgery. Int J Med Robot 2006;2:341-9. 51. Rizun PR, McBeth PB, Louw DF, Sutherland GR. Robot-assisted neurosurgery. Semin Laparosc Surg 2004;11:99-106. 52. Hongo K, Kobayashi S, Kakizawa Y, Koyama J, Goto T, Okudera H, Kan K, Fujie MG, Iseki H, Takakura K. Neurobot: telecontrolled micromanipulator system for minimally invasive microneurosurgery-preliminary results. Neurosurgery 2002;51:985-8. 53. Goto T, Hongo K, Kakizawa Y, Muraoka H, Miyairi Y, Tanaka Y, Kobayashi S. Clinical application of robotic telemanipulation system in neurosurgery. Case report. J Neurosurg 2003;99:1082-4. 54. Hongo K, Goto T, Miyahara T, Kakizawa Y, Koyama J, Tanaka Y. Telecontrolled micromanipulator system (neurobot) for minimally invasive neurosurgery. Acta Neurochir Suppl 2006;98:63-6. 55. Rodrigues NN Jr, Mitre AI, Lima SV, Fugita OE, Lima ML, Stoianovici D, Patriciu A, Kavoussi LR. Telementoring between Brazil and the United States: initial experience. J Endourol 2003;17:217-20. 56. Tian Z, Lu W, Wang T, Ma B, Zhao Q, Zhang G. Application of a robotic telemanipulation system in stereotactic surgery. Stereotact Funct Neurosurg 2008;86:54-61. 57. Lum MJ, Rosen J, King H, Friedman DC, Donlin G, Sankaranarayanan G, Harnett B, Huffman L, Doarn C, Broderick T, Hannaford B. Telesurgery via Unmanned Aerial Vehicle (UAV) with a field deployable surgical robot. Stud Health Technol Inform 2007;125:313-5. 58. Rosen J, Lum M, Trimble D, Hannaford B, Sinanan M. Spherical mechanism analysis of a surgical robot for minimally invasive surgery–analytical and experimental approaches. Stud Health Technol Inform 2005;111:422-8. 59. Jacobs LK, Shayani V, Sackier JM. Determination of the learning curve of the AESOP robot. Surg Endosc 1997;11:54-5. 60. Mohr FW, Onnasch JF, Falk V, Walther T, Diegeler A, Krakor R, Schneider F, Autschbach R. The evolution of minimally invasive valve surgery–2 year experience. Eur J Cardiothorac Surg 1999;15:233-238. 61. Nathan CO, Chakradeo V, Malhotra K, D’Agostino H, Patwardhan R. The voice-controlled robotic assist scope

62.

63. 64. 65.

66.

67.

68. 69.

70.

71.

72.

73.

74.

75.

76.

38

holder AESOP for the endoscopic approach to the sella. Skull Base 2006;16:123-31. Ali SM, Reisner LA, King B, Cao A, Auner G, Klein M, Pandya AK. Eye gaze tracking for endoscopic camera positioning: an application of a hardware/software interface developed to automate Aesop. Stud Health Technol Inform 2008;132:4-7. Moran ME. Robotic surgery: urologic implications. J Endourol 2003;17:695-708. Hazey JW, Melvin WS. Robot-assisted general surgery. Semin Laparosc Surg 2004;11:107-12. Proske JM, Dagher I, Franco D. Comparative study of human and robotic camera control in laparoscopic biliary and colon surgery. J Laparoendosc Adv Surg Tech A 2004;14:345-8. Alessandrini M, De Padova A, Napolitano B, Camillo A, Bruno E. The AESOP robot system for video-assisted rigid endoscopic laryngosurgery. Eur Arch Otorhinolaryngol 2008;265(9):1121-3. Lunca S, Bouras G, Stanescu AC. Gastrointestinal robotassisted surgery. A current perspective. Rom J Gastroenterol 2005;14:385-91. Advincula AP, Falcone T. Laparoscopic robotic gynecologic surgery. Obstet Gynecol Clin North Am 2004;31:599. Nguan C, Kwan K, Al Omar M, Beasley KA, Luke PP. Robotic pyeloplasty: experience with three robotic platforms. Can J Urol 2007;14:3571-6. Giorgi C, Eisenberg H, Costi G, Gallo E, Garibotto G, Casolino DS. Robot-assisted microscope for neurosurgery. J Image Guid Surg 1995;1:158-63. Giorgi C, Sala R, Riva D, Cossu A, Eisenberg H. Robotics in child neurosurgery. Childs Nerv Syst 2000;16:832-4. Siemionow M, Ozer K, Siemionow W, Lister G. Robotic assistance in microsurgery. J Reconstr Microsurg 2000;16:643-9. Nimsky C, Rachinger J, Iro H, Fahlbusch R. Adaptation of a hexapod-based robotic system for extended endoscopeassisted transsphenoidal skull base surgery. Minim Invasive Neurosurg 2004;47:41-6. Rachinger J, Bumm K, Wurm J, Bohr C, Nissen U, Dannenmann T, Buchfelder M, Iro H, Nimsky C. A new mechatronic assistance system for the neurosurgical operating theatre: implementation, assessment of accuracy and application concepts. Stereotact Funct Neurosurg 2007;85:249-55. Togawa D, Kayanja MM, Reinhardt MK, Shoham M, Balter A, Friedlander A, Knoller N, Benzel EC, Lieberman IH. Bone-mounted miniature robotic guidance for pedicle screw and translaminar facet screw placement: part 2–Evaluluation of system accuracy. Neurosurgery 2007;60:129-39. Menciassi A, Quirini M, Dario P. Microrobotics for future gastrointestinal endoscopy. Minim Invasive Ther Allied Technol 2007;16:91-100.

597

44 Stereotactic and Image Guided Craniotomy E. C. Parker . P. J. Kelly

Introduction In 1906, Robert Henry Clarke and Victor Horsley described the stereotactic method, which consisted of systems for defining the brain in Cartesian coordinates and accessing subcortical structures by means of a three-dimensional positioning device based on that coordinate system [1–2]. Although Clarke had suggested to Horsley that stereotaxis could be used to treat brain tumors, Horsley could not see applications beyond laboratory investigations in animals. Subcortical stereotaxis was not employed in humans until Spiegel and Wycis [3] introduced it in 1947 for functional neurosurgery. Nonetheless, Spiegel and Wycis also predicted the application of stereotaxis for the management of human brain neoplasms. The general availability of L-dopa in the late 1960s brought about a precipitous decline in the number of functional stereotactic procedures performed worldwide until the advent of computed tomography (CT in the 1970s led to renewed interest in stereotactic surgery and prompted many neurosurgeons to rethink their approaches to common intracranial tumors [4–7]. In contrast to projection radiography and ventriculography, which was employed in functional stereotactic procedures, CT was a natural three dimensional data source for tumor stereotaxis that could easily be incorporated into a stereotactic coordinate system. In addition, with CT scanning surgeons could actually see the intracranial tumor target volume directly and

#


define its margins. Later, magnetic resonance imaging (MRI) was rapidly incorporated into stereotactic database acquisition techniques. CT-based, and later MRI-based, stereotactic biopsy procedures for the diagnosis of intracranial tumors became commonplace [4,8–10]. These imaging-based biopsy procedures usually target a point whose coordinates are derived from stereotaxic CT scanning or MRI. In addition to biopsy, point stereotasis can be used to center a craniotomy over a superficial lesion or to find a deep lesion. However, the use of point-in-space stereotactic techniques to identify tumor margins is cumbersome and inconvenient. Volumetric stereotactic methods were developed to facilitate the intra operative identification of CT-and MRI-defined tumor borders within a stereotactically defined surgical field and to maintain spatial orientation within irregularly shaped neoplasms [11–14]. In contrast to point-in-space stereotaxis, which requires mathematics no more complicated than addition, subtraction, multiplication, and simple trigonometry, volumetric stereotaxis is mathematically complex and requires interpolations, integrations, elaborate image processing, and reconstructions. Our first volumetric stereotactic procedures were performed utilizing inexpensive manual methods [7]. However, those methods proved to be cumbersome, timeconsuming, and impractical on a day-to-day basis. The incorporation of an operating room computer system rendered these procedures practical and time- and cost-efficient [11,13–15].

680

44

Stereotactic and image guided craniotomy

This chapter will describe the instrumentation and the current methodology for and results of computer-assisted stereotactic extirpation of intra- axial lesions.

Methods Volume in Space Patients undergo stereotactic imaging studies with their heads fixed in CT/MRI-compatible stereotactic head frames. The frames are applied under local anesthesia and are secured to the patient’s skull by means of four flanged carbon fiber pins that are inserted through drill holes made in the out table of the skull into the diploe. Detachable micrometers are used to measure the distance from the end of the carbon fiber pins to the vertical supports of the stereotactic head holder. This provides a fixed reference so that the head holder can be removed after data acquisition and replaced for surgery when convenient. Stereotactic CT scanning is performed as follows: the base ring of the stereotactic head holder fits into a CT table adaptation plate, and a localization system indexes into the base ring. The localization system consists of carbon fiber rods arranged in the shape of the letter N that are located on each side of the patient’s head and anteriorly. Each CT slice then exhibits nine reference marks from which stereotactic coordinates for any pixel on that slice may be calculated. Similarly, MRI examinations are performed utilizing a localizing system that in principle resembles the one developed for CT, except that the localizing system for MRI contains N-shaped localizing devices bilaterally, anteriorly, posteriorly, and superiorly to take advantage of the multiplanar imaging capability of MRI. Stereotactic angiography is performed using a biplane digital angiography (DA) table. Again, an adaptation plate is used for fixation of the stereotactic head holder. The DA localization

system consists of Lucite plates, each containing nine reference markers located bilaterally, anteriorly, and posteriorly, which create 18 reference marks on each anteroposterior and lateral DA image. By using the known location of these reference markers in space, one can determine magnification and stereotactic coordinates for any point in space and cross-correlate CT- and MRI-defined stereotactic target points and interpolated lesional volumes in the correct position on each DA image. In practice, orthogonal and 6-degree stereoscopic pairs are obtained for each stereotactic angiogram. Deep vascular segments identified on the stereotactic angiogram delineate the stereotactic position of all the major sulci and fissures of the cerebral hemispheres. The computer-generated CT, MRI, and DA data are transferred to the operating room computer system (COMPASS configured dual display Linux workstation, Compass International, Rochester, MN). By means of a menu-based program with an intuitive graphic interface (Admiral; Compass International, Rochester, MN) and a mouse subsystem, the surgeon simply traces around the contours of the tumor on serial CT slices and MRI images. The computer program then suspends those slices within a threedimensional image matrix that corresponds directly to the stereotactic coordinate system of the stereotactic frame. An interpolation program creates intermediate slices between the digitized slices and then fills them in with 1-mm cubic voxels, thus creating, for the CT- and MRI-defined limits of the lesion, separate volumes in abstract stereotactic space. This volume can be sliced perpendicular to the intended surgical viewline to present to the surgeon the appearance of the lesion as it will be encountered at surgery. Alternatively, it can be represented by a shaded graphics algorithm, rotating in space to provide the surgeon with a conceptualization of the lesion as a threedimensional volume. The lesional volume scaled to the proper image size also can be displayed in


the correct location in the stereoscopic angiogram or on a map of the brain sulci and fissures derived from that angiogram. Finally, all these data can be displayed in a shaded graphics rendition of the patient’s skull that has been extracted from the stereotactic CT scan. Such displays are used to plan the stereotactic trajectory to an intracranial lesion to approach and extract a tumor in the safest manner. The best and safest surgical approach to the lesion is one that: (1) preserves important vascular and neuroanatomic structures, (2) traverses nonessential brain parenchyma in a direction parallel to major white matter fiber tracts and, (3) approaches the tumor along its longest axis.

44

. Figure 44-1 Schematic of a simplified COMPASS stereotactic frame, which includes a circular stereotactic holder (a), a 160mm-radius arc quadrant (b), and a three-axes slide mechanism (c) that attaches to a semi-permanent floor stand or a standard operating table. The hand cranks shown in the drawing provide mechanical backups to three-axes computer-controlled stepper motors (not shown)

Surgical Procedures These procedures employ a COMPASS stereotactic frame (> Figure 44-1), occasionally a headsup video display terminal that is attached to the operating microscope (> Figure 44-2), a carbon dioxide laser system (for deep tumors only), and, for the resection of deep tumors, various custom extra-long microsurgical instruments that have been adapted for these procedures. The COMPASS stereotactic frame is basically a Cartesian robotics system in which the patient’s head, while fixed in the stereotactic head holder, is moved in X, Y, and Z space by a stepper motorcontrolled three- dimensional slide system to position the intracranial target volume in the isocenter of a fixed arc quadrant [16]. Surgical trajectories are expressed in terms of setting on the arc quadrant: the Collar (angle from the horizontal plane) and Arc (angle from the vertical plane) angles. Stereotactic coordinates automatically detected by linear optical encoders on each axis of the three-dimensional slide system are relayed to the host computer system in the operating room. Stereotactic coordinates calculated by the computer are executed by means of the stepper motors that are controlled from a

remote location. Electronic and manual backup systems are provided for each automated feature in this system. Computer-generated images of the imagingdefined tumor volume sliced perpendicular to the surgical approach trajectory are displayed on video monitors in the operating room. The slice images can also be projected into a heads-up display unit mounted on the operating microscope, scaled to the exact size of the surgical field viewed through the operating microscope as they are superimposed on the surgical field. Thus, during these procedures, the surgeon views not only the surgical field but also a computergenerated rendition of the CT and MRI defined

681

682

44


. Figure 44-2 Heads-up display device mounts on a standard operating microscope. This projects computergenerated slice images into one of the oculars of the microscope and allows translation and scaling for precise superimposition of CT/MRI-derived computergenerated tumor slice images onto the surgical field

tumor and its boundaries at any given level in the stereotactic surgical field. The carbon dioxide surgical laser is only useful in vaporizing tissue from deep-seated tumors that are approached and removed through a stereotactically directed 140-cm-long cylindrically shaped retractors 2 cm in diameter (> Figure 44-3) Computer-assisted stereotactic resections can be performed in superficial and deep-seated lesions. In superficial lesions, a circular trephine is turned on a stereotactically placed cranial pilot hole centered over the lesion. The boundaries of the trephine craniotomy with a known configuration and size serves as a reference structure for the indexing of the scaled image within the heads-up display of the operating microscope (> Figure 44-4) or displayed on video monitors in the operating room. The computer-generated

. Figure 44-3 Cylindrical 2-cm-diameter stereotactic retractor mounted on a stereotactic arc quadrant. This provides not only a means of maintaining surgical exposure to deep-seated lesions but also a fixed reference structure whose configuration, size, and location in the stereotactic surgical field are known and to which the computer-generated slice images can be related. Also shown is the dilator device that is inserted into the retractor and used to dilate a subcortical incision

image of the trephine with respect to the CT/ MRI-defined tumor volume is superimposed over the actual trephine in the surgical field. Thus, the computer-generated tumor slice images serve as a template that guides the dissection around subcortical lesions and facilitates identification of the plane between the lesion and the surrounding brain parenchyma. Deep-seated lesions are resected by means of a stereotactically directed cylindrical retractor that is inserted through a dilated cortical and subcortical white matter incision. The incision is made with the carbon dioxide laser and is dilated by means of the retractor-dilator system (> Figure 44-5). The configuration of the deep end of the cylindrical retractor is represented as a circle in the computer-generated slice images (> Figure 44-6) so that it can be superimposed over the actual surgical field by means of the heads-up display unit on the operating microscope. In practice, a plane is developed between the tumor and the surrounding brain tissue before any tumor is removed. Lesions that are


44

. Figure 44-4 Method employed for the stereotactic resection of a superficial tumor. A trephine opening of the skull is performed, centered on a pilot hole drilled by means of the stereotactic frame. The computer displays the position of the tumor slice in the proper position with respect to the location of the trephine at a specified distance along the viewline on the display monitor and into the heads-up display unit of the operating microscope (A). The image is scaled in the heads-up display, and the microscope is moved until the configuration of the trephine in the image display is exactly the same size as the actual trephine in the surgical field and aligns to the trephine. The surgeon then uses the tumor slice image as a template, to aid in the identification of the surgical plane between CT-and MRI-defined tumor and surrounding brain tissue. This facilitates isolation of the tumor from surrounding brain tissue. (From: Kelly, with permission [16].)

much larger than the retractor can be removed by multiple images translations on the display screen, which result in the calculation of new stereotactic coordinates. Once executed on the stereotactic slide system, these translations position a new part of the tumor under the stereotactic retractor (> Figure 44-7). In practice, a plane is established entirely around the lesion before removing it. Once this plane has been established, the tumor can be partially debulked to allow completion of the deepest aspect of the dissection on the far side of the tumor and removal of the remaining specimen through the retractor tube.

Avoidance of ‘‘Brain Shift’’ Much has been written on the problem of ‘‘brain shift’’ during stereotactically directed

craniotomies. Early on our group was also so concerned about this theoretical problem that we used to place a series stainless steel reference balls though a twist drill cranial opening and a stereotactically directed probe along the stereotactic viewline within a tumoral target volume. AP and lateral stereotactic teleradiographs document their position prior to the craniotomy [14]. With experience, we found this step unnecessary and discontined this practice; the reference balls didn’t shift in their position on serial radiographs in the vast majority of cases. We initially thought that this was because the position of deep tumors didn’t change because of the tethering effect of parenchymal blood vessels or the spatial stabilization resulting from the stereotactic cylindrical retractor. However, there was probably a much more valid reason that ‘‘brain shift’’ has not posed a significant problem in our experience. From our

683

684

44


. Figure 44-5 Method for extending and dilating a subcortical incision by using the stereotactic retractor and dilator. (a) A sulcus opened microsurgically. (b) An incision made in deep cortex with the retractor cylinder advanced. (c) A subcortical white matter incision deepened with a carbon dioxide laser. (d) An incision dilated with a dilator and a retractor advanced over the dilator. (e) A. retractor cylinder at depths of incision. (f) An incision deepened farther. (g) Retractor advanced, with the incision made to the superficial level of the tumor. (h) Superficial extent of tumor exposed. (i) Retractor advanced to superficial aspects of the tumor

first cases we have always employed the following methodological principles: 1.

The trephine craniotomy is in the least dependent position in the surgical field. The patient’s head (in the stereotactic frame) is rotated so that the vertical approach (arc) angle is close to zero. The

2.

patient is placed in reverse Trendelenberg position until the vertical approach angle is also zero. The dura is not opened until the intracranial pressure was controlled by hypocarbia, reverse Trendelenburg position, barbituates and only a small dose of hyperosmotic agents (rarely required in our experience)


44

. Figure 44-6 The stereotactic cylindrical retractor is employed during the resection of deep-seated lesions. The computer displays the configuration of a cross section of the retractor (circle) with respect to a selected slice through the tumor volume cut perpendicular to the surgical viewline. This information is displayed on a computer monitor in the operating room as well as in the heads-up display unit of the operating microscope (A). (From: Kelly, with permission [18].)

. Figure 44-7 Method of accessing the extent of a tumor by stereotactic translation. At any set of stereotactic coordinates, the position of the tumor edge with respect to the end of the stereotactic retractor is shown in the computer-generated image. (a) Accessing the posterior margin of the tumor. Slice distance = 10, x = 14, y = 45, z = 20. (b) Accessing the lateral aspect of the tumor. In the COMPASS system, the patient’s head is moved to place a different part of the tumor under the aperture of the stereotactic retractor, which is always directed at the center of the stereotactic arc quadrant. Slice distance = 10, x = 4, y = 35, z = 20

685

686

44 3.

4.


The ventricular system and tumor cysts are not entered until after the peritumoral plane had been defined surgically. In cases having intraventricular tumors, the cylindrical stereotactic retractor that is used to resect these lesions is, after incising the ependyma, rapidly advanced to the superficial aspect of the tumor in order to minimize the loss of ventricular fluid. Once placed, the retractor prevents the brain from ‘‘sagging’’ or shifting. Tumors are not ‘‘debulked’’ until a plane is developed entirely around the lesion. Here the stereotactic slice images are used to identify that plane within the surgical field. Developing the plane before debulking the tumor mass may, in fact, be contrary to the method that most tumor neurosurgeons use; traditionally surgeons tended to work from the center of the lesion to the periphery. However, if this is done, the tumor will collapse upon itself and ‘‘brain shift’’ WILL occur. Our principle is to first separate the tumor volume from surrounding brain, then remove it. This process may require a few steps in especially large tumors. Here one works in 10–15 mm ‘‘layers’’: first developing the plane down to a specific depth, then removing the tumor tissue in that layer down to that specific depth before developing the plane between tumor and surrounding brain in the next layer. The surgeon thus removes the tumor in defined layers progressing from the most superficial to the deepest.

After removing the tumor the brain does, indeed shift and the cortical surface may, indeed, have dropped a centimeter or so below the inner surface of the skull. But after the tumor is out what difference does this make? In addition, if the craniotomy is in the least dependent position (i.e., the top) of the surgical field, the ‘‘brain shift’’ will usually be along the stereotactically

defined viewline and tumor depth can be monitored by a measured (and from the stereotactic viewline calculations, known) distance from the cortical surface to the superficial slice (aspect) of the tumor.

Patient Selection Selective and accurate resection of any CTor MRI-defined intracranial volume can be performed by employing imaging-based computer-assisted volumetric stereotactic methods. Although the target volume can be an intracranial lesion, volumetric resection techniques were most frequently applied to the most common intraaxial lesions: glial neoplasms in eloquent brain regions. Stereotactic serial biopsy studies have shown that glial neoplasms frequently have two elements: tumor tissue and isolated tumor cells that infiltrate brain parenchyma. The tumor tissue component of high-grade gliomas is most accurately defined by the volume that exhibits contrast enhancement. However, tumor tissue is low-grade (non-pilocytic) gliomas usually is indistinguishable from infiltrated parenchyma on CT and MRI; both are hypodense on CT and do not usually exhibit contrast enhancement. Stereotactic serial biopsy is the only reliable method by which tumor tissue that is hypodense on CT or has prolonged T2 signal characteristics on MRI can be differentiated from infiltrated parenchyma in low-grade (nonpilocytic) astrocytomas, oligodendrogliomas, and mixed gliomas. Stereotactic volumetric resection of infiltrated parenchyma defined by CT/MRI is advisable only in nonessential brain regions. In eloquent brain areas, stereotactic resection is appropriate for the glial tumor tissue component of high-grade glial neoplasms, pilocytic astrocytomas, and low-grade CT-hypodense gliomas in which a stereotactic serial biopsy procedure has confirmed tumor tissue only.


Specific Lesion Types On the basis of this overall experience, we have been able to draw certain conclusions about surgical approaches for specific anatomic lesions and the selection of appropriate patients for computer-assisted volumetric stereotactic resection. We have developed certain technical maneuvers that are useful in the stereotactic removal of various lesions, depending on their histology and anatomic location.

High-Grade Glial Tumors Computer-assisted stereotactic resection can be used to remove all CT or MRI defined contrastenhancing portions of high-grade glial neoplasms from neurologically important subcortical areas with acceptable levels of mortality and morbidity [12,14,17,18]. Postoperative studies usually demonstrate an absence of contrast enhancement around the surgical defect (> Figure 44-8). Nevertheless, mean postoperative survival of our patients harboring grade 4 astrocytomas treated with postoperative external-beam radiation therapy (50–65 Gy) was 50.6 weeks. This compares favorably to a consecutive series of patients with grade 4 gliomas who underwent radiation therapy after biopsy alone (mean survival, 33 weeks). However, after resection, new areas of contrast enhancement on CT scanning developed in lowdensity areas surrounding the surgical defect within 6–9 months of the procedure. Death in the majority of these cases was therefore due to tumor recurrence and progression. However, quality of survival is better than resection of the lesion and radiation therapy than it is after biopsy alone and radiation. The mean survival time of 50.6 weeks in our patients with grade 4 astrocytomas in central and deep-seated locations (historically associated with poor survival and high surgical morbidity and mortality) is slightly better than the survival

44

times of 37 weeks quoted in other series of the literature [19–22]. Nonetheless, there are two major reasons why our results cannot be compared to historical controls. First, a high percentage of the patients in the other series had lesions in the frontal and temporal lobes, which are more amenable to radical surgical resection by lobectomy [17,22]. By contrast, many of the lesions we treated were centrally located and deep-seated. Second, the survival statistics quoted above after stereotactic resection involve grade 4 astrocytomas only. Readers should exercise caution in comparing these survival statistics to those reported in other series for ‘‘glioblastomas,’’ which in some series include a significant number of grade 3 (Kernohan) neoplasms with a better prognosis. In most historical grading schemes for astrocytomas, the presence or absence of necrosis separates glioblastomas from anaplastic astrocytomas, respectively, but in the Kernohan classification scheme; necrosis can be found in grade 3 and grade 4 astrocytomas. Endothelial proliferation and frequency of mitoses are used to differentiate between grade 3 and grade 4 tumors. Therefore, in most histological classification schemes, glioblastomas can contain some Kernohan grade 3 tumors mixed in with grade 4 neoplasms. Therefore, mean survival in a series of ‘‘glioblastomas’’ (which include grades 3 and 4 astrocytomas) should be longer than mean survival in a series of pure grade 4 astrocytomas. Computer-assisted volumetric stereotactic resection allows safe and complete resection of the contrast-enhancing mass lesion in high-grade gliomas. However, preoperative stereotactic MRI (especially the T2-weighted image) in high-grade glial tumors always demonstrates much larger areas of abnormality than are indicated by contrast enhancement on CT scanning [9,10]. Examination of stereotactic serial biopsy specimens obtained in patients with high-grade gliomas from these MRI-defined abnormalities outside the contrast-enhancing tumor mass reveals a

687

688

44


. Figure 44-8 Preoperative (a and b) and postoperative (c and d) contrast-enhanced MRI in a patient with deep parietal GBM. Patient was neurologically stable postoperatively

larger area of intact edematous brain parenchyma infiltrated by aggressive isolated tumor cells [8–10,22–24]. In fact, this parenchyma usually extends as far as, and in some cases beyond, the area of signal prolongation abnormality on T2-weighted MRI [9,10]. It would be technically possible, using volumetric stereotaxis, to resect the volume defined by the MRI abnormality, and this in theory would substantially prolong postoperative survival [25,26]. However,

unacceptable neurological deficits would result from removal of the intact, albeit infiltrated, parenchyma. Therefore, in grade 4 astrocytomas, resection of the volume of tissue defined by contrast enhancement permits the most aggressive reduction of tumor burden that allows preservation of neurological function. Radiotherapy and chemotherapy can be used to their best advantage in a patient with the least possible tumor burden.


These modalities are given to kill the isolated tumor cells that reside in the intact parenchyma that is left behind. However, until we have a means for selectively and safely targeting these isolated tumor cells, we can expect few significant changes in the survival of these patients. A similar problem exists for patients with grade 3 astrocytomas, mixed gliomas, and oligodendrogliomas. Although the cellular elements in these tumors are not as mitotically active as are those in grade 4 tumors, isolated tumor cells also infiltrate intact and surrounding edematous parenchyma [9,10] and defy surgical attempts to cure them. In addition, grade 3 gliomas, particularly astrocytomas, tend to have larger infiltrative components with respect to tumor tissue components than do grade 4 lesions. Therefore, the benefit of resecting a relatively small tumor tissue mass in the face of a large volume of infiltrated parenchyma is questionable; thus, for the most grade 3 lesions, surgeons tend to recommend biopsy over stereotactic resection. In stereotactically resected grade 3 gliomas, tumor recurrences are seen later in the patient’s postoperative course than in the case with grade 4 astrocytomas but recur in the same spatial pattern as that described for grade 4 tumors. Nevertheless, as with grade 4 astrocytomas, the procedure can remove all the solid tumor tissue in selected neoplasms with significant contrastenhancing volume on stereotactic CT scanning.

Low-Grade Astrocytomas The resectability of low-grade astrocytomas depends on the degree of histological circumstances. In adults, these tumors usually are manifest by an area of low density on CT and prolongation of signal on T2 weighted MRI [9]. Stereotactic serial biopsy studies of these socalled fibrillary astrocytomas reveal that the tumor is composed almost entirely of infiltrated intact parenchyma with little tumor tissue

44

proper [9,10]. Therefore, resection of the tumor by stereotactic craniotomy involves resection of intact but infiltrated parenchyma defined by the low-density areas on CT and T2 weighted hyperintensity on MRI. However, in important brain areas, this results in a postoperative neurological deficit, and stereotactic resection is therefore rejected as an option [14,18]. In some cases, the lesion is confined to expendable brain tissue, such as the posterior portion of the superior frontal convolution (> Figure 44-9). These lesions can be resected in their entirety. Pontine astrocytomas are usually fibrillary and are not well circumscribed. They are best biopsied using stereotactic techniques and then treated with radiation therapy. However, radiation therapy can demarcate some of these lesions from pontine parenchyma, creating a zone of neovascularity between necrotic tumor centrally and edematous (usually infiltrated) parenchyma peripherally. The area of central necrosis can be resected stereotactically without inherent neurological risk if the zone of contrast enhancement extends to the floor of the fourth ventricle or far laterally into the middle cerebellar peduncle, facilitating the approach.

Pilocytic Astrocytomas Pilocytic astrocytomas, which tend to occur in children and young adults, are histologically circumscribed. Despite the fact that many are located in the thalamus and other important subcortical locations, they can be resected completely by computer-assisted stereotactic technique with excellent postoperative results [14,18,27,33] (> Figure 44-10). These lesions exhibit prominent enhancement on CT or MR imaging with gadolinium, and the histological borders are defined accurately by the contrast enhancement. As with higher grade gliomas that demonstrate contrast enhancement, this enhancing region contains no functional tissue

689

690

44


. Figure 44-9 Preoperative (top) and postoperative (bottom) gadolinium-enhanced MRI studies in a patient with a grade 2 mixed glioma involving the posterior portion of the left superior frontal convolution. The patient had no neurological deficit preoperatively or postoperatively. Most of the tumor consisted of intact edematous parenchyma infiltrated by isolated tumor cells

and can be removed without neurological deficit aside from that associated with damage to any normal brain that must be traversed to reach the tumor.

Metastatic Tumors It is generally thought that metastatic tumors are histologically circumscribed and that there is usually no problem identifying the plane between tumor and edematous brain. In some cases, however, this interface may not be clear, especially when bleeding is encountered. In fact, reported surgical series of metastatic tumors removed at nonstereotactic craniotomy report a certain percentage of patients with incomplete resections [28–31].

Many surgeons have had the unsettling experience of trying unsuccessfully to locate deep subcortical metastatic lesions during a conventional craniotomy. Metastatic tumors usually are located at the gray-white junction subcortically. They can be located superficially near the crown of a gyrus. They also can be located at the gray-white junction in the depths of a deep sulcus and can be difficult to find with conventional cramotomy. Other tumors may be deep to the insular cortex, deep to the mesial occipital cortex, or under the cortex of the interhemispheric fissure. Stereotactic techniques can be advantageous in the resection of superficial metastases as well as deeply situated lesions [32]. First, stereotactic point localization helps center small cranial trephines directly over superficial lesions


44

. Figure 44-10 Thalamic pilocytic astrocytomas in two patients. Anterior thalamic tumors are approached through the anterior limb of the internal capsule (a and b). Dorsal posterior tumors may be reached via the superior parietal lobule (c and d)

(the trephine needs be no larger than the crosssectional area of the neoplasm). The approach is therefore selective and direct, and no more brain than absolutely necessary need be exposed. With volumetric stereotaxis and intraoperative image

displays, identification of the plane between tumor and brain is straightforward and simple. These lesions are readily resectable by either frameless or frame based stereotactic techniques (> Figure 44-11).

691

692

44


. Figure 44-11 Preoperative (a) and postoperative (b) MRI in a 62-year old man with a metastatic tumor (non-small cell lung cancer) located deep to the motor strip. In this case the tumor was approached by splitting the posterior aspect of the superior frontal sulcus. For metastatic tumors, frameless stereotaxis provides efficient intraoperative navigation, aiding in placement of scalp and bone openings as well as optimizing the trajectory of approach

Our postoperative morbidity for stereotactic resection of centrally located and deep-seated metastatic tumors (mortality, 0; morbidity, 4.3%) compares favorably with that association with conventional craniotomy for these lesions in the past (mortality, 11%) [27,28,30–32]. External-beam radiation therapy has followed surgery in most of our patients to treat possible microscopic metastatic lesions that are not visible on CT. In a 5.5-year experience with computer-assisted stereotactic resection of intracranial metastases at the Mayo Clinic, we have had no known local recurrences of the tumor as determined on serial postoperative CT scans.

Meningioma Meningiomas generally have a very obvious and discrete tumor/brain interface and are often superficially located. Stereotactic techniques can still be quite useful in their removal, however. The general principles of optimizing bony and dural openings

to minimize exposure of normal brain are often especially important when operating on extraaxial masses. Accurate intraoperative navigation often allows opening the scalp with a small linear incision and placing the borders of a craniotomy flap immediately outside the superficial limits of the tumor. When the brain is ‘‘tight’’ because of edema or simply from tumor related mass effect, any exposed brain may quickly herniate out of the dural opening and be strangulated. A properly placed dural incision will be directly over this tumor/brain interface. This serves the purpose of protecting the surrounding brain tissue as well as aiding in the resection. Intracranial pressure will often help force the tumor out through a properly sized dural opening, allowing more rapid dissection of the tumor from the surrounding arachnoid. Although similar spatial information can be obtained by using ultrasound to guide the dural incision, stereotactic techniques provide more precise and usable feedback. The usefulness of stereotaxis to meningiomas and other tumors located at the skull base is of less


clear benefit. Anatomical landmarks remain the primary cues in locating such tumors as well as important nearby neural and vascular structures. Frameless stereotaxis is often employed for these cases as it is occasionally helpful to retain orientation in the face of highly distorted anatomy.

44

. Figure 44-12 Preoperative (left) and postoperative (right) CT scans in a 29-year old woman with a mesencephalic arteriovenous malformation (AVM) with a hemorrhage. The hematoma and AVM were resected with the 2-cmdiameter stereotactic retractor. The patient was lethargic and had a syndrome of Weber preoperatively. She made an excellent neurological recovery after evacuation of the hematoma and resection of the AVM

Vascular Malformation Closed stereotactic needle biopsy of superficial or deep-seated circumscribed lesions that demonstrate intense contrast enhancement on CT and no perilesional ‘‘edema’’ can be dangerous. These lesions could represent an occult vascular malformation even though arteriography may fail to demonstrate vascularity consistent with an AVM. A postoperative hemorrhage can occur after biopsy of a cryptic AVM or cavernous hemangioma. Computer-assisted stereotactic microsurgical resection provides an alternative to closed stereotactic biopsy and observation. Cryptic AVMs and cavernous hemangiomas are well-circumscribed lesions that can be completely removed stereotactically with relatively low risk. A by-product of establishing the histology is that a cessation or significant reduction of seizures, when present, usually results. Small deep-seated active AVMs also may be resected with similar techniques (> Figure 44-12). The position of the feeding vessels is established in the three-dimensional surgical-planning matrix and is approached and clipped or coagulated before the remainder of the lesion is dissected away from the surrounding parenchyma.

Intraventricular Lesions Intraventricular landmarks can be used to maintain surgical orientation to locate lesions during conventional craniotomies for intraventricular tumors. There usually are few problems with

spatial orientation in patients with large lateral ventricles. However, more difficulty can be encountered staying oriented in small- or normalsized ventricles or, in some case, even finding the ventricle. A more limited but direct approach to intraventricular lesions can be made stereotactically. Brain and ventricular incisions need to be large enough only to remove the lesion. Thus, intraventricular lesions are removed through a 1.5-in trephine and a 2-cm cylindrical retractor (> Figure 44-13 and Video). Large third ventricular lesions usually are approached through the right lateral ventricle. One fornix can be incised to extend the stereotactic retractor into the third ventricular lesion, where an internal decompression of the lesion is performed with a carbon dioxide laser until only a thin rim of the capsule remains. The computer display of the cross sections of the digitized tumor volume are extremely useful in this step, as a surgeon, knowing where the tumor stops and the third ventricular wall begins, can be aggressive within the tumor with no risk of extending through the capsule and damaging the walls of the third ventricle. After this internal

693

694

44


. Figure 44-13 Preoperative (a and b) and postoperative (c and d) MRI in a woman with a solid intraventricular tumor. The lesion was totally resected utilizing a 2-cm stereotactic retractor and multiple translations. A gross total excision of the lesion was accomplished. The patient did not require a shunt. A video of this surgery is provided on the accompanying DVD

decompression, the retractor is withdrawn to the level of the roof of the third ventricle and the capsule is carefully dissected from the walls of the third ventricle. The tumor capsule can be contracted by using the defocused laser, which facilitates the dissection of the capsule from the wall of the third ventricle.

Colloid cysts are approached through the lateral ventricle and the foramen of Monro in which the cyst shows the greatest extension. The approach features an anterior trephine craniotomy (about at the frontal hairline), splitting of the superior frontal sulcus, and exposure of the cyst in the foramen of Monro by


means of the 2-cm diameter stereotactic retractor (> Figure 44-14). This approach does not violate eloquent brain tissue and provides an anterior vantage point for dealing with the attachment of the cyst. We have resected 28 colloid cysts in the manner without permanent complication.

44

the representation of a tumor volume in stereotactic space but also demonstrate the relationships of normal vascular and neuroanatomic structures to that volume. In general, the technique allows more thorough removal of intracranial lesions and has less morbidity for patients with centrally located and deep-seated lesions. Without volumetric stereotaxis three things are possible:

Discussion 1. Intracranial mass lesions (tumors) are volumes in space that are readily apparent on review of contiguous CT and MR slice images. However, translation of this three-dimensional information from the imaging studies (CT and MRI) to the three-dimensional surgical operating space in the patient’s head during conventional surgical procedures is difficult and imprecise. In addition, computer reconstructions of stereotactically acquired CT and MRI data not only allow

2.

A surgeon can get lost attempting to find the tumor, and brain tissue is damaged unnecessarily. This can result in a neurological deficit and prolonged and expensive rehabilitation efforts. A surgeon cannot tell where tumor ends and normal brain tissue begins. Thus, there is some risk that the surgeon will resect normal brain tissue along with the tumor. In important brain areas, this will result in a neurological deficit.

. Figure 44-14 A giant colloid cyst in a 56-year old man presenting with obstructive hydrocephalus memory disturbance. Preoperative (a) and postoperative (b). The lesion had a solid interior and was partially calcified. Marked improvement in his recent memory over the preoperative level was noted at the 3-month postoperative examination

695

696

44 3.


A surgeon performs a subtotal removal of the lesion. Much tumor remains behind and will recur sooner and require another operation later on.

1.

Volumetric stereotaxis provides the following specific advantages: 1. 2. 3.

4.

It allows the surgeon to find the lesion. It imparts a concept of the three-dimensional shape of the lesion that is to be removed. It allows preoperative surgical simulation and surgical approach or trajectory planning with respect to the configuration of the lesion and normal brain and vascular anatomy that must be preserved. Thus, the safest and most effective surgical approach may be selected. It indicates by means of a scaled real-time display, interactive software, and stereotactic instrument where tumor ends and normal brain begins.

2.

3.

Volumetric stereotaxis has major advantages for the patient: 1.

2.

3.

The smallest possible skin incision, craniotomy, and brain incision can be made. This minimizes injury to normal brain tissue. Since the surgeon knows exactly where tumor ends and normal brain begins, a more complete tumor removal can be accomplished with much less risk to surrounding brain tissue. The postoperative neurological results are better than those associated with conventional (nonstereotactic, nonvolumetric) surgical techniques.

Finally, volumetric stereotaxis costs less than standard neurosurgical procedures, since patients get out of the hospital faster, do better neurologically, and return to work earlier. In a practical sense, volumetric stereotaxis saves third-party payers money because:

4.

Less money is spent on intensive care unit charges and postoperative hospital days. Total hospital charges, including surgical fees for patients with astrocytic brain tumors undergoing computer-assisted stereotactic volumetric resection procedures, accounted for approximately 67% of the total hospital charges for conventional surgical procedures in similar patients (Kelly, unpublished data). Volumetric stereotactic procedures require less time in the operating room (2–3 h less in most cases) than do conventional neurosurgical procedures for brain tumors, because the procedures are simulated on a computer system beforehand and can proceed efficiently as planned. This saves money on operating room charges. ‘‘Inoperable’’ tumors (by conventional surgical techniques) can be resected with volumetric stereotactic resection procedures. Frequently, these are deep-seated relatively benign tumors in children and young adults. Many of these tumors can be cured with volumetric stereotaxis, which saves money wasted on radiotherapy and chemotherapy that is not effective for these lesions and on rehabilitation and terminal care as the tumor progressively disables and kills the patient. Neurological results are better, fewer patients require rehabilitation programs, and patients return to work sooner.

Future Directions for Computer-Assisted and Volumetric Stereotactic Surgery In our experience in the development of volumetric and computer-assisted stereotactic surgery, we have seen a tremendous potential for computer assistance of surgery in general. First, it is clear that computers can be used to monitor


and display to a surgeon the position of surgical instruments in a stereotactically defined work envelope. This will allow more minimally invasive and endoscopic surgery not only in subarachnoid and intraventricular spaces but also for intraaxial target volumes. However, many other techniques will affect the future of neurosurgery in general and surgery for brain tumors and other lesions in particular. These technologies may include electronics, robotics, lasers, and other technologies adapted from industry and the military. They will have the greatest impact in stereotactic neurosurgery, a three-dimensional and mathematically precise surgical discipline that can best exploit techniques (such as endoscopy) borrowed from other fields. Most important, surgical computer systems coupled to surgical instrumentation will make significant advances possible. Since this chapter was originally published, frameless stereotactic equipment and procedures have evolved and matured, and computational power available in workstation and laptop computers has increased dramatically. Most stereotactic brain tumor operations performed today utilize some form of frameless stereotaxy instead of a frame-based system. Modern frameless systems clearly possess sufficient accuracy to allow the safe and effective resection of both intra- and extra-axial tumors. The volumetric techniques described here have been adapted to an electromagnetic tracking based frameless system, the Cygnus PFS (Compass International, Rochester, MN). As with the COMPASS, the target lesion can be digitized by outlining the contours on individual imaging slices. Cross-sectional lesion contours can then be viewed along the axis of the stereotactic probe. For most types of cases, the benefits of volumetric stereotaxis should be realized whether the system utilizes a headframe or some frameless technology. In practice, however, most tumor resections done with the assistance of a frameless system are probably performed using only point in space stereotaxis. In truth, this is easily sufficient for

44

planning the scalp, bone, and dural opening. Once a tumor with an obvious border, such as a meningioma, metastasis, or even a high grade glial tumor, is found, it is not usually difficult to follow the tumor plane and resect the tumor. The stereotactic system can then be used to confirm orientation or proximity to some anatomic structure of interest during tumor resection. For primary, low grade tumors, however, the authors continue to feel that frame based volumetric techniques provide the most efficient and reliable means of surgical navigation. The chance of a poor registration is virtually eliminated, and the visual representation of the tumor contours superimposed on the craniotomy outline provides an intuitive reference for locating and following the tumor border. Additionally, the ability to define the target and surgical trajectory and therefore the location of the craniotomy prior to surgery makes the actual procedure very efficient. It eliminates the need to frequently check the position of a probe or some other registered instrument and compare this with the desired location, a process that can become tedious. Furthermore, deep-seated lesions such as thalamic tumors and intraventricular lesions are most amenable to resection with a cylindrical retractor system like the one described here. After more than 3,000 computer-assisted stereotactic procedures, the methods for computerassisted stereotactic and volumetric surgery described in this chapter are well-developed, and the instrumentation and software are quite mature. Technical support personnel are not required for the day-to-day performance of these procedures. However, continued development of computer-assisted surgery can best be accomplished with full-time technical people. Computer programming is a tedious, painstaking endeavor. Few neurosurgeons will ever have the patience, much less the ability, to write and debug a complex computer program (even though various object-oriented development programs may allow us to manipulate existing programs and customize them for our own purposes).

697

698

44


Dedicated engineers and computer scientists should become part of academic neurosurgical departments. They will provide the tools for surgery in the future. Surgeons will provide the needed guidance by telling them what we need and what is useful.

References 1. Clark RH, Horsley V. One a method of investigating the deep ganglia and tracts of the central nervous system (cerebellum). Br Med J. 1906;2:1799-1800. 2. Horsley V, Clark RH. The structure and function of the cerebellum examined by a new method. Brain 1908;31:45-124. 3. Spiegel EA, Wycis HT, Marks M, et al. Stereotaxic apparatus for operation on the human brain. Science 1947;106:349-50. 4. Apuzzo MLJ, Savshin JK. Computed tomographic guidance stereotaxis in the management of intracranial mass lesions. Neurosurgery 1983;12:277-85. 5. Brown RA. A computerized tomography-computer graphics approach to stereotaxic localization. Neurosurgery 1979;50:715-20. 6. Goerss S, Kelly PJ, Kall B, et al. A computer tomographic stereotactic adaptation system. Neurosurgery 1982;10: 375-9. 7. Kelly PJ, Alker GJ Jr. A stereotactic approach to deep seated CNS neoplasms using the carbon dioxide laser. Surg Neurol. 1981;15:331-4. 8. Daumas-Duport C, Monsaigngeon V, Szenthe L, et al. Serial stereotactic biopsies: a double histological code of gliomas according to malignancy and 3-D configuration, as an aid to therapeutic decision and assessment of results. Appl Neurophysicol. 1982;45:431-7. 9. Kelly PJ, Daumas-Duport C, Kispert DB, et al. Imagingbased stereotactic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 1987;66:865-74. 10. Kelly PJ, Daumas-Duport C, Scheithauer BW, et al. Stereotactic histologic correlations of computed tomography and magnetic resonance imaging defined abnormalities in patients with glial neoplasms. Mayo Clin Proc. 1987;62:450-9. 11. Kelly PJ, Alker GJ Jr, Georgss S. Computer assisted stereotactic laser microsurgery for the treatment of intracranial neoplasms. Neurosurgery 1982;10:324-31. 12. Kelly PJ, Alker GJ Jr, Kall B, et al. Precision resection of intra-axial CNS lesions by CT-based stereotactic craniotomy and computer monitored CO2 laser. Acta Neurochir (Wien) 1983;68:1-9. 13. Kelly PJ, Kall BA, Goerss SJ. Transposition of volumetric information derived from computed tomography scanning into stereotactic space. Surg Neurol. 1984;21:465-71.

14. Kelly PJ, Kall B, Goerss S, et al. Computer-assisted stereotaxic resection of intra-axial brain neoplasms. J Neurosurg 1986;64:427-39. 15. Kelly PJ, Kall BA, Goerss SJ. The results of CT based computer assisted stereotactic resection of metastatic intracanial tumors. Neurosurgery 1988;22:7-17. 16. Kelly PJ, Georss SJ, Kall BA. Evolution of contemporary instrumentation for computer-assisted stereotactic surgery. Surg Neurol. 1988;30:204-15. 17. Devaux BC, O’Fallon JR, Kelly PJ. Resection, biopsy and survival in malignant glial neoplasms: a retrospective study of clinical parameters, therapy, and outcome. J Neurosurg. 1993;78:767-75. 18. Kelly P. Volumetric stereotactic surgical resection of intra-axial brain mass lesions. Mayo Clin Proc. 1988; 63:1186-98. 19. Frankel SA, German WJ. Glioblastoma multiforme: Review of 219 cases with regard to natural history, diagnostic methods, and treatment. Neurosurgery 1958;15:489-503. 20. Gehan EA, Walker MD. Prognostic factors for patients with brain tumors. NCR Monogr. 1977;46:189-95. 21. Hitchcock E, Sato F. Treatment of malignant gliomata. J Neurosurg. 1964;21:497-505. 22. Jelsma R, Bucy PC. The treatment of glioblastoma multiforme of the brain. Neurosurgery 1967;27:38-400. 23. Burger, PC, Dubois PJ, Schold SC Jr, et al. Computerized tomographic and pathologic studies of the untreated, quiescent, and recurrent glioblastoma multiforme. Neurosurgery 1983;59:159-68. 24. Daumas-Duport C, Scheithauer BW, Kelly PJ. A histologic and cytologic method for the spatial definition of gliomas. Mayo Clin Proc. 1987;62:435-49. 25. Hoshino T. A commentary on the biology and growth kinetics of low grade and high grade gliomas. J Neurosurg. 1984;27:388-400. 26. Hushino T, Barker M, Wilson CB, et al. Cell kinetics of human gliomas. Neurosurgery 1972;37:15-26. 27. McGirr SJ, Kelly PJ, Scheithauer BW. Stereotactic resection of juvenile pilocytic astrocytomas of the thalamus and basal ganglia. Neurosurgery 1987;20:447-52. 28. Haar F, Patterson RH Jr. Surgery for metastatic intracranial neoplasms. Cancer 1972;30:1241-5. 29. MacGee EE. Surgical treatment of cerebral metastases from lung cancer. The effect on quality and duration of survival. J Neurosurg. 1971;35:416-20. 30. Raskind R, Weiss SE, Manning JJ, et al. Survival after surgical excision of single metastatic brain tumors. AJR 1971;111:323-8. 31. Yardeni D, Reichenthal E, Zucker G, et al. Neurosurgical management of single brain metastasis. Surg Neurol. 1984;21:377-84. 32. Van Eck JHM, Go KG, Ebels EJ. Metastatic tumors of the brain. Psychiatr Neurol Neurochir. 1965;68:443-62. 33. Moshel Y, Link M, Kelly P. Stereotactic volumetric resection of thalamic pilocytic astrocytomas. Neurosurgery 2007;61:66-75.

42 Stereotactic and Image-Guided Biopsy J. B. Elder . A. P. Amar . M. L. J. Apuzzo

Introduction and History Image-guided stereotactic biopsy for histopathologic diagnosis of cranial lesions has become a standard component of the neurosurgical armamentarium [1–3]. The word ‘‘stereotactic’’ derives from the Greek word ‘‘stereos’’ for ‘‘three dimensions,’’ and the Latin word ‘‘tactus’’ for ‘‘to touch’’ [4,5]. In 1908, Horsley and Clarke reported the first stereotactic device in the English literature, which was used to access the dentate nucleus in the cerebellum of monkeys [6]. Nearly 40 years later, in 1947, stereotactic techniques were introduced in humans by Spiegel and Wycis, who used their system for ablative neurosurgical procedures [7,8]. Nearly simultaneously, Leksell developed a separate stereotactic system in 1949 based on the concept of the arcquadrant. By combining the use of stereotactic equipment, a quantitative anatomy atlas and radiographic imaging, these innovators established the concepts on which current techniques and methods are based. Prior to stereotactic techniques, biopsies were conducted during open craniotomies or via free-handed needle aspiration using guidance from indirect radiographic images such as ventriculography and angiography. Over the last 50 years, achievements in mathematics, physics and computing technology have rendered phrenology virtually obsolete, and allow today’s neurosurgeon to visualize and potentially sample the brain and its pathologies with sub-millimeter resolution [9]. With continued advances, stereotactic image-guided biopsies have progressed from subjective techniques whose success #


depended largely on the skill and experience of the neurosurgeon, towards an increasingly precise scientific exercise more reliant on measurable factors such as patient- and lesion-specific variables. Early human stereotactic techniques were primarily used for functional neurosurgical procedures such as the treatment of Parkinsonian tremor. Although mortality was greatly reduced compared to open procedures (16% vs. 2%), subsequent advances in medical treatments largely replaced most surgical interventions [9,10]. Reports regarding stereotactic methods for tissue sampling became more common in the 1960s [11,12]. However, stereotactic techniques struggled to gain broad acceptance during this time due to difficulties with image-localization provided by angiography and ventriculography. The advent of the computed tomography (CT) era in the 1970s heralded the renaissance of stereotactic neurosurgery. This technology provided improved visualization of intracranial lesions and allowed the use of patient-specific anatomic data, thus avoiding the problems associated with standardized atlases [13]. Subsequent coordination of the data provided by CT imaging with known reference points in three-dimensional space allowed for the target localization and coordination with the stereotactic apparatus necessary for successful neuronavigation [14,15]. This major development led to increased clinical applications of stereotactic image-guided techniques to intracranial biopsy procedures, such as for neoplastic lesions. Clinical reports, beginning in 1973 with a series of 31 patients who underwent stereotactic biopsy for deep-seated intracranial neoplasms, helped CT-guided stereotactic biopsy gain wider

646

42

Stereotactic and image-guided biopsy

acceptance among neurosurgeons [16]. Subsequently, the late 1970s and early 1980s saw the use of and indications for stereotaxy for lesion biopsy increase, and this technique has now become a standard neurosurgical procedure [17,18]. Since gaining wider acceptance, stereotactic methods have been refined, instruments and devices have improved, and radiographic techniques offer a wider range information and greater detail. This chapter discusses current techniques, tools and concepts governing stereotactic and image-guided biopsy of intracranial lesions. The indications, methods and complications associated with this procedure are reviewed. Specific stereotactic systems are discussed in other chapters, but a brief, general discussion of framebased and frameless systems is offered. Unique considerations based on location or pathology are discussed where they may differ from standard practice. Historical techniques, methods and instruments are also mentioned where their knowledge is felt to be an important element of understanding current stereotactic image-guided practices. Finally, some ideas regarding future tools, directions and concepts in stereotactic brain biopsy are presented.

Indications For patients with an intracranial lesion, the first goal in medical management is to establish a diagnosis. Over the last 30 years, the implementation of new cranial imaging modalities has increased the diagnostic power of imaging alone. However, there continues to be a need for procedures that allow safe acquisition of tissue samples for histopathologic characterization. This is partly because some lesions cannot be definitively characterized on imaging. Previous reports have indicated that a significant percentage of histologic diagnoses made using tissue obtained with stereotactic biopsy were unsuspected based

on radiographic imaging prior to surgery [19]. Some authors report having to alter therapy due to unexpected histopathologic findings obtained via stereotactic biopsy [1,20]. Even if the radiographic studies correctly diagnose the lesion, characterization using traditional histologic classification methods can today be supplemented with information from electron microscopy, as well as molecular and genetic assays [21]. Each of these techniques requires tissue, which indicates a continued role for stereotactic biopsy in the foreseeable future. Histologic characterization using tissue obtained with stereotactic biopsy may have other significant advantages beyond lesion diagnosis. If operative resection is warranted, information from histopathologic examination may help guide surgical planning. For infectious pathologies, antibiotic sensitivities can be assayed using the biopsy tissue to optimize medical management. Patients requiring stereotactic biopsy may also benefit from other types of stereotactic intervention that can occur during the same procedure. For example, patients with liquid components to their lesion, such as pus or blood, may benefit from stereotactic drainage after the biopsy has been performed. In other cases, placing a catheter into a cytic cavity at the time of surgery may allow post-operative, percutaneous drainage for cysts expected to require chronic drainage. These types of disease-specific decisions and actions would be difficult without surgical sampling of the lesion. Once the need for a tissue specimen has been determined, the next decision is whether to obtain the material via open craniotomy or stereotactic image-guided biopsy. The majority of brain lesions revealed using current imaging modalities can be safely sampled and classified histopathologically using standard stereotactic biopsy procedures with a high diagnostic yield [18,22]. In general, stereotactic biopsy, rather than open craniotomy, is indicated if the lesion is surgically inaccessible, deep-seated or in a


functionally critical area. Examples include lesions in the motor cortex, basal ganglia, corpus callosum or brainstem. Open surgery for lesions in these areas could possibly result in an unacceptable neurologic deficit. Multifocal lesions are likely more appropriate for biopsy alone due to the presumed morbidity associated with attempted resection of multiple lesions. Also, patients with a lesion that would be better treated using noninvasive methods such as chemotherapy or radiation, such as lymphoma or germ cell tumor, should undergo stereotactic biopsy rather than craniotomy to histologically confirm the suspected diagnosis. Most other types of neoplastic brain lesions should be maximally resected [23]. Patient-related factors such as significant medical illness and extreme age may contribute to selection of stereotactic biopsy instead of craniotomy as the operative intervention (> Table 42‐1). In commonly cited large case series reporting stereotactic biopsies of a brain lesions, the most common indications based on postoperative diagnosis were neoplastic diseases [18,24,25]. This included primary central nervous system (CNS) neoplasms such as glioma, as well as metastatic tumors. Other reported etiologies included a variety of intracranial pathologies including infectious processes such as abscess and encephalitis,

42

vascular pathologies such as hemorrhage and cavernous malformation, demyelinating conditions such as multiple sclerosis, and inflammatory conditions such as sarcoidosis [26]. Recommended contraindications to performing a stereotactic biopsy vary based on the literature source, and should be considered on a patient-specific basis. Factors such as HIV status, lesion location, suspected diagnosis and patientrelated factors such as age, comorbidities and treatment preference may play a role in determining the best treatment. In general, lesions associated with significant mass effect should not undergo stereotactic biopsy due to the likelihood of swelling and postoperative neurologic worsening after biopsy. Lesions that are possibly vascular pathologies, such as arteriovenous malformations, should not undergo stereotactic biopsy due to the risk of hemorrhage. For some lesions, radiographic diagnoses can be confirmed by serum or cerebrospinal fluid analysis, which may obviate the need for confirmatory tissue diagnosis prior to planning definitive therapy. Some contraindications may not be specific to stereotactic biopsy. For example, patients with hematologic disorders or difficulties with bleeding may need temporary correction of their disorders prior to undergoing the procedure. Certain medication regimens, such as

. Table 42-1 Indications and contraindications for performing stereotactic image-guided brain biopsy versus open craniotomy

Location

Number of lesions Size Suspected diagnosis

Patient

Possible Indications

Possible Contraindications

Surgically inaccessible Deep-seated (basal ganglia, thalamus, brainstem) Eloquent area Multifocal

Superficial Amidst vascular structures

Very small Infection Lymphoma Germ cell tumor Inflammatory pathology Any for which histologic, molecular or genetic information will affect management Too ill for open resection Extreme age

Very large Gliomatosis Vascular pathology Radiographic diagnosis can be confirmed by less invasive method Multiple lesions in setting of metastatic disease Too ill for any surgical intervention Coagulopathy

647

648

42


aspirin and plavix, may need to be discontinued temporarily prior to surgery. Other contraindications within specific scenarios are discussed later in the chapter.

Technique and Instruments A number of frame-based and frameless stereotaxy systems appropriate for image-guided stereotactic biopsy are available. Each is described and discussed in a separate chapter in this book. Therefore, only a brief overview of our protocol for stereotactic brain biopsy is presented. Continued development of stereotactic techniques has been conducted at our institution since 1981 with the prototype arc-quadrant Brown-Roberts-Wells (BRW) system and then the Cosman-Roberts-Wells (CRW) system (Radionics, Burlington, MA). Our institution employs the Cosman-Roberts-Wells (CRW) stereotactic system (Radionics, Inc., Burlington, MA, USA) for frame-based CT-guided stereotactic biopsies, and the Leksell G-Frame stereotactic system (AB Elekta Instruments, Stockholm, Sweden) for functional neurosurgery procedures and radiosurgery procedures. Another popular frame-based system is the Brown-Roberts-Wells system (Radionics, Inc., Burlington, MA, USA), which was the predecessor of the CRW system.

Frame-based Stereotactic Biopsy The first step for patients undergoing stereotactic brain biopsy is application of the CRW base frame. Pin sites are selected so that they are outside the imaging plane of the biopsy target, and not in a position to interfere with the anticipated trajectory of the biopsy probe. After the patient receives intravenous sedation, the scalp at the intended pin sites is infiltrated with local anesthetic. Pins are then applied through the scalp and to the skull until ‘‘finger tight.’’ The pins should not be over-tightened and

should be the only part of the frame touching the head. Compression of soft tissue structures such as the cheeks or nose by other parts of the frame may result in pressure necrosis. The second step is image acquisition. Our preference for most frame-based stereotactic brain-biopsies is a contrast enhanced CT. This is acquired after attaching a localizer with N-shaped fiducial bars to the base frame. Although CT resolution is lower than that of MRI, it is typically adequate for most biopsy procedures. Also, CT acquisition time is significantly faster than for MRI. Disadvantages may include artifact associated with the pins, hypersensitivity to the contrast dye, and difficulty targeting lesions that are poorly contrast enhancing. The biopsy target is selected based on multiple factors. The target should be the location most likely to yield diagnostic tissue. Our practice is to select a contrast-enhancing region well inside the borders of the lesion. The trajectory of the biopsy probe should avoid eloquent neural anatomy and vascular structures. Data points are read from the scanner display and documented by both the attending and resident neurosurgeons. This data is then entered into a stereotactic computer in a similar manner. This computer generates the final stereotactic coordinates, which are applied to the stereotactic arc on a phantom base. A stereotactic cannula placed on the arc should coincide with the target coordinates set onto the base. This phantom targeting process is a method of ensuring the accuracy of the coordinates and components of the stereotactic frame. The stereotactic frame is then secured to the base frame. After sterilizing the operative field, the scalp overlying the entry point is identified using a probe fixed to the stereotactic frame and then infiltrated with local anesthetic. A 1 cm incision is made in the scalp. The arc is then rotated into position and a probe passed via the arc to the skull. The probe is then removed and a burr hole is created at the site indicated by the probe using


a drill guide attached to the stereotactic frame. The dura is punctured and a blunt obturator is inserted into the target site. The desired biopsy instrument is then inserted for tissue sampling. We typically use cupped forceps, but a variety of biopsy instruments may be used, including screw aspirators, needle core instruments, and the side cutting needle. Several passes with the needle may be required to achieve diagnosis [27]. Indeed, some work indicates that the diagnostic accuracy of the procedure improves as the number of biopsy samples taken increases [28]. For lesions that appear heterogeneous on imaging, the needle should be directed to the highest grade portion of the lesion (high contrast enhancement on CT or MRI, hypermetabolism on PET) [4]. Techniques for sampling the lesion may vary depending on the suspected pathology. For example, a tract biopsy can be planned if the lesion is suspected of having areas of different histologic grades. A tract biopsy involves taking multiple biopsies along a tract through the lesion. This technique may optimize histologic grading of a lesion by decreasing the influence of heterogeneity within a tumor [29,30]. Also, for peripherally enhancing lesions suspicious for high grade neoplasms such as glioblastomas, sampling from the margin reduces the incidence of a histologic diagnosis of ‘‘necrosis.’’ Alternatively, tissue at the periphery of the lesion may diagnosed histologically as ‘‘gliosis.’’ After confirmation from the pathologist that adequate diagnostic tissue has been taken, the cannula is removed. The wound is irrigated, and the scalp reapproximated. After removal of the stereotactic frame and base frame, a head CT without contrast is obtained. This allows verification of the biopsy target, typically by demonstrating a small amount of air at the previously identified target site. The CT also evaluates for hemorrhage. Typically, patients remain in the recovery room until they return to their neurologic baseline. They are then transferred to the ward, and discharged the next morning if there are no new problems. Patients with preexisting

42

morbidity or who undergo a complicated biopsy are admitted to the intensive care unit. In general, the vast majority of patients are discharged within 24 h of their procedure. As will be discussed later, postoperative imaging and hospital admission are two areas of debate in the postoperative management of patients undergoing elective stereotactic biopsy [31,32].

Frameless Stereotactic Biopsy Frameless stereotactic techniques initially developed in the 1980s represent an alternative to frame-based systems. The development of this technique has depended largely on advances in computing technology and processing power. Improved cranial imaging permitted the necessary accuracy of three-dimensional volumetric rendering of the brain required for reliable surgical planning. Advances in computing power enabled real-time visualization of cranial anatomy and orientation of a pointing device in three dimensions. Over the last 20 years, decreasing costs of computing equipment and improvements in the underlying technologies have resulted in more wide-spread use of this technology [3,33,34]. The techniques for performing a frameless stereotactic brain biopsy are discussed in greater detail in other chapters. In general, though, the goals of each step are similar. Rather than applying a stereotactic frame to the patient’s skull, fiducial markers or anatomic landmarks serve as reference points. If fiducial markers are used, they are affixed to the skin in a pattern optimal for the anatomical location of the underlying pathology. The word ‘‘fiducial’’ stems from the Latin word meaning ‘‘trust,’’ and implies the faith placed on these markers to accurately maintain their anatomic identity. The patient then undergoes cranial imaging and this data is entered into a computer. The use of anatomic landmarks as reference points allows the imaging to occur a longer time before surgery.

649

650

42


The reference points are registered as fixed points in space using a pointing device, an external passive or active infrared camera, and the computing interface of the three-dimensional digitizer. Most modern frameless neuronavigation systems are nonlinked, meaning the positional probe is not directly linked to the positional detector. The output from this data entry is the ability to use the pointing device to identify the optimal entry point, trajectory and depth for performing the biopsy. Compared to frame-based systems, fiducials or anatomic landmarks are used in lieu of the localizer attached to the stereotactic frame. Data entry of fiducial bars on the frame is replaced with registration of reference points using the pointing device and infrared optical imaging camera. Thus, the preparation is conceptually quite similar. Fixed points in space with known location relative to cranial anatomy (including the target lesion) are reported as three-dimensional coordinates to a software package that permits surgical planning of the biopsy in terms of entry point, trajectory and depth. One distinction is that the biopsy entry point and trajectory can be visually evaluated intraoperatively after registering the reference points. This allows some degrees of freedom in planning the surgical approach based on the anatomic location of the lesion. Typically, the frameless surgical navigation system is used to position an instrument holder, which, after being fixed in the desired location, is used to stabilize the biopsy instrument. Early reports of frameless neuronavigation systems reported accuracy approaching 4 mm [35,36], whereas recent reports indicate accuracy of approximately 1 mm [37]. The frameless stereotaxy systems used at our institution are the StealthStation-TREON (Medtronic, Louisville, CO) and BrainLAB (BrainLAB, Feldkirchen, Germany) neuronavigation systems. We most commonly use them during functional neurosurgery procedures and open craniotomies, rather than for stereotactic biopsy.

Although frame-based systems represent the ‘‘gold standard’’ in terms of stereotactic brain biopsy, frameless systems have been increasingly utilized and investigated since their inception [30,38]. Some authors maintain that the frameless techniques are gradually replacing frame-based procedures in the same way that freehand brain biopsies were replaced by frame-based stereotactic techniques [3]. Others assert that the three-dimensional accuracy of frameless techniques using skull-applied fiducials can exceed that of stereotactic frames [3]. Initially, frameless systems were used for planning and guidance in open craniotomies. More recently, frameless biopsy techniques and instruments have been refined making this option more practical. Recent work has compared frameless stereotactic biopsy to frame-based techniques, and evaluated for differences in diagnostic yield and complications. A large series in 1999 reported rates of diagnostic biopsy (96.3%), neurologic morbidity (1.4%) and death (1.0%) that were comparable to those reported for frame-based procedures [2]. However, results for posterior fossa biopsy were significantly worse. Other work has also reported that frameless procedures showed no significant differences in terms of diagnostic yield or permanent morbidity when compared to frame-based biopsies. The authors noted that frameless techniques were potentially advantageous for larger or cortical lesions, whereas frame-based stereotaxy was possibly more effective for smaller or deep-seated lesions [38]. For lesions at least 2-cm in diameter, frameless biopsy techniques are likely equal to frame-based systems in terms of targeting accuracy, diagnostic results and complications [39]. The time in the operating room is typically longer for frameless biopsy procedures compared to frame-based biopsies, although the time spent with the head immobilized is usually shorter (> Table 42‐2) [27,39]. Various adjuncts to frame-based and frameless stereotaxy have been developed and incorporated


. Table 42‐2 Comparison of frame-based versus frameless stereotactic biopsy

Region imaged Image reconstruction Length of cranial immobilization Trajectory Intraoperative time

Frame-based

Frameless

Focal at level of lesion None

Entire head

Longer

Threedimensional Shorter

Fixed Shorter

Free Longer

(Adapted from Bernstein Berger [27,39])

into neurosurgical procedures including biopsy. This includes coordination of multiple imaging modalities, such as ultrasound and endoscopy, into the procedure for real-time verification of target localization. MR images can also be coregistered with CT images prior to biopsy using cranial landmarks, which is reported to increase accuracy [40]. Robot-assisted frameless stereotactic biopsy has also been investigated as another alternative to frame-based systems [41].

Types of Image-Guidance As imaging modalities have improved in terms of sensitivity and specificity, their incorporation into stereotactic procedures has improved the diagnostic accuracy and safety of biopsies, and decreased the need for repeat biopsies. This includes CT technology capable of finer detail, as well as other imaging modalities such as magnetic resonance imaging and positron emission tomography. In addition, advancements in the instruments used for stereotactic biopsy has minimized the technical difficulty and decreased the morbidity of these procedures. Over the last 30 years, computed tomography (CT) has been established as a mainstay in

42

image-guided stereotactic biopsy. Prior to CT, neurosurgeons had to infer the location of intracranial lesions based on displacement of blood vessels using angiography, or distortion of ventricular anatomy using ventriculography [3]. CT technology was initially used to direct freehanded intracranial procedures, thus providing image guidance without stereotactic localization [42,43]. In fact, some authors reported that the diagnostic yield and complications were comparable to those for CT-guided stereotactic procedures [44]. Free-hand CT-guided biopsies were often performed in the CT suite. After placement of a burr hole, a biopsy instrument was advanced and repeated CT scans were taken to determine the proximity of the instrument to the lesion. Naturally, this technique had greater success in large, superficial lesions for which a larger margin for error existed in choosing the trajectory. Despite numerous case series regarding stereotactic CT-guided biopsy, arguments for image-guided free-hand techniques continued to appear in the literature [45]. Ultimately, however, literature regarding stereotactic image-guided brain biopsies demonstrated this technique to be the new standard in terms of diagnostic yield, morbidity and mortality [1,22,46,47]. Eventual acceptance of stereotactic imageguided biopsy as a superior tool for obtaining tissue for histopathologic diagnosis of a brain lesion occurred concurrently with widespread implementation and evaluation of CT technology. Researchers realized that mathematical data from CT images could allow localization of a lesion in three-dimensions relative to predetermined reference points [14,48]. This led to the development of concepts behind the head frame and fiducials. The locations of the fiducials were fixed relative to the frame and their locations relative to the lesion could be decoded on each image slice into mathematical coordinates (e.g., Cartesian coordinates). Further refinements in stereotactic techniques, such as target and trajectory selection, have been aided

651

652

42


by improvements in CT imaging [3]. For large lesions with distinct target areas, a CT with contrast using standard scan thickness and interscan spacing is adequate for selecting a biopsy site. Smaller lesions may require a decrease in the inter-scan spacing and alteration of the scan thickness so that the target area of the lesion can be identified. Magnetic resonance imaging technology, introduced in the 1980s, enhanced the diagnostic power of non-invasive cranial imaging. Some lesions difficult to visualize using CT can be seen in great detail using standard MRI techniques. The introduction of magnetic resonance (MR) technology fostered interest in evaluating MR-guided stereotactic biopsy as a possibly more accurate alternative to CT-guided procedures [49]. In general, advantages of MR over CT typically include finer detail of brain anatomy and the ability to identify lesions not visualized using standard contrast enhanced CT. Also, the risks of gadolinium are much less than those of CT contrast agents. These properties have resulted in MR as the study of choice for open procedures in which intraoperative neuronavigation is used [37]. For stereotactic biopsies, however, acquiring an MR requires significantly more time than acquiring a CT. Also, previous reports indicated concern due to magnetic susceptibility artifacts causing anatomic distortion, and found a discrepancy of 2 mm when comparing CT and MR coordinates [49]. Other work has shown increased accuracy with coregistration of CT and MR images [40]. Ultimately, the role of MR in stereotactic biopsy depends partly on surgeon preference and partly on how well the target lesion is visualized using both studies. Brain lesions may have significant heterogeneity on imaging that complicates selection of a target. In these cases, accurate target sampling is often crucial for making the correct histologic diagnosis. For example, radiation necrosis can appear radiographically similar to recurrence of a neoplasm. Advanced imaging techniques such

as MR spectroscopy and positron emission tomography (PET) may help differentiate these lesions radiographically. Alternatively, these adjuvant imaging techniques can be coordinated with CT and/or MR and used in target selection for stereotactic biopsy. Currently, the majority of image-guided stereotactic brain biopsies are performed with CT- or MR-guidance. However, these techniques fail to provide a tissue diagnosis in approximately 5% of cases [1]. Some authors feel that inaccuracies associated with CT- or MR-guidance could result in misdiagnosis or incorrect histologic grading [9]. To address these issues, supplemental imaging modalities have been investigated for their usefulness in improving diagnostic accuracy and yield. One example is positron emission tomography, which evaluates the metabolic profile of the target tissue. Data from this imaging modality can be used to identify a target for stereotactic brain biopsy by identifying the abnormal metabolic regions in brain tumors. One study performed stereotactic biopsies using combined CT- and PET-guidance. The authors reported that targets defined using PETwere always diagnostic, whereas 17% of the targets defined using CT were nondiagnostic. Furthermore, some regions that appeared normal on CT yielded diagnostic tissue with PET-guided stereotactic biopsy [50,51]. These results suggest that PET-guidance may increase the diagnostic yield of stereotactic biopsy with no increase in procedure-related morbidity [52]. Similar improvements in accuracy and diagnostic yield have been observed when combining MR- and PET-guidance for stereotactic brain biopsies in pediatric patients [53]. For glioma biopsies, a recent report recommended using 11C-methionine (MET) as the tracer rather than 18F-fluorodeoxyglucose (FDG) because MET provides a more sensitive signal [54,55]. PET is most useful in high-grade lesions because the study depends on increased metabolism to show differential uptake, and lower grade lesions may not demonstrate increased uptake.


Real-time visualization of the target lesion cannot be achieved using standard image-guided stereotactic techniques unless the operating room is equipped with imaging equipment. This creates concern regarding the effects that ‘‘brain shift’’ may have on specific target localization. Slight alteration of the location of the target lesion relative to the reference point used for stereotaxy may occur for a number of reasons. After imaging, surgical creation of the entry point for the biopsy probe may result in brain relaxation and alteration in the underlying anatomy. Alternatively, administration of anesthetic agents, diuretics or steroids may affect the relative location of the lesion. If the imaging occurs a longer time before surgery, changes in the lesion itself may shift the three-dimensional coordinates of the target. These concerns could be addressed by real-time imaging offered by intra-operative CT- or MR-guided stereotactic biopsy. Alternatively direct visualization using endoscopy or ultrasound may confirm target location. Each technique has been employed for intracranial biopsy, but each also has limitations that prevent universal application in the same fashion as CT- or MR-guided stereotactic biopsy. Early experience with the use of intraoperative CT-guidance for stereotactic biopsies was reported in 1982 [39]. Advantages of intraoperative imaging included the potential for real-time verification that a lesion was accurately targeted and that the biopsy caused no immediate hemorrhage. Intraoperative MRI offers the potential for enhanced anatomic detail, but presents additional difficulties in terms of compatibility of operative equipment. Endoscopy is another category of imaging technology that can be used to assist with intracranial biopsy, and has been described for a variety of intracranial pathologies. Larger series typically discuss both biopsy and resection of intra- or paraventricular lesions such as pineal region tumors [56]. However, some studies suggest that although the rate of serious complications is lower with

42

endoscopic biopsy, this technique is less accurate in terms of diagnosis than frame-based stereotactic biopsy procedures [57]. More recently, simultaneous frameless image-guided and endoscopic neuronavigation was applied to infants with intraventricular pathology [58]. The results indicated that such combined methods could prove advantageous in specific situations. The ability to perform the procedures using frameless neuronavigation was important for avoiding unnecessary procedure-related morbidity. Although most commonly used for intraventricular lesions, endoscopes may also be helpful in verifying complete evacuation of cystic lesions such as abscesses [59]. Additional measure such as third ventriculostomy may take place simultaneously with endoscopic biopsy procedures [57]. Ultrasound has been used as image-guidance in a variety of neurosurgical settings, including ultrasound-guided free-hand biopsy of cortical lesions [45]. The ultrasound probe can be fixed in a specific position and images obtained intraoperatively can be superimposed on preoperatively acquired MR or CT images in order to track brain shift [60]. Comparisons between different imaging techniques used as guidance for lesion biopsy are inconclusive, but some work suggests that frameless MR-guidance may have higher diagnostic accuracy than ultrasound-guided techniques [61].

Complications Although stereotactic biopsy is minimally invasive, the procedure has distinct risks that must be weighed for each patient against the benefits of obtaining histologic diagnosis. The incidence of morbidity due to stereotactic biopsy ranges from 1.0–6.5%, and mortality rates range from 0 to 1.7% in commonly referenced large series [1,25,62]. Patient-specific factors may place varying importance on each risk, and must be considered accordingly. Additionally, a preoperative

653

654

42


discussion of the risks and advanced directive with the patient and family is important in the unlikely event that the patient is neurologically devastated after the procedure. One potentially severe complication of stereotactic biopsy is hemorrhage. Up to 60% have asymptomatic small hemorrhages [63]. In general, though, the risk of morbidity due to hemorrhage is approximately 1%. However, this risk is higher in malignant pathologies, possibly due to neovascularization, and approaches 6% in patients with GBM [25]. Minimization of the number of biopsy specimens taken and passes with the biopsy probe will help reduce the risk of hemorrhage. Preoperative risk assessment in each patient should help identify patients with higher risk of hemorrhage, such as those in whom a higher grade lesion is suspected. If brisk bleeding is encountered from the biopsy probe, the bleeding may be allowed to continue until it stops spontaneously [64]. If brisk bleeding continues for more than a few minutes, intravenous thrombin can be administered [65]. If the postoperative CT shows a large hemorrhage and the patients has deteriorated neurologically, an immediate craniotomy for hematoma evacuation is likely warranted. Another possible complication of stereotactic biopsy is neurologic worsening not associated with hemorrhage. This neurologic deterioration may be associated with focal neurologic symptoms, such as motor weakness, or an altered mental status without focal signs. There may not be an obvious clinical or radiographic cause for this deterioration, and literature regarding this observation is largely anecdotal. In patients with large tumors, postoperative brain swelling after biopsy may worsen neurologic condition [25]. Most patients recover over several days to a few weeks, but some do not. In such cases, the neurosurgeon must rule out treatable causes of the neurologic worsening [4]. A non-diagnostic biopsy is another negative outcome that must be considered prior to

stereotactic biopsy. Failure to achieve histological or microbiological diagnosis based on the obtained tissue can occur in 8.1% of cases according to a large series from 1995 [27]. Another report showed that the diagnosis based on the first biopsy sample was inaccurate in 33% of cases, but this decreased to 11% by the fourth biopsy specimen [66]. Factors associated with a non-diagnostic biopsy may include immunocompromised patients, non-neoplastic lesions and non-enhancing lesions [66]. If the initial biopsy specimens are non-diagnostic, the instruments and coordinates must be rechecked for accuracy. Based on the planned target and histopathology results from the initial specimens, the biopsy depth may be slightly adjusted for further sampling. For example, if the histology results show necrosis, the depth of the biopsy probe could be reduced to obtain specimens closer to the margin of the lesion. The opposite corrective measure could be taken if the initial histologic examination shows gliosis or normal brain parenchyma.

Pathology The purpose of performing a stereotactic imageguided biopsy is to achieve histopathologic diagnosis. Only a small amount of tissue is obtained with such techniques. Using the cup forceps described above, biopsy specimens are typically 1–2 mm3. Therefore, collaboration with an experienced neuropathologist is critical. In addition, all relevant clinical and radiographic information must be communicated to the pathologist in order to make best use of the small amount of tissue. The initial evaluation of a biopsy specimen commonly involves a smear, except when the specimen is especially firm and will not smear on a slide [22]. The smear involves placing one or more specimens on a glass slide and using a second slide to crush the specimen. The specimen


is then stained with hematoxylin and eosin (H&E) and examined using light microscopy. If examination fails to provide a diagnosis, a frozen section is performed. This technique involves examination of thin slices of the specimen stained with H&E. Either method of quick section examination, smear or frozen section, can predict the final histologic diagnosis in up to 90% of cases, but the primary goal of such techniques is to ensure that the specimens contain diagnostic tissue, even if a definitive diagnosis cannot be made using quick section [67]. If both the smear and frozen section are non-diagnostic, additional biopsy specimens may be required. At some point, if continued samples are non-diagnostic using quick section techniques, the surgeon will have to re-evaluate the risks and benefits of continued sampling. Ultimately, after conclusion of the biopsy procedure, the remaining tissue samples are fixed in formalin and undergo more definitive evaluation depending on the diagnosis made during quick section. For example, tumors will undergo immunohistochemistry staining for specific antigens such as glial fibrillary acid protein (GFAP) and epithelial membrane antigen (EMA). Infectious lesions may undergo a variety of culturing techniques to isolate the infectious organism and assess antimicrobial efficacy and resistances. Current stereotactic image-guided biopsy techniques in conjunction with an experienced neuropathologist provide the correct diagnosis in up to 95% of cases. In our experience with specimens that were non-diagnostic, necrosis was seen in 45% and inflammation was observed in 41% [1]. Other non-diagnostic results can include gliosis and granuloma. In addition to failure of diagnosis, the histologic results can fail to accurately diagnose the histologic grade of a neoplasm. An example would be a histologic diagnosis of an anaplastic astrocytoma after stereotactic biopsy that is upgraded to glioblastoma based on pathology results after subsequent open resection of the lesion. This type of nonperfect

42

correlation between stereotactic biopsy diagnosis and final diagnosis of the resected tumor would only be proven if the patient underwent a second surgical procedure after their initial stereotactic biopsy. Thus the incidence of this error is difficult to assess. In our series of 30 patients who underwent craniotomy for resection of a lesion after having stereotactic biopsy, the histologic diagnosis achieved with stereotactic biopsy directed clinical management appropriately in 28 patients. However, an exact correlation between histopathology results was observed in only 19 of the 30 patients [22].

Locations Stereotactic image-guided biopsy as described above can be conducted for lesions in nearly any intracranial location [68]. However, the safety and efficacy of the procedure in some anatomic locations has been debated. A few of these are discussed here.

Pineal Region Open surgical exploration of the pineal region can be associated with significant morbidity. However, the diverse collection of possible diagnoses mandates histo pathologic evaluation to guide the treatment plan [69]. Stereotactic approaches to the pineal region are considered by some to carry greater risk of hemorrhage than stereotactic biopsy in other parts of brain due to the proximity of deep draining veins and propensity of some pineal region tumors to bleed. Early work in the 1970s established stereotactic biopsy as a an option in the management of lesions of the pineal/third ventricle region [16,20]. Later work confirmed that stereotactic biopsy of pineal and para-third ventricle lesions could safely guide further medical management

655

656

42


[1,70]. A large series involving stereotactic biopsy of pineal region lesions demonstrated successful diagnosis in 94% of 370 patients, with a mortality rate of 1.3% and neurologic morbidity of 7.8%. These percentages are comparable to those for other regions of the brain and show that stereotactic image-guided biopsy is a safe and effective method for achieving histologic diagnosis of a pineal region lesion [71]. To avoid injuring veins in the pineal region, the entry point of the biopsy probe should be at least 3 cm anterior to the coronal suture and 3 cm lateral to midline.

Posterior Fossa and Brainstem Lesions located within the brainstem and deep cerebellum represent a broad spectrum of pathologies, and are often a challenge in terms of diagnosis and management. Most cannot be surgically resected, and open biopsy carries a high risk of morbidity and mortality [72]. As with most pathologies, ideal treatment would be guided in part by histologic diagnosis [73]. Stereotactic biopsy of brain stem lesions avoids complications associated with open surgical techniques and is associated with minimal morbidity and mortality rates [74]. Comparatively, it is a minimally invasive and safe technique that provides adequate tissue for histopathologic diagnosis in adults and pediatric patients [75,76]. Furthermore, by providing histologic diagnosis, the technique allows optimization of the patient’s future medical management [77]. Stereotactic biopsy of brainstem lesions can be performed via supratentorial approach or an infratentorial, transcerebellar approach under local anesthesia and intravenous sedation [77]. The infratentorial approach proceeds through a sub-occipital burr hole with the patient in a semi-sitting or prone position. A technique that may reduce the risk of venous air embolism is to place the burr hole while the patient is prone,

and then perform the biopsy with the patient in a sitting position [75]. The supratentorial approach may be performed either through a precoronal frontal burr hole for brain stem lesions, or through a parieto-occipital, transtentorial trajectory for cerebellar lesions. For brain stem biopsies, as the biopsy needle penetrates the brainstem, vital signs such as heart rate, respiratory rate and blood pressure should be carefully monitored. Alterations in the vital signs could indicate compression of brain stem nuclei and the instrument should be withdrawn until the vital signs return to baseline. Lesions in the brain stem and posterior fossa may not be as amenable to CT-guided stereotactic biopsy due to bone artifact, and MR-guided procedures may improve diagnostic yield. Other work showed that combined MR and PET imaging improved the radiologic evaluation of brain stem lesions, but could not replace the clinical guidance provided by histologic diagnosis, which can only be achieved through surgical sampling. This work also demonstrated that combined MR- and PET-guidance improved the accuracy and diagnostic yield of stereotactic biopsies of brain stem lesions [78].

Skull Base Skull base lesions most commonly undergo open resection. Minimally invasive procedures for tissue diagnosis alone are uncommon. However, in some situations patients who are in poor medical condition and cannot tolerate open craniotomy may benefit from stereotactic biopsy. Although little literature exists, reports indicate that stereotactic biopsy can be safely and accurately performed for the diagnosis of skull base lesions [79,80]. This has been demonstrated for pathologies of the cavernous sinus, jugular foramen, and clivus. A transoral stereotactic approach to the second cervical vertebral body has also been described [81].


Other Considerations Pediatric Histologic diagnosis of an intracranial lesion is especially important in children given the higher sensitivity of their immature brain to radiation and certain types of chemotherapy. Although dedicated pediatric studies regarding stereotactic imageguided biopsy are less numerous than those for adult or mixed patient populations, reports indicate that the technique is a safe and efficacious method for obtaining tissue, even in eloquent areas such as the brain stem [82]. In addition, the possibility of preventing unnecessary adjunctive therapy is worth the risks associated with the procedure if radiographic studies are not definitively diagnostic [83].

HIV/AIDS HIV is associated with several opportunistic intracranial pathologies that can occur in the setting of immunosuppression. Many of these are treatable if diagnosed accurately. Infectious etiologies such as toxoplasmosis are much more likely in this subset of patients. Toxoplasmosis is so common that some authors have recommended an empirical antitoxoplasmosis trial as the first step prior to considering stereotactic biopsy if the cranial imaging is suspicious for toxoplasmosis [96]. A number of other infectious possibilities have been reported in various case series and reports. Because optimal management varies for each type of infection, and because patients are often in poor condition medically, obtaining tissue is critically important to optimizing treatment. Diagnoses such as progressive multifocal leukoencephalopathy are rare in the general population, but are a common etiology of brain lesions in HIV patients. Lymphoma is much more common in HIV patients, and is a common diagnosis

42

after stereotactic biopsy [84]. In one series, lymphoma accounted for 46% of all lesions, followed by progressive multifocal leukoencephalopathy (23%) and toxoplasmosis (15%) [85]. However, the same series reported a high rate (10%) of morbidity and mortality for stereotactic biopsy in HIV patients. Another study reported 2.9% mortality and 8.4% morbidity associated with stereotactic biopsy procedures in HIV patients [86]. Thus, although cerebral lesions are seen in a significant percentage of patients with AIDS, the high complication rate is an important consideration prior to performing a stereotactic biopsy in an HIV patient.

Cystic Lesions Some cases will warrant not only stereotactic biopsy, but also drainage of an associated cystic lesion. In these instances, the target should be selected such that after biopsy of the enhancing border, the probe can be advanced using the same trajectory into the cyst for drainage. Biopsy of the cyst wall should occur first so as not to distort the three-dimensional orientation of the lesion. After diagnostic biopsy is confirmed by the pathologist, the probe can be advanced the appropriate distance for cyst drainage [1].

Postoperative Management Stereotactic biopsies at our institution are inpatient procedures, though the vast majority of patients admitted electively are discharged within 24 h. At some centers, stereotactic biopsy is an outpatient procedure. This practice is reportedly safe for patients with supratentorial lesions if there was no notable intraoperative bleeding or evidence of such on postoperative cranial imaging, and if the postoperative neurologic status of the patient is unchanged [87]. These authors recommended non-enhanced brain CT 2 h after

657

658

42


the biopsy to verify the absence of hemorrhage larger than 1 cm. Other work has debated the necessity of postoperative imaging to rule out hematoma. In one report, postoperative CT scans did not affect management in the majority (90%) of patients. The authors recommended imaging only for patients with a new postoperative neurologic deficit, or in whom significant bleeding was encountered during surgery [32].

Spine Just as with cranial lesions, certain clinical or radiographic considerations may argue for biopsy rather than surgical resection or non-invasive management for patients with spinal lesions. In these cases, spine lesions can often be successfully sampled with image guidance alone, such as with fluoroscopy or CT-guidance. However, application of stereotactic principles may improve the diagnostic yield from these procedures. Increasingly, modern neuronavigation techniques are being used in spine procedures such as the insertion of instrumentation. Recent reports demonstrate the efficacy of stereotactic devices in spine biopsy procedures [88]. Continued refinement of these techniques may further the development of stereotactic image-guided biopsy of spine lesions.

Non-neoplastic Lesions The majority of intracranial lesions that undergo stereotactic biopsy are tumors [18]. In larger series, non-neoplastic lesions typically comprise up to 20% of diagnoses [89]. These lesions are often infectious, demyelinating or inflammatory disorders. However, the reported diagnostic accuracy of stereotactic biopsy for non-neoplastic lesions is typically lower than that for tumors [90]. This is

possibly due to a greater variability in the pathologic processes underlying non-neoplastic lesions. Despite an increased risk of diagnostic inaccuracy, stereotactic biopsy is a safe and likely underused technique in the management of non-neoplastic cranial lesions that has clear potential benefits in guiding treatment [91,92].

Future Much like the development of stereotactic techniques to this point, continued advances will be influenced to a large degree by technological advances in a broad spectrum of sciences that includes physics, mathematics, computing, materials science and molecular biology. Future innovations will continue progress towards minimizing invasiveness while maximizing efficiency, accuracy and the amount of information gleaned from the biopsy specimens. Advances in intraoperative image guidance may help address concerns regarding brain shift, and lead to increased accuracy and safety of intracranial biopsy procedures. Many large academic centers have operating rooms equipped with MR or CT technology [93]. However, the resulting anatomic detail is typically inferior to scanners in the radiology suites and may not be suitable for real-time image guidance for small or non-enhancing lesions. Intraoperative imaging combined with compatible surgical navigation instruments can allow real-time visualization of the biopsy target. This technique could be useful for very small lesions or those located within or deep to eloquent tissue or vascular structures [94]. Future improvements could include refinements that render these machines less cumbersome while offering higher resolution that allows intraoperative targeting of very small lesions. Another option for real-time target verification is the incorporation of endoscopy and ultrasound technology into the stereotactic biopsy procedure. Combining endoscopy or ultrasound


devices with stereotactic equipment has been described as a useful imaging adjunct for some biopsy procedures. Continued minimalization of each technology may yield further benefits. One target for modification could be the biopsy probe itself. Integrating ultrasound capabilities into the tip of the probe could enhance the diagnostic accuracy and confidence of stereotactic biopsies by providing intraoperative visual confirmation of target localization. Smaller endoscopes, and flexible versions of both equipped with actuators, could allow for steering of the tip if the real-time feedback indicates that the trajectory is inaccurate or if initial histologic results are non-diagnostic. Linking such real-time images to three-dimensional reconstructions of preoperative images could occur with heads-up or holographic display technology, thus maximizing the benefit of each technology. The biopsy probe could also be ‘‘functionalized’’ in other ways. Advanced tumor therapies will be partly dependent on genetic and molecular subtyping of the lesion, and may be implemented at the time of biopsy using the same surgical access. Instantaneous biochemical feedback could provide real-time information about the type of tissue being encountered by the probe [95]. Biologic assays could detect molecular abnormalities such as high lactate or genetic defects such as a 19q deletion. The probe could also sense the presence of intravenous contrast agents such as gadolinium or alterations in local pressure. Analysis of these data could yield guidance regarding optimal biopsy location, and offer early insight into the histologic and biomolecular profile of the lesion. As imaging modalities continue to improve in terms of their resolution and diagnostic capability, the role for stereotactic tissue biopsy may change. Traditional histopathology and immunohistochemistry techniques might not be necessary if future imaging technologies can non-invasively diagnose the lesion with equal or greater specificity. However, new biomolecular and genetic assays important for treatment

42

of intracranial lesions continue to be developed. Thus, for the foreseeable future, there will likely continue to be a need to acquire tissue from intracranial locations via the least invasive method possible.

Conclusion Stereotactic image-guided biopsy is the least invasive technique for obtaining issue from an intracranial lesion. Lesions of nearly all cranial regions can be sampled with minimal trauma to surrounding neural structures. Frame-based and frameless systems available today are easy to use and provide high diagnostic accuracy with minimal morbidity. Although the diagnostic power of non-invasive imaging techniques is likely to increase in the future, advances in histopathologic and genotyping techniques will perpetuate the need for neurosurgeons to be familiar with stereotactic principles and methods.

References 1. Apuzzo ML, Chandrasoma PT, Cohen D, Zee CS, Zelman V. Computed imaging stereotaxy: experience and perspective related to 500 procedures applied to brain masses. Neurosurgery 1987;20:930-7. 2. Barnett GH, Miller DW, Weisenberger J. Frameless stereotaxy with scalp-applied fiducial markers for brain biopsy procedures: experience in 218 cases. J Neurosurg 1999;91:569-76. 3. Germano IM. Advanced techniques in image-guided brain and spine surgery. New York: Thieme; 2002. 4. DeAngelis L, Gutin P, Leibel S, Posner J. Intracranial tumors: diagnosis and treatment. London: Martin Dunitz; 2002;411. 5. Gildenberg PL. Stereotactic versus stereotaxic. Neurosurgery 1993;32:965-6. 6. Horsley V, Clarke R. The structure and function of the cerebellum examined by a new method. Brain 1908;31:45-124. 7. Gildenberg PL. Spiegel and Wycis – the early years. Stereotact Funct Neurosurg 2001;77:11-16. 8. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50.

659

660

42


9. Black PM, Loeffler JS. Cancer of the nervous system. Philadelphia, PA: Lippincott Williams & Wilkins; 2005. 10. Kelly PJ. Applications and methodology for contemporary stereotactic surgery. Neurol Res 1986;8:2-12. 11. Housepian EM, Pool JL. The accuracy of human stereoencephalotomy as judged by histological confirmation of roentgenographic localization. J Nerv Ment Dis 1960;130:520-5. 12. Kalyanaraman S, Gillingham FJ. Stereotaxic biopsy. J Neurosurg 1964;21:854-8. 13. Iskandar BJ, Nashold BS Jr. History of functional neurosurgery. Neurosurg Clin N Am 1995;6:1-25. 14. Bergstrom M, Greitz T. Stereotaxic computed tomography. AJR Am J Roentgenol 1976;127:167-70. 15. Chen TC, Rabb C, Apuzzo ML. Complex technical methodologies and their applications in the surgery of intracranial meningiomas. Neurosurg Clin N Am 1994;5:261-81. 16. Conway LW. Stereotaxic diagnosis and treatment of intracranial tumors including an initial experience with cryosurgery for pinealomas. J Neurosurg 1973;38:453-60. 17. Lunsford LD, Coffey RJ, Cojocaru T, Leksell D. Imageguided stereotactic surgery: a 10-year evolutionary experience. Stereotact Funct Neurosurg 1990;54–55:375-87. 18. Yu X, Liu Z, Tian Z, Li S, Huang H, Xiu B, Zhao Q, Liu L, Jing W. Stereotactic biopsy for intracranial spaceoccupying lesions: clinical analysis of 550 cases. Stereotact Funct Neurosurg 2000;75:103-8. 19. Friedman WA, Sceats DJ Jr, Nestok BR, Ballinger WE Jr. The incidence of unexpected pathological findings in an image-guided biopsy series: a review of 100 consecutive cases. Neurosurgery 1989;25:180-4. 20. Pecker J, Scarabin JM, Vallee B, Brucher JM. Treatment in tumours of the pineal region: value of sterotaxic biopsy. Surg Neurol 1979;12:341-8. 21. Bouvier C, Roll P, Quilichini B, Metellus P, Calisti A, Gilles S, Chinot O, Fina F, Martin PM, Figarella-Branger D. Deletions of chromosomes 1p and 19q are detectable on frozen smears of gliomas by FISH: usefulness for stereotactic biopsies. J Neurooncol 2004;68:141-9. 22. Chandrasoma PT, Smith MM, Apuzzo ML. Stereotactic biopsy in the diagnosis of brain masses: comparison of results of biopsy and resected surgical specimen. Neurosurgery 1989;24:160-5. 23. Kondziolka D, Lunsford LD. The role of stereotactic biopsy in the management of gliomas. J Neurooncol 1999;42:205-13. 24. Apuzzo ML, Sabshin JK. Computed tomographic guidance stereotaxis in the management of intracranial mass lesions. Neurosurgery 1983;12:277-85. 25. Bernstein M, Parrent AG. Complications of CT-guided stereotactic biopsy of intra-axial brain lesions. J Neurosurg 1994;81:165-8. 26. Grand S, Hoffmann D, Bost F, Francois-Joubert A, Pasquier B, Le Bas JF. Case report: pseudotumoral brain lesion as the presenting feature of sarcoidosis. Br J Radiol 1996;69:272-5.

27. Soo TM, Bernstein M, Provias J, Tasker R, Lozano A, Guha A. Failed stereotactic biopsy in a series of 518 cases. Stereotact Funct Neurosurg 1995;64:183-96. 28. Jain D, Sharma MC, Sarkar C, Deb P, Gupta D, Mahapatra AK. Correlation of diagnostic yield of stereotactic brain biopsy with number of biopsy bits and site of the lesion. Brain Tumor Pathol 2006;23:71-5. 29. Daumas-Duport C, Monsaingeon V, Szenthe L, Szikla G. Serial stereotactic biopsies: a double histological code of gliomas according to malignancy and 3-D configuration, as an aid to therapeutic decision and assessment of results. Appl Neurophysiol 1982;45:431-7. 30. Woodworth G, McGirt MJ, Samdani A, Garonzik I, Olivi A, Weingart JD. Accuracy of frameless and framebased image-guided stereotactic brain biopsy in the diagnosis of glioma: comparison of biopsy and open resection specimen. Neurol Res 2005;27:358-62. 31. Boulton M, Bernstein M. Outpatient brain tumor surgery: innovation in surgical neurooncology. J Neurosurg 2008;108:649-54. 32. Warnick RE, Longmore LM, Paul CA, Bode LA. Postoperative management of patients after stereotactic biopsy: results of a survey of the AANS/CNS section on tumors and a single institution study. J Neurooncol 2003;62:289-96. 33. Barnett GH, Kormos DW, Steiner CP, Weisenberger J. Intraoperative localization using an armless, frameless stereotactic wand. Technical note. J Neurosurg 1993;78:510-4. 34. Watanabe E, Watanabe T, Manaka S, Mayanagi Y, Takakura K. Three-dimensional digitizer (neuronavigator): new equipment for computed tomography-guided stereotaxic surgery. Surg Neurol 1987;27:543-7. 35. Gumprecht HK, Widenka DC, Lumenta CB. BrainLab vectorvision neuronavigation system: technology and clinical experiences in 131 cases. Neurosurgery 1999;44:97-104; discussion 104-5. 36. Muacevic A, Uhl E, Steiger HJ, Reulen HJ. Accuracy and clinical applicability of a passive marker based frameless neuronavigation system. J Clin Neurosci 2000;7:414-8. 37. Jung TY, Jung S, Kim IY, Park SJ, Kang SS, Kim SH, Lim SC. Application of neuronavigation system to brain tumor surgery with clinical experience of 420 cases. Minim Invasive Neurosurg 2006;49:210-5. 38. Woodworth GF, McGirt MJ, Samdani A, Garonzik I, Olivi A, Weingart JD. Frameless image-guided stereotactic brain biopsy procedure: diagnostic yield, surgical morbidity, and comparison with the frame-based technique. J Neurosurg 2006;104:233-7. 39. Bernstein M, Berger MS. Neuro-oncology: the essentials. New York: Thieme; 2000. 40. Cohen DS, Lustgarten JH, Miller E, Khandji AG, Goodman RR. Effects of coregistration of MR to CT images on MR stereotactic accuracy. J Neurosurg 1995;82:772-9. 41. Willems PW, Noordmans HJ, Ramos LM, Taphoorn MJ, Berkelbach van der Sprenkel JW, Viergever MA,


42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Tulleken CA. Clinical evaluation of stereotactic brain biopsies with an MKM-mounted instrument holder. Acta Neurochir (Wien) 2003;145:889-97; discussion 897. Hahn JF, Levy WJ, Weinstein MJ. Needle biopsy of intrancranial lesions guided by computerized tomography. Neurosurgery 1979;5:11-15. Levy WJ, Hahn JF. Inexpensive plastic brain biopsy needle for use in the computed tomographic scanner. Neurosurgery 1980;7:606-7. Goldstein S, Gumerlock MK, Neuwelt EA. Comparison of CT-guided and stereotaxic cranial diagnostic needle biopsies. J Neurosurg 1987;67:341-8. Di Lorenzo N, Esposito V, Lunardi P, Delfini R, Fortuna A, Cantore G. A comparison of computerized tomography-guided stereotactic and ultrasound-guided techniques for brain biopsy. J Neurosurg 1991;75:763-5. Gomez H, Barnett GH, Estes ML, Palmer J, Magdinec M. Stereotactic and computer-assisted neurosurgery at the Cleveland clinic: review of 501 consecutive cases. Cleve Clin J Med 1993;60:399-410. Lee T, Kenny BG, Hitchock ER, Teddy PJ, Palividas H, Harkness W, Meyer CH. Supratentorial masses: stereotactic or freehand biopsy? Br J Neurosurg 1991;5:331-8. Brown RA. A computerized tomography-computer graphics approach to stereotaxic localization. J Neurosurg 1979;50:715-20. Kondziolka D, Dempsey PK, Lunsford LD, Kestle JR, Dolan EJ, Kanal E, Tasker RR. A comparison between magnetic resonance imaging and computed tomography for stereotactic coordinate determination. Neurosurgery 1992;30:402-6; discussion 406-7. Levivier M, Goldman S, Pirotte B, Brucher JM, Baleriaux D, Luxen A, Hildebrand J, Brotchi J. Diagnostic yield of stereotactic brain biopsy guided by positron emission tomography with [18F]fluorodeoxyglucose. J Neurosurg 1995;82:445-52. Pirotte B, Goldman S, Brucher JM, Zomosa G, Baleriaux D, Brotchi J, Levivier M. PET in stereotactic conditions increases the diagnostic yield of brain biopsy. Stereotact Funct Neurosurg 1994;63:144-9. Pirotte B, Goldman S, Bidaut LM, Luxen A, Stanus E, Brucher JM, Baleriaux D, Brotchi J, Levivier M. Use of positron emission tomography (PET) in stereotactic conditions for brain biopsy. Acta Neurochir (Wien) 1995;134:79-82. Pirotte B, Goldman S, Salzberg S, Wikler D, David P, Vandesteene A, Van Bogaert P, Salmon I, Brotchi J, Levivier M. Combined positron emission tomography and magnetic resonance imaging for the planning of stereotactic brain biopsies in children: experience in 9 cases. Pediatr Neurosurg 2003;38:146-55. Pirotte B, Goldman S, Massager N, David P, Wikler D, Lipszyc M, Salmon I, Brotchi J, Levivier M. Combined use of 18F-fluorodeoxyglucose and 11C-methionine in 45 positron emission tomography-guided stereotactic brain biopsies. J Neurosurg 2004;101:476-83.

42

55. Pirotte B, Goldman S, Massager N, David P, Wikler D, Vandesteene A, Salmon I, Brotchi J, Levivier M. Comparison of 18F-FDG and 11C-methionine for PET-guided stereotactic brain biopsy of gliomas. J Nucl Med 2004;45:1293-8. 56. Oi S, Shibata M, Tominaga J, Honda Y, Shinoda M, Takei F, Tsugane R, Matsuzawa K, Sato O. Efficacy of neuroendoscopic procedures in minimally invasive preferential management of pineal region tumors: a prospective study. J Neurosurg 2000;93:245-53. 57. O’Brien DF, Hayhurst C, Pizer B, Mallucci CL. Outcomes in patients undergoing single-trajectory endoscopic third ventriculostomy and endoscopic biopsy for midline tumors presenting with obstructive hydrocephalus. J Neurosurg 2006;105:219-26. 58. Mangano FT, Limbrick DD Jr, Leonard JR, Park TS, Smyth MD. Simultaneous image-guided and endoscopic navigation without rigid cranial fixation: application in infants: technical case report. Neurosurgery 2006;58: ONS-E377; discussion ONS-E377. 59. Longatti P, Perin A, Ettorre F, Fiorindi A, Baratto V. Endoscopic treatment of brain abscesses. Childs Nerv Syst 2006;22:1447-50. 60. Giorgi C, Casolino DS. Preliminary clinical experience with intraoperative stereotactic ultrasound imaging. Stereotact Funct Neurosurg 1997;68:54-8. 61. Jain D, Sharma MC, Sarkar C, Gupta D, Singh M, Mahapatra AK. Comparative analysis of diagnostic accuracy of different brain biopsy procedures. Neurol India 2006;54:394-8. 62. Lunsford LD, Martinez AJ. Stereotactic exploration of the brain in the era of computed tomography. Surg Neurol 1984;22:222-30. 63. Kulkarni AV, Guha A, Lozano A, Bernstein M. Incidence of silent hemorrhage and delayed deterioration after stereotactic brain biopsy. J Neurosurg 1998;89:31-5. 64. Kelly PJ. Tumor stereotaxis. Saunders; Philadelphia, PA: 65. Chimowitz MI, Barnett GH, Palmer J. Treatment of intractable arterial hemorrhage during stereotactic brain biopsy with thrombin. Report of three patients. J Neurosurg 1991;74:301-3. 66. Brainard JA, Prayson RA, Barnett GH. Frozen section evaluation of stereotactic brain biopsies: diagnostic yield at the stereotactic target position in 188 cases. Arch Pathol Lab Med 1997;121:481-4. 67. Colbassani HJ, Nishio S, Sweeney KM, Bakay RA, Takei Y. CT-assisted stereotactic brain biopsy: value of intraoperative frozen section diagnosis. J Neurol Neurosurg Psychiatry 1988;51:332-41. 68. Chandrasoma P, Apuzzo MLJ. Stereotactic brain biopsy. New York: Igaku-Shoin; 1989. 69. Blakeley JO, Grossman SA. Management of pineal region tumors. Curr Treat Options Oncol 2006;7:505-16. 70. Apuzzo ML, Chandrasoma PT, Zelman V, Giannotta SL, Weiss MH. Computed tomographic guidance stereotaxis

661

662

42 71.

72. 73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.


in the management of lesions of the third ventricular region. Neurosurgery 1984;15:502-8. Regis J, Bouillot P, Rouby-Volot F, Figarella-Branger D, Dufour H, Peragut JC. Pineal region tumors and the role of stereotactic biopsy: review of the mortality, morbidity, and diagnostic rates in 370 cases. Neurosurgery 1996;39:907-12; discussion 912-4. Villani R, Gaini SM, Tomei G. Follow-up study of brain stem tumors in children. Childs Brain 1975;1:126-35. Boviatsis EJ, Voumvourakis K, Goutas N, Kazdaglis K, Kittas C, Kelekis DA. Stereotactic biopsy of brain stem lesions. Minim Invasive Neurosurg 2001;44:226-9. Rajshekhar V, Chandy MJ. Computerized tomographyguided stereotactic surgery for brainstem masses: a riskbenefit analysis in 71 patients. J Neurosurg 1995;82: 976-81. Mathisen JR, Giunta F, Marini G, Backlund EO. Transcerebellar biopsy in the posterior fossa: 12 years experience. Surg Neurol 1987;28:100-4. Valdes-Gorcia J, Espinoza-Diaz DM, Paredes-Diaz E. Stereotactic biopsy of brain stem and posterior fossa lesions in children. Acta Neurochir (Wien) 1998;140:899-903. Boviatsis EJ, Kouyialis AT, Stranjalis G, Korfias S, Sakas DE. CT-guided stereotactic biopsies of brain stem lesions: personal experience and literature review. Neurol Sci 2003;24:97-102. Massager N, David P, Goldman S, Pirotte B, Wikler D, Salmon I, Nagy N, Brotchi J, Levivier M. Combined magnetic resonance imaging- and positron emission tomography-guided stereotactic biopsy in brainstem mass lesions: diagnostic yield in a series of 30 patients. J Neurosurg 2000;93:951-7. Patil AA, Leibrock LG, Kumar PP, Aarabi B. Stereotactic approach to skull-base lesions. Skull Base Surg 1991;1: 235-9. Rosenfeld JV, Wallace D, Klug GL, Danks A. Transnasal stereotactic biopsy of a clivus tumor. Technical note. J Neurosurg 1992;76:878-9. Patil AA. Transoral stereotactic biopsy of the second cervical vertebral body: case report with technical note. Neurosurgery 1989;25:999-1001; discussion 1001-2. Roujeau T, Machado G, Garnett MR, Miquel C, Puget S, Geoerger B, Grill J, Boddaert N, Di Rocco F, Zerah M, Sainte-Rose C. Stereotactic biopsy of diffuse pontine lesions in children. J Neurosurg 2007;107:1-4. Pincus DW, Richter EO, Yachnis AT, Bennett J, Bhatti MT, Smith A. Brainstem stereotactic biopsy sampling in children. J Neurosurg 2006;104:108-14.

84. Chappell ET, Guthrie BL, Orenstein J. The role of stereotactic biopsy in the management of HIV-related focal brain lesions. Neurosurgery 1992;30:825-9. 85. Luzzati R, Ferrari S, Nicolato A, Piovan E, Malena M, Merighi M, Morbin M, Gerosa M, Rizzuto N, Concia E. Stereotactic brain biopsy in human immunodeficiency virus-infected patients. Arch Intern Med 1996;156:565-8. 86. Skolasky RL, Dal Pan GJ, Olivi A, Lenz FA, Abrams RA, McArthur JC. HIV-associated primary CNS lymorbidity and utility of brain biopsy. J Neurol Sci 1999;163:32-8. 87. Kaakaji W, Barnett GH, Bernhard D, Warbel A, Valaitis K, Stamp S. Clinical and economic consequences of early discharge of patients following supratentorial stereotactic brain biopsy. J Neurosurg 2001;94:892-8. 88. Chakeres D, Slone W, Christoforidis G, Bourekas E. Real-time CT-guided spinal biopsy with a disposable stereotactic device: a technical note. AJNR Am J Neuroradiol 2002;23:605-8. 89. Brucher JM. Neuropathological diagnosis with stereotactic biopsies. Possibilities, difficulties and requirements. Acta Neurochir (Wien) 1993;124:37-9. 90. Whiting DM, Barnett GH, Estes ML, Sila CA, Rudick RA, Hassenbusch SJ, Lanzieri CF. Stereotactic biopsy of non-neoplastic lesions in adults. Cleve Clin J Med 1992;59:48-55. 91. Bhatia R, Tandon P, Misra NK. Inflammatory lesions of the basal ganglia and thalamus: review of twenty-one cases. Neurosurgery 1986;19:983-8. 92. Waltregny A, Maula AA, Brucher JM. Contribution of stereotactic brain biopsies to the diagnosis of presenile dementia. Stereotact Funct Neurosurg 1990;54‐55:409-12. 93. Truwit CL, Hall WA. Intraoperative magnetic resonance imaging-guided neurosurgery at 3-T. Neurosurgery 2006;58:ONS-338–45; discussion ONS-345–36. 94. Hall WA, Liu H, Truwit CL. Navigus trajectory guide. Neurosurgery 2000;46:502-4. 95. Reisner LA, King BW, Klein MD, Auner GW, Pandya AK. A prototype biosensor-integrated image-guided surgery system. Int J Med Robot 2007;3:82-8. 96. Scheld M, Whitely RJ, Marra CM. Infections of the Central Nervous System 3rd edition. Published by Lippincott Williams & Wilkins; 2004.

51 Stereotactic Approaches to the Brain Stem L. U. Zrinzo . D. G. T. Thomas

Background The stereotactic concept was first developed by Horsley and Clarke in 1905 to allow precise surgical navigation within the monkey cranium [1]. Applying these principles to clinical practice in 1947, Spiegel and Wycis ushered in a new era where surgical access to deep seated and highly eloquent brain areas could be achieved with minimal morbidity and mortality [2]. These stereotactic techniques were swiftly applied to pathologies and procedures involving the brain stem [3]. In the decades that followed, the introduction of cross-sectional imaging led to the first imagedirected stereotactic brain stem procedures [4,5]. With judicious application, stereotactic brainstem approaches remain a powerful clinical tool. Biopsy of brainstem masses provide a safe and reliable method of obtaining a histological diagnosis, aspiration of cysts, blood clots and abscesses provide therapeutic relief, and brainstem targets are increasingly being considered in functional procedures for pain control and movement disorders.

Role in the Management of Brainstem Lesions Providing a Histological Diagnosis Dorsally exophytic brainstem lesions are generally subjected to debulking by an open procedure [6]. However, stereotactic approaches should be considered for ventral midline exophytic lesions,

#


focal lesions within the substance of the brainstem and diffuse intrinsic lesions (> Figure 51-1). Image directed stereotactic brainstem biopsy was first reported by Gleason et al. in 1978 [5]. Since this first report, some authors have documented an above average complication rate from brainstem biopsy when compared to that from supratentorial sites [7–9]. However, proponents of this technique refute this and claim that stereotactic biopsy is safe and provides a histological diagnosis on which to base treatment protocols and provide prognostic information [9–11]. Indeed, a recent meta-analysis of stereotactic biopsy in 293 consecutive brainstem masses revealed one mortality (0.3%), 1% permanent and 4% temporary morbidity (including hemiplegia, cranial neuropathies, diplopia and lethargy) with a 94% diagnostic rate on first attempt rising to 96% with repeated sampling [12]. This high diagnostic yield and low morbidity rival those at other sites within the brain [10,13–17]. CT and MRI directed stereotactic targeting are both used in brainstem biopsies. However, MRI provides better anatomical definition and may allow subtle advantages in distinguishing areas of contrast enhancement that may minimize sampling errors. Initial concerns of image distortion affecting MRI targeting accuracy were initially resolved by using CT/MRI fusion protocols. However, this issue can now be addressed with strict attention to MRI quality control and correction of MR distortion [18]. The concerns over sampling error and possible erroneous diagnosis have led some authors to combined MRI and PET-guided

790

51

Stereotactic approaches to the brain stem

. Figure 51-1 Coronal T1 with contrast and axial T2 MR images of brainstem lesion. Biopsy revealed low grade glioma

stereotactic biopsy to improve the diagnostic yield of the representative sample [19,20]. After the introduction of MRI, it was suggested that the higher resolution afforded by this modality could reliably establish the diagnosis of high grade glioma when a diffuse pontine lesion was visualized, particularly in the pediatric age group [21–24]. This led a number of authors to recommend that such patients should receive radiotherapy empirically, without a tissue diagnosis and that stereotactic biopsy should be reserved for patients with focal brainstem lesions. However, a recent study has shown that sensitivity of MR imaging for diagnosis of pediatric brainstem tumors was high (0.94) but specificity was low (0.43) [25]. Another study reports on 24 children with a suspected diagnosis of high grade glioma on MR imaging; the diagnosis was overturned in two by stereotactic biopsy with a significant alteration in the management plan [26]. In summary, in the absence of a tissue diagnosis, patients with ‘‘false positive’’ MRI findings would receive inappropriate therapy directed against high grade lesions and appropriate treatment would be omitted. In many pediatric centers current practice is to treat diffuse pontine lesions

empirically as presumed high grade gliomas. However, tissue diagnosis is likely to play an increasingly important confirmatory role in the management of all pediatric brainstem tumors as new treatment protocols are developed [15]. It has been amply shown that in the adult population, histological diagnosis does not always agree with the preoperative assessment based on clinical presentation and radiological findings. Unexpected findings can occur in over 15% of cases [10,17]. Even groups supplementing diagnostic imaging with PET find discordance between the predictive value of imaging in the detection of malignancy and histological diagnosis and document a concordance of 63% for MRI and 79% for combined MRI and PET [20]. The literature contains a long list of brainstem pathologies where the correct diagnosis was confirmed only after stereotactic sampling. These include: low grade glioma, metastasis, lymphoma, ependymoma, gangliocytoma, pineoblastoma, epidermoid cyst, epidermal cyst, gangliosidoses, angioma, granuloma, vasculitis, leucoencephalopathy, radionecrosis, demyelination, encephalitis, amoebiasis, tuberculoma and pyogenic abscess [10,16,17,20,27–32]. In a number of these reports


51

. Figure 51-2 Coronal FLAIR and axial T1 with contrast MR images of brainstem lesion. Biopsy excluded tumor and showed granulomatous inflammation

high grade glioma was the top radiological differential and biopsy made a significant impact on prognosis and patient management. Given the small but not insignificant risks of stereotactic biopsy versus the risk of an incorrect diagnosis and management without biopsy, it is perhaps not surprising that, even in the adult literature, there is debate as to the best course of action when presented with a brainstem lesion. Samadani et al. propose a decision analysis method based upon the probability of an incorrect radiological diagnosis leading to suboptimal empirical treatment, weighed against the probability of stereotactic biopsy leading to a suboptimal outcome (defined as a surgical complication, non-diagnostic sample or sampling error) [13]. The complex interplay of probabilities ensures that one cannot be dogmatic about the role of stereotactic biopsy in brainstem. However, the wide variety of adult brainstem pathologies, the relative rarity of primary gliomas in this age group and our personal experience of a low complication rate, leads the present authors to advise stereotactic biopsy of brainstem lesions unless there are surgical contraindications (> Figure 51-2).

Tumor Cyst Aspiration Aspiration of cysts associated with brainstem tumors can temporarily alleviate neurological symptoms [16,33–35]. Placement of an Ommaya reservoir may allow repeated aspiration of recurring cysts without further surgical intervention. Additionally, it allows for instillation of intracavitary instillation of radioisotopes should repeated aspirations fail to achieve cyst control [35].

Brachytherapy Interstitial brachytherapy of brainstem lesions has been reported in the literature and is held by some authors to be an alternative or an adjunct to external beam radiotherapy [36–39]. Stereotactic interstitial therapy can be performed will minimal surgical complications and it is thought that I-125 low-dose rate implants may limit the risk of radiation-induced necrosis. Although radiological response to this therapy has been documented, the role of interstitial brachytherapy for these lesions remains undefined and its use in the neurosurgical community remains limited [38].

791

792

51


Stereotactic Aspiration of Brainstem Abscess Pyogenic abscess of the brainstem is rare and prior to 1974 was invariably fatal. The pons is the most frequently affected region. Haematogenous and direct spread from adjacent structures are the commonest causes. However, in over a third of cases no source is identified [40–42]. Stereotactic aspiration allows confirmation of the diagnosis since this is not always obvious by means of CT or MR imaging alone [28]. Drainage of pus, reduction in mass effect and the possible identification of the offending organism together with its antibiotic sensitivity, are additional advantages and may provide an informed choice of antimicrobial agents. Indeed, there is a suggestion that stereotactic aspiration of brainstem abscess in combination with antimicrobial treatment may be associated with a superior functional outcome that medical treatment with or without open surgical drainage [28]. Both transfrontal and suboccipital transcerebellar approaches have been described in treating brain abscesses [28,43–49]. Avoiding ventricular penetration when draining an abscess assumes even greater importance as this may not only lower the risk of hemorrhage but may also be important in avoiding ventriculitis.

Role in Functional Neurosurgery Brainstem structures have been targeted in the management of chronic pain with procedures including stereotactic mesencephalic tractotomy and deep brain stimulation of the periaqueductal gray [50–52]. The brainstem is also a potential target in the treatment of movement disorders with the pedunculopontine nucleus (PPN) holding promise as a potential target for deep brain stimulation (DBS) in parkinsonian patients with gait disturbance and postural instability refractory to other treatment modalities [53–55]. This elongated

neuronal collection in the lateral pontine and mesencephalic tegmental reticular zones is unfamiliar territory to most functional neurosurgeons [56]. MRI guided stereotactic targeting is an important technique in identification and targeting of appropriate anatomical structures in Functional Neurosurgery [57–59]. MRI protocols that allow localization of the PPN will be relevant to groups evaluating the clinical role of PPN DBS (> Figure 51-3) [60].

Technical Considerations Surgery can be performed under local or general anesthesia. Numerous frames have been used in stereotactic procedures on the brainstem including the Leksell, Riechert and BRW/CRW frames. CT and MRI are the main modalities used in image directed targeting with PET also being used in an attempt to improve sampling accuracy of diagnostic biopsies. In general MRI provides superior resolution in the posterior fossa and, when possible, is the preferred imaging modality for the authors. Modern MRI equipment, strict quality assurance and the use of frames with fiducials close to the skull (e.g., Leksell frame) reduce the effect of geometric distortion [62]. Three-dimensional planning software now allows an accurate assessment of both target and trajectory prior to surgery. Transgression of multiple pial or ependymal surfaces places the associated vascular structures at risk. Therefore, a trajectory confined to brain parenchyma would, theoretically, reduce the risk of hemorrhage. CSF loss from a penetrated ventricular system may result in increased brain shift and our own research has suggested that targeting accuracy may be reduced with trajectories that pass close to or through the ventricle walls (as yet unpublished results). A blunt round-tipped probe advanced slowly to the target without precession allows minimal disruption of brain tissues as they are displaced around the probe rather than being transected.


51

. Figure 51-3 Axial MRI images through the inferior colliculi reformatted in a plane perpendicular to the midline of the fourth ventricular floor. The T1 image on the left shows excellent contrast between brain and CSF but little internal tissue contrast. The middle image is a Proton Density MRI where gray matter appears hyperintense (bright); white matter appears hypointense (dark). Identifiable anatomical structures are labeled on the right. CTT, central tegmental tract; DSCP, decussation of the superior cerebellar peduncles; LL, lateral lemniscus; ML, medial lemniscus; PAG, periaqueductal gray; PPN, pedunculopontine nucleus; PT, pyramidal tract; SN, substantia nigra; STT, spinothalamic tract. Reproduce with kind permission from Oxford University Press [61]

An ipsilateral transfrontal entry point provides access to the mesencephalon and midline regions of the pons [17]. A contralateral transfrontal entry point has also been described that allows access to more laterally placed pontine lesions without having to traverse the ventricular system [63]. Both approaches allow the patient to remain supine during surgery, in a similar position to that in which images are traditionally acquired thus preventing error due to positional brain shift. In this region, a burrhole can be placed without painful muscle dissection and twist drill holes can be planned to avoid sulci. The transtentorial route has been virtually abandoned and is not employed by the authors because of the increased risk of hemorrhage and trajectory deviation by traversing the tentorium and several other anatomical surfaces. The suboccipital transcerebellar approach is often used to access brainstem lesions [4,26,27,64–67]. Care must be taken to ensure that the frame is placed low enough to allow the lesion to be visualized and to physically allow the required trajectory with a particular frame [66]. Semi-recumbent, lateral and prone positions have been described to provide access, some of which may limit the possibility of awake

surgery. This approach provides the shortest distance to the desired target [66,67]. Twist drill holes provide the convenience of minimal tissue disruption and at this site minimize pain associated with more excessive muscle stripping. However, one is then committed to that particular trajectory and access for haemostatic control of the dura and pia is restricted. With supratentorial entry points, planning software can minimize this risk by avoiding sulci. However, such planning over the folia of the cerebellar surface may not be possible. Nonetheless, there is no evidence that the risk of hemorrhage depends upon whether entry is via a burrhole or a twist drill hole [8].

Concluding Remarks Stereotactic procedures provide surgical access to deep seated eloquent brain areas with minimum morbidity and mortality and are thus ideally suited to surgery of the brainstem. Judicious clinical application and meticulous surgical planning and technique will exploit the benefits and minimize the risks of this powerful surgical tool.

793

794

51


References 1. Horsley V, Clarke RH. The structure and functions of the cerebellum examined by a new method. Brain 1905;28 (1):1-98. 2. Spiegel EA, et al. Stereotaxic apparatus for operations on the human brain. Science 1947;106(2754):349-50. 3. Ward AA, Jr. Trends in the application of stereotaxy to the brain stem. Clin Neurosurg 1958;6:223-39. 4. Mathisen JR, et al. Transcerebellar biopsy in the posterior fossa: 12 years experience. Surg Neurol 1987;28(2):100-4. 5. Gleason CA, Wise BL, Feinstein B. Stereotactic localization (with computerized tomographic scanning), biopsy, and radiofrequency treatment of deep brain lesions. Neurosurgery 1978;2(3):217-22. 6. Epstein F, Wisoff JH. Surgical management of brain stem tumors of childhood and adolescence. Neurosurg Clin N Am 1990;1(1):111-21. 7. Frank F, et al. Stereotactic biopsy and treatment of brain stem lesions: combined study of 33 cases (BolognaMarseille). Acta Neurochir Suppl (Wien) 1988;42: 177-81. 8. Grossman R, et al. Haemorrhagic complications and the incidence of asymptomatic bleeding associated with stereotactic brain biopsies. Acta Neurochir (Wien) 2005;147 (6):627-31; discussion 631. 9. Kondziolka D, Lunsford LD. Results and expectations with image-integrated brainstem stereotactic biopsy. Surg Neurol 1995;43(6):558-62. 10. Kratimenos GP, Thomas DG. The role of image-directed biopsy in the diagnosis and management of brainstem lesions. Br J Neurosurg 1993;7(2):155-64. 11. Thomas DG, et al. Computer-directed stereotactic biopsy of intrinsic brain stem lesions. Br J Neurosurg 1988; 2(2):235-40. 12. Samadani U, Judy KD. Stereotactic brainstem biopsy is indicated for the diagnosis of a vast array of brainstem pathology. Stereotact Funct Neurosurg 2003;81(1–4):5-9. 13. Samadani U, et al. Stereotactic biopsy of brain stem masses: decision analysis and literature review. Surg Neurol 2006;66(5):484-90; discussion 491. 14. Chico-Ponce de Leon F, et al. Stereotactically-guided biopsies of brainstem tumors. Childs Nerv Syst 2003;19 (5–6):305-10. 15. Pincus DW, et al. Brainstem stereotactic biopsy sampling in children. J Neurosurg 2006;104(2 Suppl):108-14. 16. Rajshekhar V, Chandy MJ. Computerized tomographyguided stereotactic surgery for brainstem masses: a riskbenefit analysis in 71 patients. J Neurosurg 1995;82 (6):976-81. 17. Kratimenos GP, et al. Image directed stereotactic surgery for brain stem lesions. Acta Neurochir (Wien) 1992;116 (2–4):164-70.

18. Menuel C, et al. Characterization and correction of distortions in stereotactic magnetic resonance imaging for bilateral subthalamic stimulation in Parkinson disease. J Neurosurg 2005;103(2):256-66. 19. Aker FV, et al. Accuracy and diagnostic yield of stereotactic biopsy in the diagnosis of brain masses: comparison of results of biopsy and resected surgical specimens. Neuropathology 2005;25(3):207-13. 20. Massager N, et al. Combined magnetic resonance imaging- and positron emission tomography-guided stereotactic biopsy in brainstem mass lesions: diagnostic yield in a series of 30 patients. J Neurosurg 2000;93(6):951-7. 21. Albright AL, et al. Magnetic resonance scans should replace biopsies for the diagnosis of diffuse brain stem gliomas: a report from the Children’s Cancer Group. Neurosurgery 1993;33(6):1026-9; discussion 1029–30. 22. Albright AL. Tumors of the pons. Neurosurg Clin N Am 1993;4(3):529-36. 23. Epstein F, McCleary EL. Intrinsic brain-stem tumors of childhood: surgical indications. J Neurosurg 1986;64 (1):11-15. 24. Cartmill M, Punt J. Diffuse brain stem glioma. A review of stereotactic biopsies. Childs Nerv Syst 1999;15 (5):235-7; discussion 238. 25. Schumacher M, et al. Magnetic resonance imaging compared with biopsy in the diagnosis of brainstem diseases of childhood: a multicenter review. J Neurosurg 2007; 106(2 Suppl):111-19. 26. Roujeau T, et al. Stereotactic biopsy of diffuse pontine lesions in children. J Neurosurg 2007;107(1 Suppl):1-4. 27. Coffey RJ, Lunsford LD. Stereotactic surgery for mass lesions of the midbrain and pons. Neurosurgery 1985;17 (1):12-18. 28. Bavetta S, et al. Brainstem abscess: preoperative MRI appearance and survival following stereotactic aspiration. J Neurosurg Sci 1996;40(2):139-43. 29. Nassogne MC, et al. Unusual presentation of GM2 gangliosidosis mimicking a brain stem tumor in a 3-year-old girl. AJNR Am J Neuroradiol 2003;24(5):840-2. 30. Rajshekhar V, Chandy MJ. Tuberculomas presenting as isolated intrinsic brain stem masses. Br J Neurosurg 1997;11(2):127-33. 31. Valdes-Gorcia J, Espinoza-Diaz DM, Paredes-Diaz E. Stereotactic biopsy of brain stem and posterior fossa lesions in children. Acta Neurochir (Wien) 1998;140 (9):899-903. 32. Lowichik A, et al. Leptomyxid amebic meningoencephalitis mimicking brain stem glioma. AJNR Am J Neuroradiol 1995;16(4 Suppl):926-9. 33. Akhtar N, Lewis TT. Management of midbrain cyst with repeated CT guided aspiration. Neuroradiology 1994;36 (8):642-3. 34. Hood TW, McKeever PE. Stereotactic management of cystic gliomas of the brain stem. Neurosurgery 1989;24 (3):373-8.


35. Giovanini MA, Mickle JP. Long-term access to cystic brain stem lesions using the Ommaya reservoir: technical case report. Neurosurgery 1996;39(2):404-7. 36. Mundinger F, et al. Long-term outcome of 89 low-grade brain-stem gliomas after interstitial radiation therapy. J Neurosurg 1991;75(5):740-6. 37. Matsumoto K, et al. Stereotactic brachytherapy for a cystic metastatic brain tumor in the midbrain. Case report. J Neurosurg 1998;88(1):141-4. 38. Chuba PJ, et al. Permanent I-125 brain stem implants in children. Childs Nerv Syst 1998;14(10):570-7. 39. Julow J, et al. Iodine-125 brachytherapy of brain stem tumors. Strahlenther Onkol 2004;180(7):449-54. 40. Hall WA. Infectious lesions of the brain stem. Neurosurg Clin N Am 1993;4(3):543-51. 41. VanGilder JC, Allen WE, III, Lesser RA. Pontine abscess: survival following surgical drainage. Case report. J Neurosurg 1974;40(3):386-90. 42. Jamjoom ZA. Solitary brainstem abscess successfully treated by microsurgical aspiration. Br J Neurosurg 1992; 6(3):249-53. 43. Nakajima H, et al. Successful treatment of brainstem abscess with stereotactic aspiration. Surg Neurol 1999;52 (5):445-8. 44. Nauta HJ, et al. Brain stem abscess managed with computed tomography-guided stereotactic aspiration. Neurosurgery 1987;20(3):476-80. 45. Fuentes S, et al. Management of brain stem abscess. Br J Neurosurg 2001;15(1):57-62. 46. Fujino H, et al. Cure of a man with solitary abscess of the brain-stem. J Neurol 1990;237(4):265-6. 47. Kalarostaghi AH, et al. Polymicrobial brain stem abscess due to Streptococcus anginosus and Actinomyces species. J Clin Neurosci 1999;6(5):415-18. 48. Rajshekhar V, Chandy MJ. Successful stereotactic management of a large cardiogenic brain stem abscess. Neurosurgery 1994;34(2):368-71; discussion 371. 49. Rossitch E, Jr, et al. The use of computed tomographyguided stereotactic techniques in the treatment of brain stem abscesses. Clin Neurol Neurosurg 1988;90 (4):365-8. 50. Amano K, et al. Stereotactic mesencephalotomy for pain relief. A plea for stereotactic surgery. Stereotact Funct Neurosurg 1992;59(1–4):25-32. 51. Fountas KN, et al. MR-based stereotactic mesencephalic tractotomy. Stereotact Funct Neurosurg 2004; 82(5–6):230-4. 52. Young RF, Chambi VI. Pain relief by electrical stimulation of the periaqueductal and periventricular gray matter. Evidence for a non-opioid mechanism. J Neurosurg 1987;66(3):364-71. 53. Stefani A, et al. Bilateral deep brain stimulation of the pedunculopontine and subthalamic nuclei in severe Parkinson’s disease. Brain 2007;130(Pt 6):1596-607.

51

54. Nandi D, et al. Deep brain stimulation of the pedunculopontine region in the normal non-human primate. J Clin Neurosci 2002;9(2):170-4. 55. Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine nucleus for Parkinson’s disease. Neuroreport 2005;16(17):1883-7. 56. Zrinzo L, Zrinzo LV, Hariz M. The pedunculopontine and peripeduncular nuclei: a tale of two structures. Brain 2007;130(Pt 6):e73; author reply e74. 57. Hirabayashi H, Tengvar M, Hariz MI. Stereotactic imaging of the pallidal target. Mov Disord 2002;17 Suppl 3: S130-4. 58. Hariz MI, et al. A quick and universal method for stereotactic visualization of the subthalamic nucleus before and after implantation of deep brain stimulation electrodes. Stereotact Funct Neurosurg 2003;80 (1–4):96-101. 59. Vayssiere N, et al. Comparison of atlas- and magnetic resonance imaging-based stereotactic targeting of the globus pallidus internus in the performance of deep brain stimulation for treatment of dystonia. J Neurosurg 2002;96(4):673-9. 60. Muthusamy KA, et al. Connectivity of the human pedunculopontine nucleus region and diffusion tensor imaging in surgical targeting. J Neurosurg 2007;107 (4):814-20. 61. Zrinzo L, et al. Stereotactic localization of the human pedunculopontine nucleus: atlas-based coordinates and validation of a magnetic resonance imaging protocol for direct localization. Brain 2008;131:1588-98. 62. Yu C, et al. A phantom study of the geometric accuracy of computed tomographic and magnetic resonance imaging stereotactic localization with the Leksell stereotactic system. Neurosurgery 2001;48(5):1092-8; discussion 1098–9. 63. Amundson EW, McGirt MJ, Olivi A. A contralateral, transfrontal, extraventricular approach to stereotactic brainstem biopsy procedures. Technical note. J Neurosurg 2005;102(3):565-70. 64. Abernathey CD, Camacho A, Kelly PJ. Stereotaxic suboccipital transcerebellar biopsy of pontine mass lesions. J Neurosurg 1989;70(2):195-200. 65. Guthrie BL, Steinberg GK, Adler JR. Posterior fossa stereotaxic biopsy using the Brown-Roberts-Wells stereotaxic system. Technical note. J Neurosurg 1989;70 (4):649-52. 66. Neal JH, Van Norman AS. Transcerebellar biopsy of posterior fossa lesions using the Leksell gamma model stereotactic frame. Neurosurgery 1993;32(3):473-4; discussion 474–5. 67. Spiegelmann R, Friedman WA. Stereotactic suboccipital transcerebellar biopsy under local anesthesia using the Cosman-Roberts-Wells frame. Technical note. J Neurosurg 1991;75(3):486-8.

795

53 Technical Aspects of Image-Guided Neuroendoscopy J. D. Caird . J. M. Drake

Neuroendoscopy has become an established neurosurgical technique permitting neurosurgeons to address deep-seated intracranial pathology under direct vision. The advent of image-guided surgery has enhanced the potential of neuroendoscopy beyond the boundary of direct vision, thus minimizing trauma to the brain. Image-guided neuroendoscopy does not replace frame-based stereotactic surgery where access to deep parenchymal structures for the purposes of biopsy or electrode placement are concerned. The experience of image-guided neuroendoscopy at our institution has been intracranial using a rigid, zero-degree Aesculap neuroendoscope in combination with Brainlab Neuronavigation.

Since Horsley and Clarke reported their use of a stereotactic frame device to study the cerebellum in monkeys in 1906 there have been innumerable advances in stereotactic surgery. The earliest frame-based devices utilized boney reference points; with the advent of air ventriculography and cerebral angiography, image guided stereotaxy using intracranial anatomic reference points was feasible. Amongst the many contributors to the field, Leksell is notable for linking stereotactic surgery with modern imaging modalities [3,4]. Frameless stereotaxy developed as a concept during the 1980s but has only become widespread in neurosurgical practice since the mid 1990s. Its application to neuroendoscopy is very recent and continues to progress [5–7].

Background Dandy pioneered the field of neuroendoscopy in the early 1920s and coined the term ‘‘ventriculoscope’’ in a brief article in which he described the use of an endoscope to inspect the lateral ventricles of two children with hydrocephalus [1]. Limitations at that time with regard to compatible endoscopic instrumentation restricted the potential use of the endoscope as a working surgical instrument. Present-day neuroendoscopy has been made possible through successful attempts at miniaturization of instrumentation. Fukushima is credited with pioneering the technique of neurendoscopic biopsy of intraventricular tumor in 1978 [2]. Since that time, many refinements in instrumentation have taken place in an effort to optimize the working neuroendoscope. #


Surgical Technique of ImageGuided Neuroendoscopy Preoperative volumetric MRI or CT sequences are a pre-requisite for image-guided surgery with CT being useful to define sphenoid bone and air sinus anatomy. Whilst we have had some experience with transphenoidal neuroendoscopic surgery, the majority of our cases are concerned with the management of hydrocephalus, intraventricular tumors and cysts. The surgical approach for neuroendoscopy is clearly dictated by the location of the intraventricular pathology. A frontal entry point for ventricular access placed slightly anterior to the coronal suture and approximately 3 cm lateral to the midline is considered standard practice. Where two

808

53

Technical aspects of image-guided neuroendoscopy

or more procedures such as endoscopic third ventriculostomy (ETV), septostomy, tumor biopsy, or shunt insertion are considered, two or more trajectories may be required; we have found it practical in this setting to fashion a small craniotomy through a curvilinear skin incision, with two separate dural incisions to accommodate the endoscope for each approach. Simultaneous use of two trajectories using both rigid and steerable neuroendoscopes has been described in the ‘‘multi-axial’’ approach to the pineal region and floor of the third ventricle [8]. As with shunt surgery, neuroendoscopy demands meticulous attention to detail, and should be carried out in a skilled and expeditious fashion. Body wash and shampoo the night before and again before surgery with an antiseptic solution such as Chlorhexidine has been the practice at our institution, but is not mandatory. In the operating room the patient is positioned under general endotracheal anesthesia. The head is generally positioned in the neutral position in a pinned head clamp with some neck flexion or table tilt ‘‘head up’’ to prevent excessive CSF losses unless shunt insertion during the same procedure is anticipated, in which case the head needs to be rotated to the appropriate side. For babies and children with a patent fontanelle or fragile cranium, including those with osteogenesis imperfecta, a pinned head clamp may be undesirable in which case pre- and intraoperative ultrasound may well suffice. Prophylactic antibiotic administration is highly recommended; we use 30 mg kg 1 of cefazolin at induction. The image-guidance reference arc is applied to the head clamp and surface registration obtained with either the pointer wand or with laser registration. Using the pointer and off-set application, the surgical trajectories required may be planned. The midline and coronal suture landmarks, and skin incisions are marked on the patient, bearing in mind that a shunt valve should not be directly under a wound if possible. The hair is then clipped, not shaved,

to assist with wound closure and dressing application. The eyes are taped closed after registration and the skin is meticulously prepped with an iodine or chlorhexidine solution. Disposable, adhesive drapes are used to cover the patient and the operating table entirely except for a small area of skin required for incision. An iodineimpregnated transparent adhesive is applied to the prepped skin. At this juncture, it is prudent to establish that the irrigation solution, either Ringer’s lactate or normal saline, is warmed to 37 C and that the endoscope and light-source are fully operational. The video camera is passed through a sterile plastic sheath and attached to the endoscope. With the light-source connected, white-balancing can be performed. It is important to check that the scope is correctly orientated and focused, for which purpose a suture packet can be used. It is useful at this point to ensure that the endoscopic instruments (monopolar/bipolar diathermy, biopsy forceps, alligator forceps, balloon catheters) are functional prior to ventricular access. Both the endoscope and the image-guidance screens should be positioned for easy viewing to avoid having to turn one’s head. For navigation purposes, a reference arc is clamped to the endoscope side irrigation port (to prevent crimping of the working channel) and registered with the image-guidance device (> Figure 53-1). The skin incision and burrhole or minicraniotomy are fashioned and the dura coagulated at the optimal entry point as determined with image-guidance. A cruciform durotomy sufficient to permit passage of the endoscope outer casing, generally 6 mm in diameter, is made and the pia and cortex are coagulated and incised with a size 15 or 11 scalpel blade. Once the ventricle has been entered, the obturator can be withdrawn and the endoscope introduced by the surgeon, with the operative assistant holding the neuroendoscope motionless. It is important to recheck the orientation


. Figure 53-1 Instrument calibration device with Aesculap endoscope barrel at registration. Note reference arc clamped to endoscope irrigation port to prevent crimping of working channel

of the scope at this stage as it may have altered in the setting up process. Prior to entering the ventricle, the anticipated depth of cortex to be traversed can be measured with the navigation ‘‘off-set’’ function and the endoscope trajectory can be followed with ultrasound (> Figure 53-2). A sensation of slight resistance is usually encountered at the ependyma followed by a slight ‘‘give.’’ Once the endoscope tip has been confirmed visually to be intraventricular, the irrigation entry port should be opened, ensuring the irrigation egress port is patent to avoid a net increase in CSF volume. The operative assistant can now ‘‘drive’’ the scope within the ventricle, the first step being to identify a constant landmark such as the foramen of Monro by following choroid plexus antero-medially and to reference the foramen with the image-guidance. The surgeon is now in a position to perform the desired image-guided endoscopic procedure; the use of a rigid endoscope holding device permits the operative assistant to partake in the endoscopic instrumentation. Over-head monitors displaying the endoscopic image and image-guidance sideby-side are strongly advocated to avoid having to turn one’s head to see different screens at this delicate stage (> Figure 53-3 and > Figure 53-4).

53

. Figure 53-2 Trajectory confirmed after drilling of burr-hole. Depth of cortical mantle to be traversed can be calculated with ‘‘off-set’’ mode and confirmed with ultrasound

Where ETV is contemplated, it should be done prior to any further procedure such as tumor biopsy to avoid bloody CSF obscuring the view of the floor of the third ventricle. It has been our practice when performing ETV, to very carefully perforate the floor of the third ventricle using the metal stylet from a 35 cm ventriculostomy catheter. Once perforated, the stoma can be widened by passing a closed alligator forceps into the stoma and gently opening it, or by passing a Fogarty balloon catheter and slowly dilating the balloon within the stoma. Where tumor biopsy is concerned, consideration must be given as to whether the lesion in truly intraventricular or subependymal; the latter may be occult on ventriculoscopy and it is here that image-guidance can be indispensable. It is preferable to avoid diathermy at the biopsy site which may cause thermal artifact in the pathology specimen. Furthermore, where a choroid plexus tumor is suspected on preoperative imaging, craniotomy may be more prudent to avoid uncontrollable intraventricular hemorrhage. On completion of the neuroendoscopy, the scope is carefully withdrawn, irrigating

809

810

53


. Figure 53-3 Single split screen showing multi-axial neuronavigation and endoscope view for septostomy

throughout, until free of the cortex. The corticotomy site may be plugged with a piece of hemostatic foam to prevent further CSF loss or oozing from the cortex. In doing so, one must be mindful that such material may migrate and has been known to cause failure of an ETV by plugging the stoma [9]. If sufficient intraventricular hemorrhage has occurred, the scope tract and ventricle may be examined with ultrasound prior to wound closure. If, following hemostasis, the CSF is bloodstained at the end of the procedure it may be preferable to leave an external ventricular drain in situ. The scalp is closed in two layers, using absorbable sutures for galea and an absorbable monofilament suture for skin.

Applications of Image-Guided Neuroendoscopy Neuroendoscopy has become an effective tool in the treatment of obstructive hydrocephalus,

intracranial cysts, and a variety of intraventricular and periventricular mass lesions [10,11]. The exponential improvement in the technology of frameless and wireless image-guidance systems has all but superceded the requirement for frame-based neuroendoscopy. Even in the case of small ventricles, intraoperative ultrasound permits real-time definition of the ventricular system and image-guidance can facilitate ventricular cannulation [12]. Image-guidance excels in the planning stage of the surgery as once the patient is anaesthetized, the neuroendoscope trajectory can be delineated and the burrhole position marked out on the cranium. This is particularly crucial for cases of intraventricular tumors where the ventricles may be distorted and perhaps both septum pellucidotomy and tumor biopsy are envisaged; optimal burrhole position is frequently more anterior and lateral than one would have anticipated. Image-guided neuroendoscopy has been utilized for a variety of intracranial conditions,


53

. Figure 53-4 Ceiling-mounted display allows surgeons to focus on their instrumentation at critical moments without having to turn their heads

including pathology of the lateral, third, and fourth ventricles, transnasal approaches to the sellar and parasellar lesions, arachnoid cysts of the anterior, middle and posterior fossa, and vascular lesions including cavernous malformations and cerebral aneurysms. At our institution, we utilize neuroendoscopy predominantly for endoscopic third ventriculostomy, intraventricular tumor biopsy, cyst fenestration and the management of complex hydrocephalus where septum pellucidotomy and/or shunt placement is indicated. Furthermore, it has proven useful in the conversion of long-term shunted hydrocephalus to ETV where extraction of an old ventricular catheter may be imprudent without direct vision and diathermy. We do not routinely employ image-guidance where ETV alone is considered. The accuracy of neuroendoscopic biopsy of pineal region tumors and intraventricular germinomas has been reported as high as 89 and 100% respectively [13,14]. It has been suggested that the higher yield from germinomas compared

with pineal tumors of various sub-types has been due to the fact that germinomas are generally exophytic within the ventricle, whereas gliomas are frequently subependymal; we postulate that the concomitant use of image-guidance for these cases could improve the histologic yield from subependymal lesions where obscuration of direct vision by bloody CSF following the first biopsy sample does not preclude further image-guided tissue sampling. A retrospective analysis of endoscopic tumor biopsy at our institution suggests an acceptable degree of histologic certainty in diagnosis is achieved in only 69% of cases, with 31% of biopsies yielding problematic or uninterpretable samples [15].

Pitfalls and Complications of Neuroendoscopy Rigid skull fixation during image-guided surgery is preferable but not an absolute necessity; there

811

812

53


may be occasions where intraoperative adjustment of head position is desirable such as during shunt placement. Pinned head clamps are undesirable with very young children due to the risk of skull fracture or distortion and epidural hematoma formation; trans-fontanelle or trans-cortical ultrasonography is invaluable in infants and young children undergoing neuroendoscopy. Alternatively, non-pinned head holders are available in these circumstances. Where the patient’s head is placed in a nonneutral position including for septostomy and shunt placement, particular care must be taken to ensure that the neuroendoscope is correctly orientated (‘‘up is up’’) as the usual intraventricular landmarks are not in the neutral anatomical position. Furthermore, attachment of the imageguidance reference arc to the barrel of the neuroendoscope may distort the barrel making passage of instruments through the working channel difficult. At our institution, we have found that the reference arc can be clamped to one of the irrigation ports of the neuroendoscope barrel (Aesculap) without hindering passage of either the camera or the working instruments. Neuroendoscopy incurs a risk of intraventicular hemorrhage, particularly where biopsy of a mass is undertaken. Similarly, the dissemination of tumor cells along the surgical tract is a rare but recognized complication of neuroendoscopic biopsy [16]. Where hemorrhage in the ventricle or from the cortical tract is encountered, copious saline irrigation or the endoscopic bipolar device may provide effective hemostasis. Alternatively, local tamponade with an inflated Fogarty balloon catheter may be employed. Where ETV is anticipated with or without fenestration or biopsy procedure, the likelihood of successful CSF drainage diminishes with younger age, with failure in infants younger than 1 month of age reaching 75%. One and 5 year success rates (i.e., not requiring further CSF diversion) are reported as 65 and 52% respectively.

Complications following ETV including CSF leak, meningitis, hemorrhage, seizure, injury to cerebrum/cranial nerves has collectively been reported at 13.6%, with isolated reports of precipitous decline and death from ETV failure [17]. A mathematical model to determine the likelihood of ETV success has been developed and is being prepared for publication [18].

Ultrasound Directed Neuroendoscopy Intraoperative ultrasound is considered invaluable in our experience and is employed routinely in our neuroendoscopic cases. It allows real-time inspection of the ventricular system and the endoscope tract following withdrawal of the endoscope where bleeding may have been a concern. It also permits reliable placement of a ventricular catheter, even through the stoma of a septum pellucidotomy. It is employed where image-guidance is unavailable or fails, or where rigid head fixation is undesirable. We also use it in conjunction with neuronavigation if it is felt that excess CSF egress may render image registration inaccurate. In infants, ultrasound may be used to image the ventricular system via the anterior fontanelle whereas for older children and adults, its use generally requires a larger than standard 16 mm burrhole. This is generally fashioned with a highspeed drill to allow placement of the burrhole probe flush against the dura.

Future Dimensions and Conclusion Although image-guided neuroendoscopy is currently limited torigidneuroendoscopes,technologic advances will, no doubt, permit image-guided tracking of flexible neuroendoscopes. Intraoperative MRI or CT may permit real-time updating


of registration data to allow for brain-shift during surgery; 3-D ultrasound has already been developed with this purpose in mind. Further developments in integrated monitor display units may decrease the volume of equipment required in the operating room and permit the surgeons to concentrate only on one screen displaying both navigation and endoscope views. Image-guided neuronavigation and intraoperative ultrasonography have exponentially expanded the horizon of neuroendoscopic surgery, offering surgeons a greater sense of security in the pre and intraoperative decision making processes in the treatment of complex neurosurgical pathology. Useful company websites for neuroendoscopic, neuronavitgation and ultrasound devices. www.medtronic.com www.brainlab.com www.stryker.com www.codman.com www.aesculapusa.com www.karlstorz.com www.aloka.com

References 1. Dandy WE. An operative procedure for hydrocephalus. Bull Johns Hopkins Hosp. 1922;33:189-90. 2. Fukushima T. Endoscopic biopsy of intraventricular tumors with the use of a ventriculofiberscope. Neurosurg 1978;2:110-3. 3. Clarke RH, Horsley V. On a method of investigating the deep ganglia and tracts of the central nervous system (cerebellum). Br Med J. 1906;1799–1800. 4. Leksell L, Jernberg B. Stereotaxis and tomography. A technical note. Acta Neurochir Wien. 1980;52:1-7.

53

5. Watanabe E, Watanabe T, Manaka S, Mayanaqi Y, Takakura K. Three dimensional digitizer (neuronavigator): new equipment for computed tomographyguided stereotaxic surgery. Surg Neurol. 1987;27:543-7. 6. Golfinos J, Fitzpatrick B, Smith L, Spetzler R. Clinical use of a frameless stereotactic arm: results in 325 cases. J Neurosurg. 1995;83:197-205. 7. Mayberg M, La Presto E, Cunningham E. Image-guided endoscopy: description of technique and potential applications. Neurosurg Focus. 2005;19(1):E10. 8. Oi S, Kamio M, Joki T, Abe T. Neuroendoscopic anatomy and surgery in pineal region tumors. J Neuro-Onc. 2001;54:227-86. 9. Edwards R, Dirks P. Gelfoam obstruction of endoscopic third ventriculostomy. Case illustration. J Neurosurg. 2006;105:154. 10. Abbott R. The endoscopic management of arachnoidal cysts. Neurosurg Clin N Am. 2004;15:9-17. 11. Hellwig D, Grotenhuis J, Tirakotai W, Riegel T, Schulte D, Bauer B, Bertalanffy H. Endoscopic third ventriculostomy for obstructive hydrocephalus. Neurosurg Rev. 2005;28:1-34. 12. Souweidane M. Endoscopic surgery for intraventricular brain tumors in patients without hydrocephalus. Neurosurgery 2005;57:312-8. 13. Yamini B, Refai D, Rubin G, Frim D. Initial endoscopic management of pineal region tumors and associated hydrocephalus: clinical series and literature review. J Neurosurg. 2004;100:437-41. 14. Shono T, Natori Y, Morioka T, et al. Results of long-term follow-up after neuroendoscopic biopsy and third ventriculostomy in patients with intracranial germinomas. J Neurosurg. 2007;107:193-8. 15. Depreitere B, Dasi N, Rutka J, Dirks P, Drake J. Endoscopic biopsy for intraventricular tumors in children. J Neurosurg:Pediatrics 2007;106:340-6. 16. Haw C, Steinbok P. Ventriculoscope tract recurrence after endoscopic biopsy of pineal germinoma. Paediatric Neurosurg. 2001;34:215-7. 17. Drake JM. Endoscopic third ventriculostomy in pediatric patients: the Canadian experience. Neurosurgery 2007;60:881-6. 18. Kulkarni AV. Personal communication. 2007.

813

36 The History, Current Status, and Future of the StealthStation Treatment Guidance System R. Bucholz . L. McDurmont

History of the StealthStation Treatment Guidance System The StealthStation neuronavigation system was specifically developed to facilitate broad adoption of stereotactic techniques across the entire spectrum of neurosurgical procedures. It has evolved into a system that can improve the accuracy and efficacy of a broad spectrum of procedures performed by a variety of surgical specialists. Stereotactic surgery was first developed by Sir Victor Horsley and RH Clarke [1] in 1908 to improve the accuracy of functional procedures as performed in laboratory animal investigations. In 1945, Spiegel and Wycis developed the first human application, and approximately a year later performed the first stereotactic procedure on a patient with Huntington’s chorea [2]. In 1953, Cooper found that small lesions within the basal ganglion could ameliorate the symptoms of Parkinson’s disease [3], and soon thereafter the most frequent application of stereotactic technique was the production of lesions to enhance function of patients with movement disorders, a field called functional stereotactic neurosurgery. With the advent of medical treatment for Parkinson’s disease, the frequency of use of stereotactic technique declined rapidly during the latter half of the twentieth century [4], and in most centers it was employed for biopsy procedures only. Even though the accuracy of stereotaxis was widely accepted, the difficulty of performing stereotactic interventions resulted in the technique #


being used only in select procedures in which accuracy was of paramount importance. Therefore, the vast majority of neurosurgical procedures were not performed stereotactically, a situation that persisted even after the advent of threedimensional imaging techniques such as computed tomographic (CT) scanning and, subsequently, magnetic resonance imaging (MRI). These imaging modalities were capable of providing detailed anatomical information to the surgeon intraoperatively, which could be highly useful in the performance of routine cranial procedures. Even though stereotactic procedures could couple this information to such operations, they were not employed due to the difficulty in using such techniques.

Integrating Imaging into the Intra-operative Space The goal of our development team at Saint Louis University was to integrate advances in imaging seamlessly into our neurosurgical technique. For that to occur, two issues had to be addressed. First, the coordinate system in which the threedimensional images were obtained preoperatively had to be registered to a reference system established during surgery, a process termed registration. Second, instrumentation had to be tracked within the surgical space during performance of the procedure. Both of these issues required the appropriate use of computational power. Images obtained from CT scanners were saved as large

544

36

The history, current status, and future of the StealthStation treatment guidance system

data files reflecting the detailed information obtained by even early units. A high degree of computational power was needed to manipulate these files, and the development of a solution to these issues had to wait until computers with sufficient power to perform these tasks became sufficiently inexpensive to be available to our team, and subsequently, the operating surgeon. Further, although computer networks were slowly becoming available, for our effort, and for the vast majority of operating rooms, there was also a requirement that the computers used for the system be sufficiently small to enable the system to be brought into the operating room during the procedure, and then removed to allow other procedures to be performed in the same room. It became evident that, for our effort, the only viable solution consisted of a computer on a cart with sufficient power to manipulate the large threedimensional image data sets, coupled with a three-dimensional digitizer of sufficient accuracy and robustness to be used in a sterile operative field. It became possible to satisfy both of these requirements in the late 1980s, leading to the development of our system and first application of a generalized solution in 1990. In the early to mid 1980s, computers capable of the manipulation and display of diagnostic images were large, requiring multiple equipment racks, and had significant power and cooling requirements. Propelled by the continuing miniaturization of computer power, and fueled by the rapid advancement of gaming and graphic applications, personal computers (PCs) were available by 1987 that met these requirements while being small enough in be placed in a cart. The first PC specifically altered for surgical use was the Heilbrun Stereotactic Model One, using an IBM PC with a graphics card capable of the intraoperative display of diagnostic images [5]. An added benefit of this system was its use of a common stereotactic apparatus, the BRW system, to establish a surgical coordinate system. We acquired the first production model of this

computer with the specific intent to employ it within a navigational system that would use an alternative means of tracking surgical instruments rather than relying on the arc-based BRW system.

The First Surgical Navigation Prototype The Model One was originally developed to replace the small laptop-style calculator that came with the BRW system, and to allow the required stereotactic calculations to be performed in the operating room rather than using the computer associated with the scanning device. Our team used the display capabilities of the system to serve as the display device for our navigational system. The Model One was modified by the group in Utah to display on the pre-operative scans the position of an instrument whose position was known within the BRW coordinate system. For our first prototype the registration of the preoperative images, and the establishment of an intraoperative coordinate system, would be handled by the BRW frame, as this stereotactic system had already established its accuracy and was clinically accepted and approved. The remaining requirement was for the tracking of instruments within the coordinate system established by the BRW frame. This required the use of a three-dimensional (3D) digitizer to produce these coordinates, and the development of instrumentation which could be tracked by the digitizer system employed. Further, additional software was needed to transfer these coordinates from the digitizer software into the visualization software from the Utah group. A major benefit of the Model One was the use of an International Business Machines (IBM) PC, running a standard Microsoft operating system (MS-DOS). Even at that early stage in the use of PCs, we were able to leverage our knowledge of this operating system, coupled with software routines available


for the platform, to rapidly program the first prototype. An integral aspect of our development was that all hardware and software components of the system had to be modular in nature to allow us to upgrade or change a specific component without interfering with the operation of the remainder of the system. This design constraint was based on our appreciation of the speed of development of both computer capability and digitization technology, and the recognition that the system should be capable of benefiting from the replacement of any component that was rendered obsolete by the rapid pace of technological evolution. Therefore, the digitization hardware and software was separate from the visualization hardware and software, and prototypes could be updated quickly to take advantage of new technology.

Digitization in Navigation At the time of initial development of the first prototype, the only digitizers that were sufficiently accurate for generation of instrument coordinates were mechanical, electromagnetic or acoustic in nature. Mechanical digitizers, using angle detectors mounted in jointed arms to determine the position of a surgical probe, had already been employed in surgical navigational systems by Watanabe [6]. Although accurate, these mechanical arms had serious ergonomic drawbacks, and their intraoperative use was similar to operating with a device that resembled, and handled like, a dentist’s drill. It was our impression that manipulating such a system within the confines of the deep and narrow exposures routinely employed during neurosurgery would be problematic, and it was therefore not the system of choice for general application. Electromagnetic digitizers determine position using a magnetically sensitive probe placed within a generated magnetic field. In the late 1980s these systems used magnetic fields which were easily distorted by the presence of ferromagnetic

36

substances. Given that the vast majority of surgical instruments are made of such substances, it was our impression that such a system would never be sufficiently accurate for use in neurosurgery. Therefore, at the time of our initial development, acoustic digitizers appeared to be the best devices for surgical use. Acoustic digitizers measure the time of flight of a sound pulse generated by an emitter (such as spark gap emitter) to a microphone. Given the speed of sound, the distance of the emitter from the microphone can be determined. Using an array of microphones, the position of the emitter can be triangulated and determined within a three-dimensional space. As these digitizers require an unobstructed path from the emitter to the microphone, a simple one-emitter solution could not be used for surgery as the instrument being localized would be within the body of the patient. The solution consisted of positioning at least two emitters along the handle of the instrument, with each emitter being outside the body of the patient with an unobstructed view of the microphone array. By designing and fabricating an instrument with emitters located at known distances from the tip of the instrument, and locating the position of each emitter, the position of the probe tip could be calculated. Given these advantages, acoustic digitization technology appeared to be ideal for neurosurgical applications. The only issue that remained was the location of the microphone array; and, given the geometry of the surgical field, it became apparent that locating the array within the surgical lights would maximize the chances of the emitters being seen by the array. Once the decision had been made as to the digitalization technique, it was necessary to decide what to track in the surgical field. Prior navigation devices had tracked the eye of the surgeon by connecting digitalization technology to the microscope. An example of an eye-tracking system was Kelly’s Compass system [7], consisting of a mechanical digitizer built into the microscope

545

546

36


support to display position using a framed stereotactic system. The same concept was used by Roberts to track sound emitters attached to a surgical microscope using an acoustic digitizer [8]. Roberts described using the system to perform spinal surgery which precluded the use of a stereotactic frame. The system therefore employed a frameless spinal registration technique using markers (called fiducials) applied to the skin over the area of interest prior to obtaining pre-operative images. This first implementation of a frameless stereotactic system was hindered by the inaccuracy in registration caused by the deformation of the spine, and movement of the fiducials relative to the spine, between imaging and surgery. This inaccuracy resulted in infrequent use of frameless registration for spine surgery; however, the frameless solution appeared to be highly applicable to cranial surgery, given the thinness of the soft tissue overlying the skull, as opposed to that over the spine. With less soft tissue, the inaccuracy of the registration process could be minimized by using multiple fiducials and averaging out inaccuracies caused by local distortions. In order to be applicable across a broad spectrum of surgeries, including those that did not involve the use of the microscope, we felt that . Figure 36-1 Sonic base ring for BRW frame

the concept of tracking the surgical instrument, as suggested by Watanabe [9], should be coupled to the use of a sonic digitizer, as employed by Roberts. We proceeded with the modification of a surgical instrument by attaching sound emitters. Given that a bipolar coagulating forceps was one of the most commonly used instruments during a cranial procedure, and that this instrument already had a cable connected to it, the first tracked instrument so modified was a forceps with two emitters attached in a known geometry, produced by Karl Storz in 1990. It was also necessary to track the position of the body part undergoing surgery (> Figure 36-1). To speed development and to insure accuracy of the first prototype, we decided to register the images using the BRW frame and N-bar fiducials employed by the Heilbrun system. The BRW stereotactic frame attached to the patient prior to imaging has three attachment points for placement of the N-bar fiducial cage. This cage is attached during imaging and then removed, and in traditional stereotactic surgery an arc system is attached to the frame to allow biopsy and creation of lesions. Rather than using this arc system, we created a reference system of three sound emitters attached to a ring with three


frame attachment balls. This was constructed in the Saint Louis University machine shop. This reference system attached to the same points as the usual arc system, and to simplify calculations the emitters were positioned in the base plane of the BRW coordinate system (z ¼ 0). Software provided by PixSys of Boulder, Colorado, generated the coordinates of the tip of the instrument within the BRW coordinate system, coordinates that could then be transferred to the Heilbrun routines to display position using their modified PC. Using an early version of a multitasking operating system called DesqView (Quarterdeck), the routine that generated positional coordinates was operated in one block of memory and was then transmitted, via a scripting macro, to the localization routine in another block of memory. The first prototype indicated the forceps position by selecting the closest axial CT image and placing a blue dot over the location of the tip of the surgical probe. The system required 5 s to display position after activating the system using a floor switch. The graphics card employed in the PC was not capable of generating coronal or sagittal projections, and could not produce interpolated axial images between the axial CT images. Therefore, it was

36

imperative to obtain a pre-operative CT scan with thin slices to minimize error in z localization; the system employed generated slices 1.5 mm thick to address this issue. The accuracy of the system was extensively investigated prior to clinical application by using the phantom base supplied with the BRW system. The sonic localizing ring was attached to the phantom base, and the forceps were rigidly attached to the pointer of the base, with the tip of the forceps touching the tip of the pointer. This technique essentially converted the phantom base into a calibration jig. For these experiments, the microphone array was suspended above the phantom base at a distance consistent with that achievable in the operating room. The pointer was moved through a variety of specific coordinates using the Vernier scales in the base, and we verified that the sonic localizer reported the same coordinates throughout a volume which would normally contain the head of a patient (> Figure 36-2). We then wanted to check the accuracy of the system with a patient within the ring to make certain that the presence of the head did not cause a distortion in the system; and we wanted to make certain that ambient noise did

. Figure 36-2 BRW phantom base with sonic base ring and sonic forceps attached

547

548

36


not render the system inoperative. These concerns were addressed by conducting experiments on patients with attached frames who were undergoing stereotactic radiosurgery using a linear accelerator at Saint Louis University. For these early cases using radiosurgery, a stereotactic floor stand was attached to the BRW patient ring and used to support the patient’s head. The target to be treated was entered into the Vernier scales of the floor stand, which would bring the target into the center of the LINAC beam. Prior to treating the patient, the sonic reference arc would be attached to the ring, the floor stand coordinates would be set to an anatomical landmark on the surface of the patient, and the room’s lasers were used to verify the accuracy of the floor stand in identifying the anatomical landmark. The landmarks were then touched to make certain that the forceps identified the landmark on the display screen of the system. As the LINAC produced a great deal of noise, this was a robust test of the system which indicated, prior to any use of the system in the operating room, that the sonic system was relatively immune to ambient noise and was accurate in the presence of human anatomy attached to the ring. This was demonstrated with both

CT and MRI scans being used for the imaging technique.

Clinical Use, Testing, and Refinement of Model One From July 1991 through January 1992, we used the first prototype to perform nine sonically navigated cranial procedures under a research protocol approved by the Institutional Review Board (IRB) of Saint Louis University. Procedures were chosen that could be completed in a normal fashion should the system fail. For each of these procedures, a BRW frame was attached prior to placing the patient under anesthesia, and imaging was performed using either CT or MRI with the appropriate localizer attached to the ring. In the operating room, the microphone array was attached to a stanchion supporting the operating room lights and centered above the patient’s head (> Figure 36-3). The patient was then brought into the operating room, placed under anesthesia, and prepped. Prior to making the incision, the emitter-equipped base ring was attached to the frame. This allowed the arc system to be used to

. Figure 36-3 Microphone array over operative field and sonic workstation


confirm the location produced by the sonic system. Points were marked on the scalp of the patient after draping, and the coordinates of each point determined by attaching the arc system, pointing to the spot using the arc, and then placing the arc on the phantom base. By moving the pointer of the phantom base to the tip of the arc, the coordinates of each point on the scalp could be determined. We could then determine the accuracy of the prototype throughout the course of the procedure by pointing to these three points with the forceps. The error of the system could be determined by measuring the distance of the coordinates of the forceps when pointing to these calibration spots from the actual coordinates as determined by the arc system; we found the accuracy of the system to be within 1.5 mm, and felt that the utility of the system had been validated. However, during use, the system would occasionally display a position for the forceps that was outside the patient’s head or far removed from where the procedure was occurring. Upon reviewing the triangulation data coming from the timeof-flight calculations generated by the sonic digitizer routines, it became apparent that, during erroneous localization, the time of flight from an emitter to the microphone array was much longer than usual for correct localizations. We were obviously dealing with echo issues that were inherent to acoustic digitizers. Sound coming from the emitter could bounce off a reflective surface and be picked up by the microphone, causing the emitter to be localized far from its actual location. We attempted to mitigate echoes during surgery by using blankets on the walls, positioning equipment so that flat sides would not be perpendicular to the surgical field, and positioning personnel strategically to absorb echoes. In spite of these measures, it became apparent that an acoustic solution would not be practical for a wide variety of operating rooms which had highly reflective surfaces such as tiled walls. Fortunately, at this pivotal juncture in the development process, optical digitizers became commercially available.

36

Optical digitizers employed either reflected or emitted points of light focused upon an array of optical cameras usually consisting of charged coupled device (CCDs). In a manner analogous to human stereoscopic localization, the cameras, separated by a known fixed distance, could determine the position of the source of light by comparing the pixels on the CCDs illuminated by the light coming from the object (> Figure 36-4). To avoid becoming confused by light emanating from the surgical field, these devices generally used light of a highly controlled wavelength in the infrared spectrum, and filters were placed over the cameras to allow only this wavelength to reach the CCDs. The first optical digitizers to become available employed light emitted from infrared light-emitting diodes (IR LEDs). To change our prototype to an optically based system, all that was needed was to mount LEDs in place of the sound emitters on the forceps, to replace the emitters on the reference ring with LEDs, and to suspend the camera array where the microphone array had been positioned for the first prototype. As with the sonic system, determination of the position of the tip of a probe within the body of the patient required at least two IR LEDs to be mounted on the handle of a surgical instrument, the geometry of the instrument being programmed into the system. Given this similarity, many of the same software routines from the acoustic prototype were used for this new optically based system, and the optically based navigational system was ready for clinical trials by the end of 1991. Aesculap created the first commercially available LED forceps for the second prototype in 1991 (> Figure 36-5). It also became apparent after clinical use of the sonic system that indication of surgical position using a single axial image, although useful, did not convey a complete appreciation of location in the superior/inferior dimension. Therefore, the surgeon was forced to create an appreciation for location in all three dimensions. Although other systems in use during this period

549

550

36


. Figure 36-4 Original optical cameras

. Figure 36-5 Operative picture showing closeup of forceps and base ring with LED

were capable of producing sagittal and coronal images by reformatting the axial images, the computation power needed to render such images in a reasonable time-frame far surpassed that available when using an IBM PC-based

system. Therefore, we made the decision to create a new system that was capable of reformatting images continuously, based on a system by the leader in computer graphics at that time, Silicon Graphics Incorporated (SGI).


Integrating Frameless Localization One of the goals of the initial design of the first prototype was frameless localization, as pioneered by Roberts in his use of skin-mounted fiducials. However, due to the inaccuracy of skin-based fiducials, as witnessed by Roberts in spine applications, and the computational requirements for such a registration, this goal was not realized in the initial prototype. It was thought that the inaccuracy of skin-based fiducials would be minimized in cranial applications due to the proximity of scalp-based markers to the rigid body of the skull, which would limit the amount of movement of fiducials relative to the intracranial anatomy between imaging and surgery. However, there remained the possibility of sliding or slipping of any fiducial attached to the scalp due to the natural movement plane of the subgaleal space. One way to address this source of error was presented by Allen [10] in his proposal for skullbased fiducials. These markers were screwed directly into the skull, thereby eliminating movement between the marker as seen on imaging and the anatomy within the skull. The markers were designed to hold a variety of contrast agents that could be seen by the imaging technologies employed, such as CT, MRI, or even positron emission tomography (PET) [11]. These fiducials would eliminate error due to scalp movement, but would not correct for movement in the position of the brain relative to the skull between imaging and surgery. Although it was believed that the brain did not move significantly within an intact skull, the movement of the brain following opening of the skull became grounds for the future development of intraoperative imaging. Skull-based fiducials were demonstrated to be highly accurate and potentially employable by our system; however, this solution, requiring a small but potentially painful procedure prior to imaging, was felt to be necessary only in those applications requiring the highest accuracy.

36

For the much wider variety of intracranial procedures commonly performed, we felt that a multiplicity of scalp-based markers would be sufficient, as long as the registration routine employed by the navigation system compared distances between markers to determine which markers had moved between imaging and surgery.

Second Generation Navigational System After the initial clinical cases with the first sonicbased prototype, it became clear that the PCbased Model One was not sufficiently powerful to generate the visualization necessary or to implement ‘‘frameless’’ registration based on point fiducials. The Model One required approximately 5 s to display the pre-operative CT slice with overlying cross-hairs, a time-frame that was unacceptable in the intra-operative situation. This was true for both the optical and sonicbased versions of the initial prototype. Therefore, even though the optical system was available for clinical trials late in 1991, we had reached the conclusion that considerable programming and engineering support was needed to develop a completely new second-generation prototype, along with experience in programming SGI computers. Dr. Kenneth R. Smith, MD, who at the time was the Director of Neurosurgery at Saint Louis University School of Medicine, referred us to a development team at Southern Illinois University at Edwardsville (SIUE) headed by his nephew, Kurt Smith, PhD, an assistant professor at that institution. Dr. Smith’s team was developing a digital audio recording and editing device called the StealthStation. The team consisted of students working at the lab and included Kevin Frank, Kurt Noltensmeyer and Paul Kessman, all of whom had considerable experience working with the Silicon Graphics processors that were at the heart of the StealthStation. After

551

552

36


an initial meeting and discussions, the focus of his group was expanded to include the development of a navigational system. After a year of development work, the audio device was abandoned, but the name of that recording device was salvaged and applied to the new navigational system. It should be noted that the name of the system was initially selected to appeal to the potential users of the audio device, such as rock and pop musicians, but was subsequently found to appeal to surgeons as well. Given the intensity of the development work, SIUE granted credit to the students working on this project at an off-site location, and a separate corporate entity, Stealth Technologies, was formed. A suitable Silicon Graphics Incorporated processor with sufficient capability to handle the more complex frameless registration algorithms and visualization demands was chosen, taking advantage of the then unparalleled computer graphics capabilities of this system. The initial software was actually patterned after software used for visualization of oil fields to allow more efficient placement of oil wells (> Figure 36-6).

In early 1992, the IRB protocol was modified to allow the use of the new optical system, and initial procedures were performed with a framed registration solution, the sound emitters on the ring being retrofitted with LEDs. After extensive lab-based testing, the first frameless optical procedure was performed at Saint Louis University Hospital on 21 Feb 1992. After the development and refinement of multiple prototypes, an optical ring was produced by Stealth Technologies, and the Neurostation, the second-generation navigational system, was available for clinical testing.

Early Additions to the Surgical Navigation Device In October 1993, collaboration was begun with Moeller-Wedel and the SIUE group, now Surgical Navigation Technologies, on the development of a computer-tracked microscope. In February 1994, the first microscope-tracked cranial procedure was performed at St. Louis University Hospital using

. Figure 36-6 Workstation showing hybrid system of a PC and SGI computer working together


LEDs mounted on a bracket attached to the microscope, essentially using the microscope as an additional instrument, or pointer replacement, for the Neurostation. The robotic features of the microscope, consisting of auto focus and auto position, as well as video input into the Neurostation and calibration verification, were all co-developed by Moeller-Wedel and Surgical Navigation Technologies in 1995 (> Figure 36-7). Kevin Foley, MD, of the University of Tennessee in Memphis, approached the development team in 1994 to explore the potential for a spinal application of the system. With the frameless cranial solution fully developed, Dr. Foley suggested using anatomic spinal landmarks exposed within the surgical field to serve as fiducials for the registration process. Registration would therefore not occur at the beginning of the procedure, as with the cranial application, but after the exposure of these points during surgery. This called for a modification of the LED reference array, and the first spinal array was created by mounting LEDs onto a locking pliers which was clamped onto the spinous

. Figure 36-7 Operative microscope with LED tracking array

36

process of the vertebra undergoing surgery. The first frameless stereotactic procedure on the spine was performed by Dr. Foley using the system on 20 July 1994. With the success of this procedure, it was apparent that the Neurostation had applications in both cranial and spine procedures, and as there was concern over complications associated with pedicle screw placement [12] during this period, financial support for further development was obtained from Sofamor Danek, a leading manufacturer of pedicle screws. With this funding, a third prototype was developed that was easier to use by surgical teams outside Saint Louis University and Memphis. Sixteen units of this first prototype, marketed under the name StealthStation, were sold throughout the world, and those systems were used to support a Food and Drug Administration (FDA) application for the device, which was granted in 1996. Sofamor Danek wholly acquired Surgical Navigation Technologies in 1996, and they actively marketed the StealthStation system until 1999, when Sofamor Danek was acquired by Medtronic.

553

554

36


Current Status of the StealthStation Treatment Guidance System From 1999 to 2008, the StealthStation system has realized continuous advances due to faster computers and graphics processors, new software algorithms, increased focus on user-centric design, and incorporation of intra-operative imaging. Usage has expanded to include a number of additional cranial, spinal and ENT procedures. As a framework for discussion of both the current and future status of the StealthStation system, we have chosen the following essential elements of contemporary navigation: 1. 2. 3. 4. 5. 6. 7.

Preoperative imaging Patient and image registration Intra-operative tracking Visualization System control Effectors Intra-operative imaging

Preoperative Imaging The process of transferring preoperative images into the navigational system was viable yet cumbersome with the early prototypes and commercial systems. The initial protocol for importing images involved the use of unwieldy tapes or optical disks to transfer raw scanner media into the system. As Digital Imaging and Communications in Medicine (DICOM) became the standard mode for distribution of images, importation became more facile. While DICOM remains a preferred transfer medium, the StealthStation system has been evolved to import many different data types via a variety of media, comprising the Ethernet, compact disk (CD), medical imaging archiving systems (PACS) and, in the near

future, flash drives attached via the universal serial bus (USB). A broad spectrum of imaging modalities from radiological scanners can be input into the StealthStation system. The most common imaging modalities used in navigation are MRI and CT. Other anatomic modalities include computed tomographic angiography (CTA) and magnetic resonance angiography (MRA). At the start of the twenty-first century, functional imaging modalities such as functional MRI (fMRI), positron emission tomography (PET), and single photon emission computed tomography (SPECT) became prominent, and commercial neuronavigation products, including the StealthStation system, began to allow importation and manipulation of these modalities as well.

Patient and Image Registration Once image data is imported into the StealthStation system, the user has the option of coregistering two or more volumetric medical images into a common coordinate system so that they can be fused and visualized together. Advanced image-processing algorithms made StealthMerge possible in 1998. This multimodality image fusion software permitted precise correlation and merging of multiple image data sets (e.g., CT fusion with MR, PET fusion with CT, etc). This software allows surgeons a combined view of complementary image information (e.g., bony anatomy from CT with soft tissue from MR, or functional information from PETwith anatomy from MR). This fusion and visualization provides additional information to surgeons, allowing greater confidence during stereotactic procedures. The ability to create surgical plans has been available from the earliest incarnation of the Model One. These plans are often used to represent a desired outcome during surgery, such as the path to be traveled by a biopsy needle or the implanted location of a pedicle screw. Over


the past several years, surgical planning tools have begun to make increasing use of information about the therapies to be delivered or the procedures to be performed. For example, planning capabilities in the StealthStation system’s spinal software incorporate three-dimensional models of each of the pedicle screws sold by Medtronic Spine and Biologics. These models allow preplanning of screw placement and confirmation of correct sizing and screw type selection. Software used for performing biopsies incorporates information about the location of the cutting window on the biopsy needle to ensure accurate placement of the needle and resulting tissue sample.

Intra-operative Tracking Intra-operative tracking is the position measurement technology that provides the basis for real-time feedback of surgical instrument locations during a procedure. For many years, intra-operative tracking in the StealthStation system has been built upon commercially available optical measurement systems from Northern Digital, Inc. (NDI). There are two widely adopted methods of optical measurement in navigation: active and passive, both of which have been employed within the StealthStation system. The infrared LEDs that were used at the inception of navigation are still in use today and are activated by an electrical signal. Passive tracking involves markers made of a retro-reflective material, which reflect infrared light emitted by the optical tracker. In both cases, the optical tracker measures the positions of individual markers attached to an instrument and, based on a model of the instrument stored within the system, can infer the location of the entire instrument, usually with sub-millimetric accuracy. Optical tracking has at least two weaknesses: the need for a line of sight between the instrument and the tracker; and the requirement for

36

rigidity between the markers and the portions of the instrument to be tracked. Within the past 5 years, the StealthStation has incorporated a second form of tracking which uses electromagnetic energy rather than infrared light. Electromagnetic tracking has become a viable and valuable method of position measurement, especially as the applications of navigation have expanded and the number of less invasive procedures being performed has increased. Electromagnetic energy passes through the body and is therefore not susceptible to line-of-sight issues. The electromagnetic tracker can be much smaller than a functionally equivalent optical tracker, thus opening navigation to a host of new applications. Electromagnetic trackers have been manufactured that are approximately 1 mm in diameter and approximately 5 mm in length, whereas a similar optical device may be over 100 mm in length. In addition, electromagnetic tracking provides the ability to track the distal end of a flexible instrument, which expands the applicability of navigation into the realm of various soft tissue procedures and more reliable placement of ventricular shunts. These systems differ from the initial electromagnetic tracking systems available early on in the development of the system in that they have improved preservation of accuracy in the presence of ferromagnetic substances.

Visualization As discussed previously, the preoperative images must be mapped to the physical anatomy of the patient via a registration process. Registration allows for real-time, intra-operative visualization of preoperative imaging data based on the position and orientation of one or more tracked surgical instruments relative to the actual patient anatomy. Registration requires the establishment of a spatial transformation between the respective coordinate spaces of the intra-operative

555

556

36


tracker and the image. The most common and longest-standing method for patient registration for the system as described above is paired-point matching, called PointMerge. This approach can use either fiducial markers that are attached to the patient prior to the preoperative scan or anatomical landmarks to establish corresponding landmarks in the system’s coordinate systems for the image and the patient. Based on this data, computer algorithms can compute the spatial transformation between the two coordinate systems. Recent advances in PointMerge include the automatic identification of fiducial markers from image data, as well as the automatic establishment of correspondence between points in the image and patient space. This latter enhancement allows the user to touch fiducial points in any order without the need to manually specify which patient point corresponds to which image point. Another popular method of registration in cranial procedures uses a surface matching technique. This approach extracts the surface of the patient’s face from preoperative images, and requires the user to collect corresponding surface data on the actual patient anatomy using a specially designed surgical instrument. This Tracer registration algorithm allows for accurate registration without the use of fiducial markers, even in situations when the patient’s face is not directly visible to the optical localizer. The registration process for preoperative CT images in spinal procedures is a clinically challenging task that has been simplified in the StealthStation system through the use of fluoroscopy. The FluoroMerge registration algorithm uses ‘‘virtual fluoroscopy’’ techniques to perform patient registration. Using these techniques, a synthetic fluoroscopic image can be created from the preoperative CT images (often referred to as a digitally reconstructed radiograph or DRR). The concept is to acquire two actual fluoroscopic images of the spine from different directions. Using a mathematical model of the fluoroscope’s image formation process, the computer can then determine where

a ‘‘virtual’’ fluoroscope would need to be positioned in order to generate DRRs that match the acquired fluoroscopic images. Since the fluoroscope is being tracked by a localizer, the mathematical solution to this problem results in the computation of the transformation between the CT images and the patient. The result is an automated registration.

System Control A core requirement of neuronavigation is the ability to display clinically relevant data from which a surgeon can make decisions relevant to the surgical procedure. Such decisions often result in minimizing damage to healthy tissue, reducing surgical procedure times by selecting optimized paths to a target and making smaller incisions, or increasing the chances of a more complete tumor resection. Early iterations of the StealthStation provided simple grey-scale display of CT and MR images in coronal, sagittal and axial cross sections, along with a 3D surface reconstruction. As graphics and processors improved, the ability to display more complex ‘‘volume rendered’’ 3D images was developed. More recently, there has been a focus on designing visualization capabilities that are best suited to the clinical task at hand. For example, when evaluating the trajectory of a biopsy needle, it is possible to display a ‘‘look-ahead’’ view, which shows the anatomy that the needle will pass through as it continues along the trajectory. A similar view shows the remaining distance to the target, as well as any deviation from the planned trajectory. The ability of the surgeon to visualize anatomy from a myriad of sources, such as multimodality anatomic or functional images, real-time intra-operative fluoroscopy, ultrasound, MR or CT, allows the surgeon to minimize the risk of trauma to surrounding tissue, minimize incision size and displacement, and, ideally, reduce operative times.


Over the 18 years since the stereotactic surgery revival, navigation has become an integral part of most neurosurgical practices and has been integrated into the curricula of internships, residencies and medical schools. In order to maintain and increase relevance during this time of rapid proliferation, the StealthStation system needed to evolve with the advances in other areas of medicine, such as imaging, communications and operating room control. In 2005, Medtronic Navigation embarked on an extensive study of navigation system usability and workflow and the application of user-centric design, resulting in several ergonomic and workflow enhancements to the StealthStation system. The result of this study was the creation of the ‘‘Synergy Experience’’ for the StealthStation system, which incorporated innovations in display technologies, user interface devices, and software user interfaces. In addition to designing products for the surgeon, much consideration was given to the operating room support staff and their needs. The resulting products are greatly simplified and intuitive compared to their predecessors, and incorporate procedure specific workflows, wireless mice and built-in surgeon customization to help minimize operating room time and workflow, and maximize efficiency. For example, it is possible to pre-define operative room layout, instrument usage, visualization options and software workflow on a procedure-by-procedure basis for each surgeon that uses the system. In addition, features such as peripheral connectivity dashboards allow support staff to ensure that connections to other devices in the operating room (e.g., surgical microscopes, fluoroscopes, etc.) are operating properly before the surgeon even enters the room.

Effectors As intra-operative tracking technology progressed, the need to track all manner of surgical instru-

36

ments, from simple to complex became imperative, in order to offer surgeons an unprecedented visual window into the surgical space. As discussed previously, the first instruments to be tracked consisted of a coagulating forceps followed by several simple instruments for bone cutting and suction. Further developments enabled the tracking of many spinal surgery tools, such as drills, screw systems and delivery instruments for inter-vertebral devices. Using a tracked surgical microscope, an ultrasound probe, or even an intra-operative imaging system such as the PoleStar1 iMRI System or the O-Arm1 Intra-operative Imaging System provides real-time guidance to accommodate changes in anatomy that occur during the course of a procedure (> Figure 36-8). As surgeons became more proficient with neuronavigation and refined the art, a potential drawback to using frameless approaches for biopsies and functional neurosurgery was the lack of a rigid frame to provide support. The StealthStation Navigus device uses a disposable trajectory guide that is mounted directly to the skull with titanium screws. This device allows the surgeon to insert a probe into the guide and pivot it via a joystick-like mechanism to achieve the proper trajectory based on images of the brain as displayed on the workstation. The trajectory is then fixed through a locking ring. Another way to address this issue is via the attachment of the Vertek arm directly to the operating room table. This flexible arm can be locked in position once the target trajectory has been selected. As surgeries and techniques evolve, applications of such rigid fixation devices will expand beyond biopsies and shunt placement into drug delivery, tumor removal, biotechnology and neurostimulation.

557

558

36


. Figure 36-8 Intraoperative MRI system

Intra-operative Imaging The overarching goals of twenty-first-century navigation include reduced invasiveness, improved outcomes, increased efficiency and productivity, and more cost-effective clinical solutions. The incorporation of intra-operative imaging into the practice of surgery provides immediate and dynamic images to support clinical decisionmaking in the operating room. In addition, when coupled with navigation, intra-operative imaging facilitates the radical simplification of the patient registration process, allowing for automatic registration of images to the patient with no action required by the user. As the surgery is performed, surgeons have the capability to visualize changes in the patient’s anatomy and react immediately to those changes as they acquire and review updated, patient-specific medical images during the procedure. The first intra-operative imaging device incorporated into the StealthStation system was based on ultrasound. The SonoNav application allowed for side-by-side viewing of real-time ultrasound with preoperative CT or MR. Several years

later, virtual fluoroscopy was introduced in the form of the FluoroNav application. This innovation allowed the acquisition of multiple fluoroscopic images and subsequent navigation based on these images, thus enabling multi-planar fluoroscopic capabilities. In early 2003, the PoleStar1 iMRI System was introduced. This MR imaging technology was designed specifically for cranial neurosurgery to help surgeons monitor tumor tissue throughout the course of a resection, thereby increasing the likelihood of more complete resections. The O-arm1 intra-operative 2D and 3D fluoroscope was designed to provide 3D X-ray-based imaging for bony anatomy. This system, which incorporates robotic positioning, a unique gantry that opens for positioning around operating room beds, a flat panel detector, automatic registration and many other innovations, has recently gained popularity in spine surgery. This intra-operative imaging device has tremendous promise for future applications (> Figure 36-9). The StealthStation system has always been a platform for research. With the addition of the StealthLink networking capabilities, a con-


. Figure 36-9 Intraoperative O-arm1

duit between external research tools and the StealthStation is possible. Many exciting projects are currently in development using this collaborative tool. Enhancing surgical performance and enabling new patient treatment options are the hallmark of the all-encompassing navigation systems of the modern-day operating room. As the future unfolds, the StealthStation system is becoming a node on the information highway, offering an integrated digital infrastructure for surgeons as well as hospital managers and support staff. Hospitals and surgeons have access to an integrated navigation and intra-operative imaging hub that empowers them to make data-driven decisions in the OR as well as to maximize cost efficiencies in today’s cost-conscious healthcare marketplace.

Future Development The system should be considered as being still in its adolescence. The StealthStation is being actively developed to address new procedures being performed in neurosurgery and to take advantage of advances in new technologies which will improve the functionality of the device (> Figure 36-10).

36

. Figure 36-10 Current S7 system

The remaining portion of this chapter will focus on neurosurgical applications only, but some of the most exciting developments with the system are occurring in specialties other than neurosurgery. To reiterate our framework for discussion, these are the essential elements by which we will discuss contemporary navigation: 1. 2. 3. 4. 5. 6. 7.

Preoperative imaging Patient and image registration Intra-operative tracking Visualization System control Effectors Intra-operative imaging

Preoperative Imaging Imaging modalities available and useful to neurosurgery continue to develop, as has been the case

559

560

36


since the revival of stereotactic surgery was initiated in the late 1980s. Most recent advances have focused on the imaging of brain function, via magnetoencephalography (MEG), and functional MRI (fMRI). Although these modalities have been partially implemented in the StealthStation system, further development is in progress to fully integrate them and allow their use on a routine basis, and to make such use practical for physicians without requiring extensive imaging support staff. As the use of these imaging technologies becomes more commonplace and review of the appropriate images prior to a surgical intervention becomes necessary, the concept of a planning station becomes more appealing. Ideally, it would be best to not only analyze the images needed to plan the procedure, but also to actually perform and merge the actions required for the procedure using the preoperative images. This would require not only fusion of different imaging modalities, but also segmentation of key structures on these images and implementation of specific surgical plans which, once perfected, would have to be registered to the surgical act. Further, it would be optimal to plan, and practice, the procedure with three-dimensional vision, or stereoscopy, in order to produce the most realistic environment for generating a valid plan. Such a comprehensive integration of a surgical planning/ practice system into the intra-operative navigation system is being developed in cooperation with Volume Interactions of Singapore. This stereoscopic planning system uses a magnetic threedimensional digitizer to track the movements of the surgeon as tissue is removed while displaying any form of medical imaging in a co-registered three-dimensional virtual body. The final surgical plan is then transmitted to the StealthStation using Stealth Link, described in the preceding section. The combined system enables the visualization of all key structures and targets for the procedure prior to encountering them in surgery, and the stereoscopic projection allows precise

surgical paths to be generated and followed by the operator.

Patient and Image Registration The registration techniques described above employ either artificial or anatomical landmarks seen on preoperative imaging to register those images to the surgical workspace. With the increased availability and use of intra-operative imaging, an alternative approach could be the use of images acquired in the operating room to perform the registration process. For example, an intra-operative CT scanner, rendering a detailed representation of the skull, could be employed in a process identical to the aforementioned image fusion process with the intraoperative MRI scanner to perform an image fusion between the intra-operative images and the preoperative images. Critical to such an application would be a means by which to correlate the intra-operative image to the coordinate system of the procedure. This could be accomplished by rigidly attaching the skull to the CT scanner or, to allow the CT scanner to be moved out of the way during the surgical intervention, a tracker device could be placed on the skull prior to imaging to allow the anatomy to be tracked after the CT scanner has been retracted. This technique would allow any preoperative diagnostic image of sufficient resolution for navigation to be employed intra-operatively without the application of fiducials prior to imaging. The accuracy of the registration process should be equivalent to, or surpass, that of using either a stereotactic frame or skull-based fiducials. This technique would thereby eliminate the discomfort of skull-based fiducials, and lower the barrier to use of image guided techniques for a variety of surgical interventions, as any diagnostic image could be used for navigation.


36

Intra-operative Tracking

Visualization

The vast majority of navigational devices currently employ two relatively mature tracking technologies, electromagnetic and optical. It does not appear that either technology will eliminate the other in the near future, as each has benefits and drawbacks that render them complementary to each other rather than essentially equivalent. For example, the ability of electromagnetic devices to track effectors within the body of the patient that are not rigidly attached to any tool outside the body renders electromagnetic localization the technique of choice for endovascular procedures. This benefit is so critical in these procedures that the need and expense of re-engineering the surgical workspace to eliminate ferromagnetic instruments is easily justified. However, in open procedures in which a variety of instruments must be employed, the relative robustness of optical localization will probably remain the technique of choice unless electromagnetic localization can be made immune to the presence of ferromagnetic substances within the surgical field. One possible alternative, or adjunct, to these tracking technologies is the use of intra-operative imaging to track effectors. For example, during placement of a stimulator electrode, intraoperative CT or ultrasound could be used to visualize the tip of the electrode as it is inserted into the region of interest. However, even with this simple example, the use of continuous imaging to navigate may expose both patient and surgeon to excessive amounts of ionizing radiation, and intense use of ultrasound frequently results in procedure delays. Therefore, even in this example, some sort of simple tracking of the effectors would seem to be preferable to a continuous imaging solution. Verification of position using intra-operative imaging would seem to be best used to confirm the correct position of an effector after it is navigated into position using standard tracking technologies.

The sheer amount of information becoming available about the structure and function of a specific patient’s brain will soon overload the ability of the clinician to visualize the entirety of the information at a glance. What is needed is method within the StealthStation system by which to selectively de-emphasize certain information when not applicable to the particular task at hand. For example, MEG information is not particularly germane during dissection of the deeper portion of a tumor, whereas DTI information becomes far more relevant. To avoid ‘‘information overload’’ of the surgeon during the procedure, the system would have to be ‘‘situationally aware’’ about the task at hand and suppress information not of use to the surgeon at that point in the procedure. This awareness of the system could be part of the planning procedure; in addition to practicing the procedure using the preoperative images, the system could learn, based on the information selected by the surgeon during the process, what information should be displayed at specific points during the operation.

System Control As surgeons become more reliant on navigational systems to perform surgery, and as the capabilities and complexity of these systems increase, control of the system will become increasingly demanding. The more the surgeon becomes dependent upon the system, the more interactions will occur between the surgeon and the system. The issue of system control has not been adequately addressed by any current navigational device. Use of a touch screen display beneath a sterile transparent drape, while it can increase usability, is problematic in terms of concerns about sterility as well as the decreased contrast

561

562

36


in images resulting from the touch screen overlay. Indeed, this problem of contrast has led to the use in the operating room of liquid crystal display (LCD) panels without touch screen capabilities, dedicated to providing the best image for the surgeon and leaving control of the system to another device or person. Computer mice have also been employed in the operating room, but sterilization of these electronic devices can be problematic, and there are few flat surfaces in the operating room that can be used as a base for the mouse. Computer mice designed to be used in free air have also been tried, however these tools leave much to be desired from an ergonomic perspective. Given these restraints, most busy neurosurgical facilities rely on an individual outside the sterile field to control the system, which adds both cost and complexity to routine use of a navigational system. In looking for a solution to this issue of system control, it is best to reflect on what has worked well in the past in operating rooms, and that is the use of voice commands. Almost every communication that occurs in a conventional operating room is verbal in nature, and it would seem logical to employ the same mechanism to control the navigational system. Voice control systems have been experimented with in the operative environment and have been generally rejected on the basis of their lack of reliability, probably due to the high background noise in the operating room, and the large vocabulary needed to respond effectively to an operating surgeon. Voice control systems usually fall into two broad categories: those that require training to a specific voice, and those that do not, literally working ‘‘out of the box’’ and responding to a speaker without prior exposure to that voice. In order to maintain reasonable accuracy and speed of response, non-educable systems would usually have a limited vocabulary, forcing the user to remember specific words. With the advent of increased computer power in the

operating room, and the proliferation of electronic devices, a system that learns the surgeon’s words and adapts to the surgeon, rather than the other way around, would seem to be the ultimate answer to making voice system control a viable solution.

Effectors In spite of considerable developments in visualization, as delineated in this chapter, the effectors used by surgeons, consisting of instruments and devices that manipulate tissue to achieve a desired effect, have remained relatively unchanged over the past decade. The forceps, scalpels, biopsy instruments, and drills that have been used for decades are still being used with more knowledge of the patient’s anatomy. Although some new effectors are in common use, such as ultrasonic aspirators [13] and bipolar coagulators [14], it is safe to say that a surgeon from 30 years ago would be comfortable using the effectors found in the current neurosurgical operating room. This static situation is not the case across all surgical specialties. Very complex and expensive robots are currently and routinely employed in urological procedures. Modern-day abdominal surgery has experienced a renaissance with the employment of new endoscopic surgery techniques which has in turn required the development of a complete set of effectors capable of working through small surgical channels [15]. The application of these new effectors has required significant retraining on the part of surgeons in these fields, an educational effort justified by the benefits afforded by using these technologies. These benefits include fewer complications in robotic urological surgery [16] and less invasiveness coupled with shorter recovery times in endoscopic versus open abdominal surgery. Given these advances in other surgical specialties, it is surprising that similar technological advances in


effector design have not been apparent in neurosurgery, with its demand for the ultimate accuracy and precision in movement. It is of interest that the application of these new effector technologies has generally occurred in those fields in which patients routinely pick and choose their surgeon due to the elective nature of the surgical intervention. Therefore, the adoption of such interventions such as endoscopic cholecystectomy has been driven by patients who can and do exert considerable pressure on surgeons to use techniques that maximize effectiveness and minimize impact on the patient. One could imagine that if neurosurgical procedures were more common, and more elective, patients would exert similar pressure on neurosurgeons to adopt new effector technologies. Perhaps the most obvious effector technology that could be used in neurosurgery is the use of robotic devices. Although the term robot implies a device that moves on its own without necessarily having human control, the term has been broadened to describe a device that can assist a human in performing a task that is difficult to perform without assistance. An area of intense interest is the concept of a scaling robot [17], in which the movements of the surgeon in traditional dimensions can be scaled to a smaller, even microscopic level. For example, the very small movements necessary to clip an aneurysm could be performed by the surgeon on a human scale and translated to the microscopic level by a scaling robot. This type of application has already been employed in the urological robotic application previously mentioned. A significant factor holding back neurosurgical robotic applications is the very small corridors employed to reach neurosurgical targets, which are not compatible with the size of robotic effectors currently employed. At many engineering centers, this limitation is being overcome by miniaturization of robotic devices that have been specifically designed for neurosurgical

36

applications rather than being simply borrowed from some other field. The term effector could also be employed for non-mechanical agents that achieve a desired therapeutic effect. Recently, there has been considerable attention in neuro-oncology to the use of intra-tumoral injection of chemotherapeutic agents to address the infiltrative nature of malignant glial neoplasms [18]. This area of medicine, termed convection-enhanced delivery, has proposed the use of a variety of agents for this purpose, many of which have molecules with an attachment point specific to antigens seen on the cell membrane of malignant cells, and a killing moiety which once inserted into the cell through the attachment point results in the killing of the cell. Although the protocols investigating these agents have yet to demonstrate efficacy, there is already a commercially approved chemotherapeutic wafer which has been shown to have some effect on the progression of glioblastoma multiforme [19], and it is to be expected that additional efficacious agents will be found in the future. It is also to be anticipated that the benefit accrued with these agents will be maximized if the agent is placed in the optimal location, and that navigation will have a role in this process. Another area of effector development is neural stimulation. Significant clinical efficacy has been found with localized stimulation for pain of spinal origin [20], Parkinson’s disease [21], and essential tremor [22]. Studies are currently planned on the use of stimulation for the treatment of depression [23]. Stimulation techniques are still in their infancy, and the electrode arrays currently employed are orders of magnitude larger than the structures they stimulate; indeed, as currently employed, stimulation simply drives the implanted structure to the point where the neurotransmitters are depleted, and the net effect is akin to a sort of reversible functional lesion [24]. As our knowledge of the nervous system evolves, and specific targets are found to exert a

563

564

36


highly specific effect, it can be imagined that microscopic stimulation devices that can already be manufactured using integrated chip technology could be inserted to achieve a desired effect. The rate limiting factor is the knowledge of exactly where to place the device, and how to get the device to that point; both answers will be found in the further development of navigational devices like the StealthStation system.

Intra-operative Imaging As mentioned in the previous sections of this chapter, intra-operative imaging has become routinely employed in more complex procedures to improve the overall accuracy of surgical interventions. The implementation of these imaging technologies introduces additional complexities and delays into procedures already known for their complexity and length. For example, intraoperative MRI, particularly using high-field magnetic devices, imposes significant issues upon the surgeon in conducting a typical cranial procedure. It is to be anticipated that with further development the difficulty in using current imaging technologies will be reduced, and, as the threshold to use is reduced, the application of intra-operative imaging will increase. Of greater interest is the advent of new ways to ‘‘image’’ or ‘‘visualize’’ the surgical environment to further improve the surgical act. These techniques may or may not fall into what is normally considered imaging; however, as they do impact the overall situational knowledge of the surgical environment, they can be considered as a form of imaging insofar as that term is used to imply a device that improves knowledge of the surgical situation. One technique that is already being used is the application of fluorescence. By injecting a contrast agent that attaches to specific tissues within the surgical field, and exposing the agent to specific wavelengths of illumination, the tissue

so labeled will exhibit fluorescence to differentiate itself from surrounding, non-labeled tissue. A major issue with this technique is that surgery cannot proceed during specific wavelength illumination; therefore, in order to make maximal use of this technology, the illuminated field needs to be back-projected into the surgeon’s view of the field under normal illumination conditions. The StealthStation system could implement this process by tracking a device viewing the surgical field, such as a microscope, taking an image of field, and then superimposing the illuminated field onto the field visualized under normal illumination. As agents evolve to label specific components of the brain (normal or abnormal), it can be anticipated that surgical interventions will be improved by this additional knowledge gained during surgical exposure. Another ‘‘imaging’’ technique could employ sampling and analysis of tissue with respect to the specific properties of the tissue. For example, if the target tissue had specific chemical characteristics, then analysis with a miniature mass spectrometer could allow resection based upon the presence of that chemical characteristic. Alternatively, analysis of tissue DNA, perhaps made possible by DNA chip technology, could allow resection based upon the elaboration of specific genetic traits, which could prove quite useful particularly in the resection of malignant glial neoplasms. Finally, navigation through the brain could be performed using the specific signals present within the brain. Current neurophysiologic techniques, using large macro electrodes, must undergo tremendous development to be scaled to the environment at which the brain routinely operates. By using microscopic electrodes, on the scale of a neuron, signals could be recorded that when analyzed (probably requiring computational power rivaling that of the brain) would ‘‘show the way’’ to implant the microscopic effectors discussed in the prior section.


Conclusions The StealthStation system has seen tremendous improvement and development in the nearly two decades of its existence. The improvements in surgical technique made possible by the device will only be exceeded by further development of the device. It can be foreseen that, in the near future, almost every neurosurgical device will employ some form of navigation to improve its efficacy and minimize the complications associated with the surgery. This is perhaps the most exciting time to be a surgeon in the field, as these improvements in surgery will redefine what a surgeon can accomplish, and in the process, literally redefine what it means to be a neurosurgeon. Any views, ideas, opinions or historical matter expressed in this paper are solely attributed to the authors. None of the material in this paper should be construed as the views or representations of any third party, including in particular Medtronic, Inc. or any of its divisions or subsidiaries, and such third parties assume no legal liability for the accuracy, completeness, or usefulness of any information contained herein.

7.

8.

9.

10.

11.

12.

13.

14. 15.

References 16. 1. Horsley V, Clark RH. The structure and functions of the cerebellum examined by a new method. Brain 1908;31:45-124. 2. Spiegel EA, Wycis HT, Marks M, Lee AS. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 3. Gildenberg PL. Whatever happened to stereotactic surgery? Neurosurgery 1987;20:983-7. 4. Heilbrun MP, McDonald P, Wiker C, Koehler S, Peters W. Stereotactic localization and guidance using a machine vision technique. Stereotact Funct Neurosurg 1992;58(1–4):94-8. 5. Heilbrun MP, Brown RA, McDonald P. Real-time threedimensional graphic reconstructions using Brown-RobertsWells frame coordinates in a microcomputer environment. Stereotact Funct Neurosurg 1985;48:7-10. 6. Watanabe E, Watanabe T, Manaka S, Mayanagi Y, Takakura K. Three-dimensional digitizer (neuronaviga-

17.

18. 19.

20.

21. 22.

36

tor): new equipment for computed tomography-guided stereotaxic surgery. Surg Neurol 1987;27(6):543-7. Kelly PJ. Computer-assisted stereotaxis: new approaches for the management of intracranial intra-axial tumors. Neurology 1986;36(4):535-41. Friets EM, Strohbehn JW, Roberts DW Improved accuracy for the frameless stereotaxic operating microscope. In Proceedings of the 14th Annual Northeast Bioengineering Conference, Durham, New Hampshire; 1988. p. 47-9. Watanabe E, Watanabe T, Manaka S, Mayanagi Y, Takakura K. Three-dimensional digitizer (neuronavigator): new equipment for computed tomography-guided stereotaxic surgery. Surg Neurol 1987;27(6):543-7. Maciunas RJ, Fitzpatrick JM, Galloway RL, Allen GS. Beyond stereotaxy: extreme levels of application accuracy are provided by implantable fiducial markers for interactive image-guided neurosurgery. In: Maciunas RJ, editor. Interactive image-guided neurosurgery. Park Ridge, IL: AANS; 1993. p. 259-70. Maurer CR, Jr, Fitzpatrick JM, Wang MY, alloway RL, Jr, Maciunas RJ, Allen GS. Registration of head volume images using implantable fiducial markers. IEEE Trans Med Imaging 1997;16(4):447-62. Laine T, Lund T, Ylikoski M, Lohikoski J, Schlenzka D. Accuracy of pedicle screw insertion with and without computer assistance: a randomised controlled clinical study in 100 consecutive patients. Eur Spine J 2000; 9(3):235-40. Brock M, Ingwersen I, Roggendorf W. Ultrasonic aspiration in neurosurgery: Neurosurg Rev 1984;7(2–3): 173-7. Bulsara K, Sukhla S, Nimjee S. History of bipolar coagulation. Neurosurg Rev 2006;29(2):93-6. Fork F. Endoscopy of the upper gastrointestinal tract: radiology of the stomach and duodenum. Berlin: Springer; 2008. p. 29-72. Shah N, Hemal A, Menon M. Robot-assisted radical cystectomy and urinary diversion. Current Urol Rep 2005;6(2):122-5. Hockstein N, Gourin C, Faust R, Terris D. A history of robots: from science fiction to surgical robots. J Robotic Surg 2007;1(2):113-18. Joshi S, Ornstein E, Bruce J. Targeting the brain. Neurocrit Care 2007;6(3):200-12. Lawson H, Sampath P, Bohan E, et al. Interstitial chemotherapy for malignant gliomas: the Johns Hopkins experience. J Neuro-Oncol 2007;83(1):61-70. Deer T. Current and future trends in spinal cord stimulation for chronic pain. Curr Pain Headache Rep 2001; 5(6):503-9. Watts C. Deep brain stimulation for Parkinson’s disease. Acta Neurochirurgica 2008;150(1):97. Pilitsis JG, Bakay RAE. The future of deep brain stimulation. In: Tarsy D, Vitek JL, Starr PA, Okun MS, editors. Deep brain stimulation in neurological and

565

psychiatric disorders. Clifton, UK: Humana Press; 2008. p. 571-91. 23. Greenberg B. Deep brain stimulation in depression: background, progress, and key issues. In: Tarsy D, Vitek JL, Starr PA, Okun MS, editors. Deep brain stimulation in

neurological and psychiatric disorders. Clifton, UK: Humana Press; 2008. p. 511-29. 24. Stojanovic M. Stimulation methods for neuropathic pain control. Curr Pain Headache Rep 2001;5(2):130-7.

46 Virtual Reality in the Operating Room P. L. Gildenberg

Modern imaging techniques have made the computer reconstruction of three-dimensional volumetric anatomy and pathological targets, such as tumors routine. They can be derived from preoperative CT or MRI scans, or the merging of both, sometimes with the addition of other modalities. The problem still to be solved is What is the most efficient and intuitive way to present this valuable three-dimensional information to the surgeon during an operation? It is virtual reality image guided surgery. In virtual reality surgery, the target tumor as well as significant surrounding or overlying structures are visualized on a monitor as a localized opaque image of the structure, superimposed on a real time video of the surgical field. The images displayed are those of significance to the immediate surgical need, under control of the surgeon by use of the surgical instrument presently in hand. The display of necessary information and none other can be selected. The depth of the image under resection can be determined by the position of the resection instrument and that image can be frozen or updated by means of a foot pedal. In addition, auditory signals can be added to inform the surgeon about the actual position of the instrument in hand in relation to the desired position. The technology to achieve virtual reality surgery is straight forward. A video camera is localized in space, using the same image guidance system as localizing the anatomy or any Electronic supplementary material Supplementary material is available in the online version of this chapter at http:// dx.doi.org/10.1007/978-3-540-69960-6_46 and is accessible for authorized users. #


other instrument. The optical characteristics of the camera are known, so the view that the camera sees can readily be registered to the surgical field. The monitor simultaneously displays both the image from the camera and a virtual reality computer generated image of the intended target or other anatomy accurately superimposed. The same registration can be used with an endoscope, which is essentially a video camera with specialized optics attached. In order to be useful for intraoperative image guidance, the three-dimensional anatomy and target volume (1) is reconstructed from a series of two-dimensional imaging slices into a three-dimensional volume, (2) is registered to the patient’s anatomy and image guidance system for localization, (3) is under direct control of the surgeon, (4) is interactive with the surgical instruments, which can be used to control the imaging, (5) is updated to represent the remaining tumor and the part of the tumor being resected at the time, and (6) is compatible with techniques under development to account for shift of the brain or other tissue. The question becomes how can one present such three-dimensional information to the surgeon while limited by two-dimensional displays without making it necessary for the surgeon to change views repeatedly between the surgical field and the monitor or to wear stereoscopic glasses or other modification of vision? Current image guidance techniques identify from preoperative imaging studies not only intracranial pathology but landmarks that may be used to register the patient’s head to the image guidance apparatus and guide the surgeon to the target.

726

46

Virtual reality in the operating room

Landmarks may be based on surface contours or may be based on the localization of several fiducial marks which may be taped to the head prior to the preoperative CT or MR scan. The image guidance computer constructs a volumetric virtual image of the head and face, the fiducials taped to the head, the internal anatomy, and whatever mass is identified as the target. The volumetric image is registered to the patient’s head, which is registered to the same stereotactic space as the image guidance system, which allows the surgeon to identify where a probe or instrument is in relation to the patient’s head and its contents. Presently used monitors present twodimensional pictures. Stereoscopic or 3-D images displayed on a 2-D monitor may use an artificial perspective to provide a simulation of a threedimensional world. Even with stereoscopic viewing techniques, depth perception may not be accurate enough for surgical guidance, and wearing stereoscopic glasses makes it difficult to see the surgical field. The configuration of the image display in presently commercially available image guidance systems is very similar and has not changed appreciably since image guided surgery was introduced almost two decades ago. In the present generation of image guidance systems, the images that are displayed on the monitor generally show three planes which intersect at right angles at the tip of the pointer the surgeon holds in his or her hand. The surgeon must hold the pointer at the point of reference on or within the head, look away from the operative field to the monitor, and then, while looking away from the surgical field, reconstruct in his or her mind the three-dimensional virtual concept of the brain to determine the location of the tip of the pointer in relation to the head, and then look back to the surgical field. A fourth image on the monitor may be a rendered image of the head or slices or wedges through the head, which has the appearance of being three-dimensional, but is actually a two-dimensional display. How can a system be organized so the surgeon can see both the localizing information on the

monitor and the surgical field at the same time? There are two possible ways to show the localizing information and the surgical field on the same monitor – one can move the localizing information to a view of the surgical field (as with the Compass system or heads-up display through an operating microscope [1]), or move a real-time picture of the surgical field to the monitor, which is done routinely in endoscopic surgery, which does not have image guided localization of the surgical field. Present day endoscopes contain a video camera to display the surgical field on a monitor, which the surgeons looks at to guide the surgery. The surgeon can perform image guided surgery while looking at a similar monitor, which contains the needed localization information in addition to a real-time view of the surgical field. Just the video camera display of the endoscope is used in virtual reality surgery, and that is localized in space by the same image guided surgery presently used. A localized picture of a large surgical field is displayed, along with a virtual reality image of, for instance, the tumor that lies beneath the surface of the tissue, the blood vessels that may be of concern to the surgeon, other eloquent avoidance areas or structures, or anatomical structures that may serve as landmarks to the surgeon. Alternatively, the surgical field can be visualized on a video monitor that has the localizing information displayed in virtual reality superimposed on the real time view of the operative field. The surgeon sees a picture of the localizing information superimposed on a picture of the real anatomy, along with a real-time image of the surgical tools and tissue.

How does Virtual Reality Image Guidance Compare with Conventional Image Guidance? In conventional image guided brain tumor surgery, the images obtained by preoperative CT


or MRI scanning are reconstructed into a threedimensional image of the skull and its contents. The surgery ordinarily begins with the registration ritual, to indicate the location of the head in three-dimensional space, even before the surgical field is sterilized and draped. A system of applying sterile field localizing fiducials is mounted so localization reference information remains available within the sterile field. The entry point is localized with the pointer, using the three-dimensional localization information derived from the preoperative scan. The pointer is then set aside and the scalp is incised, the bone flap opened and the approach is designed. Often the pointer and all the localization information within the computer are not used again until final confirmation of the extent of resection, so that valuable and available localization information is underutilized or not used at all during the resection. At other times, the surgeon may use localization repeatedly as resection is done. Each time, the surgeon interrupts the surgery to set down the resection instrument, pick up the pointer, place the poinert at a point of interest, look back and forth between the surgical field and the monitor to determine how accurately he or she may have approximated the position of the pointer, decide how the surgery should proceed, put down the pointer, pick up the resection instrument, look back to the surgical field, and continue the surgery. The inefficiency of this exercise discourages the surgeon from using image guidance frequently and repeatedly throughout tumor resection. At the conclusion of resection, the surgeon may apply the pointer to the base of the resection cavity to estimate how complete the resection may have been. In virtual reality image guided brain tumor surgery, preoperative planning is done on previously obtained merged images, just as in image guided surgery. Any scanning modality that can be input is acceptable, and many times multiple modalities are merged to obtain the maximum

46

targeting information possible. CT and/or MRI may be used, depending on which shows the target best. Functional MRI can be used to detect eloquent areas to be avoided. Threedimensional angiography can be incorporated, or, if that is not available, major vessels can be detected on enhanced MRI studies. To use tumor resection as an example, the intended resection line around the tumor can be indicated automatically as the tumor surface, and then edited by the surgeon, or the surgeon can map the line of resection manually on each slice. The various structures can be color coded – for instance, green for the tumor, red for arteries, blue for venous structures, etc. The volume of the ventricle can be displayed, either for additional orientation, to plan the surgery to avoid entering the ventricle, or an approach to a mass within the ventricle. Eloquent areas and other avoidance structures can also be indicated. In virtual reality tumor surgery, the entire tumor volumein-space is addressed, rather than just the pointin-space that is used in conventional image guided surgery. In virtual reality surgery, the video camera has fiducials so the position and direction can be localized, so the video image is also registered to the head. The resection instrument also can be localized with fiducials mounted on the handle out of the grasp of the surgeon, so the computer can determine where the tip of the instrument is in relation to the tumor. The video camera is positioned stereotactically just out of the line of sight of the surgeon, either with a stereotactic head frame (> Figure 46-1) or with a frameless system, so the computer knows the exact location of the video camera in space. Consequently, it can accurately determine the location of the virtual image, which then matches with the real-time video image. The video camera shows a real time image of the surgical field. Repositioning the video camera allows the surgeon to see an alternate approach to deep structures, so superficial vascular structures can be avoided. Structures can be visualized as either

727

728

46


. Figure 46-1 The video camera is mounted on the CRW1 frame aimed at the surgical field. The camera can be raised or lowered to show just the size of the field on the monitor

a solid mass or just an outline of the surface of the mass, which for a tumor generally represents the line of resection. The scalp is visualized in the monitor. An image of the tumor within the brain is projected on the surface of the scalp, with perspective adjusted so the actual size is seen, along with other selected structures (> Figures 46-2 and > Figures 46-3a). Last minute adjustments for the approach can be made, for instance, to avoid vascular structures, which can be displayed along with the tumor, to use the best area of the scalp, or to obtain an ideal opening in the skull. With a sterile pen, the surgeon outlines the tumor on the scalp while looking at the monitor (Video A). This allows him or her to design the smallest scalp incision that will provide sufficient access, so brain shift is lessened and the surrounding cortex remains protected, so there is less chance for a neurological deficit from damage to the surrounding cortex, which has led to a perceptibly faster post-operative course. A smaller than conventional opening may be used, so most often a lazy-S incision rather than a flap is sufficient. The scalp is incised and retracted. The tumor is then projected on the bone flap, and

. Figure 46-2 The tumor is projected onto the scalp and outlined with a sterile pen just prior to planning the scalp incision. The small window in the corner shows a lateral projection of the tumor, and the size of the tumor at the slice indicated is shown on the scalp in the monitor

the bone opening is verified. The bone and dura are opened, just enough to facilitate the intended surgical approach, leaving the surrounding brain protected outside the exposure. After the brain is exposed, the monitor shows a live image of the surgical field and resection instruments. The surgeon operates by


46

. Figure 46-3 Selected slices during tumor resection. (a) The tumor projected on the scalp. (b) The tumor lies just beneath the cortex. (c) The tumor during resection, about half way through. (d) The tumor is gone from the display. The outline is the ventricle, which was purposefully avoided so intracavitary chemotherapy could be administered at the conclusion of resection. Video A: The tumor is projected on the scalp and outlined with a sterile pen prior to planning the incision. Video B: The cortex is incised to approach the tumor just beneath. Video C: At a deeper level, the edge of the tumor lies just lateral to the edge of the craniotomy opening

looking primarily at the monitor mounted just above the surgical field, not unlike endoscopic surgery, although the actual surgical field is available for direct vision at any time. Since the resection instrument is also localized stereotactically, it can be used just like the pointer in conventional image guidance. Thus, the relationship between the tip of the resection instrument and the tumor volume is always apparent to the computer, which is necessary if the surgeon wishes to have an audio signal showing where the tip is, as described below. The point where the tumor will first be met is projected on the image of the surgical field, and a tract to the tumor is made. If the tumor is on the surface, good correspondence between the virtual and actual image should be noted. If the tumor is deep to the surface, the entry point on the cortex is selected as the point where the tumor is closest to the surface (> Figure 46-3b, Video B). If an eloquent area or blood vessel overlies the tumor in that

approach, the best entry point can be visualized by moving the video camera so there is no obstruction to the approach to the tumor, although that is usually anticipated and corrected during the pre-surgical planning. During surgery, planes bearing the outline of slices of the tumor at right angles to the surgeon’s eye view can be displayed. The location of the displayed slice in relation to the depth of the tumor is shown graphically in a small window at the corner (> Figures 46-2 and > 46-3). As the tumor is approached, just the closest edge of the tumor is displayed to guide the surgeon most efficiently to the tumor beneath the surface (> Figure 46-3b). Resection can be done by resecting along the line that indicates the surface of the identifiable tumor at the level being resected or the pre-planned line of resection. As the surgery proceeds deeper, the resection instrument is used to select the outline at which depth to be displayed. Since the resection instrument is localized by image guidance, an audible

729

730

46


signal can also be used to indicate whether the tip is within or outside of the tumor or at the proper line of resection (> Figures 46-4 and > 46-5). The tract to the tumor is developed and two opposing self-retaining retractors used to maintain access. The depth of the tumor beneath the surface can be seen in the lateral projection window. The size of the tract can be smaller than the tumor diameter, because it is possible to work more laterally at deeper levels. When the tract reaches the tumor, the self-retaining retractor . Figure 46-4 The tumor monitor shows the outline of the tumor at the level being resected and the resection instrument as held in the field

. Figure 46-5 Tumor tissue is resected beneath the edge of the craniotomy opening

blades can be bent to afford a larger access at the depth of the tumor. When a centimeter or more depth of the resection line is mobilized completely around the tumor, the debulking can be carried down to that level (> Figure 46-3c, Video C). The virtual image is upgraded to show the next centimeter or more of the tumor and the process is repeated. Each time the virtual tumor outline is readjusted to the level being dissected. When the virtual outline of the tumor disappears from the monitor, the resection is complete (> Figure 46-3d). The video camera that captures the live image can localized also with conventional image guidance, which provided more flexibility in the localization of the camera and the ability to have more than one instrument or device localized, such as the resection instrument or ultrasound transducer, which make it possible to use an audio signal for additional guidance. The development of virtual reality image guidance began with the use of the Radionics X-Knife software and the CRW1 frame [2,3]. The volumetric target was constructed with the same protocol as used for a radiosurgery target. I originally termed the system the ‘‘Exoscope,’’ since it involved working with a video image, similar to endoscopic surgery, but with the endoscope mounted outside the body. When the endoscope was omitted, it became ‘‘Videotactic surgery.’’ When the stereotactic frame was omitted, it became ‘‘virtual reality image guided surgery.’’ The first artificial targets used to prove both the concept and the accuracy were five volumes, a sphere, two pyramids, and two cubes, each 2.5–3.0 cm maximum dimension. They were secured within a skull, which was secured into a CRW head ring. The N-shaped fiducial array was attached, and a CT scan was done. The identifiable targets were the tip of the pyramid, the closest point on the surface of the sphere, and each accessible corner of the cube. The volumetric display of each of the test objects were produced with the XKnife software that was


designed to produce volumes that were targets for stereotactic radiosurgery. The CRW arc was secured to the head ring, and the pointer was directed to each of the target points. The endoscope was attached to the probe holder on the arc and coordinates were adjusted to show each of the test objects in turn. The real time image was turned off so only the virtual image was seen on the monitor. The surgeon touched each of the target points guided by the image on the monitor, and the observer measured any discrepancy between the hand-held pointer and each target point. Measurements were taken with the arc at a variety of angles, simulating various surgical approaches. In all cases, the target points were approached with accuracy within two mm in all cases and a mean of 1.5 mm. Once the accuracy of Videotactic surgery was demonstrated in the laboratory, it was used in surgery, at first with targets that were clearly visible even without stereotactic guidance, and later to guide resection of gliomas with very irregular shapes. I have performed 74 craniotomies for brain tumors using this system of virtual reality guided surgery [3]. The first generation software required a CRW frame. In all cases, the postoperative scan revealed that the intended resection had been accomplished. There were no complications related to the use of virtual reality image guidance. The patients had a shorter post-operative recovery than in my prior craniotomies, and the length of hospital stay averaged one day shorter. There are significant benefits of using virtual reality image guidance to obtain optimal tumor resection. There is reason to believe from this pilot study that addressing the tumor as its entire mass with virtual reality image guidance, rather than a series of points-in-space, may lead to a better controlled and more complete resection. Dissecting along the pre-planned resection line at the first encounter of the tumor conceivably allows the

46

surgeon to remove more of a glioma, which, may provide an improvement in length of survival if more than 95% of tissue identified on MR scan has been resected [4–6]. Excision of solid tumors en bloc with virtual reality image guided surgery may prevent dural seeding of tumor that may cause a recurrence of the tumor [7]. Many neurosurgeons fail to take advantage of the full use of image guidance because of potential brain shift. Some brain shift invariably occurs when the dura is opened and resection and retraction are done. Even so, localization reference to the pre-operative study can often provide a better determination of the location of the tumor and anatomical avoidance structures than the surgeon can estimate without image guidance, and the surgeon can minimize this inaccuracy and accommodate for it. Several means of minimizing brain shift can be used, as outlined in Chapter D-15 [8]. (1) The opening should be at the highest point to minimize loss of cerebrospinal fluid and consequent settling of the brain within the cranial cavity. (2) Aspiration of cerebrospinal fluid should be minimized for the same reason. (3) Retraction should be applied symmetrically, so as not to push the brain in one direction or another. (4) When the tumor is encountered, the plane at the edge of the tumor should be developed before gutting or decompressing the tumor itself, so that the plane of dissection is established while the brain or tumor shift is at the minimum, just the opposite as indicated in most non-guided surgery. A number of techniques are available to identify brain shift so the pre-operative scan can be updated during surgery. Most involve taking a CT or MRI during surgery. This is helpful not only to localize the tumor, but also to estimate whether the resection has been complete or whether additional resection is indicated. However, most operating rooms do not have that capability. Intraoperative ultrasound has been used to identify the target mass prior to dissecting the

731

732

46


approach to it [9]. Most often the ultrasound image is used without integrating the imaging data with pre-operative studies, the same as it would be in non-imaged procedures. There have been attempts to localize the ultrasound transducer stereotactically with the same type of fiducials used on the image-guided pointer, and this possibility lends itself readily to the virtual reality system described herein, so that the ultrasound image is also localized to the same stereotactic space and may be merged with pre-operative images, that may be morphed into the dimensions of the intraoperative ultrasound. The quality and information content of three-dimensional intraoperative ultrasound is improving rapidly, and may one day soon become the modality of choice to correct for brain shift. Endoscopic surgery: Other present systems rely on the surgeon operating from a video image. Endoscopic surgery could be improved significantly by registering the endoscope in space, using the same technology as described herein. Once the endoscope were localized, the field of view of the endoscope could be determined, and the surface that first lies in the field of view could be used with the virtual reality targeting information. In the configuration that shows the view of the endoscope on the monitor, localizing information can be superimposed on the view of the surgery, which would provide three-dimensional localization to the endoscope. DaVinci robotic surgery is performed while looking at a stereoscopic high definition image of the surgical field as seen by two video cameras attached to an endoscope. The surgeon looks with each eye at a stereo monitor at a location within the operating room away from the patient. The surgeon not only has an excellent stereoscopic view of the surgical field, but has access to hand and foot controls to guide two or three tools introduced through access ports. This is somewhat akin to operating with a surgical microscope, which provides excellent threedimensional images of the surgical field. If the

microscope is image guided, the target volume can be localized, and a localizing view of the circumference of the target can be superimposed on the view through the microscope with a heads up display. However, the microscope field of view may not encompass the entire circumference of the target tumor, or even a portion of the surface, which makes a microscopic display of a large tumor problematic. Since much tumor surgery does not require a microscope, the use of image guided microscopy may not be appropriate for many tumors. Audiotactic surgery: One advantage to the virtual images containing both the localization of the anatomy and the location of the surgical resection instrument is that auditory signals can be used to indicate the position of the tip of the instrument in relation to the intended resection line. For instance, a steady tone can be used if the instrument is within the tumor boundary and an interrupted tone if the instrument lies outside the tumor. The pitch of the tone can increase as the resection line is approached, so the highest pitched continuous tone indicates that the instrument is along the intended line of resection. In addition to tones, computer generated words can be used, just as in GPS automobile navigation, so the computer can warn of a ‘‘blood vessel on the right in 5 mm.’’ For training purposes, the attending surgeon can program all the admonitions he or she would tell the resident as the surgery proceeds. The pedicle of a vertebra can be drilled, with the surgeon looking at the operative field, rather than the guidance system, all the time directed by an audible signal. Thus, tumor surgery can be facilitated by adding a video camera, localized with image guidance and looking at the surgical field. Virtual images of the tumor and relevant anatomy can be graphically integrated with this video image, mostly using software already present in most image guided systems. By localizing the resection instrument to the same image guided space, the relationship of the instrument to the tumor


provides information that can be transmitted to the surgeon with an audible signal. Disclosure: The author holds patents on Videotactic, Audiotactic, and image guided endoscopic technology.

References 1. Rousu JS, Kohls PE, Kall B, Kelly PJ. Computer-assisted image-guided surgery using the Regulus Navigator. Stud Health Technol Inform 1998;50:103-9. 2. Gildenberg PL, Labuz J. Stereotactic craniotomy with the exoscope. Stereotact Funct Neurosurg 1997;68:64-71. 3. Gildenberg PL, Labuz J. Use of a volumetric target for image-guided surgery. Neurosurgery 2006;59:651-9. 4. Hentschel SJ, Sawaya R. Optimizing outcomes with maximal surgical resection of malignant gliomas. Cancer Control 2003;10:109-14.

46

5. Sawaya R. Neurosurgery issues in oncology. Curr Opin Oncol 1991;3:459-66. 6. Sawaya R, Hammoud M, Schoppa D, Hess KR, Wu SZ, Shi WM, Wildrick DM. Neurosurgical outcomes in a modern series of 400 craniotomies for treatment of parenchymal tumors. Neurosurgery 1998;42:1044-55. 7. Suki D, Abouassi H, Patel AJ, Sawaya R, Weinberg JS, Groves MD. Comparative risk of leptomeningeal disease after resection or stereotactic radiosurgery for solid tumor metastasis to the posterior fossa. J Neurosurg 2008;108:248-57. 8. Kelly PJ. Tumor stereotaxis. Philadelphia, PA: Saunders; 1991. 9. Hernes TA, Ommedal S, Lie T, Lindseth F, Lango T, Unsgaard G. Stereoscopic navigation-controlled display of preoperative MRI and intraoperative 3D ultrasound in planning and guidance of neurosurgery: new technology for minimally invasive image-guided surgery approaches. Minim Invasive Neurosurg 2003;46:129-37.

733

68 Cyberknife: Clinical Aspects F. C. Henderson Sr . W. Jean . N. Nasr . G. Gagnon

CyberKnife Stereotactic Radiosurgery Treatment of neoplasms of the spinal column and spinal cord can be most challenging. Spinal tumors are usually large, and surrounded by radio-sensitive structures – the spinal cord, heart, kidneys, airway and gastrointestinal system – the increased radio-sensitivities of which allows for a very small margin of error. Furthermore, the mobility and potential instability of the spine can introduce motion and potential difficulties with positioning during treatment. Malignancies of the spine are common. Of the 560,000 cancer related deaths each year in the USA, most result from metastatic disease [1]. The bone is the third most common metastatic site, after lung and liver [2], of which the spine is the most common target. The estimate of 100,000 new spinal metastases in the United States each year is probably low; the number of cases is far larger if one considers multiple lesions in the same patient. Bone pain is the most common pain syndrome requiring treatment in cancer patients, and is typically more severe than that due to visceral metastases. Patients with bone metastases survive longer, and become symptomatic earlier than those with visceral metastases. Furthermore, complications arise in 1/3 patients with bone metastases, and are usually associated with significant morbidity [1]. Treatment options for the spine are somewhat limited, because of its functional requirements as well as the significant internal or adjoining critical structures. The spine has several functions which it must perform in the healthy state – structural support, weight distribution, a basis for attachment of the appendicular skeleton, #


flexibility, protection of the spinal cord and nerve roots, hematopoeisis and mineral storage. In the diseased state, these functions are compromised, resulting in significant pain, functional impairment and deteriorating quality of life. Therapy is based upon the length of anticipated survival. Survival depends primarily on histology. Median survivals are 29.3 months for metastatic prostate cancer, 22.6 months for breast, 11.8 months for renal cell, and 3.6 months for lung cancer [3]. Survival is also a function of the number and location of other sites [4,5,29], performance status and a host of other parameters. Surgery is indicated primarily for decompression of neurological structures, stabilization of the spine and for severe, pain resulting from ‘‘micro-instability’’ – weakened structures that cause severe pain with the normal activities of daily living. Secondary indications for surgery include need for biopsy, the occasions in which resection of a solitary lesion may effect cure, or in those cases of recurrent lesion when all other treatment modalities have been exhausted. Stereotactic irradiation may be an effective adjunct to surgery. When severe pain, instability or neurological compromise is not present, the goals of stereotactic irradiation are to maximize tumor control, and where possible to achieve tumor ablation, while maintaining quality of life. This is accomplished by maximizing radiobiological efficacy and minimizing complications due to radiation toxicity.

What is the CyberKnife? The CyberKnife is a stereotactic radiosurgery system with a 6-MV X-band linear accelerator

1112

68

Cyberknife: clinical aspects

(LINAC) capable of delivering 600 monitor units (MU)/min. The robot delivers collimated beams 5–60 mm in diameter at target. The LINAC is mounted on a fully articulated robotic arm capable of six degrees of freedom allowing full rotational and translational movements (> Figure 68-1). Real time image-guidance is provided by two ceilingmounted diagnostic X-ray tubes with corresponding orthogonal floor-mounted amorphous silicon detectors which periodically acquire images during treatment to track bony skull, spinal anatomy or implanted metal fiducials. The beam is dynamically brought into alignment during treatment to account for patient/ target movement. Non-isocentric treatment allows the delivery of highly conformal and homogenous radiation doses to complex target volumes. The Gamma Knife uses isocentric planning, wherein the radiation beams intersect in a sphere, creating spherical dose distributions which are stacked adjacently to cover the lesion. The CyberKnife uses a linear algorithm to ‘‘paint out’’ any complex shape (> Figure 68-2). Steep dose gradients limit the radiation dose to surrounding normal

structures. Multiple treatment paths are available using approximately 110 nodes in space (robot positions) and 12 different pointing directions from each node, providing approximately 1,320 beam directions for selection. Flexibility to move with six degrees of freedom, end-to-end accuracy of less than 1 mm, non-isocentric targeting, and real-time image guidance allow the robot to accommodate moment-to-moment changes in patient position. The CyberKnife is particularly well-suited for treatment of spinal tumors [6–10].

Radiobiological Considerations Direct and Indirect Effects of X-irradiation The primary site of injury following irradiation appears to be the nucleus, and there is overwhelming evidence that DNA is the principle target of irradiation. X-irradiation causes breaks in chromosomes. The broken ends of the chromosomes may then rejoin in their normal position, may join

. Figure 68-1 The CyberKnife is a miniature linear accelerator (a) attached to a robot that moves with 6 degrees of freedom around the patient. The X-irradiation is collimated to beams ranging from 6 to 50 mm at the target. The patient position is adjusted in real-time to accommodate patient movement


68

. Figure 68-2 The difference between CyberKnife and Gamma Knife. (a) Gamma Knife uses converging beams to create spherical dose or distribution at selected isocenters. This ‘‘isocentric’’ planning may result in hot spots in regions of overlap and cold spots between the edges of the isocenters. (b) CSRS uses a linear algorithm that selects any of 1,300 beams to ‘‘paint out’’ the lesion; CyberKnife also uses an inverse treatment planning to avoid critical structures in a forward planning mode. While hot spots also occur in CK/SRS, the linear algorithm is more useful for complex and non spherical shapes

to other broken chromosomes, or may fail to rejoin; the results are chromatid aberrations (before telophase) or chromosomal aberrations (after telophase). Chromosomal injury may be sublethal, potentially lethal or lethal. Sub-lethal cell injury is potentially reparable, but may accumulate with other sub-lethal injuries with further irradiation to become lethal injury. Repair of sub-lethal chromosomal injury is the repair of double strand breaks, and usually occurs within 5 h of irradiation, between radiation fractions [1]. Sub-lethal chromosomal repair is dependent upon cell metabolism, oxygen and nutrients. Potentially lethal injury can be modified in the post-irradiation environment: inhibition of mitosis can allow chromosomal repair. Lethal injury is unredeemable. Most single stranded DNA injuries produced by irradiation are repaired, the opposite DNA strand serving as the template. Double strand breaks which are separated along the chromosome can also be repaired in the same manner. When the double strand breaks occur close together, a ‘‘mitotic cell death’’ usually ensues. Whilst many aberrations of the DNA stand are possible, there are three which are lethal combinations: the ring, the dicentric and the anaphase bridge [11]. Tumor cells have a broad range of radiosensitivity – the more radiosensitive cells

manifesting relatively more apoptotic changes, and the more radio-resistant tumors almost exclusively mitotic cell death. While cell death may be the result of induced apoptosis, most cell death in tumors is the result of mitotic cell death. Cells die when they attempt mitosis with injured chromosomes. Cell kill is the result of the abnormal recombination of chromosomes resulting from two double-strand breaks and the formation of lethal chromosomal aberrations. An important factor in fractionated irradiation is tumor repopulation. Tumor doubling time is approximately 28 days for prostate cancer, 14 days for breast cancer, and 4 days for head and neck cancers. Tumor cell repopulation, though an important factor when considering multifraction regimens with rapidly growing tumors, is not an important factor when considering stereotactic irradiation in one to five fractions.

Early and Late Effects of Irradiation The goal with staged radiosurgery is to maximize lethal cell kill and to minimize lethal injury to normal tissues. The effects of irradiation should be considered in terms of early and late effects. Early effects include injury to most neoplastic

1113

1114

68


cells and to normal tissues like skin, hair follicles and esophagus; they occur within a month of treatment. Late effects include peritumoral brain necrosis and transverse myelitis, and are due to loss of functional parenchymal cells; the delay, due to slow parenchymal cell turnover, arises 6 months to 5 years after treatment. The magnitude of early effects is related to the number of fractions and to the overall time over which they are administered. The magnitude of Late effects on the other hand is determined by the size of the individual fractions. In general, therefore, multi-session treatments will enhances early effects (tumor cell kill), and decreases late effects(CNS injury). Radiation in general is complicated by the presence of significant dose-limiting structures within and surrounding the spine: the spinal cord, spinal nerve roots, peripheral nerves, esophagus, bowel, kidneys, lung, and heart. The dose tolerances of central nervous critical structures have been determined through bitter experience: brain necrosis occurs with >4,000 cGy in 10 fx; optic myelitis with a single fraction >800 cGy; radiation myelitis in thoracic cord with 3,500 cGy in 10 fx administered over 12 cm of spinal cord length. With standard fractionation, there is a Figure 68-4). The a/b ratio, is that dose in Gray, where linear (a) cell kill equals quadratic (b) cell kill. The a/b ratio reflects

68

. Figure 68-4 The negative logarithmic curve of tumor cell survival, showing the upper linear portion – the alpha cell kill – and the lower curved portion of the curve representing the quadratic beta cell kill. The a/b ratio is that dose in Gray where the a cell kill is equal to the b component of cell kill. The a/b ratio is low for the CNS and tumors such as chordoma, and high for most adenocarcinomas

the susceptibility of a given tissue to injury from irradiation to a hypofractionated regimen versus standard fractionation: neuroblastoma, most lung cancers and skin have a high a cell kill, and therefore, high a/b ratios, approximating 10 Gy, whereas meningioma, chordoma, prostate cancer, and normal central nervous system are considered ‘‘slow reacting,’’ have a low a cell kill, and therefore a low a/b ratio, less than 3. The CNS is thought to have an a/b ratio of 2.2 Gy. In general, radiation sensitive tumors, such as neuroblastoma are sensitive to irradiation, and rapidly respond. Slow reacting cells, such as the CNS, respond more slowly. The a/b ratio offers a method to compare fractions of different size. The Linear quadratic Formalism, described by Fowler, allows comparison of the relative biological equivalent dose (BED) of one fractionation regimen with another [3].

The Linear Quadratic Formulation Biological effect E of a given dose D is given by the cell kill due to the linear component a D + cell kill due the quadratic component bD2.

1115

1116

68


Therefore, E ¼ aD þ bD2 This rearranges to BED ¼ ndð1 þ d=a=bÞ [3]:

[TV] (> Figure 68-5). The New conformity index takes into account the location of the prescription volume with respect to the target volume by dividing the ratio of [PV] to [TVPV] by the coverage. New conormity INDEX

Mathematically, the formulation begins to break down for hypo-fractionation (1 cm triggers an e-stop). Exact repositioning of the patient with each session is thus not required, and multisession treatments pose no difficulty [10]. The beam to target accuracy has been defined at this institution and others to be 0.5–0.7 mm.

Rationale for Multi-Session Treatment Recognizing that the radiobiology of high dose, single fraction irradiation remains unclear and falls outside of the domain of present radiobiological theory, we have chosen use multi-session treatments (fractionation) to achieve greater conformity and safety through the temporal partitioning of dose and exploitation of radiobiological differences between target and normal tissue. Our treatment strategy, to minimize late effects, and maximize tumor control probability, is based on seven principles. First, the low a/b ratio of normal brain and spinal cord predicts that for a given dose to

68

tumor, the adjacent CNS tissue will see a higher biological equivalent dose (BED); conversely, administering the total dose in 3–5 fractions, will decrease the BED to the CNS. Thus, the inherent difference of a/b ratio between tumor and CNS can theoretically be exploited with staging of treatment (administering in multi-sessions). Second, normal CNS tissue more faithfully undergoes repair of DNA injury than does tumor. Third, fractionated irradiation theoretically results in improved cell kill through ‘‘cell reassortment’’: cells passing through the cell cycle tend to hold up in the late G2 and M phase as a result of DNA injury, and the resulting expression of check-point genes which prevent spindle formation until DNA repair is completed; it is in these stages – M and late G2 – that tumor cells are most sensitive to irradiation. Fourth, to some extent, irradiation results in opening up of arterioles and increased oxygenation of the tumor, which imparts greater radio-sensitivity after the first dose. Fifth, small errors introduced into the treatment will tend to average out with fractionation. Sixth, treatment times are faster; and seventh, early side effects are less. The general dose strategies at GUH are shown > in Table 68-1. Generally, we treat the clinical treatment volume (CTV), which includes the gross tumor volume (GTV) plus the surrounding margin of tissue at risk for microscopic disease. The average volume of tissue irradiated is 60 cm3, with a range of 1–850 cm3. The tumors occupy all levels of the spine with a proclivity to the cervicothoracic and thoracolumbar junctions.

Neuro-Surgical Considerations The neurosurgeon is an integral member of the radio-surgical team. The surgeon is responsible for determination of which patients are reasonable candidates for surgery as opposed to radiation, assessment of neurological and spinal stability, and availability if the patient deteriorates during Radiosurgical treatment. The neurosurgeon probably best understands the urgency of the spinal

1117

1118

68


lesion, and should have input into the planning and timing of radiosurgery. The contouring should be performed by the neurosurgeon, who best understands the anatomy, likely regions of spread, and location of critical structures such as the spinal cord within the spinal canal. As the surgeon steps into the realm of radiation delivery, he assumes a new responsibility for following the patient after irradiation. Though initially engaged in treatment of a central nervous system cancer, the neurosurgeon, en passant, will inevitably encounter other lesions and become involved in treatment of non-CNS, medical problems that arise.

Illustrative Cases Case #1 Eighty-eight-year-old Caucasian female presented with a 1 year history of increased difficulty walking. . Table 68-1 Dose/fraction (cGy)

# fractions

800

3

800

5

700

3

500

5

Examples of use Untreated spine: (met. breast, thyroid, colon, renal, bladder, melanoma, non small cell lung, small cell lung) Chordoma, chondrosarcoma, osteogenic sarcoma Retreatment of spinal lesions after external beam (met. renal, nsc lung, sc lung, breast, giant cell, melanoma, colon, cervix, bladder) Benign spinal lesions (neuroma, schwannoma, meningioma, hemangioma, ependymoma). Gross disease of less resistant histology (leukemia, lymphoma)

She developed progressive bilateral lower extremity weakness, foot drop, and sensory loss and eventually became wheel-chair bound. She also reported mid-lower back pain. At that time, she also suffered an acute myocardial infarction. MRI imaging of the spine on 12/03 revealed a 1.5 1.3 1.4 cm diffusely enhancing intradural, extramedullary mass at the level of T10–11. There was evidence of adjacent dural enhancement, and imaging characteristics were consistent with spinal meningioma (> Figure 68-6). The lesion was located posteriorly in the spinal canal with evidence of significant, but chronic, cord compression. On examination, she reported poor quality of life and a pain level of 90/100. She was unable to stand; strength was 2/5 in both lower extremities and an incomplete sensory level to pinprick was evident at L1. Steroids were initiated. Given her history of severe cardiac disease s/p CABG, recent MI, and severe peripheral artery disease, she was not felt to be an acceptable candidate for surgical resection. Tracking fiducials were placed percutaneously. . Figure 68-6 Sagittal MRI of the thoracic spine (T2 Weighted, providing a myelogram effect) shows the spinal cord severely compressed by an intradural, extra-medullary dural based tumor at the T11 level. The patient was in severe pain and densely paraparetic


The T10 meningioma was treated with the CyberKnife with a dose regimen of 2,500 cGy (5 sessions of 500 cGy) prescribed to the 80% isodose line (> Figure 68-7). Treatment was completed over the course of 7 days, and was extremely

68

well-tolerated without acute toxicity. Her VAS pain score improved dramatically from 90 prior to treatment to 10/100 at 1 month after treatment, and she was able to stand and take a few steps. She remains pain free at 3-year follow up.

. Figure 68-7 Treatment plan for T11 meningioma in the Illustrative Case #1.Clockwise from top left: the axial view through T11 shows that the tumor (red line with yellow dots) is almost completely covered by the 80% IDL (orange line); the sagittal view of the reconstructed CT shows the contour in this case occupying most of the spinal canal; the coronal view of the reconstructed CT showing that the tumor contour occupies most of the spinal canal: the various beam trajectories of the CyberKnife

1119

1120

68


Her SF-12 scores improved from 20 (MCS) and 22 (PCS) before treatment to 37 (MCS) and 31 (PCS) at 1 month after treatment and 57 (MCS) and 37 (PCS) at 2-year follow up. She is now neurologically intact and her ability to walk has returned to normal.

Case #2 A 27-year-old woman was diagnosed with Ewing’s sarcoma of the right T7 rib in the paraspinal region. She underwent conventional, external beam irradiation encompassing the rib and right side of the spine at the T6, 7, 8 levels in a foreign country. The patient did well for 2 years before suffering a recurrence in the spine at T7 (> Figure 68-8). She was retreated with external beam from T2 to T10 with 10 300 cGy. Approximately 1 year later, she presented to the emergency room at GU with severe pain (10/10) and progressive spastic paraparesis. As she had already undergone irradiation . Figure 68-8 Sagittal MRI (T1 weighted, contrast enhanced), showing recurrent Ewing’s sarcoma, previously irradiated. There is spinal cord compression, maximal at T8, but extending within the posterior longitudinal ligament from T7 to T9

beyond the tolerance dose for a 5% risk of myelitis, further irradiation was considered to be unsafe. Therefore, a surgical plan was entertained. Subsequent spinal angiography identified the Artery of Adamkiewicz at the T7 level, raising concerns about the safety of vertebrectomy to remove the recurrent tumor. Therefore, the patient underwent a posterior decompression, bilateral resection of the posterior and lateral structures (the facets and pedicles) and stabilization. Fiducials were placed at the same time, to allow tumor localization for CK/SRS. Even though the patient had been irradiated twice, and had received well over what was considered to be a safe radiation dose, her severe neurological condition militated for SRS. The surgeon performed the contouring of the tumor and critical structures, including the spinal cord (> Figure 68-9b). The treatment plan (> Figure 68-9a) was generated. An ideal plan would be 4 8 Gy to the sarcoma and margin. In this case, however, a lower dose was elected to lessen the risk of spinal cord necrosis. CyberKnife SRS was administered over 5 days (regimen of 5 450 cGy for a total dose of 22.5 Gy to the 70% IDL). Treatment was completed without complication. At 2 months there was no neurological deficit, no pain, and the patient had returned to work as a computer engineer. At 6 months, there was no evidence of the spinal sarcoma on MRI (> Figure 68-10) and PET. Unfortunately, the tumor recurred several months later, at 9 months, and was associated with multiple metastases. She died 16 months after CK/SRS treatment.

Results of CK/SRS Treatment at GUH Patient Selection Georgetown University Hospital acquired the CyberKnife system in 2002 and immediately applied the technology to treatment of spinal tumors. Each patient was prospectively evaluated by a multidisciplinary team. Two hundred patients with a


68

. Figure 68-9 (a) The CyberKnife treatment plan demonstrating the surgeon’s contour of the recurrent sarcoma tumor wrapped around the spinal cord, extending from the T6 toT8 level. The spinal cord (25% of the volume) received less than 1,600 cGy. (b) The radio-surgical plan of Illustrative Case#2. CT scan at the T8 level showing the X-ray beams wrapping around the spinal cord to minimize cord toxicity. Clockwise from top left: the treatment plan in axial view, sagittal view, beam entry pattern with tumor shown in red and critical structure (spinal cord) in green; the coronal view (lower left) of CT. Approximately 500 beams were used in this very complex plan to maximize conformity and minimize cord toxicity to the CNS

1121

1122

68


. Figure 68-10 Sagittal MRI (T2 weighted view) of the thoracic spine showing remission of the sarcoma 6 months later. The high signal of the vertebrae indicates previous irradiation as well as the recent CyberKnife SRS; the high signal reflects increased fatty replacement of the hematopoietic region (cancellous portion) of the vertebrae

total of 274 spinal tumor sites with both primary and metastatic tumors of the spine who met eligibility criteria for CyberKnife treatment were consecutively enrolled, and treated with the CyberKnife. Forty nine patients presented with primary tumors of the spine and 151 patients with metastatic disease. All spinal levels were treated, although thoracic levels predominated. No patients were excluded on the basis of pathologic subtype. The most common tumor types treated were breast cancer, non-small cell lung cancer, and sarcoma. Of the 274 tumor sites, 125 had received prior conventional irradiation to the site that was to undergo SRS. CyberKnife SRS was used in the initial management in 118, or following surgery in 19.

Pain Seventy six precent of patients reported some degree of pain prior to treatment, the mean

pain score being 40/100 with medication. Most patients were managed with narcotic analgesics and NSAIDS. A paired t-test was used to analyze change in pain. Statistically significant overall pain relief was seen at 1 month (p < 0.001), 12 months (p = 0.003), 24 months: (p = 0.001), 36 months (p = 0.0003) (> Figure 68-11).

Quality of Life Patients were followed with quality of life surveys. The mean physical component of quality of life remained stable, regardless of whether patients were treated for benign disease or malignant disease. Scores for the mental component of quality of life improved significantly (p.01). Mean PCS and MCS scores of patients treated with CyberKnife as initial management were not significantly different to those undergoing retreatment after previous radiotherapy (> Figure 68-12a,b).


68

. Figure 68-11 A durable and statistically significant overall pain relief was seen at 1 month (p < 0.001), 12 months (p = 0.003), 24 months: (p = 0.001), 36 months (p = 0.0003)

. Figure 68-12 (a) The mean physical component of quality of life showed a statistically significant improvement (p. 01). The improvement occurred and was similar in magnitude in both benign and malignant disease trended upward, but did not reach statistical significance. (b) Scores for the mental component of quality of life improved significantly (p. 01)

1123

1124

68


Survival Overall, the Kaplan Meier curve for those patients undergoing CK/SRS as a first modality enjoyed a survival of 15 months (> Figure 68-13). Those treated with CK/SRS as a salvage modality survived a mean of 10 additional months.

the center of the treatment field, there was no evidence of late effects or treatment related myelitis.

Treatment of Intracranial Lesions with CyberKnife Brain Metastases

Complications Most patients experience minimal or no side effects. Most acute complications were selflimited and mild. The most commonly reports toxicities were fatigue, nausea, esophagitis, dysphagia, and transient diarrhea. One patient developed breakdown at a surgical site requiring surgery for debridement and re-closure of the wound, one patient developed an intraperitoneal infection, and two patients who developed vertebral fractures in the irradiated spine. One of these had been previously irradiated with external beam. With exception of minor radicular symptoms from nerve roots passing through . Figure 68-13 Median survival for all metastatic lesions was 15 months

Although the development of brain metastases is often viewed as an end-stage event of cancer, modern aggressive management of brain metastases has significantly improved survival and functional outcome compared to statistics oftcited in the literature. The majority of patients with controlled intracranial metastases will die from systemic disease rather than from recurrence and progression of these metastases. Surgical resection was considered gold standard treatment of single metastasis to non-eloquent brain areas. However, physicians and patients alike are now choosing more often to treat these tumors with radiosurgery. Patients with cancer have usually gone through many physically taxing treatments by the time they develop brain metastasis, and the non-invasive nature of radiosurgery provides significant comfort and reassurance. Radiosurgery is the first-line treatment of brain metastases smaller than 4 cm in diameter in eloquent areas, especially when edema is minimal. The addition of whole-brain radiotherapy (WBRT) decreases tumor recurrence rate, but has not been correlated with better survival. Radiosurgery seems to be effective in controlling tumors that are so-called ‘‘radioresistant,’’ such as melanoma, renal call carcinoma and sarcoma, for which WBRT is rather ineffective. However, radiosurgery is seldom effective against cystic brain metastasis, and surgical treatment must be considered for these if they progress in size. Radiosurgery utilized as a boost to the tumor region after WBRT has also been shown to reduce the risk of tumor recurrence or progression.


Nishizaki et al. [20] recently reported results from CyberKnife treatment of multiple or large brain metastases. At 44 weeks median follow up, the local tumor control rate was 83%. No patients died from their intracranial disease, even though 21 patients were treated with additional CyberKnife sessions for new metastases during their follow up. The authors concluded that fractionated treatment with the CyberKnife is highly useful for controlling large brain metastases in patients with advanced cancer. Our dose-regimens are as follows: 1. 2. 3.

4.

Boost after WBRT: 14–18 Gy in single session For a single met (3 cm), or lesions considered complex by virtue of their shape, size or location: 2,400 cGy in three treatments Retreatment of complex lesions: 1,600–2,500 cGy in five sessions

Vestibular Schwannoma Although surgical resection remains the gold standard treatment of vestibular schwannomas, stereotactic radiosurgery is a reasonable alternative for most patients with small to moderate tumors, not exceeding 3 cm in the largest dimension and which do not compress the brainstem. Because it is less invasive than surgical therapy, it is also quickly becoming most popular choice of treatment for the internet-savvy patients, who first seek webbased information, and then hi-tech medical care. Although the longevity of its effect is still questionable beyond 15–20 years, the efficacy of radiosurgery on vestibular schwannoma is clear. Most reports in the literature cite a 5–10-year tumor control rate between 95 and 98%. The risk of brainstem, trigeminal or facial nerve injury has been reduced to 1% or less with modern treatment techniques. The possibility of hearing-preservation

68

is much more variable, and is dependent on the patient’s pre-treatment hearing status, and the technique of radiosurgical treatment. In a recent report, fractionated treatment achieved a hearing preservation rate 2.5-fold higher than singlesession radiosurgery (Andrews [21]). One of the best hearing-preservation rates reported in the literature comes from Chang et al. [22] who utilized the CyberKnife to treat vestibular schwannomas with marginal doses between 18 and 21 Gy given in three sessions. The tumor control rate in that study was similar to previous reports, but the hearing preservation rate of 74% in patients with serviceable hearing is noteworthy.

Cavernous Sinus Meningioma Given the cerebrovascular and cranial nerve morbidity of surgical resection of meningiomas in the cavernous sinus, radiosurgery is now the first-line treatment of small-to-moderate size tumors in this region. These tumors usually have a characteristic appearance on MRI images: brightly enhancing, almost always homogenous appearing tumors originating from the cavernous sinus; thus, surgical biopsy is rarely necessary. In addition to its efficacy in halting tumor progression, radiosurgery is also highly effective in reversing cranial nerve deficits when given promptly after the appearance of symptoms. The risk of injury to the optic apparatus can be kept reasonably low if there is a 2–3 mm gap between tumor and the optic chiasm, and if the chiasmal dose is 8 Gy or less for single-session treatment. Hasegawa et al. [23] recently reported their long-term outcomes for gamma knife treatment for these tumors. The actuarial 5- and 10-year tumor control rate is 94 and 92% respectively. Forty six percent of patients experienced improvement in their neurological function, whereas only 12% experienced any cranial nerve morbidity. We recently reported the experience from our own institution, treating meningiomas of the

1125

1126

68


skull base with the CyberKnife. Forty four percent of the patients had meningiomas in the sphenocavernous region, and all patients were treated with 25 Gy to the tumor margin in five treatment sessions. With this regimen, the short-term tumor control rate was 97% and only 6% of patients experience deterioration of any neurological function. Even with large tumors directly in contact with the optic apparatus, the risk of visual complication was acceptably low with the CyberKnife regimen of 25 Gy in five sessions. Schwannomas of the cavernous sinus are also treated with the same dose (> Figure 68-14a, b, c).

In this case, a neuroma recurred 6 year s/p gross total resection (> Figure 68-14a). The formal visual field showed a field cut superiorly. The patient was treated with 450 cGy 5 to 80% IDL; the Rt Optic nerve dose was Figure 68-14c). There was no alteration of vision over the follow-up period of 4.5 years. There have been no neurological deficits as a result of CyberKnife irradiation of benign lesions with multi-session treatments in the cavernous sinus.

. Figure 68-14 (a) CT, axial view through the orbits, contrasted. An irregular shaped lesion (Schwannoma) is contrast enhanced, and occupying the right optic neural foramen. (b) The formal visual field testing at the time of presentation shows a superior hemianopsia (black). (c) The visual field testing 6 months after irradiation, showing a minimal visual field deficit


Non-Neurosurgical Tumors The authors note that the CyberKnife has been very effective in the treatment of non-neurosurgical tumors. At GUH, CyberKnife is routinely used with great efficacy in the treatment of cancers of the head and neck, urogenital organs, pancreas, lung, and sarcomas of the extremities. A case is presented (> Figure 68-15a,b,c) that is emblematic of lung cases performed at GUH. The patient, a woman with poor lung function, underwent a Bx, yielding a diagnoses

68

of NSCLC. She was intermittently O2 dependent, and clearly not a surgical candidate. Fiducials were placed into posterior chest wall. She was contoured (> Figure 68-15b), and treated with CyberKnife (800 cGy 3). At 6 months, there was no evidence of tumor (> Figure 68-15c). Nasopharyngeal tumors are excellent candidate lesions for CyberKnife. At GUH, we have treated 10 patients treated to date with external beam irradiation (5,040 cGy in standard fractionation) followed by CyberKnife boost

. Figure 68-15 (a) Bx+ NSCLC in a woman with poor lung function, intermittently O2 dependent. Not a surgical candidate. Fiducials were placed into posterior chest wall. (b) Treatment plan of the lesion, encompassing the gross tumor volume and a margin. (c) No evidence of the primary lung cancer 6 months after irradiation

1127

1128

68


. Figure 68-16 Contouring of a nasopharyngeal tumor in preparation for CyberKnife irradiation. Large series have demonstrated excellent results using conformal external beam with a CyberKnife boost to the tumor and margin

structures, and of the physicist, who creates the treatment plan. A weekly planning session to discuss cases and treatment plans is a useful adjunct to daily discussions.

References

(200 cGy 10). There has been no evidence of disease in six at 3 years; three were lost to follow up, and one had a local recurrence and distant metastases. Contouring of a nasopharyngeal tumor in preparation for CyberKnife irradiation is shown (> Figure 68-16). Stanford University has noted excellent results: their patients enjoyed a 100% local control of nasopharyngeal carcinoma in 23 patients treated with a CyberKnife radiosurgical boost [24].

Conclusion Stereotactic radiosurgery is an interdisciplinary endeavor. The best treatment results from active input from the surgeon, who provides perhaps the best understanding of the anatomy and of course surgical management of the tumor, of the radiation oncologist who provides the greatest insight into dosing strategies and understanding of critical dose to previously radiated critical

1. Ang AK, Thames HD, Van der Kogel AJ, Van der Schueren E. Is the rate of repair of radiation induced sublethal damage in rat spinal cord dependent on the size per fraction? Int J Radiat Oncol Biol Phys 1987;13 (4):557-62. 2. Brenner DJ, Martel MK, Hall EJ. Fractionated regimens for stereotactic radiotherapy of recurrent tumors in the brain. Int J Radiat Oncol Biol Phys 1991;21(3):819-24. 3. Fowler JF, Joiner MC, Williams MV. Br J Radiol 1983;56 (668):599-601. 4. Walther HE. Krebsmetastasen. Basel: Bens Schwabe Verlag; 1948. 5. Weichselbaum RR, Beckett MA, Vijayakumar S, et al. Radiobiological characterization of head and neck and sarcoma cells derived from patients prior to radiotherapy. Int J Radiat Oncol Biol Phys 1990;19:313-9. 6. Chang SD, Main W, Martin DP, Gibbs IC, Heilbrun MP. An analysis of the accuracy of the cyberknife: a robotic frameless stereotactic radiosurgery system. Neurosurgery 2003;52(1):140-7. 7. Degen JW, Gagnon GJ, Voyadzis JM, McRae DA, Lunsden M, Dieterich S, Molzahn I, Henderson FC. CyberKnife stereotactic radiosurgical treatment of spinal tumors for pain control and quality of life. J Neurosurg Spine 2005;2:540-9. 8. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA, Kalnicki S, Welch WC. CyberKnife frameless stereotactic radiosurgery for spinal lesions: clinical experience in 125 cases. Neurosurgery 2004;55(1):89-98. 9. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA, Kalnicki S, Welch WC. CyberKnife frameless singlefraction stereotactic radiosurgery for benign tumors of the spine. Neurosurg Focus 2003;14(5):e16. 10. Ryu SI, Chang SD, Kim DH, Murphy MJ, Le QT, Martin DP, Adler JR, Jr. Image-guided hypo-fractionated stereotactic radiosurgery to spinal lesions. Neurosurgery 2001;49:838-46. 11. Hall EJ. Radiobiology for the radiologist: DNA breaks and aberrations, pp. 22–229. 12. Wigg DR, Koschel K, Hodgson GS. Tolerance of the mature human nervous system to photon irradiation. Br J Radiol 1981;54:787-98. 13. Schultheiss TE, Kun LE, Ang KK, Stephens LC. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995;31(5):1093-112. 14. Nieder C,MilasL, Ang KK. Tissue tolerance to reirradiation. Sem Radiat Oncol 2000;10:200-9.


15. Nakamura JL, Verhey LJ, Smith V, Petti PL, Lamborn KR, Larson DA, Ware WM, McDermott MW, Sneed PK. Dose conformity of gamma knife radiosurgery and risk factors for complications. Int J Radiat Oncol Biol Phys 2001;51(5):1313-9. 16. Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. Technical note. J Neurosurg 2000;93 Suppl 3:219-22. 17. Stafford SL, Pollock BE, Leavitt JA, Foote RL, Brown PD, Link MJ, Gorman DA, Schomberg PJ. A study on the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003;55(5):1177-81. 18. Friedman WA, Foote KD. Linear accelerator radiosurgery in the management of brain tumours. Ann Med 2000; 32(1):64-80. 19. Naoi Y, Cho N, Miyauchi T, Iizuka Y, Maehara T, Katayama H. Usefulness and problems of stereotactic radiosurgery using a linear accelerator. Radiat Med 1996;14(4):215-9. 20. Nishizaki T, Saito K, Jimi Y, et al. The role of CyberKnife radiosurgery/radiotherapy for brain metastases of multiple

21.

22.

23.

24. 25. 26. 27. 28.

68

or large size tumors. Minim Invasive Neurosurg. 2006;49 (4):203-9. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and fractionated stereotactic radiotherapy for the treatment of acoustic schwannomas: comparative observations of 125 patients treated at one institution. Int J Radiot Onc Biol Phy 2001;50:1265-78. Chang SD, Gibbs IC, Sakamoto GT et al. Staged stereotactic irradiation of acoustic neuroma. Neurosurgery 2005;56(6):1254-61. Hasegawa T, Kida Y, Yoshimoto M et al. Long-term outcomes of Gamma Knife surgery for cavernous sinus meningioma. J Neurosurg 2007;107(4):745-51. Tate DJ, et al. IJROBP 1999;45:915-21. Harrington KD. Orthopaedic management of metastatic bone disease. pp. 283–307. Herbert D. Prediction of response in radiation therapy 1988:400–513. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics 1999. CA Cancer J Clin 1999;49(1):8-31. Markman M. Early recognition of spinal cord compression in cancer patients. Cleve Clin J Med 1999;66(10): 629-31.

1129

60 CyberKnife: Technical Aspects J. R. Adler . D. W. Schaal . A. Muacevic

Robotic Radiosurgery – Overview The design of the CyberKnife System (Accuray Incorporated, Sunnyvale, CA, USA) is based on principles of radiosurgery that have been in clinical practice for over 30 years. The practice of radiosurgery involves the precise application of an ablative dose of radiation to a defined target volume while protecting the surrounding healthy tissue. During radiosurgery, many radiation beams from different directions intersect in the tumor region, where they accumulate to a total dose, while the surrounding healthy tissue receives only a small fraction of the radiation. Until recently, the Gamma Knife system (Elekta AG, Stockholm, Sweden) was considered the standard instrument for neuro-radiosurgical applications. In addition, some centers use linear accelerators (LINAC) for radiosurgical procedures that are most commonly used for conventional radiation therapy but can be altered for radiosurgical treatment. Detailed physics testing is part of this alteration process because radiosurgical applications demand significantly higher quality and precision than conventional radiotherapy indications. Gamma Knife and conventional LINACs both require the application of an invasive stereotactic frame onto the patient’s head to achieve the desired accuracy of +/ 1 mm. The revolutionary development of the frameless CyberKnife technology, which combines integrated image guidance and robotic technology, has led to a paradigm shift in radiosurgery. During CyberKnife radiosurgery, instead of a stereotactic frame, real-time intraoperative imaging is used to establish the tumor position with

#


reference to skeletal anatomy or implanted fiducial markers. In many clinical circumstances, the combination of image guidance and robotics has numerous advantages over conventional approaches to radiosurgery. For the first time in history, a truly non-invasive and pain-free radiosurgical treatment is available. Moreover, whenever deemed necessary, the treatment can be delivered in several fractions or sessions. This has the potential to be safer for the treatment of lesions in highly sensitive areas, i.e., meningiomas of the optic system, acoustic neuromas, or larger lesions [1–3]. In this chapter, we describe the technological background of CyberKnife robotic radiosurgery and its component parts.

Registration: Patient Alignment and Target Localization The primary control point of the CyberKnife operational system is a graphical user interface delivery system that initiates and monitors operations of the different components. During treatment delivery, the software monitors the system status and safety controls, reports errors, manages the patient database and records treatment data log files for post-treatment assessment and analysis. The target localization system (TLS) is composed of two orthogonally positioned diagnostic X-ray sources and two corresponding amorphous silicon plates. They provide near real-time digital X-ray images of the patient in the treatment position. During patient setup, the target position is determined (relative to nearby bony anatomy

950

60

Cyberknife: technical aspects

or, in extracranial treatments, implanted fiducials) by comparison of the left anterior oblique (LAO) and right anterior oblique (RAO) X-rays, with digitally reconstructed radiographs (DRRs) in the same LAO and RAO positions derived from the preoperative CT scans. The system uses a movable treatment table to position the patient before treatment. This operating table can be moved automatically along five axes (three translations and two rotations). The final rotational movement (yaw) is applied manually. The newly developed robotic couch (RoboCouch Patient Positioning System, Accuray Incorporated) can be adjusted entirely robotically (i.e., based on image information and without intervention by the user) in all 6 axes. > Figure 60-1 illustrates the overall system configuration.

The TLS is activated repeatedly throughout the treatment to verify the position of the target. Again, intraoperative X-rays are registered in real-time to a library of pre-operative DRRs that samples the full range of motion of the target center in 6 degrees of freedom (see > Figure 60-2). The registration process verifies that the real-time images represent an acceptable position within the range sampled by the reference images. It interpolates the actual position and orientation of the real-time image with respect to the reference, and sends positional correction data to the robot. Patient movements within certain limits (10 mm in x, y or z; 1 pitch and roll, 3 yaw) can be compensated for automatically by the robot. If a movement exceeds these limits, an error warning is triggered

. Figure 60-1 A recent configuration of the CyberKnife System. Orthogonal X-ray sources project to in-floor amorphous silicon detectors. A compact 6-MV LINAC is manipulated by an industrial 6-axis robot (KUKA Robotics Corp, Augsburg, Germany). The RoboCouch is a robotic patient positioning system. The Synchrony camera collects positional information from light-emitting diodes attached to a patient vest for use in tracking of moving extracranial tumors


60

. Figure 60-2 A library of digitally reconstructed radiographs (DRRs) is generated from the treatment planning CT. Each of these images, which approximate an oblique projection, emulates a unique pose of the patient’s anatomy

causing the system to pause for patient realignment. By verifying the position of the target frequently throughout the treatment in this manner, the CyberKnife System ensures the accurate delivery of radiation to the target without framebased targeting and immobilization. The clinically relevant accuracy of the entire system has been determined at several centers using thin-slice CT to be sub-millimetric for intracranial treatments [4,5]. The same measure of accuracy can also be achieved when ablating spinal lesions, whether tracking using implanted markers [6] or without these markers, but using the Xsight Spine Tracking System (Accuray Incorporated) [7,8]. Despite being frameless, the CyberKnife emulates an important feature of the Gamma Knife, its ability to deliver many beams from many, noncoplanar orientations. Unlike the Gamma Knife however, the CyberKnife is not restricted to treating isocenters; instead, at each position of the robot, the beam can be directed toward a different area of the target region. This design feature enables the treating surgeon to select from a large array of non-isocentric, non-coplanar beams during treatment planning, thereby creating dose distributions that conform especially well to irregularly shaped lesion volumes (> Figure 60-3) [9]. In contrast, conventional radiosurgical devices

are only capable of constructing spherical dose distributions around a discrete isocenter.

Surgical Guidance: 6D Skull Tracking and 6D Fiducial Tracking The CyberKnife enables the full, 6D range of motion of the target center to be tracked within the reference frame of the treatment planning CT. The original version of the CyberKnife System’s 6D skull-tracking algorithm was able to locate targets relative to known skeletal landmarks, such as the mastoid process [10]. Later versions extract this information from bony anatomy automatically, without relying on named landmarks [11]. The current algorithm employs a multi-phase registration strategy to achieve submillimetric targeting and tracking accuracy in near real-time [5]. The total clinical accuracy of such image-guided localization has been shown to be sub-millimetric in phantom studies using thin-slice planning CTs [4,5]. Arguably the greatest advantage of frameless, image-guided targeting and tracking is that this makes radiosurgery outside the brain possible. Extracranial application of the CyberKnife began with the groundbreaking work for lesions in the

951

952

60


. Figure 60-3 Panel a: In standard radiosurgery, dispersed isocentrical beams all intersect at a common region (light gray). Because multiple spherical volumes are needed to cover non-spherical lesions, the resulting dose distribution tends to be inhomogeneous. Panel b: Non-isocentric beams from various directions do not all intersect a single point and the dose is more homogeneous

spine by Ryu and co-workers [12]. Spinal targets were initially treated using radio-opaque fiducial markers (stainless steel screws) implanted in vertebrae adjacent to the lesion prior to preoperative imaging. In this technique, which is still necessary in certain circumstances, the fiducial tracking algorithm searches the pair of X-ray images, detects the fiducial markers, and calculates translations and rotations of the target. While only one fiducial is required to track anatomical translations, three or more are necessary to compute both translational and rotational motions [13]. CyberKnife’s total clinical accuracy for fiducial tracking in the spine is also sub-millimetric [6]. In recent years, system enhancements have facilitated fiducialbased tracking of targets throughout the body, including lesions that move with respiration, such as lung and abdominal tumors [13–17].

Spinal Tracking Without Fiducials Spinal radiosurgery is a new class of procedures designed for primary or adjuvant treatment of certain spinal disorders [18,19]. Because such large doses of radiation are administered, spinal radiosurgery, similar to its intracranial predecessor, requires extremely accurate targeting.

In contrast, the lack of precision inherent in conventional external beam radiation therapy, and the limitations of target immobilization techniques, generally preclude large single-fraction irradiation near radiosensitive structures such as the spinal cord. The frameless CyberKnife radiosurgery system has overcome these problems by using real-time image guidance which allows the paraspinous target to be tracked even in the presence of occasional patient movement. Continuous tracking and correction for motion of the spine throughout treatment is a requirement for spinal radiosurgery, because patients do move after set-up is complete [20]. Until recently, clinicians surgically implanted fiducials into the spine to track the movement of the lesion during treatment [12,21]. However, this step introduces the added surgical risks associated with an invasive surgery (albeit minor), lengthens treatment time, and reduces patient comfort. It would be ideal if it were possible to track spinal lesions using bony landmarks (similar to tracking intracranial lesions based on skull anatomy) instead of fiducials. Recently, such a fiducial-free spinal tracking system has been introduced (XsightSpine Tracking System, Accuray Incorporated). The Xsight fiducial-free localization process is performed in several stages, beginning with image


enhancement, in which DRRs and intra-treatment radiographs undergo processing to improve the visualization of skeletal structures. Prior to treatment, a region of interest (ROI) surrounding the target volume is selected based on an initial userdefined position, which is refined automatically by an algorithm that seeks to maximize the image entropy within the ROI. The resulting optimal ROI typically includes one to two vertebral bodies which form the basis of patient-motion tracking and alignment. 2D-3D image registration uses similarity measures to compare the X-ray images and DRRs, and a spatial transformation parameter search method to determine changes in patient position. A mesh is overlaid in the ROI, and local displacements in the mesh nodes are

60

estimated individually, constrained by displacement smoothness. Nodal displacements in the two images within the mesh form two, 2D displacement fields. 3D displacements of the targets and global rotations of spinal structures within the ROI can then be calculated from the two 2D displacement fields by interpolation. A screen shot from an Xsight treatment session shows that the overlaid-mesh technique is successful even in the presence of spinal instrumentation (> Figure 60-4). The main advantages of the fiducial-free system are: (1) the ability to account for nonrigid deformation, thereby improving the targeting accuracy in the situation that a patient-pose change occurs subsequent to the CT scan, and

. Figure 60-4 Xsight spine tumor tracking after dorsal transpedicular stabilization surgery. Bone tracking of the vertebra was possible even though the tumor area was overshadowed by the metal implants

953

954

60


(2) no risk of complication and increased convenience for both the patient and the clinician. The fiducial-free tracking system of the CyberKnife has been proven in end-to-end phantom tests and simulations, using existing CT-image sets of the spine, to be accurate to within about 0.5 mm [7,8].

system has been shown under laboratory conditions to result in highly precise radiation delivery without patient restraints [17,22,23]. Its utility in treating patients has also been documented in multiple institutions [13,15–17,24].

CyberKnife Team Robotic Motion Compensation The Synchrony Respiratory Tracking System (Accuray Incorporated) is one of the most advanced components of the CyberKnife System’s product configuration. It is only briefly described here as it is mainly used for non-neurosurgical, extracranial indications (i.e., lung, liver and pancreatic tumors). With Synchrony, tumor movement is detected by means of implanted fiducials, usually small gold seeds that are introduced into or around the tumor under CT fluoroscopy or via a bronchoscope. Waiting a few days after placement allows the fiducials to settle and edema to subside, after which a preoperative planning CT scan is made so as to define the 3D operative volume. During patient setup, these fiducials are used to align the DRRs generated from the planning CT with the X-ray images obtained in the treatment room. In the meantime, multiple light emitting diodes (LEDs) positioned on the patient’s chest or abdomen are monitored in real time (32 frames per second) by cameras mounted on the ceiling. A correlation model between the internal and external markers is created prior to treatment and is updated and verified throughout the treatment. The correlation between the positions of the external LEDs and internal fiducials enables the robot to move the LINAC synchronously with the patient’s breathing as it relates to the motion of the tumor inside. Displacement errors higher than a user-prescribed value based on an analysis of tumor motion triggers the system to pause so that adjustments can be made before continuing with the treatment. This

High quality radiosurgical applications are becoming increasingly more complex. Different treatment applications require a range of medical professionals, such as surgical specialist, radiation oncologists, radiologists, and specially trained medical physicists, to choose, plan, and conduct the most effective and safe procedure. Intracranial and spinal treatments should be reviewed by experienced neurosurgeons capable of understanding the complex topographical relationships of cranial and spinal anatomy and pathology. Orthopedic surgeons may also be helpful to support spinal treatments. The contribution of experienced imaging experts for optimal selection and interpretation of radiologic studies will often significantly enhance the quality of a radiosurgical procedure. Non-invasive robotic radiosurgery is a rapidly emerging interdisciplinary field that is opening new horizons in the area of cancer treatment and perhaps beyond, into select non-neoplastic disorders.

References 1. Adler JR Jr, Gibbs IC, Puataweepong P, et al. Visual field preservation after multisession cyberknife radiosurgery for perioptic lesions. Neurosurgery 2006;59:244-54; discussion 244‐54. 2. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation for acoustic neuroma. Neurosurgery 2005;56:1254-61; discussion 1253‐61. 3. Giller CA, Berger BD, Fink K, et al. A volumetric study of CyberKnife hypofractionated stereotactic radiotherapy as salvage for progressive malignant brain tumors: initial experience. Neurol Res 2007;29(6):563-8. 4. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the CyberKnife: a robotic frameless


5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

stereotactic radiosurgical system. Neurosurgery 2003;52: 140-6; discussion 146‐7. Fu D, Kuduvalli G, Mitrovic V, et al. Automated skull tracking for the CyberKnife image-guided radiosurgery system. In: Reinhardt JM, Pluim JP, editors. SPIE medical imaging: image processing. vol 5744. San Diego, CA: SPIE; 2005. p. 366-77. Yu C, Main W, Taylor D, et al. An anthropomorphic phantom study of the accuracy of Cyberknife spinal radiosurgery. Neurosurgery 2004;55:1138-49. Ho AK, Fu D, Cotrutz C, et al. A study of the accuracy of Cyberknife spinal radiosurgery using skeletal structure tracking. Neurosurgery 2007;60:147-56. Muacevic A, Staehler M, Drexler C, et al. Technical description, phantom accuracy, and clinical feasibility for fiducial-free frameless real-time image-guided spinal radiosurgery. J Neurosurg Spine 2006;5:303-12. Yu C, Jozsef G, Apuzzo ML, et al. Dosimetric comparison of CyberKnife with other radiosurgical modalities for an ellipsoidal target. Neurosurgery 2003;53:1155-62; discussion 1153‐62. Adler JR. Frameless radiosurgery. In: Goetsch SJ, De Salles AAF, editors. Stereotactic surgery and radiosurgery. vol 17. Madison, Wisconsin: Medical Physics Publishing; 1993. p. 237-48. Murphy MJ. An automatic six-degree-of-freedom image registration algorithm for image-guided frameless stereotaxic radiosurgery. Med Phys 1997;24:857-66. Ryu SI, Chang SD, Kim DH, et al. Image-guided hypofractionated stereotactic radiosurgery to spinal lesions. Neurosurgery 2001;49:838-46. Collins BT, Erickson K, Reichner CA, et al. Radical stereotactic radiosurgery with real-time tumor motion tracking in the treatment of small peripheral lung tumors. Radiat Oncol 2007;2:39. Brown WT, Wu X, Fayad F, et al. CyberKnife radiosurgery for stage I lung cancer: results at 36 months. Clin Lung Cancer 2007;8:488-92.

60

15. Le QT, Loo BW, Ho A, et al. Results of a phase I doseescalation study using single-fraction stereotactic radiotherapy for lung tumors. J Thorac Oncol 2006;1:802-9. 16. Pennathur A, Luketich JD, Burton S, et al. Stereotactic radiosurgery for the treatment of lung neoplasm: initial experience. Ann Thorac Surg 2007;83:1820-4; discussion 1824‐5. 17. Muacevic A, Drexler C, Wowra B, et al. Technical description, phantom accuracy, and clinical feasibility for single-session lung radiosurgery using robotic imageguided real-time respiratory tumor tracking. Technol Cancer Res Treat 2007;6:321-8. 18. Degen JW, Gagnon GJ, Voyadzis JM, et al. CyberKnife stereotactic radiosurgical treatment of spinal tumors for pain control and quality of life. J Neurosurg Spine 2005; 2:540-9. 19. Gerszten PC, Burton SA, Ozhasoglu C, et al. Radiosurgery for spinal metastases: clinical experience in 500 cases from a single institution. Spine 2007;32:193-9. 20. Murphy MJ, Chang SD, Gibbs IC, et al. Patterns of patient movement during frameless image-guided radiosurgery. Int J Radiat Oncol Biol Phys 2003;55: 1400-8. 21. Gerszten PC, Ozhasoglu C, Burton SA, et al. Evaluation of cyberknife frameless real-time image-guided stereotactic radiosurgery for spinal lesions. Eur J Cancer Suppl 2003;1:S151. 22. Seppenwoolde Y, Berbeco RI, Nishioka S, et al. Accuracy of tumor motion compensation algorithm from a robotic respiratory tracking system: a simulation study. Med Phys 2007;34:2774-84. 23. Wong KH, Dieterich S, Tang J, et al. Quantitative Measurement of CyberKnife Robotic Arm Steering. Technol Cancer Res Treat 2007;6:589-94. 24. Nuyttens JJ, Prevost JB, Praag J, et al. Lung tumor tracking during stereotactic radiotherapy treatment with the CyberKnife: Marker placement and early results. Acta Oncol 2006;45:961-5.

955

71 Focused and Conventional Radiation for Acoustic Nerve Tumors R. Den . S. H. Paek . D. W. Andrews

Introduction Acoustic neuromas, also referred to as vestibular schwannomas, are benign tumors commonly arising from the transition zone between central oligodendroglial cells and peripheral schwann cells within the vestibular portion of cranial nerve VIII. They account for approximately 80–90% of all cerebellopontine angle tumors in adults and have a median age of diagnosis of approximately 50 years [1]. The management of vestibular schwannomas has evolved over the past half century from a microneurosurgical modality to a non-invasive approach using focused radiation either as a single treatment stereotactic radiosurgery or as a fractionated course fractionated stereotactic radiotherapy. Stereotactic radiosurgery or SRS has been widely regarded as the standard of care for small to intermediate-sized tumors with the current standard dose of 12 Gy prescribed to the 50% isodose line. This current dose represents two decades of dose reductions that have resulted in decreased incidence of cranial neuropathies and improved hearing preservation rates while maintaining a high tumor control rate. A parallel development has occurred using fractionated stereotactic radiotherapy or FSR. FSR treatment courses may be as short as a week or as long as 5–6 weeks depending on the daily dose schedule. Fractionation has been thought to provide several radiobiological advantages over single fraction treatment resulting in higher rates of hearing preservation. This chapter is a #


review of acoustic neuromas and their treatment with the use of FSR.

Epidemiology In a review of 1,400 temporal bones, Schuknecht found an incidence of occult acoustic neuromas in 0.57%. According to the 1991 National Institute of Health Consensus Statement, an estimated 2,000–3,000 new cases of unilateral acoustic neuromas are diagnosed in the United States each year – an incidence of about 1 per 100,000 per year [2]. There has been an increasing incidence of asymptomatic lesions detected, which has been posited to be secondary to the increasing use of intracranial imaging modalities such as MRI or high resolution CT [3]. This indicates that the vast majority of acoustic neuromas that exist never become clinically evident, because of very slow or arrested growth. Five percent of all diagnosed tumors are associated with neurofibromatosis type 2 [4], an autosomal dominant disorder which results in bilateral acoustic neuromas and multiple meningiomas. This disorder is characterized by mutations in the gene product merlin, located on chromosome 22q12 [5].

Pathology Acoustic neuromas are benign tumors that arise from neural crest-derived Schwann cells. They

1152

71

Focused and conventional radiation for acoustic nerve tumors

are generally well circumscribed, encapsulated masses which on microscopic examination have two growth patterns: a high cellularity zone and low cellularity zone classified as Antoni A, compact tissue with spindle cells in palisades, and Antoni B, loose tissue with cyst formation [6]. Since tumors do not arise from the neurons, nerves are generally displaced by the tumor; however, they may become entrapped in the capsule. Malignant degeneration to fibrosarcoma is quite rare, with only case reports in the literature.

suggested that tumors smaller than 1 cm could be watched, especially in elderly patients. Bederson et al. compared patients who required surgery to those who did not require intervention and found that a larger initial size (2.7 cm vs. 2.1 cm) and a faster initial growth rate (7.9mm/year vs. 1.3 mm/year) were significantly prognostic [7]. Fucci et al. in a review of 119 patients found that tumors greater than 2 cm at presentation were more likely to grow than tumors less than 2 cm [19]. Currently, the underlying mechanisms, which explain the differential tumor growth seen in various patients has not been elucidated.

Natural History There are several reports, which detail the natural history of tumor growth in patients with sporadic acoustic neuromas [7–15]. Silverstein et al. demonstrated a growth rate of 0.2 cm per year in seven patients under surveillance [16]. Mirz et al. reported on 64 patients with acoustic neuromas, and over the median follow-up between 5 months and 15 years, 14 tumors (22%) regressed, 35 tumors did not grow or had only minimal growth (growth rate up to 1 mm/year), whereas 15 tumors grew > 1 mm/year [17]. Walsh et al. investigated the natural history of 72 patients with a radiological diagnosis of unilateral acoustic neuromas [15]. Over the median follow-up period of 37.8 months, they reported that significant tumor growth (total growth >1 mm) in 36.4%, no or insignificant growth (0–1 mm) in 50%, and negative growth ( Table 71-1. The literature reports of acoustic neuromas published from 1988 to 1998, document the average rates of trigeminal, facial, and cochlear neuropathies as 34, 33, and 40% respectively. This led to a reduction in marginal SRS dose prescription to the current standard of 12 Gy. Tumor control rates were not compromised until the prescribed dose was reduced to below 10 Gy [32]. While, rates of trigeminal and facial neuropathies were dramatically reduced, hearing preservation was suboptimal with rates ranging from 33 to 83%. This led to the investigation of fractionated radiation therapy as opposed to single dose treatment regimens. The hypothesis as mentioned above is that optimal dose prescription balances tumor kill and normal tissue survival. Multiple smaller doses of radiation can achieve an equivalent tumor effect while limiting toxicity. Fractionation allows for the cellular repair mechanisms within normal tissue time to correct any damage to DNA

1155

1156

71


. Table 71-1 Summary of published experience with reliable reports of audiometry

Reference

N

Isodose prescription

Gamma knife radiosurgery Hirsch [30] 126 18–25 Gy Flickinger [31] 85 14–20 Gy (18 median) Foote [32] 36 16–20 Gy Kondziolka [33] 162 12–20 Gy (16.6 mean) 1988–1998 mean results Andrews [34] 69 12 Gy Karpinos [35] 96 10–24 Gy (14.5 mean) Regis [27] 104 12–14 Gy Iwai, Y [36] 51 8–12 (12 median) Flickinger [37] 313 12–13 Gy Van Eck [38] 78 13 Gy Hasegawa [39] 74 13 Gy 1999–2005 mean results Fractionated stereotactic radiotherapy Kagei [40] 39 36–44 Gy 2 Gy/fraction + 4 Gy boost Andrews [34] 56 50 Gy 2 Gy/fraction Chung [41] 72 45 Gy 1.8 Gy/fraction 1999–2005 mean results

Rate of cranial neuropathy

Tumor control rate

V

VII

VIII (m/s/yrs*)

91% 97% 100% 98% 96 2

21% 29% 59% 27% 34 8

15% 30% 67% 21% 33 12

24% (m,4.7)a 46% (s,2)b 42% (s,2)b 47% (s, 5–10)a 40 11

100% 91% 100% 96% 99% 98% NS 97 1

2% 11% 4% 4% 4% 2.5% NS 5 31

1% 4% 2% 0% 0% 1.2% NS 1 12

33% (s, 0.8)b 44% (s, 4)a 50% (s, > 3)a 56% (s, 5)a 79% (s, 6)b 83% 68% (s, 7) 59 18

97%

16%

8%

78% (s, 2)b,d

97%

7%

2%

100%

7%

4%

98 2

10 5

53

81% (s, 1.2)a 71% (s, 5)a,c 85% (s, 1)b 57% (s, 2) b 69 113

*m/s/yrs reflects measurable or serviceable hearing preservation rates at follow up intervals in years a raw hearing preservation rate; bactuarial hearing preservation rate; c current follow-up for this cohort (unpublished observation); d unclear impact of 4 Gy end-boost on hearing preservation 1 p = 0.0028 vs. early Gamma Knife cohort, trigeminal neuropathy; 2 p = 0.0087 vs. early Gamma Knife cohort, facial neuropathy; 3 p = 0.0166, FSR longest follow-up hearing vs. early Gamma Knife cohort hearing

to maintain the integrity of the genome prior to a mitotic event. Further, LINAC based fractionated courses is more widely available than Gamma Knife and may be applicable to more patients.

Summary of Literature for FSR Treatment of Acoustic Neuromas Tumor control rates reported in fractionated stereotactic radiosurgery series are well above 90% irrespective of fractionation schedule [34],

[42–48]. In modern series, facial nerve preservation rates have been reported from 95 to 100% [34],[42–44],[46],[48]. There seems to be little or no difference in treatment-related facial nerve toxicity between the gamma knife series, with a possibly more conformal dose distribution, and the linac series. Trigeminal nerve preservation rates in modern series have been reported from 92 to 100%, both in LINAC and in gamma knife series and also both in single-fraction and in fractionated series [34],[42–44],[46],[48]. There are, however, no randomized studies on a comparison of single and fractionated treatments. For the


purpose of this chapter, conventional fractionation is defined as single treatments of 1.5–2.5 Gy, while hypofractionation refers to radiation fraction sizes of 3–5 Gy.

Hypofractionation Kalapurkal et al. reported on 19 patients treated with either 36 or 30 Gy in 6 fractions [49]. This patient group had large tumors with a mean pons to petrous diameter of 28 mm. The total dose was reduced after the first six patients secondary to ataxia experienced by two patients. The follow up was a median 54 months either using CT or MRI. This regimen resulted in tumor regression in ten patients and stabilization of size in nine patients. Further, of the nine patients who had hearing prior to treatment, eight had preservation of hearing and one had improvement. There was no incident of facial or trigeminal dysfunction. Poen and colleagues [44] described the FSR experience of 33 patients, of whom ten had neurofibromatosis type two and seven had prior surgery. In this cohort, the median tumor diameter was 20 mm; three patients had tumor diameters greater than 3 cm. Multiple isocenters (1–4) were used for treatment planning and radiation dose was 21 Gy administered in 3 fractions. Patient follow up was a median of 24 months and 34% of patients had tumor regression, 63% had stable disease, and one patient had tumor enlargement. For this patient, although growth was noted at 2 years after the treatment, subsequent follow up revealed tumor regression. Useful hearing preservation, GardnerRobertson class I & II, overall was 77%, 92% in sporadic cases and 67% in patients with NF2. Trigeminal neuropathy was noted in five patients and 3% of patients had facial nerve injury (HouseBrackmann Grade III). Williams [50] described the treatment of 125 patients 111 of whom received 25 Gy in 5 fractions for tumors Figure 71-2a). When analyzing outcomes for patients within each serviceable hearing cohort, a greater likelihood of maintaining G-R level 1 hearing was noted in the LDC with a trend approaching significance (p = 0.059, logrank and p = 0.007, Wilcoxon, > Figure 71-2b) and a significantly greater likelihood of maintaining G-R level 2 hearing was noted in the LDC (p Figure 71-2c). Multivariate analysis revealed higher dose class and pre-treatment G-R Grade to be highly significant factors contributing to the likelihood of hearing preservation while neither age nor tumor size had any significance. All audiometric outcomes are reflected in tabular form in > Table 71-2.

1159

1160

71


. Figure 71-2 (a): Kaplan Meier analysis of hearing outcome in patients with pre-treatment serviceable hearing, corrected for follow-up ( 165 weeks) with significant improvement in LDC (p =0.04, logrank test). Green is low dose cohort, N = 42; red is high dose cohort, N = 31 (b): Kaplan Meier analysis of hearing outcome in patients with pre-treatment Gardner-Robertson level 1 hearing, corrected for follow-up ( 165 weeks) with a trend favoring the LDC that approached significance (p = 0.059, logrank test). Green is low dose cohort (N = 28); red is high dose cohort (N = 20). (c): Kaplan Meier analysis of hearing outcome in patients with pre-treatment Gardner-Robertson level 2 hearing, corrected for follow-up ( 165 weeks) with significant improvement in LDC (p Figure 71-3). Although the mechanism for hearing loss remains unclear, it has been assumed that hearing loss could be related to the dose to the cochlear nerve, either directly or due to vascular injury from radiation, or to the cochlea, both discussed below. The mean value for published doses of 12 Gy have, in fact, resulted in serviceable hearing losses ranging from 17 to 67%, and cumulative FSR doses of 45–50 Gy have resulted in serviceable hearing losses ranging from 29 to 43% (> Table 71-1).

The Hearing Ret Formula and the Dose to the Cochlear Nerve None of the papers featured in > Table 71-1 provide uniform data about mean number of isocenters stratified by tumor sizes, and these data might build a relationship between such variables as tumor size, serviceable hearing loss, mean number isodose centers as a measure of conformality, and


. Figure 71-3 (a): Logarithmic plot of dose v. number of fractions based on published rates of hearing preservation according to the Gardner-Robertson criteria (black line); parallel line (blue) through the threshold 45 Gy point (arrowhead) established by Pan et al. [61] to establish high hearing preservation ret formula. (b): Plots of biologically equivalent doses utilizing the linear quadratic formula with an a/b ratio of 1.72; comparison with a hearing ret plot. The yellow bar signifies the threshold range of dose and fraction number yielding a high probability of hearing preservation and tumor control

dose to the cochlea. As an example, for less conformal Gamma Knife radiosurgery treatment plans, it is possible that segments of the VIII nerve are included inside the standard 50% isodose prescription volume. Therefore portions of the nerve may fall within higher dose gradient regions resulting in higher dose exposure and increased risk of serviceable hearing loss (> Figure 71-4). An important additional variable

71

at play may be the segment or length of nerve exposed, or the inclusion of the cochlea above threshold dose, discussed below.

The Hearing Ret Formula and the Dose to the Cochlea The hearing ret formula was designed around a threshold dose of 45 Gy which Pan et al. prospectively identified as the radiation dose to the cochlea associated with hearing loss [61]. In another recent analysis, Thomas and colleagues corroborated the importance of dose to the cochlea following FSR of acoustic neuroma [55]. The radiotherapy doses received by the cochlea were significantly different between the deteriorated and preserved hearing group for all cochlear dosimetric parameters measured, namely cochlear V90, V80, V50%, representing the percentage cochlear volume receiving, respectively, at least 90, 80, or 50% of the prescribed dose. This study showed that high radiation doses to the cochlea resulted in a significantly larger loss in speech reception threshold. If the percentage volume of the cochlea exposed to the V90% of the prescription dose (45 Gy) was less than 73.3%, then the median hearing loss was 10 dB. However, if the cochlear V90% was greater than or equal to 73.3%, the median hearing loss was 25 dB. We have corroborated this by measuring radiation dose delivered to the cochlea in 52 patients treated with FSR with treatment plans in which dose to the cochlea could be assessed (> Table 71-3). A mean dose of 47.3 8.7 Gy (range, 32.1– 56.8 Gy) was delivered to the cochlea in patients who lost serviceable hearing compared with a mean dose of 39.5 13.9 Gy (range, 10.7–60.8 Gy) to the cochlea in the patients who had preservation of serviceable hearing during the follow-up (p = 0.0223). Other structures at risk, including the vestibulocochlear nerve in either the internal auditory canal or the cisternal space and the cochlear nucleus

1163

1164

71


. Figure 71-4 Comparison of dose distribution to the cochlear nerve with Gamma Knife and FSR treatments (a): axial T-1 gadolinium enhanced MRI scan of right acoustic neuroma; (b): artists rendering of translucent acoustic tumor with cranial nerves VII and VIII cranial nerves adherent to the anterior and caudal surface of the tumor, coursing to internal auditory canal (c): eight shot Gamma Knife radiosurgery treatment plan with a 12 Gy prescription to the 50% isodose line (yellow) for right acoustic neuroma (d): magnified sagittal view of actual treatment plan in the distal porous acousticus (yellow line is 50% isodose prescription line; green line is 60% isodose line; magenta line is tumor surface). Assuming cochlear nerve is in 7:00 position, the nerve is within a 10% dose gradient above isodose prescription. (e): single isocenter Novalis FSR treatment plan with a 1.8 Gy prescription to the 90% isodose line for a right acoustic neuroma (tumor is pink; tiel line is 90% isodose prescription line); (f): magnified sagittal view of actual treatment plan in the distal porous acousticus (tumor is pink; tiel line is 90% isodose prescription line; yellow line is 95% isodose line). Assuming cochlear nerve is in 7:00 position, the nerve is within a 5% dose gradient above isodose prescription. (g): artists rendering of magnified sagittal cross section of intracanalicular portion of right acoustic tumor and contiguous VIII nerve at 7:00 position. The VIII nerve is within the prescribed isodose line and exposed to higher dose gradients. (h): plot of a focused radiation dose distribution with typical isodose prescriptions at 50% (Gamma Knife) and 90% (FSR). Yellow bar represents potential actual dose range within the spatial location of the cochlear nerve with 50% isodose prescription which is steeper; tiel bar represents potential actual dose range within the spatial location of the cochlear nerve with 90% isodose prescription which is shallower

were not significantly different in the two outcome groups.

Proposed Treatment Guidelines Since acoustic neuromas grow slowly, they often go undiagnosed for many years. Rosenberg

reported in his series the mean delay in diagnosis of acoustic neuroma was 7.3 years with one patient presenting with a unilateral hearing loss over a 50-year period [10] Thomsen and Tos in a review of 233 patients with acoustic neuromas found a mean delay in diagnosis of 7.1 years from the patient’s initial symptoms [12]. Charabi et al also found that in a review of 94 patients the


71

. Table 71-3 Analysis of radiation dose delivered to cochlear apparatus after FSR (n = 52)

Cochlea Dmax Dmin VII nerve Dmax Dmin Cochlear nucleus Dmax Dmin

Serviceable hearing Preserved (N = 29)

Serviceable hearing lost (N = 23)

Total

39.5 13.9 27.2 11.9

47.3 8.7* 32.3 11.0

42.9 12.4 29.4 11.7

59.3 7.3 37.6 13.6

58.2 5.2 38.1 13.0

58.8 6.5 37.8 13.2

49.6 15.2 22.3 12.0

43.0 17.8 19.2 9.8

46.7 16.5 20.9 11.1

*p-value = 0.0223

mean duration of symptoms before diagnosis was 5.3 years [64]. The typical initial symptom, gradual hearing loss, may not interfere enough with the patient’s daily living to warrant medical attention. The delay in diagnosis may also be partially due to the inability of imaging studies in the past to identify small tumors. In a patient who presents with a history of long-standing asymmetric hearing loss who has had a negative CT scan in the past, the possibility of the patient’s having an acoustic neuroma should not be discounted [10]. With the development of more sophisticated MRI images, the incidence of asymptomatic patients being diagnosed with small acoustic neuromas, especially those located within the internal auditory canal will increase. Incidental asymptomatic acoustic neuromas may be managed with observation alone. Once patients experience hearing deterioration or increase in tumor size, either radiosurgery or radiotherapy is warranted. Pretreatment audiometry is critical in establishing a baseline of overall hearing. For those patients with serviceable hearing defined as Gardner-Robertson grade I or II, fractionated radiotherapy has been associated with higher hearing preservation rates than has single fraction radiosurgery while maintaining identical tumor control rates [34]. Further, treatment with FSR can maintain

hearing within the pretreatment GardnerRobertson grade at a greater rate compared to SRS [34]. For those patients with non-serviceable hearing, the decision for SRS versus FSR is either clinician or patient dependent. Microsurgery is generally not recommended unless the tumor size is greater than 3 cm or is compressing the brainstem or after failure to maintain tumor control with radiation. The rationale, as stated above, is that microsurgery is associated with higher rates of facial and trigeminal neuropathy immediately postoperatively as well with longer term follow up. In addition, patient satisfaction and quality of life favor radiation over surgery. MRI scans are obtained at regular intervals following therapy, generally 3 months after completion of radiation. Loss of central enhancement is a common radiographic finding seen within the 6 months after treatment representing enlargement and capsular thickening. This may be transient and unlike surgery, radiation generally does not result in tumor elimination but rather shrinkage or lack of growth for slow growing tumors (> Figure 71-5). Audiometry should be monitored as well to determine precise improvement and preservation of hearing rates. Hearing improvement is the exception and some audiometric decay after treatment is the norm (> Figure 71-6). If there is no documentation of change in hearing

1165

1166

71


. Figure 71-5 Typical radiographic post-treatment response after FSR: (a): axial T1W gadolinium-enhanced image of right acoustic neuroma; (b): 6 months after treatment with central necrosis; (c): at 1 year with shrinkage; (d): at 3 years with further shrinkage

within the first 3 months of follow up, repeat audiograms should only be performed for new symptoms prior to the first annual check. If there is hearing deterioration of more than 15 dB in pure tone average or 30% in speech discrimination score, steroid therapy is recommended for 3 weeks (prednisone 60 mg daily)

until next follow-up visit at 1 month. Steroids have an anti-inflammatory effect as well as membrane stabilization effect, which has been employed in the transient management of peritumoral brain edema or transient tumor swelling in various brain tumors including both intra-axial and extraaxial tumors after radiosurgery [56,57,65].


. Figure 71-6 Composite plots of pure tone average (PTA) over time. (a): with observation; red lines denote drop in PTA below serviceable hearing during a 10 year follow-up. (b): after FSR to a total dose of 50.4 Gy cohort of patients with stable or improved within the same G-R grade were hearing improved or remained in the same G-R grade with less than 15 dB loss in PTA during the follow-up: (c); same dose cohort with documented hearing loss of greater than 20 dB with drop below serviceable hearing. Note that most cases of serviceable hearing occur within the first year

71

the tumor or the surrounding normal neurovascular structures, preventive use of steroid over the specific time period, i.e., 3–6 months after treatment, when the transient swelling or adverse effects start to appear, might reduce hearing deterioration in the patient treated with SRS or FSR. However, due to the serious side effects, their use should be very cautious and limited in short-term duration with close monitoring. Thus if there is no further progression of the hearing deterioration or tumor growth on the next follow up, steroid may be tapered off. Most patients tolerate treatment without incident, however, there are some significant though minor possible sequelae of treatment. Hydrocephalus in the absence of tumor progression has been reported in 3–11% of patients treated with either radiosurgery or FSR [34,35]. Hydrocephalus occurs at a median of 1 year, is associated with treatment of larger tumors, may resolve spontaneously or require shunting and is suspected to be secondary to proteinaceous debris blocking the flow of CSF. Tinnitus and vertigo may also become worsened after treatment.

Conclusions The treatment of vestibular schwannomas by FSR is effective, and based on literature review, is associated with decreased rates of cochlear neuropathy compared to either microsurgery or SRS. For patients with serviceable hearing, therefore, FSR should be recommended. Since both Gamma Knife and FSR treatments are widely practiced, a prospective randomized trial will be necessary to establish a standard of care. There have been several reports that document that hearing has improved after the application of steroids in patients who had experienced hearing deterioration during the followup after SRS [66,67]. Since steroids can attenuate the adverse effects caused by the radiation to

References 1. Propp JM, McCarthy BJ, Davis FG, Preston-Martin S. Descriptive epidemiology of vestibular schwannomas. Neuro Oncol 2006;8:1-11.

1167

1168

71


2. Cohen RJ. Acoustic neuroma: summary of the NIH consensus. Md Med J 1992;41:1128-30. 3. Lin D, Hegarty JL, Fischbein NJ, Jackler RK. The prevalence of ‘‘incidental’’ acoustic neuroma. Arch Otolaryngol Head Neck Surg 2005;131:241-4. 4. Eldridge R, Parry D. Vestibular schwannoma (acoustic neuroma). Consensus development conference. Neurosurgery 1992;30:962-4. 5. Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP, Parry DM, Eldridge R, Kley N, Menon AG, Pulaski K, et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell 1993;72:791-800. 6. Woodruff J, Kourea H, Louis D, Scheithauer B. Schwannoma. In: Kleihues P, Cavenee W, editors. Pathology and Genetics: tumours of the Nervous system. Lyon: IARC Press; 2000. p. 164-6. 7. Bederson JB, von Ammon K, Wichmann WW, Yasargil MG. Conservative treatment of patients with acoustic tumors. Neurosurgery 1991;28:646-51. 8. Fucci MJ, Buchman CA, Brackmann DE, Berliner KI. Acoustic tumor growth: implications for treatment choices. Am J Otol 1999;20:495-9. 9. Mirz F, Pedersen CB, Fiirgaard B, Lundorf E. Incidence and growth pattern of vestibular schwannomas in a Danish county. Acta Otolaryngol Suppl 2000;543:30-3. 10. Rosenberg SI. Natural history of acoustic neuromas. Laryngoscope 2000;110:497-508. 11. Silverstein H, McDaniel A, Norrell H, Wazen J. Conservative management of acoustic neuroma in the elderly patient. Laryngoscope 1985;95:766-70. 12. Thomsen J, Tos M. Acoustic neuromas. Diagnostic delay, growth rate and possible non-surgical treatment. Acta Otolaryngol Suppl 1988;452:26-33. 13. Valvassori GE, Guzman M. Growth rate of acoustic neuromas. Am J Otol 1989;10:174-6. 14. Valvassori GE, Shannon M. Natural history of acoustic neuromas. Skull Base Surg 1991;1:165-7. 15. Walsh RM, Bath AP, Bance ML, Keller A, Tator CH, Rutka JA. The natural history of untreated vestibular schwannomas. Is there a role for conservative management? Rev Laryngol Otol Rhinol(Bord) 2000;121:21-6. 16. Matthies C, Samii M. Management of 1000 vestibular schwannomas (acoustic neuromas): clinical presentation. Neurosurgery 1997;40:1-9; discussion 9–10. 17. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg 1985;93:146-7. 18. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuroma surgery. Ann Otol Rhinol Laryngol 1988;97:55-66. 19. Sakamoto T, Fukuda S, Inuyama Y. Hearing loss and growth rate of acoustic neuromas in follow-up observation policy. Auris Nasus Larynx 2001;28 Suppl:S23-27. 20. Betchen SA, Walsh J, Post KD. Long-term hearing preservation after surgery for vestibular schwannoma. J Neurosurg 2005;102:6-9.

21. Danner C, Mastrodimos B, Cueva RA. A comparison of direct eighth nerve monitoring and auditory brainstem response in hearing preservation surgery for vestibular schwannoma. Otol Neurotol 2004;25:826-32. 22. Mohr G, Sade B, Dufour JJ, Rappaport JM. Preservation of hearing in patients undergoing microsurgery for vestibular schwannoma: degree of meatal filling. J Neurosurg 2005;102:1-5. 23. Gormley WB, Sekhar LN, Wright DC, Kamerer D, Schessel D. Acoustic neuromas: results of current surgical management. Neurosurgery 1997;41:50-8; discussion 58–60. 24. Somers T, Offeciers FE, Schatteman I. Results of 100 vestibular schwannoma operations. Acta Otorhinolaryngol Belg 2003;57:155-66. 25. Forster DM, Kemeny AA, Pathak A, Walton L. Radiosurgery: a minimally interventional alternative to microsurgery in the management of acoustic neuroma. Br J Neurosurg 1996;10:169-74. 26. Chan AW, Black P, Ojemann RG, Barker FG, II, Kooy HM, Lopes VV, McKenna MJ, Shrieve DC, Martuza RL, Loeffler JS. Stereotactic radiotherapy for vestibular schwannomas: favorable outcome with minimal toxicity. Neurosurgery 2005;57:60-70; discussion 60–70. 27. Regis J, Pellet W, Delsanti C, Dufour H, Roche PH, Thomassin JM, Zanaret M, Peragut JC. Functional outcome after gamma knife surgery or microsurgery for vestibular schwannomas. J Neurosurg 2002;97:1091-100. 28. Elshaikh M, Ljungman M, Ten Haken R, Lichter AS. Advances in radiation oncology. Annu Rev Med 2006;57:19-31. 29. Hirsch A, Noren G, Anderson H. Audiologic findings after stereotactic radiosurgery in nine cases of acoustic neurinomas. Acta Otolaryngol 1979;88:155-60. 30. Hirsch A, Noren G. Audiological findings after stereotactic radiosurgery in acoustic neurinomas. Acta Otolaryngol 1988;106:244-51. 31. Flickinger JCLunsford LD, Coffey RJ, Linskey ME, Bissonette DJ, Maitz AH, Kondziolka D. Radiosurgery of acoustic neurinomas. Cancer 1991;67:345-353. 32. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis PS. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001;95:440-49. 33. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med 1998;339:1426-33. 34. Andrews DW, Suarez O, Goldman HW, Downes MB, Bednarz G, Com BW, Werner-Wasik M, Rosenstock J, Curran WJ, Jr. Stereotactic radiosurgery and fractionated stereotactic radiotherapy for the treatment of acoustic schwannomas: comparative observations of 125 patients treated at one institution. Int J Radiat Oncol Biol Phys 2001;50:1265-78. 35. Karpinos M, Teh BS, Zeck O, Carpenter LS, Phan C, Mai WY, Lu HH, Chiu JK, Butler EB, Gormley WB, Woo SY.


36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

Treatment of acoustic neuroma: stereotactic radiosurgery vs. microsurgery. Int J Radiat Oncol Biol Phys 2002;54:1410-21. Iwai YYK, Shiotani M, Uyama T. Radiosurgery for acoustic neuromas: results of low-dose treatment. Neurosurgery 2003;53:282-8. Flickinger JC, Kondziolka D, Niranjan A, Maitz A, Voynov G, Lunsford LD. Acoustic neuroma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004;60:225-30. van Eck AT, Horstmann GA. Increased preservation of functional hearing after gamma knife surgery for vestibular schwannoma. J Neurosurg 2005;102 Suppl:204-6. Hasegawa T, Kida Y, Kobayashi T, Yoshimoto M, Mori Y, Yoshida J. Long-term outcomes in patients with vestibular schwannomas treated using gamma knife surgery: 10-year follow up. J Neurosurg 2005;102:10-16. Kagei KSH, Suzuki K, et al. Small-field fractionated radiotherapy with or without stereotactic boost for vestibular schwannoma. Radiother Oncol 1999;50:341-7. Chung HT, Ma R, Toyota B, Clark B, Robar J, McKenzie M. Audiologic and treatment outcomes after linear accelerator-based stereotactic irradiation for acoustic neuroma. Int J Radiat Oncol Biol Phys 2004;59:1116-21. Fuss M, Debus J, Lohr F, Huber P, Rhein B, EngenhartCabillic R, Wannenmacher M. Conventionally fractionated stereotactic radiotherapy (FSRT) for acoustic neuromas. Int J Radiat Oncol Biol Phys 2000;48:1381-7. Lederman GLJ, Wertheim S, Fine M, Lombardi E, Wronski M, Arbit E. Acoustic neuroma: potential benefits of fractionated stereotactic radiosurgery. Stereotact Funct Neurosurg Suppl 1997;69:175-82. Poen JC, Golby AJ, Forster KM, Martin DP, Chinn DM, Hancock SL, Adler JR, Jr. Fractionated stereotactic radiosurgery and preservation of hearing in patients with vestibular schwannoma: a preliminary report. Neurosurgery 1999;45:1299-1305; discussion 1305–1297. Song DY, Williams JA. Fractionated stereotactic radiosurgery for treatment of acoustic neuromas. Stereotact Funct Neurosurg Suppl 1999;73:45-9. Szumacher E, Schwartz ML, Tsao M, Jaywant S, Franssen E, Wong CS, Ramasesham R, Lightstone AW, Michaels H, Hayter C, Laperviere NJ. Fractionated stereotactic radiotherapy for the treatment of vestibular schwannomas: combined experience of the Toronto-Sunnybrook Regional Cancer Centre and the Princess Margaret Hospital. Int J Radiat Oncol Biol Phys 2002;53:987-91. Varlotto JM, Shrieve DC, Alexander E, III, Kooy HM, Black PM, Loeffler JS. Fractionated stereotactic radiotherapy for the treatment of acoustic neuromas: preliminary results. Int J Radiat Oncol Biol Phys 1996;36: 141-5. Van Leeuwen JP, Cremers CW, Thewissen NP, Harhangi BS, Meijer E. Acoustic neuroma:correlation among tumor size, symptoms, and patient age. Laryngoscope 1995;105: 701-7.

71

49. Kalapurakal JA, Silverman CL, Akhtar N, Andrews DW, Downes B, Thomas PR. Improved trigeminal and facial nerve tolerance following fractionated stereotactic radiotherapy for large acoustic neuromas. Br J Radiol 1999;72:1202-7. 50. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. Int J Radiat Oncol Biol Phys 2002;54:500-4. 51. Sawamura Y, Shirato H, Sakamoto T, Aoyama H, Suzuki K, Onimaru R, Isu T, Fukuda S, Miyasaka K. Management of vestibular schwannoma by fractionated stereotactic radiotherapy and associated cerebrospinal fluid malabsorption. J Neurosurg 2003;99:685-92. 52. Selch MT, Pedroso A, Lee SP, Solberg TD, Agazaryan N, Cabatan-Awang C, DeSalles AA. Stereotactic radiotherapy for the treatment of acoustic neuromas. J Neurosurg 2004;101 Suppl 3:362-72. 53. Combs SE, Volk S, Schulz-Ertner D, Huber PE, Thilmann C, Debus J. Management of acoustic neuromas with fractionated stereotactic radiotherapy (FSRT): longterm results in 106 patients treated in a single institution. Int J Radiat Oncol Biol Phys 2005;63:75-81. 54. Sakamoto T, Shirato H, Takeichi N, Aoyama H, Fukuda S, Miyasaka K. Annual rate of hearing loss falls after fractionated stereotactic irradiation for vestibular schwannoma. Radiother Oncol 2001;60:45-8. 55. Thomas C, Di Maio S, Ma R, Vollans E, Chu C, Clark B, Lee R, McKenzie M, Martin M, Toyota B. Hearing preservation following fractionated stereotactic radiotherapy for vestibular schwannomas: prognostic implications of cochlear dose. J Neurosurg 2007;107:917-26. 56. Pan DH, Guo WY, Chung WY, Shiau CY, Liu RS, Lee LS. Early effects of Gamma Knife surgery on malignant and benign intracranial tumors. Stereotact Funct Neurosurg 1995;64:19-31. 57. Kalapurakal JASC, Akhtar N, Laske DW, Braitman LE, Boyko OB, Thomas PR. Intracranial meningiomas: factors that influence the development of cerebral edema after stereotactic radiosurgery and radiation therapy. Radiology 1997;204:461-5. 58. Parsons JT, Bova FJ, Fitzgerald CR, Mendenhall WM, Million RR. Radiation optic neuropathy after megavoltage external beam irradiation: analysis of time-dose factors. Int J Radiat Oncol Biol Phys 1994;30:755-63. 59. Paek SH, Chung HT, Jeong SS, Park CK, Kim CY, Kim JE, Kim DG, Jung HW. Hearing preservation after gamma knife stereotactic radiosurgery of vestibular schwannoma. Cancer 2005;104:580-90. 60. Massager N, Nissim O, Delbrouck C, Devriendt D, David P, Desmedt F, Wikler D, Hassid S, Brotchi J, Levivier M. Role of intracanalicular volumetric and dosimetric parameters on hearing preservation after vestibular schwannoma radiosurgery. Int J Radiat Oncol Biol Phys 2006;64:1331-40. 61. Pan CC, Eisbruch A, Lee JS, Snorrason RM, Ten Haken RK, Kileny PR. Prospective study of inner ear radiation

1169

1170

71


dose and hearing loss in head-and-neck cancer patients. Int J Radiat Oncol Biol Phys 2005;61:1393-402. 62. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neuropathy after irradiation of meningioma. Radiology 1992;185:71-6. 63. Andrews D, Bednarz G, Downes B, Werner-Wasik M. Fractionated stereotactic radiotherapy: acoustic neuromas and other benign tumors. In: Chin L. and Regine W. editors. Principles and practice of stereotatic radiosurgery. New York: Springer Science and Business Media, LLC; 2008. p. 289-98. 64. Charabi S, Thomsen J, Mantoni M, Charabi B, Jorgensen B, Borgesen SE, Gyldensted C, Tos M. Acoustic neuroma (vestibular schwannoma): growth and surgical

and nonsurgical consequences of the wait-and-see policy. Otolaryngol Head Neck Surg 1995;113:5-14. 65. Souhami L, Olivier A, Podgorsak EB, Villemure JG, Pla M, Sadikot AF. Fractionated stereotactic radiation therapy for intracranial tumors. Cancer 1991;68: 2101-8. 66. Aronzon ARM, Bigelow DC. The efficacy of corticosteroids in restoring hearing in patients undergoing conservative management of acoustic neuromas. Otol Neurotol 2003;24:465-8. 67. Nedzelski JMCR, Kassel EE, Rowed DW, Tator CH. Is no treatment good treatment in the management of acoustic neuromas in the elderly? Laryngoscope 1986; 96:825-9.

66 Gamma Knife: Clinical Aspects L. Steiner . C. P. Yen . J. Jagannathan . D. Schlesinger . M. Steiner

Radiosurgery was defined by Lars Leksell as the technique of destroying intracranial targets through the intact skull with the use of highly focused ionizing beams. As a pupil of Olivecrona, Leksell witnessed both the successes and the failures of the pioneering endeavors of his teacher and concluded that neurosurgery should be made less traumatic. He felt that the stereotactic method introduced by Clarke and Horsley [1] in laboratory research and applied clinically by Spiegel and associates [2] was an avenue for function-preserving and less-invasive surgery. He then developed his own stereotactic system for open intervention and adapted his elegant concept of the arc device in building ‘‘Gamma Knife’’ as a tool for neurosurgeons (> Figure 66-1). Originally, Leksell’s aim was to use Gamma Knife for the treatment of functional disorders by producing necrotic lesions in specific nuclei or pathways of the brain. Subsequent development proved that Gamma Knife could be used in the management of brain tumors and cerebral vascular malformations. Subnecrotic doses were found to trigger cellular reactions in tumor cells and vasculature, leading to tumor shrinkage or control and obliteration of vascular malformations. Leksell’s original definition of radiosurgery was thus modified to include the destruction of intracranial targets and induction of desired biological effects in target tissue by the use of a single high dose of focused ionizing beams through the intact skull. The terms radiosurgery and Gamma Knife often are considered misnomers, however Leksell deliberately settled on the ambiguous terms because they articulated his concept of a tool that would be available to neurosurgeons as an #


alternative to the scalpel for neurosurgical intervention. In his autobiography, Leksell wrote: ‘‘I have in my hand a new type of brain surgery, an operative system, a more sophisticated and less risky surgical procedure based on progressively improving imaging of the brain and on mechanical accuracy and modern physics, a necessary addition to classical bloody surgery’’ [3]. As a neurosurgeon with a mathematical mind, Leksell was convinced that the advancement of neurosurgery depended on the adoption of advances in other technical fields. In using ionizing beams in the radiosurgical apparatus he merely supplemented the ‘‘physical agents’’ used by neurosurgeons in their various surgical tools. Well aware of the resistance the neurosurgical establishment would mount against the new idea, he was determined to emphasize that the term Gamma Knife symbolized the neurosurgical nature of the tool. The current scope of Gamma Knife radiosurgery has been expanded to include arteriovenous malformations (AVMs), dural arteriovenous fistulas, pituitary adenomas, craniopharyngiomas, meningiomas, vestibular schwannomas, gliomas, metastatic tumors, and indications in functional neurosurgery such as intractable pain, trigeminal neuralgia, movement disorders, and intractable epilepsy. There is a growing body of data with longterm follow-up on large series of patients that are helping to define the role of Gamma Knife in the neurosurgical management of these conditions. Radiosurgery remains an integral part of neurosurgery, and with the increasing number of centers with radiosurgical facilities Leksell’s intention that a neurosurgeon should be able to choose between scalpel and Gamma Knife for a given case is being fulfilled. However, the increased availability of

1038

66

Gamma knife: clinical aspects

. Figure 66-1 Professor Lars Leksell and the first patient with vestibular schwannoma treated with Gamma Knife (1970)

in a higher incidence of lethal damage to cells, enhancing the biological efficacy of the radiation. Differential responses are based on the rate of proliferation of cells, resulting in increased sensitivity of endothelial, glial and subependymal plate cells. Vascular obliteration also seems to play a role in the death of tumor cells as well. Several of these effects are governed by volumetric considerations within the treated volume.

Effects at Tissue Level

the technique raises the issue of training and standards to ensure that this neurosurgical tool is not abused.

Radiobiology Despite the extensive clinical use of Gamma Knife over the last four decades, information about the radiobiological aspects of its use, such as the effects of a single high dose of radiation on tumor tissue, normal brain, and vascular structures remained incomplete. The current state of knowledge is summarized below.

Effects at Cellular Level Fractionated radiation results in a distribution of lethal and sublethal effects across a broad radiation field. In contrast, the small field size and sharp dose fall-off of radiosurgery permits the delivery of a high volume of radiation to a precisely defined area. The relative abundance of replicating nuclear materials in tumor tissue and the higher potential for normal brain cells to repair their DNA after sublethal injury are the keys to the differential sensitivity to radiation. In radiosurgery, the use of a single dose results

The radiation doses prescribed for classical radiotherapy have been developed from decades of clinical experience. Early in the era of clinical radiotherapy, it was observed that multiple treatments (fractions) with reduced doses per fraction improved the therapeutic ratio when treating both benign and malignant tumors. Radiotherapy uses a per-fraction prescription dose which is below the lethal dose threshold of normal tissue in the volume treated so as to purposely include a ‘‘margin’’ of normal tissue around the target lesion to account for daily setup error or subclinical disease. The time between fractions then allows this normal tissue to repair sublethal damage. The radiobiological principles which govern the design of multifraction treatments do not necessarily apply to the high-dose ionizing beams as used in radiosurgery. In contrast to radiotherapy, radiosurgery specifies a precise delivery of a high single-fraction dose of ionizing beams to a defined target volume. Normal tissue is excluded from the target volume as much as possible. The steep dose gradient at the margin of the target volume assures that normal tissue receives minimal dose and tissue inside the target periphery receives a higher dose (> Figure 66-2). Thus, repair of normal tissue during the treatment is of little concern in radiosurgery. The delivery of an inhomogeneous dose to the treatment field with a higher dose at the center of a tumor (the so-called hot spot) may be desirable for


. Figure 66-2 Section of brain showing a lesion in the nucleus centrum medianum 14 months after gammathalamotomy in a patient with intractable pain from terminal cancer. The sharply delineated lesion is limited to the target area. The patient was pain-free until death

several reasons. First, it offsets the relative protection offered by the poor oxygenation of the tumor core; second, it increases the cell kill in the tumor cells adjacent to those in the hot spots due to the fact that the effect of a given dose to a population of cells is more damaging if the neighboring cells receive a high dose [4].

Cranial Nerve Sensitivity The mechanisms of radiation injury to the cranial nerves are most probably secondary to damage of small vessels and protective Schwann cells

66

or oligodendroglia. There is a difference in the tolerance of different cranial nerves with sensory nerves (optic and acoustic) tolerating the least radiation and the nerves in the parasellar region, the facial nerves and the lower cranial nerves tolerating higher doses. This may be due to the fact that both the optic and acoustic nerves are actually fiber tracts of the central nervous system and carry more complex data. Clinical experience suggests that these specialized sensory nerves do not show a capacity to recover from injury. The radiosensitivity of the cranial nerves often necessitates limits on the doses given to tumors and vascular lesions in close relation to these structures. Although the precise dose tolerance of the cranial nerves is unclear, the anterior visual pathways seem to be the least radio-resistant to single doses above 8 Gy [5–7]. Hence in this situation, the distance between nerve and the lesion being treated should be carefully assessed. It appears that the risk is related to the volume of the optic apparatus receiving the dose [8–11]. A distance of 5 mm between the tumor and the optic apparatus is desirable to achieve an optimal dose fall-off, but occasionally a distance of as little as 2 mm may be acceptable due to shielding capabilities of Gamma Knife. The tolerable distance is a function of the degree to which a dose plan can be designed to deliver a suitable radiation dose to the tumor yet spare the optic apparatus. The largest experience on the radiation tolerance of cranial nerves is available for the trigeminal and facial nerves [12]. In our series of 151 patients who underwent radiosurgery for trigeminal neuralgia, 12 patients (9%) had new-onset facial numbness after treatment [13]. Noreń and associates [12] analyzed risk factors for facial and trigeminal neuropathy in tumors receiving 12–20 Gy and concluded that the most significant factor is the length of the nerve irradiated, not the volume of tumor or dose. In treating patients with AVMs, meningiomas and secretory pituitary adenomas we have given doses of between 20–25 Gy to the cranial

1039

1040

66


nerves in the parasellar region without complications. Tishler et al. [14] noted that the maximum dose delivered to cranial nerves were related to neurologic deficits in 29 patients after LINAC radiosurgery and 33 patients after Gamma Knife surgery (GKS). Twelve new neuropathies were observed that were related to the nerves in the parasellar region, but they were all unrelated to a maximum dose in the range of 10–40 Gy. The conclusion of this study was that doses up to 40 Gy are relatively safe for nerves in the parasellar region. In our recently published paper on radiosurgery for Cushing’s disease, four of ten patients developed visual acuity reduction (two of whom also developed oculomotor nerve palsies) after repeat radiosurgery. Two additional patients who had previous fractionated radiation also developed oculomotor neuropathies [5]. Although three of these six patients had subsequent improvements in their visual deficits, these findings have convinced us to stop retreating pituitary adenoma patients with Gamma Knife until a safe cumulative dose can be determined.

or middle cerebral arteries 2–24 months after Gamma Knife irradiation with an 8-mm collimator and doses of 10–100 Gy [15]. In another experimental study, irradiation of the basilar arteries of cats by a stereotactic technique was performed with doses varying from 100 to 300 Gy in a gamma unit [16]. Histologically, vascular lesions such as vacuolization, degeneration, and desquamation of the endothelium and necrosis of the muscular coat predominated; reparatory reactions were relatively sparse and thrombosis was completely absent. Kamiryo et al. [17] irradiated the anterior cerebral arteries at the circle of Willis in rats with Gamma Knife. The maximum doses varied from 25 to 100 Gy. Occlusion of the anterior cerebral artery was observed in one rat 20 months after irradiation with 100 Gy. The changes included arterial wall thickening with fibrosis, splitting of the internal elastic membrane, and formation of a luminal organized thrombus. The differences between rat and human AVM studies were that more thrombus formation was observed in rat vessels and a higher dose was required to occlude the normal anterior cerebral artery in rats.

Effects on Normal Brain Vasculature

Effects on AVMs

The observations that the injury repair paradigm described above does not seem to extend to the normal vasculature in the region of radiosurgery and that the incidence of stenosis in major brain vasculature remains less than one percent are largely unexplained. There were only two patients in our experience who had clinical effects from such changes; one had a quadrantanopia, and the other had transient headaches secondary to vasogenic edema as a result of straight sinus occlusion. The incidence of such changes is low despite the use of radiation doses of 15–25 Gy at the periphery of the treated targets. In an experimental study on hypercholesterolemic rabbits, no changes were found in the basilar

The proliferation of intimal cells after the irradiation of AVMs was described by Cushing in 1928 [18]. Andersson and associates [19] observed intimal changes associated with thickened collagenous vessel walls and perivascular cuffing after they irradiated goat brains with proton beams. We examined nine AVMs specimens obtained 10–60 months after GKS [20]. The irradiated vessels displayed progressive changes that led to narrowing and obliteration of the lumen (> Figure 66-3). The earliest or least severe changes were endothelial damage and endothelial-intimal separation. These were followed by subendothelial and intimomedial proliferation of smooth muscle cells with elaboration of


66

. Figure 66-3 AVM vessel after GKS. Masson trichrome-stained section of a vessel from an AVM 6 months after radiosurgery (a). Note the intimal separation and the subendothelial and intimomedial proliferation of the smooth muscles. Hematoxylin and eosin-stained section of an AVM vessel after complete obliteration showing the hyalinized acellular matrix that occludes the lumen (b)

extracellular matrix components including type IV collagen; then cellular degeneration and hyaline transformation of vessel walls; and finally end-stage appearance showing complete obliteration of the vessels. The above-mentioned histopathological changes correlated with time after radiosurgery. Thus, the mechanism of obliteration of AVMs after radiosurgery is endothelial damage followed by progressive sclerosis resulting from smooth muscle proliferation and extracellular matrix deposition. The long-term effect of the process is obliteration of the vessel lumen.

Instrumentation and Technique Treatment Protocol Patients are admitted to the hospital the night before or the day of treatment. Laboratory studies and electrocardiography are performed as required. Patients who harbor AVMs require angiograms, and should have adequate renal function tests. Conditions that would preclude performing a magnetic resonance image (MRI) (i.e. pacemaker or foreign bodies that are incompatible with MRI) should be noted, so that

computed tomography (CT) scans can be used instead. For the technical aspects of GKS, please refer to Chapter 70.

Follow-up The need for adequate and thorough clinical and imaging follow-up cannot be overemphasized. A follow-up protocol should be complemented with a database of follow-up information to provide a means for analysis and conclusions at later dates. As a rule of thumb, we ask for MRI and clinical follow-up every 6 months for benign lesions and every 3 months for malignant lesions, although exceptions to this rule exist and are noted in the following sections. For AVMs, a follow-up angiogram should be performed when the AVM nidus is no longer visible on MRI. The follow-up imaging studies should be reviewed by the treating neurosurgeon and neuroradiologist to ensure the unanimity of the observation. A variety of algorithms have been implemented in software to allow the estimation of lesion volume based on MRI and CT images [21]. This can be a useful quantitative metric of treatment efficacy. However, since the margin of

1041

1042

66


error of volumetric imaging can be significant, we corroborate the volume measurements with the diameters of the lesion on follow-up imaging.

Indications, Specific Aspects of Technique, Imaging and Clinical Outcomes Benign Tumors Pituitary Adenomas The goals of radiosurgery in the treatment of pituitary tumors are: (1) to control tumor growth and (2) to normalize hormone overproduction in the case of secretory tumors. All patients suspected of harboring a pituitary tumor should undergo a complete neurologic, ophthalmologic, endocrinologic, and imaging workup. Each facet of the hypothalamic-pituitary-end organ axis should be assessed. Mild elevations in serum prolactin commonly result from a stalk effect while levels greater than 200 ng/mL suggest a prolactin-secreting adenoma. Thyroid function should be evaluated by measuring free thyroxine and thyroid-stimulating hormone. Adrenal function should be assessed by a morning serum cortisol and adrenocorticotropic hormone (ACTH) level. In cases of suspected Cushing’s syndrome, a 24-h urine free cortisol and a dexamethasone suppression test should be performed. Serum growth hormone (GH) and insulin-like growth factor (IGF)-1 levels should be measured to evaluate acromegaly. Imaging evaluation is achieved with thin sliced pre- and post-contrast MRI of the sellar and parasellar regions. CT may be useful to assess degree of sinus aeration and bony destruction. If a patient has a neurological deficit attributable to an adenoma, surgery is the initial treatment of choice for all tumors except a prolactinoma. Transsphenoidal surgery (endoscopic or microscopic) allows for the most rapid relief of mass effect and reduction in excessive

hormone levels in patients with Cushing’s disease and acromegaly [22–27]. This approach is associated with a low rate of complications in the hands of an experienced neurosurgeon [28]. In our observation of over 400 pituitary adenomas treated by experienced microsurgeons, macroadenomas that require radiosurgery remain macroadenomas even following the initial transsphenoidal resection. In 2000, Landolt et al. reported a significantly lower hormone normalization rate in acromegalic patients who were receiving antisecretory medications at the time of radiosurgery [29]. The precise mechanism by which antisecretory medications lower hormonal normalization rates is unknown, but it is thought to be related to changes in cell cycling caused by these drugs, potentially decreasing tumor cell radiosensitivity [29,30]. Our experience in treating 90 patients with acromegaly and 23 patients with prolactinomas found a relationship between withholding suppressive medications and endocrine remission in both secretory subtypes. As a result, we advise acromegaly and prolactinoma patients to hold suppressive medications 8 weeks before treatment, and resume 6–8 weeks after radiosurgery. However, as is true for much of neurosurgical practice, class I evidence is still unavailable to support this treatment approach. It is possible that these findings are confounded by the fact that the majority of patients with more aggressive tumors were on anti-secretory medications from the onset. Since a tumor may rarely enlarge quickly in the absence of suppressive medications, it is important that the decision to hold suppressive medications be performed on an individualized basis. Nonsecretory Adenomas

The primary goal of GKS in the treatment of nonsecretory tumors is to stabilize or reduce adenoma volume, especially for tumors in the parasellar region. In our material of 90 patients treated by GKS for nonsecretory tumors, tumor volume decreased in 59 patients (65.6%) (> Figure 66-4), remained


unchanged in 24 (26.7%), and increased in seven (7.8%), at a mean follow-up of 44.9 months [31]. The mean prescription dose was 18.5 Gy (Range 5–25 Gy). We found that the minimal effective prescription dose for nonsecretory tumors was 12 Gy, and doses of greater than 20 Gy did not confer an additional benefit. A total of 12 patients (19.7%) suffered a new endocrinopathy after GKS, and 25% of patients followed for more than 2 years had some hormone deficits (usually thyroid or growth hormone). Eight patients were treated with Gamma Knife as a primary therapy for medical reasons or patient preference, and in these patients, a decrease in tumor volume occurred in three (42%); five patients had stabilization of tumor volume (58%) at 34 months follow-up. In the remaining 82 patients who were treated for recurrent or residual tumors after microsurgery, a reduction in tumor volume occurred in 56 (68%), no change was detected in 19 (23%), and an increase occurred in seven (8%). The median time to tumor shrinkage on MRI was 9 months (range, 6–48 months) following GKS.

66

Tumors involving the parasellar space require special consideration. Of the 61 tumors that involved the parasellar space treated at the Lars Leksell Center, Department of Neurosurgery, University of Virginia, 39 decreased in volume (63%) and 17 remained unchanged (27%). Secretory Adenomas

Reported rates of endocrine remission following radiosurgery for functional pituitary adenomas vary greatly. The variance is likely a function of methodology, endocrine criteria used to define remission, study population, and length of followup. Most series report a higher prescription dose to patients with secreting adenomas, with a range between 20–25 Gy in most reports [5,32–34]. Because hormone normalization may be followed by relapse, we prefer the term ‘‘remission’’ over ‘‘cure.’’ > Table 66-1 summarizes the imaging and endocrinologic outcomes of 342 pituitary tumor patients treated with GKS at Lars Leksell Center. Cushing’s disease is a devastating pituitary disorder and is associated with significant morbidity and premature death. Even

Cushing’s Disease

. Figure 66-4 Sagittal T1-weighted MR image obtained before GKS demonstrates a residual nonsecretory pituitary macroadenoma following three microsurgical removals in a 34-year-old man (a). The tumor decreased significantly 30 months after GKS (b)

1043

1044

66


. Table 66-1 Gamma Knife surgery for pituitary adenomas Pathology (n=) Nonsecretory (90) Cushing’s (107)* Nelson’s (22) Prolactinoma (28) Acromegaly (95)

Remission rate (%)

MRI volume decrease (%)

MRI volume no change (%)

MRI volume increase (%)

Endocrinopathy following GKS (%)

18.5

NA

66

27

7

25

23

53

80

14

6

22

20 19

45 26

54 81

36 9

10 10

40** 28

22

53

72

15

3

34

Mean prescription dose (Gy)

*Tumor visible on MRI in 49 patients **Only ten patients with post-operative endocrine evaluation

after transsphenoidal surgery, up to 30% of patients may have persistent disease [26,28,35]. Most centers define an endocrine remission as a urine-free cortisol in the normal range associated with the resolution of clinical stigmata or a series of normal post-operative serum cortisol levels obtained throughout the day [10,36]. Reported endocrine remission rates following transsphenoidal surgery vary from 10 to 100%, with higher remission rates when radiosurgery follows surgical debulking [5,11,37–43]. In series with at least ten patients and a median follow-up of 2 years, endocrine remission rates range from 17 to 83% [41–44]. Ra¨hn and associates reported their experience at Karolinska Institute involving 59 patients with Cushing’s disease who were treated using Gamma Knife and followed for 2–15 years. The efficacy rate of the initial treatment was 50%, with retreatment eventually providing normalization of cortisol production in 76% of patients [45]. Our first published paper on Cushing’s disease evaluated 44 patients with a mean follow-up of 39 months [10]. In this study, the remission rate was observed to be 73% with only three late recurrences using a mean prescription dose of 22 Gy. More recently, we have updated these results with 107 patients treated at the Lars Leksell Center for Cushing’s disease. With more patients,

and a longer follow-up of 44 months, we observed a lower remission rate of 53%. In this latter series, the rate of remission was statistically correlated with tumor volume, but not tumor invasion into the cavernous sinus or suprasellar region [5]. The mean prescription dose in this series was 23 Gy, and was lower in patients who had prior radiation. Although the rate of endocrinopathy (22%) was similar to the previous series, this follow-up series was notable for 10 patients who experienced a relapse of Cushing’s disease, with a mean time to recurrence of 27 months. The differences in remission rates and recurrences between these two series demonstrate the importance of long-term follow-up in judging the true effectiveness of radiosurgery in secretory adenomas. As reported by others, the rate of hormone normalization after radiosurgery for Cushing’s disease appears to be difficult to predict, with remission occurring as early as 2 months and as late as 8 years [46,47]. About 70% of patients who have hormonal normalization do so within the first 2 years after radiosurgery. Patients with persistent disease should consider alternative treatments such as adrenalectomy, or repeat radiosurgery (although this may be associated with a higher rate of cranial nerve damage) [5].


A subset of Cushing’s patients does not achieve hormone normalization following microsurgery and radiosurgery, and require adrenalectomy as a ‘‘salvage’’ treatment for their disease. Although adrenalectomy is the definitive treatment for cortisol overproduction, a small subset of patients may develop Nelson’s syndrome, characterized by rapid adenoma growth, hyperpigmentation and tumor invasion into the parasellar structures [48]. There are relatively few studies detailing the results of radiosurgery for Nelson’s syndrome [25,41,49–54]. These studies report a mean prescription dose ranging from 12 to 28.7 Gy, and an endocrine remission rate ranging from 0 to 36%, however only a minority of these studies defined what was meant by endocrine remission. Our experience involves 23 patients with at least 6 months follow-up. Thirty percent of patients had a reduction in tumor size, and 60% had no change in size. A decrease in ACTH levels occurred in 67% of patients with elevated level before GKS, but normalization only occurred in four patients [52]. Pollock and Young reported on 11 patients who underwent GKS for Nelson’s syndrome. They reported control of tumor growth in 9 of 11 patients, with ACTH normalization in 4 patients (36%) [53]. Nelson’s syndrome

Just as the endocrine criteria for remission in Cushing’s disease remain the subject of debate, the criteria for remission in acromegaly have also been inconsistent. The most widely accepted guidelines for a remission in acromegaly consist of a GH level less than 1 ng/ml in response to an oral glucose challenge and a normal serum IGF-1 when matched for age and gender [55,56]. In a large series, Jezkova et al. reported a remission rate of 50% at 42 months follow-up in 96 patients with acromegaly who received radiosurgery [57]. Nearly one-third of these patients had radiosurgery as a primary treatment without previous surgical extirpation of the adenoma. Pollock et al. demonstrated a remission rate of 50% in 46 patients who were treated with Acromegaly

66

GKS for recurrent and residual tumors, with a higher remission rate in patients who were off of suppressive medications at the time of radiosurgery [58]. Our experience(accepted for publication in ‘‘Neurosurgery’’) with 135 patients with mean follow-up of 57 months demonstrates a remission rate of 59% in patients off of suppressive medications compared with 37% in patients receiving a suppressive medication (most commonly octreotide) (> Figure 66-5). In both groups, the mean prescription dose was 22 Gy (range 12–28 Gy). As a result of this finding, University of Virginia endocrinologists currently recommend a cessation of somatostatin analog medication 8 weeks before and for 8 weeks after GKS. Symptomatology of remission was noted in 85% of patients who were followed for more than 48 months, but in only 33% of patients with less than 12-months follow-up, indicating that clinical remission may take significantly longer than normalization of laboratory values. In patients with prolactinomas, the criteria to define endocrine remission are generally consistent, with most studies defining a remission as a patient who has a normal serum prolactin level while off of antisecretory medications. We use GKS as a treatment for prolactinomas after failure of medical and/or surgical treatment. Of the 23 patients treated at our institution, normalization of prolactin levels occurred in 26%, at an average time of 24.5 months, with a prescription dose of 19 Gy [34]. Consistent with the work of Landolt and Lomax, we also found that the remission rate was lower in patients receiving an anti-secretory medication at the time of GKS [30,34]. In published studies of radiosurgery for prolactinomas, the mean prescription dose has varied from 13.3 to 33 Gy, and remission rates varied from 0 to 84% [25,30,32–34,59,60]. Variations in success rate are likely related to the dose delivered to the tumor. Witt et al. noted no remission with a prescription dose of 19 Gy [11,61]. Pan et al. [62] reported a 52% endocrine Prolactinomas

1045

1046

66


. Figure 66-5 Coronal T1-weighted MR image obtained before GKS demonstrates a residual pituitary macroadenoma following a transsphenoidal resection in a 50-year-old woman with acromegaly (a) The latest follow-up MR image performed 4 years after GKS shows the tumor decreased in size (b) Clinically, the patient had hormone remission 2 years following GKS

‘‘cure’’ rate in a retrospective study of 128 patients in whom GKS was used as a first-line treatment for prolactinomas with a prescription dose of 30 Gy. This study is on a large sample size, and is interesting because GKS was used as a first-line treatment before medical therapy [63]. It remains to be seen if the favorable results experienced by this group can be reproduced.

Craniopharyngiomas Surgery has remained the mainstay in the treatment of craniopharyngiomas. However, a partial resection is associated with a high recurrence rate and gross total excision (a goal which can be achieved only in selected cases) carries an average recurrence rate of 16% [64]. Even in cases where in the surgeons’ assessment was that the tumor was totally resected, open-ended MRI follow-up will detect a high percentage of residuals/recurrences. Reducing the significant endocrine and visual morbidity that often accompanies a radical surgical approach necessitates the quest

for other modalities of treatment. Combined approaches have included partial resection with radiotherapy that results in a marked decrease in the recurrence rate. The results in 61 children who were treated for craniopharyngiomas at Children’s Hospital and the Joint Center for Radiation Therapy in Boston from 1970 to 1990 [65] provide valuable insights into the role of radiation in the management of craniopharyngiomas. The l0-year actuarial overall survival was 91% for all patients. The 10-year actuarial freedom from progression for the surgery group was 31% compared with 100% for patients treated with radiation therapy only and 86% for patients treated with surgery plus radiotherapy. There were two treatment-related deaths, both in the surgery plus radiotherapy group. A higher incidence of visual loss and diabetes insipidus was associated with the use of aggressive surgery. Five of six patients with tumors more than 5 cm experienced recurrences, while only six of thirty had a recurrence when the tumor was less than 5 cm.


A study from the Royal Marsden Hospital for a series of 173 patients with craniopharyngiomas treated between 1950 and 1986 with external beam radiotherapy either alone or after surgery [66] showed that the 10- and 20-year progression-free survival (PFS) rates were 83 and 79% and the 10- and 20-year survival rate were 77 and 66% at a median follow-up of 12 years. Survival and PFS were not found to be influenced by the extent of surgical excision. Visual field defects improved after radiotherapy in 36% of patients (38 of 106), and visual acuity improved in 30% (27 of 91). No patient developed radiation optic neuropathy. The authors concluded that limited surgery and radiotherapy achieve excellent long-term tumor control and survival with low morbidity. Personal Data

We treated 37 craniopharyngiomas in 35 patients. Of these, 3 patients had biopsies, and 22 had had prior transcranial or transsphenoidal surgeries (ranging from 1 to 6 procedures). In seven cases, intracavitary 32P or 90Yt instillation was combined with microsurgery. In one case, intracavitary instillation of 32P was performed alone following cyst aspiration. When GKS was used, the prescription doses ranged from 6 to 25 Gy (mean, 13.3 Gy). The follow-up ranged between

66

8 and 212 months with a mean of 62.5 months. Four tumors increased in size. A decrease in the solid component of the tumor was seen in 29 (> Figure 66-6) and no change was seen in 4. However, of the patients whose solid tumors decreased or remained unchanged, ten developed new or enlarged cystic component with four of them requiring further surgical resection and four receiving intracavitary 32P instillation. In total, 23 patients improved or remained stable clinically. Twelve deteriorated with ten of them died from complications of disease or surgeries. The mean 5-year survival was 71%. Review of Literature

Leksell and Liden [67] first instilled radioactive 32 P in a cystic craniopharyngioma. This patient did well initially but died 2 years later of progressive enlargement of the solid part of the tumor. Since then, intracavitary irradiation of craniopharyngioma cysts has been performed by several authors [68–77] with successful reduction in the cyst volume in 80% of patients followed for a long observation period. In our material, a total of 11 patients received intracavitary isotope instillation before or after Gamma Knife treatment. Eight of these patients had cyst shrinkage. The other three had continuous enlargement of the cyst requiring drainage.

. Figure 66-6 Residual craniopharyngioma following microsurgery on contrast enhanced T1-weighted images before (a) and 4 months after GKS (b) shows a marked decrease in the tumor size

1047

1048

66


In view of the fact that intracavitary irradiation could successfully deal with the cystic part, GKS of the solid part of craniopharyngiomas seemed to be an attractive option. Backlund’s first patient with craniopharyngioma treated with GKS died of a shunt malfunction 4 months after the procedure. The postmortem histology revealed that except for a small island of cells, no evidence of tumor was found. Subsequently, Backlund treated four other patients with a combined radiosurgery- intracavitary radiation approach, all of whom did well on follow-up to 3.5 years. Backlund [78] reported one case of visual loss in his material, and although it is not possible to pinpoint the exact cause of the damage of optic nerve, it has to be assumed that radiation was the cause of it. Coffey and Lunsford [79] also reported visual impairment after GKS in two patients.

Meningiomas The treatment of choice for meningiomas is microsurgical extirpation. This should be achieved with morbidities corresponding to the expectations of the patients. GKS should be considered for meningiomas involving locations where it is difficult to control neuronal or vascular structures, for residuals of skull base tumors following microsurgery, for tumors where complete microsurgical extirpation including the involved meninges was not achieved, for patients not fit for major surgery and for patients who refused surgery. Personal Material

From 1976 to 1987, Steiner and Lindquist treated thirty-one meningiomas with GKS at Karolinska Institute with long term follow-up between 10 and 21 years. Two-thirds of these tumors have either shrunk or remained stable. Among these cases were a few where only the vascular supply was targeted. (> Figures 66-7 and > 66-8) This approach has resulted in significant tumor shrinkage in the long term.

From 1989 to 2007, 750 meningiomas have been treated at Lars Leksell Center with GKS. The number of meningiomas treated until 2005 is 690. In a series of 206 patients with 1–6 years follow-up, a mean prescription dose of 14 Gy was used (range 10–20 Gy). There were 142 postmicrosurgical residuals and 64 primarily treated with Gamma Knife. Imaging follow-up was available in 151 patients (> Figures 66-9). Ninety four (63%) of the tumors shrank, forty (26%) remained unchanged and seventeen (11%) increased in size. Other centers report similar results [80–82]. Of 112 meningiomas involving the parasellar space, 68% shrank; 30% remained unchanged; and 2% increased in size. No adverse effects were experienced in treating meningiomas. However, in one case a tumor without histological diagnosis and with equivocal imaging characteristics in the pineal region was treated as a presumed meningioma and edema of bilateral basal ganglia occurred in this patient with associated cognitive disturbance. The patient had an incomplete recovery.

Vestibular Schwannomas In 1910, Henschen proposed to change the misnomer ‘‘acoustic neuroma’’ to ‘‘vestibular schwannoma’’ that correctly indicates the anatomic and histologic origin of this tumor [83]. Henschen also predicted that ENT surgeons would be interested in the management of this tumor. Today, both neurosurgeons and otolaryngologists are operating on vestibular schwannomas. The neurosurgeons prefer a suboccipital removal whereas otolaryngologists often prefer a translabyrinthine approach. Leksell and Steiner first applied Gamma Knife radiosurgery for vestibular schwannomas [84]. Steiner, at that time having doubts about the method, refused to be coauthor of the first report and was only mentioned in the acknowledgement. Leksell’s idea to treat vestibular schwannomas with


66

. Figure 66-7 A right parasellar meningioma treated with Gamma Knife radiosurgery. Only the nutrient vessels were targeted as defined by CT (a) and angiogram (b). The tumor remained decreased in size 18 years after the treatment (c and d)

radiosurgery turns out today to be an alternative to microsurgery. We usually use a prescription dose of 11 Gy to the isodose configuration which includes the facial nerve. Significant volumes of the tumor may in this way receive 13–15 Gy. Since March 1989–2000, we treated with Gamma Knife 200 vestibular schwannomas and from 2000 to date an additional 268 were treated.

Because we believe on principal that only series of tumors with long follow-up should be published, we published the clinical and imaging outcomes of the series we treated until 2000 [85]. In that series, follow-up periods ranging from 1 to 10 years were available in 153 patients. Followup images were analyzed using computer software to calculate lesion volumes and the clinical

1049

1050

66


. Figure 66-8 T1-weighted contrast enhanced MRI shows a left clinoidal meningioma pushing the left optic nerve medially (a). The GKS targeted a segment of the internal carotid artery with efferent nutrient vessels that supply the tumor. The carotid artery can be better defined with the source image from the MR angiography (b)

condition of the patients was assessed using questionnaires. GKS was the primary treatment modality in 96 cases and followed microsurgery in 57 cases. The volume of the tumors ranged from 0.02 to 18.3 cm3. In the group in which the Gamma Knife was the primary treatment, a decrease in volume was observed in 78 cases (81%) (> Figure 66-10), no change in 12 (12%) and an increase in volume in six cases (6%). The decrease was more than 75% in seven cases. In the group treated following microsurgery, a decrease in volume was observed in 37 cases (55%), no change in 14 (25%) and an increase in volume in six (11%). The decrease was more than 75% in eight cases. Five patients experienced trigeminal dysfunction; in three cases this was transient and in the other two it was persistent although there has been improvement. Three patients have facial paresthesias; in one it was transient lasting only six weeks; in one case, there was 80% recovery at 18 months

post-treatment. In the third case, the patient had surgery when the facial palsy occurred and the nerve was cut. During a 6 year period, hearing deteriorated in 60% of the patients. Three patients saw an improvement in hearing. No hearing deterioration was observed during the first 2 years of follow-up review. When the hearing was useful, it was preserved in 58% of the patients. Brackman [86] contends that following radiosurgery, the results of microsurgery are unfavorable. In the five cases treated by Slater and Brackman, one tumor had undergone prior microsurgery which would be as likely a cause of the intraoperative difficulties as a previous radiosurgery. Other neurosurgeons like Donlin Long and Pitts contend that they have no problem in operating cases following radiosurgery. It is clear that these discussions are around anecdotal observations only. For definitive conclusions, observations of large series will be necessary.


66

. Figure 66-9 Large left parasellar meningioma residual following microsurgery visualized on postcontrast axial and coronal MRI (a and b). Images obtained 6 months later showed that the tumor had disappeared. Repeated control axial and coronal MRI over a 5-year period showed no recurrence of the tumor (c and d)

In commenting on our results, Samii wrote that ‘‘the study is a milestone in the treatment of vestibular schwannomas and in contemporary neurosurgery,’’ however, he considered microsurgery the first choice and only in selected cases advises Gamma Knife radiosurgery [87]. Malis wrote that Gamma Knife radiosurgery ‘‘has forced reluctant neurosurgeons to consider major changes in classic thinking about the proper care of vascular malformations, cavernous sinus meningiomas and acoustic neuromas’’ [88]. He continued ‘‘I now have come to believe that over the next generation Gamma Knife ra-

diosurgery will be the mainstay of vestibular neuroma care with surgical resection being reserved for those needing urgent decompression or the very young patient’’ [88]. We contended in our report [85], ‘‘The emergence of newer approaches in the management of these and other tumors, such as gene therapy, may someday make surgery performed with the microscope or Gamma Knife obsolete. It is humbling to imagine that, in the future, the act of physically cutting out a disease process and disentangling it from normal structures, or, for that matter, treating it in situ with a burst of high

1051

1052

66


. Figure 66-10 Contrast enhanced T1-weighted MRI of a left vestibular schwannoma before (a) and 24 months after GKS (b). The tumor measured 11 cm3 at the time of GKS and decreased to a volume of 4.4 cm3

energy, will seem primitive, no matter how refined and glamorous we make it look today.’’ Long term outcomes in patients treated with radiosurgery or microsurgery should be thoroughly evaluated to determine the place of the different methods in the management of this challenging lesion.

and worsening of the disease in three (12%). On imaging, the schwannomas shrank in 12 patients (48%), remained stable in 10 patients (40%), and increased in size in three patients (12%). These results were the same in the group with upfront GKS and the group with microsurgery plus GKS. No tumor growth following GKS was observed in patients with neurofibromatosis.

Trigeminal Scwannomas Neurocytomas Trigeminal schwannomas are rare intracranial tumors. In the past, resection and radiation therapy were the mainstays of their treatment. More recently, neurosurgeons have begun to use radiosurgery in the treatment of trigeminal schwannomas because of its successful use in the treatment of vestibular schwannomas. We treated 26 patients with trigeminal schwannomas at Lars Leksell Center between 1989 and 2005 [89]. Five of these patients had neurofibromatosis. The median tumor volume was 3.96 cm3. The median prescription radiation dose was 15 Gy (range 10.2– 17 Gy). The mean follow-up period was 48.5 months. There was clinical improvement in 18 patients (72%), stable symptoms in four (16%),

Although considered benign tumors, neurocytomas have various biological behavior, histological patterns, and clinical courses. Microsurgical extirpation is the widely accepted upfront treatment. Rades and Fehlauer [90] compared 108 and 74 patients who underwent complete or incomplete resection without adjuvant radiotherapy; at 5 years they reported a tumor control rate of 85 and 46% in the former and latter groups, respectively. Although adjuvant therapy is generally not indicated if a total removal can be accomplished, a high recurrence rate after gross-total resection as assessed by surgeons has been reported [91,92].


Therefore, postoperative fractionated radiotherapy has been suggested. In the same study by Rades and Fehlauer [90], fractionated radiotherapy did not seem to provide additional benefit after complete resection. The 5-year tumor control rate with adjuvant radiotherapy increased from 46 to 83% after incomplete resection, although the treatment did not seem to improve survival rate. Given the possible adverse effect of radiotherapy on cognitive function, radiosurgery has been tried as the postoperative adjunct for neurocytoma residuals or recurrences. Between 1989 and 2004, we performed GKS in seven patients with a total of nine neurocytomas [93]. Three patients harbored five recurrent tumors after gross-total resection, three had progression of previous partially resected tumors, and one had undergone a tumor biopsy only. The mean tumor volume at the time of GKS ranged from 1.4 to 19.8 cm3 (mean 6.0 cm3). A mean prescription dose of 16 Gy (range 13–20 Gy) was prescribed to the tumor margin. After a mean follow-up period of 60 months, four of the nine tumors disappeared and four shrank significantly (> Figure 66-11). Because of secondary hemor-

66

rhage, in the remaining tumor an accurate volume could not be determined. Four patients were asymptomatic during the follow-up period, and the condition of a single patient who had residual hemiparesis from a previous transcortical resection of the tumor was stable. The patient with an intratumoral hemorrhage and rupture into ventricles required a shunt revision, and another patient died of sepsis due to a shunt infection.

Hemangioblastomas The accepted treatment for hemangioblastomas is the surgical resection of the solid component of the tumor, although radiosurgery may be a reasonable alternative for tumors in the pituitary stalk and brainstem. We have treated a total of 16 hemangioblastoma patients, 5 of whom had von Hippel Lindau disease. As with surgery, the solid portion was targeted in all of these cases. These patients were treated with a mean prescription dose of 15 Gy to the tumor margin and followed for an

. Figure 66-11 T1-weighted contrast-enhanced MR image reveals a moderately enhancing neurocytoma spanning both lateral ventricles (a). The previous surgical track can also be seen. The last follow-up image obtained 14 years later shows that the tumor has decreased in size significantly with only some residual tissue in the septum pellucidum (b)

1053

1054

66


average of 20.9 months. Twelve patients (75%) had a decrease in the solid component of the tumor, while four patients had no change (25%). It is not unusual for the cystic component of the tumor to grow regardless of the behavior of the solid portion. Six of sixteen patients (42%) in our series required surgery for expanding cysts. Some have hypothesized that the molecular basis for hemangioblastomas in patients von Hippel Lindau disease may make them even less sensitive to radiation, but we do not have enough patients to support or refute this observation.

although it should be noted that the value of the study is reduced by the fact that less than half of the patients had confirmatory pathology. We have follow-up of more than 1-year for 20 grade II astrocytoma patients (10 with biopsy confirmation). Three (15%) disappeared, ten (50%) shrank, two (10%) remained unchanged, and five (25%) increased in size. One patient died from progression of his disease at 46 months after treatment [95].

Brainstem Gliomas

Malignant Brain Tumors Low Grade Astrocytomas Heppner et al. reported 60 low-grade (I and II) astrocytomas treated at the Lars Leksell Center. The general indications for Gamma Knife radiosurgery were deep-seated tumors not easily approached surgically, or cases in which the patients insisted on GKS. Of 15 patients with grade I astrocytomas (6 with biopsy confirmation) with greater than 1-year follow-up, nine shrank after radiosurgery (60%). Five patients were pediatric and ten were adults. Tumor size was found to be a significant prognostic factor, with the best results found in patients with a tumor volume of less than 3 cm3, and in patients who had previous craniotomy and debulking. Failure of radiosurgery occurred in six cases (40%), and five of these patients were adults. Two patients had subsequent surgery, one for an increase in tumor size and one for hemorrhage and radiation-induced changes. In two patients, a cyst associated with the tumor enlarged in spite of the solid portion becoming smaller. One of these patients was the only patient to have a decline in neurologic function following GKS [94]. These overall findings support the fact that grade I astrocytomas in adult patients tend to behave like grade II tumors,

Brainstem gliomas deserve special mention. They have an indolent clinical course. Management in the past involved monitoring with open-ended imaging studies and shunt placement if cerebrospinal fluid diversion becomes required. However, more recently the taboo against treating well-defined brainstem gliomas with microsurgery and radiosurgery has been eliminated. We treated 22 patients with brainstem gliomas; 17 tumors were located in the midbrain, four in the pons, and one in the medulla oblongata. The selection criteria were a well-defined tumor and progression on imaging and/or deterioration in clinical condition. The mean tumor volume at the time of GKS was 2.5 cm3. A tissue diagnosis was available in only 11 cases (50%), and the remaining patients were treated based upon an appearance on imaging as well-defined small tumors. Mean prescription dose was 15 Gy (Range 10–18 Gy). Of 20 patients with greater than 12-months follow-up, the tumors disappeared in four patients (20%) (> Figure 66-12) and shrank in 12 patients (60%). Of these patients, one experienced transitory extrapyramidal symptoms and fluctuating impairment of consciousness (from somnolence to coma) for 6 months. Tumor progression occurred in four patients; of these four, one patient developed hydrocephalus requiring a ventriculoperitoneal shunt, two


66

showed neurological deterioration, and one patient died of tumor progression [96].

efficacy, if any, of radiosurgery in treating high grade gliomas.

High Grade Astrocytomas

Metastatic Tumors

It is difficult to accept the rationale to treat invasive and diffuse high grade gliomas with the highly focused Gamma Knife. Nevertheless, due to pressure from patients, families and referring physicians, we have treated 56 patients with malignant (grade III and IV) gliomas, with the majority of patients showing an initial decrease in tumor size. However, recurrence and progression was the general rule, with a median survival of 14 months in these cases. Because of the differences in histology and the variety of therapies and protocols available for these tumors, it is difficult to judge the benefit of GKS. However, our group, along with others [97], have observed a statistically significant prolongation of life expectancy in the group of patients undergoing aggressive multimodality treatment (e.g. including some or all of the following: radical tumor debulking, radiation therapy, chemotherapy and GKS). A carefully conducted randomized control study will ultimately be required to evaluate the

Except for solitary lesions causing mass effect, the treatment of brain metastases is primarily palliative. In the case of solitary metastases, the occurrence of long-term survival is not unheard of; but in general, the guiding philosophy is palliation, reversal of neurologic deficits and maintenance of quality of life and functional status. There has been disagreement regarding the total number and volume of tumors which can be treated using Gamma Knife radiosurgery in the instance of multiple metastases. Published data from other groups have suggested that more than three lesions should be treated using whole brain radiotherapy. However, we have experience successfully treating more than this number of tumors using radiosurgery alone in one or multiple sessions. Surgical extirpation of a solitary brain metastasis has been shown to significantly prolong survival if the primary disease is controlled. Likewise, whole-brain irradiation has been shown to be of benefit for certain tumor subtypes. These conclusions and the well-defined

. Figure 66-12 (a) Sagittal enhanced T1-weighted MRI 6 months after GKS shows a ring form enhanced tectal tumor. (b) MRI obtained 7 years after GKS shows complete tumor disappearance

1055

1056

66


margins of most metastatic lesions on neuroimaging studies make them amenable to GKS. Because of the high incidence of these lesions, the treatment of metastatic tumors has become one of the most common indications for GKS worldwide. Eight hundred and ninety-five patients have been treated at the Lars Leksell Center for metastatic brain tumors. Evaluation of our overall series demonstrates an overall control rate of 84%, with 10% of tumors completely disappearing on imaging follow-up, and 64% shrinking in size (> Figure 66-13). The median survival was 7.5 months, with most patients dying from systemic disease progression (compared with 4–6 months without treatment). In cases where adequate follow-up exists, we have evaluated the effectiveness of radiosurgery in treating various tumor subtypes (> Table 66-2). We have treated a total of 40 patients with 65 renal cell carcinoma brain deposits [98]. The average survival was 9.2 months after radiosurgery. A total of 41 tumors decreased in volume, 6 tumors disappeared and 16 tumors remained unchanged in size. Only two tumors increased in size following treatment. By contrast, in an unmatched control group of 119 patients that received whole brain radiation ther-

apy, patients had an average survival of 4.4 months [99]. Factors associated with longer survival included a higher presenting Karnofsky performance status, absence of extracranial metastases, adjuvant whole brain radiation therapy, and prior surgical resection. The size and the number of metastases did not have a significant effect on survival, although in cases of single metastases with controlled local disease, long survival could be achieved. We have treated a total of 90 melanoma patients, with a total of 133 tumors [100]. Forty tumors (30%) disappeared, 45 tumors (34%) shrank, 23 tumors (17%) remained unchanged in size, and 25 tumors (19%) grew. Mean prescription dose to the tumor margin was 19 Gy (Range, 12–23 Gy). The median survival was 10.4 months, with patients harboring a single lesion without visceral metastases having a better prognosis. The most common pathology that we have experience treating is lung carcinoma, with 903 metastases in 262 patients treated over the past 15 years. The median survival in patients treated with Gamma Knife was 15.4 months, compared with 14 months in patients treated with combined GKS and whole brain irradiation. Tumor control rates varied considerably with tumor size with an 84% control rate in tumors less than 0.5 mL; 95%

. Figure 66-13 Axial enhanced T1-weighted MR images demonstrating a metastatic deposit (21 cm3) from renal cell carcinoma (a). The lesion involved the midbrain, thalamus, and pineal gland. The tumor started to decrease 3 months after GKS (b) and disappeared 18 months after treatment (c)

Patients (nx)

86 40 262 90

Primary Pathology

Breast Renal Lung Melanoma

166 65 903 133

Tumors (n) 18 20 22 21

Mean prescription dose (Gy)

. Table 66-2 Gamma Knife surgery for brain metastases

10 (3–54) 7 (3–60) 12 (1–150) 9 (3–78)

MRI follow-up (mean/ range) (months) 12 9 21 17

% With disappearance 51 68 59 47

% With decrease

19 3 10 19

% With increase

18 20 10 17

% With no change

13 (3–60) 8 (3–64) 15 (1–160) 10.4 (1–82)

Survival (mean/ range) (months)


66 1057

1058

66


in tumors between 0.5 and 2mL; 89% in tumors between 2 and 4 mL; 93% in tumors between 4 and 8 mL; 86% in tumors between 8 and 14 mL and 87% in tumors greater than 14 mL. Age of less than 65 years, a Karnofsky performance score of greater than 70, controlled extracranial disease, and more than one GKS for multiple tumors were associated with increased survival. Eighty-six breast cancer patients with a total of 166 lesions had a median survival of 14 months after radiosurgery. High Karnofsky performance score and the presence of a single lesion correlated with prolonged survival [101]. In our series of 53 metastatic deposits in the brainstem [102], a mean tumor volume of 2.8 mL was treated with a mean prescription dose of 18 Gy. Follow-up was obtained in 37 patients (mean follow-up, 10 months). The tumors disappeared in seven patients, shrank in 22 and remained stable in three. Five patients experienced tumor growth. Of 35 patients with neurologic symptoms, amelioration was experienced by 21 patients, stabilization was seen in 11, while in three patients their neurological status deteriorated. The absence of extracranial disease was the only favorable prognostic factor. Previous whole brain radiation and the number of intracranial tumors were unrelated to survival length.

Tumors in the Pineal Region Tumors in the region of pineal gland and quadrigeminal plate can be treated by radiosurgery as long as they are relatively small and well demarcated. Backlund et al. [103] reported three pineocytomas, two ependymomas, three astrocytomas, one medulloblastoma, and three tumors of unknown histology in the pineal region treated with Gamma Knife. The average tumor diameter varied between 1 and 3 cm, and target doses of between 20 and 75 Gy were delivered to the lesions. The average duration of follow-up was 5 years. In three pineocytomas and two cases in which the biopsy had failed to provide the histological diagnosis, the

therapeutic results were excellent. In one ependymoma and two astrocytomas, the results were also good 1–3 years after the treatment. One ependymoma and one astrocytoma increased in size after treatment. A patient with a medulloblastoma and a patient with a tumor erroneously classified as a pineocytoma died 2 and 3 years, respectively, after treatment. We treated six pineocytomas, two pineoblastomas, two astrocytomas, two germinomas, one hemangioblastoma, and three tumors with unknown histology located in pineal region. Prescription doses ranging from 12 to 20 Gy were used in 12 patients treated with Gamma Knife alone. Prescription doses ranging from 6 to 15 Gy were used for four cases where GKS was used as a booster treatment. Thirteen tumors decreased in size. No side effects have been observed. Malignant tumors of the pineal region should be treated with conventional radiotherapy. However, if the size and shape of the malignant tumor permits the use of radiosurgery, this is likely advantageous because the results are comparable with those of radiotherapy without systemic side effects.

Chordomas We have treated a total of 19 patients with chordomas located in the clivus. Twelve of these tumors (63%) had cranial nerve involvement. The median prescription dose to the tumor margin was 16 Gy (Range, 12–23 Gy). We have follow-up longer than 2-years on 12 patients (mean follow-up 77 months). Five tumors (42%) shrank, four remained stable (33%), and three (25%) increased in size. With longer follow-up, however, the efficacy of radiosurgery in controlling tumor growth and/or recurrence is less convincing. Of four patients with longer than 5 years follow-up, only one had decreased in size, while the remaining three patients had local tumor recurrence. None of the patients had resolution of their cranial neuropathy.


Chondromas and Chondrosarcomas Skull base chondromas and chondrosarcomas are rare. We treated with GKS four chondromas and eight chondrosarcomas. More than 50% reduction in size was seen in two cases (both chondrosarcomas), and three cases shrank 25–50%. None progressed at follow-ups ranging from 1 to 5 years (median 3.5 years). Muthukumar et al. treated 15 patients (nine chordomas and six chondrosarcomas) with GKS. After 4 years, four of their patients had died; only two deaths were related to progression of disease, and both of these had progression outside of the treated area. Only one of the 11 surviving patients had tumor progression, and 5 had shrunk [104].

66

and proton beam therapy. The first uveal melanoma treated with Gamma Knife was in Buenos Ares. The use of GKS and its stereotactic technique requires that the eyeball is fixed relative to the stereotactic frame. This is accomplished using retrobulbar blocks and external fixative sutures which are attached to the frame. At the University of Vienna, a specially designed suture device attached to the frame is utilized [106]. We have treated five patients with uveal melanomas. At 18 months follow-up, slight shrinkage or unchanged size of the tumors were observed. When required, we place a spacer to elevate the eyelid to prevent radiation injury.

Vascular Malformations Hemangiopericytomas

Arteriovenous Malformations

Hemangiopericytomas are richly vascularized and aggressive neoplasms of mesenchymal origin. They have a predilection for both local and distant central nervous system recurrence, and a tendency to metastasize. The high recurrence rate of hemangiopericytomas during their history usually requires a combination of open surgery, radiosurgery and radiotherapy depending on the tumor’s size at the time of the recurrence. We have treated a total of 20 hemangiopericytomas in 17 patients. Mean prescription dose was 19 Gy (Range, 11–23 Gy). At a mean follow-up of 17 months post-radiosurgery, 12 tumors decreased in size, while 3 tumors remained stable. However, of the tumors that were followed for more than 36 months, 6 of 9 increased in size (67%).

After Roentgen’s discovery of X-rays, there was intense interest in the use of radiation for AVMs during the period 1914–1950, but the results were not encouraging [107–110]. This led to an almost unanimous consensus in the assessment of radiation as being worthless in the management of AVMs. With the introduction of Gamma Knife, the potential value of irradiation in vascular malformations was reassessed. Contributory factors included an increasing body of evidence that the cells constituting the vessel wall were responsive to ionizing radiation. Long-term angiographic follow-up of a small series of AVMs treated with fractionated conventional radiation by Johnson [111] in the 1950s revealed that the AVMs were obliterated in 45% of cases, although the result was never reproduced [112]. It seemed only logical that Gamma Knife be tested in the treatment of vascular malformations. In Apri1 1970, the first radiosurgical treatment for an AVM was performed by Steiner et al. at Karolinska Institute in Stockholm [113]. Since then, thousands of patients have been treated with this technique which has proven to be safe and effective.

Uveal Melanomas The most common surgical treatment for uveal melanomas is enucleation, but several centers have relatively large series in the treatment of these tumors with GKS [105,106]. Other therapeutic options include radium plaque therapy

1059

1060

66


Decision making of GKS for AVMs

The decision to treat an AVM radiosurgically should be based upon: 1. 2.

3. 4.

5. 6. 7. 8.

The natural history of the disease The presenting clinical symptoms: hemorrhage, epilepsy, headache, or the malformation being diagnosed incidentally The patient’s overall condition, including age, medical and neurological status The characteristics of the AVM, including its location, size, the number and pattern of feeding and draining vessels The presence or absence of a concurrent intracerebral hematoma The anticipated risk-benefit ratio of the available therapeutic alternatives The results of cerebral blood flow studies The relationship of the malformation to surrounding brain structures as determined by MRI or CT

In an AVM that has not hemorrhaged by the time of diagnosis, the risk of hemorrhage in the interval between the treatment and the subsequent cure of the patient is not high. In these cases radiosurgery is recommended, especially if the AVM is in a location that makes surgical excision hazardous to the patient. Patients with a prior history of hemorrhage are at a higher risk to rebleed at least in the first 2 years after the hemorrhage. For these cases microsurgery should not be discarded as an option since it provides immediate cure. Radiosurgery is also indicated in patients in whom other medical conditions preclude surgery or increase the risk of anesthesia or surgery. An issue that is sometimes discussed in the context of AVMs is the management of associated aneurysms. These aneurysms may be present on vessels that are unrelated to the malformation, and in these cases they should be managed as independent pathologies. However, they are sometimes present on a major feeding

vessel and may decrease in size as the AVM flow diminishes in response to radiosurgery. In this case, it is perhaps wiser to wait and observe this group of aneurysms until the AVM has been obliterated. A third group of aneurysms are intranidal aneurysms. These aneurysms disappear with the obliteration of the AVMs. There are some contentions that intranidal aneurysms are thin-walled structures that are potential sites of hemorrhage in the AVM and therefore should be embolized before radiosurgery [114]. This seems reasonable; however we observe cases where both the nidi and the peri- and intranidal aneurysms are obliterated by radiosurgery. Definition of Total, Subtotal and Partial Obliteration

Angiography following radiosurgery reveals that hemodynamic changes occur before changes in the size and shape of an AVM. The flow rate decreases progressively. Sometimes the sizes of the feeding arteries and outflow veins decrease as well. Total obliteration (> Figure 66-14) of the AVM after radiosurgery was defined by Lindquist and Steiner as ‘‘complete absence of former nidus, normalization of afferent and efferent vessels, and a normal circulation time on high-quality rapid serial subtracted angiography’’ [115]. Any remaining nidus, regardless of its size, represents partial obliteration (> Figure 66-15). Subtotal obliteration of an AVM (> Figure 66-16) means the angiographic persistence of an early filling draining vein but no demonstrable nidus. The early filling venous drainage suggests that the malformation has not been completely obliterated. Personal Material

We treated the first AVM with Gamma Knife in April 1970 [113]. In 1977, we reported 30 AVMs treated with Gamma Knife [116]; and up to date, 2,650 AVMs have been treated by the senior author. The presenting symptoms were hemorrhage (70%), seizure (16%), headache (5%), neurologi-


66

. Figure 66-14 Total obliteration of a midbrain AVM. Vertebral angiograms with AP and lateral views before (a and b) and 2 years after GKS (c and d). The AVM obliterated completely and patient had no neurological deficits

cal deficits (4%), and other symptoms (2%). Seventy-three percent of the malformations treated were located in deep or eloquent areas of the brain. Results

Up to March 1991, 880 patients were considered to have been optimally treated with prescription doses of 23–25 Gy. Four hundred and sixty-one of these patients had appropriate angiographic follow-up. This relatively low percentage (52%) of follow-up is explained by the fact that accord-

ing to the follow-up protocol, the majority of patients treated from March 1989 to March 1991 had not yet angiographic follow-up at the time of analysis. Complete obliteration of the AVM within 1 year after treatment occurred in 230 of 306 patients (75%). Of the 461 AVMs with satisfactory angiographic follow-up at 2 years, 369 (80%) were totally obliterated. Karlsson confirmed that the obliteration rate is dose-related. With doses less than 5 Gy, obliteration was seen

1061

1062

66


. Figure 66-15 Partial obliteration of an AVM. A left Sylvian fissure AVM is shown in AP and lateral views of left carotid angiograms before GKS (a and b). Same views 4 years after GKS show a decrease in the size of the nidus but persistent shunting through the partially obliterated malformation (arrowheads) (c and d)

in only 3% of cases. With a prescription dose of 5–14 Gy, 47% were obliterated. Increasing the dose up to 24 Gy resulted in an increase in the obliteration rate to 69%. With prescription doses of 25 Gy or more, obliteration rates of 88% have been achieved. The increasing benefit in terms of the obliteration rate of malformation levels out to a plateau at a prescription dose beyond 25 Gy (> Figure 66-17).

It is important to point out a concern that has been expressed by some authors about the alleged bias in reporting the results of radiosurgery. It is correct that angiographic follow-up of radiosurgically treated patient series is unsatisfactory. If the loss of follow-up angiography were to occur randomly, the remaining population would be a representative sample of the entire case material. However, if angiography is recom-


66

. Figure 66-16 Subtotal obliteration of an AVM. Stereotactic angiograms performed in 1991 shows a superior vermian AVM fed by bilateral cerebellar arteries and draining through the vein of Rosenthal into the straight sinus and superior vermian vein (a and b). Angiograms obtained in 1993 demonstrate the disappearance of the nidus but a persistent filling of the superior vermian vein at late arterial phase (arrowheads) (c and d)

mended only when there are no flow voids on the MRI, this introduces an inherent bias into the case material. To identify the extent of bias introduced by this fallacy, we assumed the worst-case scenario that the AVMs with MRI evidence of flow voids were still patent and added them to the unobliterated group. In the subgroup of patients who were treated with 25–30 Gy to the periphery of the malformation, this made no difference in regard to the obliteration rate. In the group of patients receiving a prescription dose of 20–24 Gy, the obliteration rate dropped

from 73.9 to 73% (statistically insignificant) [117]. Thus, although the ideal situation demands that uniform follow-up be available in all cases with an angiographic control, at least in our material, the current method provides a reliable estimate of the actual outcomes. We have published long-term neurological outcomes for 239 patients with AVMs treated with Gamma Knife between April 1970 and December 1983. Headache resolved in 65 (66%) of the 98 patients presenting with this symptom and improved in an additional 9 (9%).

1063

1064

66


. Figure 66-17 Dose-response curve of GKS for arteriovenous malformations. The curve illustrates the dose dependence of obliteration rate

Preexisting neurological deficits improved or disappeared completely after radiosurgery in 57% of the affected patients. The overall figure for residual neurological deficits was 3% [118]. Among the 247 patients analyzed and reported by us in 1992 [119], 59 had seizures in addition to their AVMs and 41 (69.4%) became seizurefree or improved markedly following GKS. An updated analysis of the outcome of epilepsy associated with AVM after GKS has revealed that of 178 patients with seizures associated with their AVMs before radiosurgery, 149 (84%) had improvement in their seizures. Of these, 110 (62%) were seizure-free, with 57 patients on no anticonvulsant therapy. Interestingly, 49 of the 110 patients who were cured of their seizures still had patent malformations at the last follow-up. Among the 317 pediatric patients, 106 had 2-year follow-up angiograms. Eighty-seven percent had total obliteration, 6% had subtotal obliteration, and 7% had only partial obliteration of the malformation. Embolization and Radiosurgery

Embolization has been used to shrink the nidus to a size suitable for radiosurgery. Howev-

er, it is important that the procedure do not fragment the nidus because doing so it increases the difficulty of localizing and targeting the multiple remnants and decreases the rate of total obliteration. Sometimes, the existence of embolic material also impedes a clear visualization of the nidus. Literature on the relative merits of various embolic materials is sparse and no single author has a large enough experience with different materials to draw valid conclusions. The two pertinent choices are particle materials and acrylic. Particles are believed to result in a higher rate of recanalization as compared to acrylic. However, exact differences in the rates of recanalization are not known. A new liquid embolic agent (Onyx) has been introduced recently. The agent is less adhesive and polymerizes slowly. This allows a better control of intranidal injection of the embolization material. However, the impact of this new agent upon radiosurgery awaits further evaluation. Microsurgery and Radiosurgery

In our material, GKS was used for residual nidus following microsurgery in 218 patients; 182 of them underwent follow-up angiography at 2 years and, 153 (84%) were found to be cured. Hemorrhage of Unobliterated AVMs

The issue of possible protection against hemorrhage in irradiated but still patent AVMs is highly controversial. We contend that patients, whether treated with microsurgery, radiosurgery or endovascular techniques, remain at risk for bleeding as long as the malformation is still patent. To assess the rate of hemorrhage, we calculated a probability estimate using both the personyears method and the Kaplan-Meier life table [118,119]. With the person-years method, the actual hemorrhage rate is similar to the value observed in the natural history. Analyzed using the Kaplan-Meier method, we found a risk of 3.7% per year up to 60 months post radiosurgery. Five years following the treatment, the life table ended in a plateau which could be


interpreted as indication of decreased risk of hemorrhage. However, this does not imply that the real risk of bleeding is negligible unless a large number of patients have been followed well into or beyond the plateau region. Karlsson studied our material to evaluate the incidence of hemorrhage during the first 2 years after radiosurgery in 1,565 patients treated before May 1992 [120]. Forty-three patients experienced hemorrhage before obliteration of their AVMs in this group. This amounts to an interval bleeding rate of 2.8% over 22 years. This risk of hemorrhage is similar to the natural history of the disease. The hemorrhage was fatal in 14 (0.9%) cases and left residual neurological deficits in 9 (0.6%). In our series of AVMs treated with GKS, we observed no recurrent hemorrhage after angiographically confirmed obliteration of a vascular malformation. In one case reported by Guo, a rebleed occurred after angiographic documentation of nidus obliteration [121]. The MRI findings suggested that hemorrhage possibly resulted from radiation-induced tissue damage. Furthermore, the histological examination of the suspected recanalized AVM revealed channels that were one-fiftieth the size of the smallest vascular channels in AVMs, making it unlikely that these were vessels with significant blood flow. This view has been further confirmed by the fact that a repeat angiogram revealed no evidence of residual malformation or recanalization. Rebleeding, in spite of posttreatment angiograms interpreted as normal, may be explained by unsatisfactory quality of the neuroimaging studies or inadequate interpretation leading to the misdiagnosis of angiographic cure [122,123]. A small residual nidus may have been missed as well because the nidus did not fill due to hemodynamic condition at the time of followup angiography. Subtotal Obliteration of AVMs (SOAVMs)

66

Subtotal obliteration of an AVM implies a complete disappearance of AVM nidus but persistence of early filling drainage veins. We reported a series of 159 patients with SOAVMs [124]. The incidence of SOAVMs was 7.6% from a total of 2,093 AVM patients who were treated with GKS and had angiographic follow-up available. The volume of the AVM nidi before GKS ranged from 0.1 to 11.5 cm3 (mean, 2.5 cm3). Sixteen patients underwent two GKS before the development of SOAVMs. The diagnosis was made after a mean of 29.4 months (range 4–178 months) following the initial GKS. Of the 175 treatments in 159 patients, the mean prescription dose was 22.5 Gy (range 15–31 Gy).

Patients and AVMs Parameters

Four patients were lost to follow-up. Of the remaining 155 patients, the clinical follow-up ranged from 5 to 185 months (mean 59.4 months). During the cumulative period of 767 patient-years (a mean of 4.9 years per patient) no SOAVM had ruptured. Follow-up angiography was performed in 90 of 136 patients in whom SOAVMs had no further treatment. These studies showed a total obliteration of the AVM as well as disappearance of the early filling vein in 66 patients (73%). Twentyfour patients (27%) had persistent SOAVMs. In patients with deeply located nidi or deep draining veins, the incidence of subsequent obliteration of SOAVMs is higher. Twenty-three patients with SOAVMs were treated with Gamma Knife targeting the proximal end of the early filling veins. In this group, follow-up angiography was performed in 19 patients, confirming disappearance of the early filling vein in 15 patients (79%) and persistent SOAVMs in four patients (21%). Compared with patients who received no further treatment, patients treated with repeat GKS had a slightly higher incidence of subsequent disappearance of the draining vein (79% compared with 73%), but the difference is not statistically significant. None of the 155 patients suffered a rupture of

Imaging and Clinical Outcomes

1065

1066

66


the lesion. This suggests that the protection from rebleeding at the stage of subtotal obliteration is significant. Our series shows that subtotal obliteration of the AVMs did not necessarily prove to be a premature stage of an ongoing obliteration, and instead might be the end point of the obliteration process. Earlier in our series, we repeated GKS for SOAVMs, targeting the proximal segment of the early filling vein. After repeat treatment, 79% SOAVMs were obliterated. However, the necessity of retreatment remains to be determined given the fact that in the whole group no hemorrhage occurred and that 73% of SOAVMs obliterated spontaneously. Repeat GKS for AVMs

The rationale for repeat GKS in AVMs is the persistence of risk of hemorrhage as long as the nidus remains patent. The alternatives for the management of a still patent AVM are embolization, microsurgery, or radiosurgery. Of these alternatives, surgical extirpation or radiosurgery proved to be more efficient than endovascular techniques. In location of difficult approach with microsurgery, the risk of morbidity and mortality may be high. In the following, we present our results of repeat GKS when the first treatment did not totally obliterate the malformations. One-hundred twenty patients underwent repeat GKS for a still-patent AVM at Lars Leksell Center between 1989 and 2003. The mean ages at initial and repeat GKS were 28.1 (range 4–70 years) and 32.4 years (range 8–74 years), respectively. The mean interval between the first and second GKS was 4.4 years (range 2–12.8 years). Thirty-seven patients underwent one or more embolization procedures (range 1–5). Incomplete surgical resection was carried out in 14 and a combined approach with embolization and microsurgery was performed in three. Fourteen patients had a hemorrhage following the initial Gamma Knife

Patients and AVMs Parameters

procedure over 539 risk years. The annual incidence of hemorrhage was 2.5%. The locations of the AVMs were in the cerebral hemispheres in 64, thalamus or basal ganglion in 34, corpus callosum in 6, brainstem in 12 and cerebellum in 4 patients. Ninety-four (78.3%) AVMs were located in eloquent area. Thirty-three (27.5%) AVMs only had superficial venous drainage, 87 (72.5%) had deep venous drainage. The causes of initial treatment failure were (1) inaccurate nidus definition at initial GKS in nine patients (7.5%); (2) failure to visualize some segments of nidus on angiography due to hemodynamic factors or still-existing hematoma in 11 patients (9.2%); (3) recanalization of prior embolized nidal compartment in six patients (5%); (4) suboptimal radiosurgical dose (less than 23 Gy) required because of a large targeted volume or a critical location of the nidi in 65 cases (54.2%); (5) subtotal obliteration of AVMs in 13 (10.8%); (6) unknown causes in 16 (13.3%). At the initial GKS, the maximum diameter of the nidus ranged from 8 to 50 mm (mean 26.6 mm), and the volume ranged from 0.1 to 24 cc (mean 4.0 cc). Mean prescription dose was 20.5 Gy (range 5–30 Gy). The maximum diameter of the retreated nidus ranged from 3 to 47 mm (mean 16.2 mm), and the volume ranged from 0.1 to 12.7 cc (mean 1.4 cc). Mean prescription dose of 19.6 Gy (range 4–27 Gy) was used at repeat treatment.

Treatment Parameters

Imaging and Clinical Outcome After Repeat GKS

Follow-up angiography was carried out after a mean of 45.9 months (range 12–189 months). The MRI or angiograms visualized residual nidus in 41 (34.2%) patients. Twelve (10%) patients refused a follow-up angiography in spite of absence of flow voids on the MRI. Angiography confirmed total obliteration of the nidus in 60 (50%) and subtotal obliteration in 7 (5.8%). Nine patients experienced 13 episodes of hemorrhage in 553 risk years after repeat


GKS (1 patient had 4 hemorrhages, 1 had 2, and 7 had 1), yielding an annual incidence of 2.4%. A lower rate of nidus obliteration was related to poor response to initial GKS, lower prescription dose, larger nidus size, and previous embolization. The clinical follow-up ranged from 18 to 237 months (mean 80.2 months). Twenty patients remained asymptomatic since their repeat GKS and forty-five improved to be symptom-free. An additional 44 improved significantly but still had residual neurological deficits. Eleven patients deteriorated; eight related to a bleed, two caused by persistent arteriovenous shunting, and one related to radiation induced changes. No GKS-related mortality has occurred in this group. One patient developed a meningioma 12 years following the initial and 7 years following the repeat GKS. Three patients developed an asymptomatic cyst at the site of the treated AVMs at 5, 7, and 7 years, respectively following the repeat treatment. GKS in Large AVMs

While satisfactory results in small and moderately sized AVMs following radiosurgery are well documented, reports on the imaging and clinical outcomes in large AVMs are sparse. The explanation presumably is that few neurosurgeons ventured the challenge or that few treated large enough series with appropriate follow-up periods. In addition, less enthusiasm to report meager results may explain the paucity of data in a field often characterized more by a deluge than scarcity of publications. The main problem with larger AVMs is due to the dependence of the obliteration response on dose and volume; this dependency requires a delicate balance in deciding an efficient dose but low enough to avoid adverse neurological deficits. The following strategies are currently available to treat large AVMs (Spetzler-Martin grade IV and V) with radiosurgery. First, one can embolize of the AVM then perform radiosurgery if the nidus shrinks to a size manageable with radiosurgery. However, embolization frequently fragments the nidus into a number of

66

segments making the radiosurgical planning difficult and increasing the probability of radiosurgery failure. Another strategy involves radiosurgery using staged dose delivery to the entire AVM volume or serial staged radiosurgery to selected volumes of the AVM. Sirin et al. used staged volumetric radiosurgery in 28 large AVMs [125]. Out of the 21 patients, seven underwent repeat radiosurgery and were eliminated from outcome analysis. Of the remaining 14 patients, 3 had total obliteration on angiograms, and 4 had no flow voids on MRI but had no follow-up angiography. Four patients had hemorrhages after radiosurgery resulting in two deaths. Worsened neurological deficits occurred in one patient. Seizure control improved in three patients and deteriorated in one. Twenty-one patients out of 28 had follow-up longer than 36 months. Sirin’s report, while flawed due to selection bias (eliminating seven cases where AVMs failed to respond to the initial treatment in other 7 patients only 3 had angiographic confirmation of nidus obliteration) still provides some pertinent information. Pan et al. [126] reported an obliteration rate of 25% for AVMs with volume larger than 15 cm3 treated with a single GKS. The obliteration rate increased to 50% at 50 months follow-up. The morbidity was 3.3%. Post-treatment hemorrhage occurred in 9.2% of cases. Miyawaki et al. observed 22 cases of radiation necrosis in a series of AVMs of comparable size to that of Pan’s series [127]. We evaluated a protocol using combined radiosurgery and microsurgery for the management of large AVMs. Radiosurgery was performed for the deep medullary portion of the AVM as a first step. The second step was planned as microsurgical extirpation of the superficial segment if the goal of the first step, obliteration of the deep segment of the AVM, was achieved. However, in no case was this goal achieved. The management of large AVMs demonstrates that every treatment has its limits. In an effort to solve the problems of the management

1067

1068

66


of large AVMs, a cautious approach is warranted pending the development of new techniques and agents for embolization, the development of new energy source for the focus beam therapy and the development of brain protectors. In very large AVMs perhaps ‘‘wait and see’’ may occasionally be the best management.

Dural Malformations Vascular malformations arising wholly or partially from the dura are amenable to treatment by GKS. We have treated 53 patients with dural fistulas, of whom 9 also had an associated AVM. Among 19 fistulas under 15 mm, 10 had adequate follow-up. There were seven (70%) cures and two subtotal obliterations (> Figure 66-18). In patients with malformations larger than 15 mm but smaller than 25 mm, 6 of 20 patients had adequate follow-up. Three were cured and one had subtotal obliteration. Of 14 patients with fistulas larger than 25 mm, follow-up was available in 7 patients. Five of these were cured, and one fistula was subtotally obliterated. There were three rebleeds. One of the malformations subsequently was obliterated and the other two were still patent. Radiation-induced changes appeared in three patients 8–12 months after radiosurgery but disappeared in all three by 18 months. No neurological deficits from these untoward effects were observed. We recommend radiosurgery for dural malformations that extend over a short distance or have few ‘‘holes.’’ Malformations over a long stretch with ‘‘multiple holes’’ should be managed by radiosurgery preceded by surgical and/or embolization.

Cavernous Malformations Cavernous malformations constitute the major portion of so called angiographically occult vascular malformations. They tend to have a benign course and the reported hemorrhage rates have been low in various natural history series, varying from 0.25 to 0.7% per year [128–130]. The main indications for aggressive treatment include repeated hemorrhages, progressive neurological symptoms or medically intractable seizures. Karlsson reported 23 cavernous malformations treated with Gamma Knife at Karolinska Institute, 16 by Steiner between 1985 and 1987 and 7 by Lindquist and Karlsson between 1988 and 1996 [131]. One was lost to follow-up. Nine of the twenty-two patients suffered a postradiosurgical hemorrhage and six developed a radiation-induced complication. MRI revealed a decrease in the size in three and no size change in the rest. The annual post-radiosurgical hemorrhage rate was 8%. There was a trend in the hemorrhage rate to decrease 4 years postradiosurgery. Higher prescription doses seemed to result in a lower risk of posttreatment hemorrhage. However, it could not be concluded whether radiosurgery changes the natural course of a cavernous malformation and the incidence of radiation-induced complications was approximately seven times higher than that expected if the same number of patients had been treated by GKS with the same dose for AVMs. Based on these disappointing results, we stopped treating cavernous malformations. In a number of publications, Kondziolka et al. and Hasegawa et al. studied the natural history of cerebral cavernous malformations and reduction of hemorrhage risk as well as long term results after radiosurgery for patients with cavernous malformations managed in the Department of Neurosurgery, Presbyterian Hospital, University of Pittsburgh [129,132,133].


66

. Figure 66-18 Total obliteration of a dural AVM following GKS. Left common carotid injection reveals the dural AVM in the area of left transverse sinus on AP and lateral projections (a and b). The AVM obliterated 2 years following GKS (c and d)

In a series of 47 patients, they observed a postradiosurgery annual hemorrhage rate of 8.8% [133] as compared to 0.6% prospective annual rate of hemorrhage in patients without a prior bleed and 4.5% annual bleed rate in patients with prior hemorrhage by studying the rate of symptomatic hemorrhage in a series of 122 nontreated patients [129]. In the group of 47 patients, they compared the post-radiosurgery hemorrhage rates with the pre-radiosurgery rate assuming that the rate could be based on an ‘‘epoch’’ starting from the first hemorrhage. This is fallacious since the malformation might have been present since birth.

Recomputed on this basis, the pre-radiosurgery annual bleeding rate comes to 5.9% which is more congruent with the expected natural history. The incidence of hemorrhage after radiosurgery appears to be higher but it is unlikely that radiosurgery increases the risk for hemorrhage. However, it seems to offer no protection or provide some protection at the high cost in terms of side effects. Hua et al., Connolly et al., and Friedman [132] while mentioning the thoughtful presentation of a controversial issue emphasized some flaws in the publications of Kondziolka et al. and Hasegawa et al. like the difference in

1069

1070

66


hemorrhage rate between the patients from Pittsburgh and series of Mariority et al. [134] or when using the study group as its own control. It is difficult to quantify the effects of radiosurgery without a true control population studied in parallel not in serial. Hua et al. mentioned that without an accurate quantification of the benefits of radiosurgery, a risk-to-benefit analysis in consideration of the complications of radiosurgery is not possible [132]. Karlsson assuming a less than 1% decrease in the annual risk for hemorrhage and a 20–30% complication rate states that it will take significantly more than 20 years before the beneficial effect from the treatment with radiosurgery exceeds the complication rate (personal communication). Pollock publishing the results of Gamma Knife radiosurgery in cavernous malformations treated in Mayo Clinic could not confirm the findings in publications he coauthored in Pittsburgh. He concludes that considering the complications of radiosurgery, its benefits may not be sufficiently great to warrant its use [135].

Venous Angiomas Based on our published results [136], we concluded that radiosurgery for venous anomalies (venous angiomas) does not fulfill the rigid criteria of minimal risk that must be set for the treatment of a lesion with a benign natural history. Thirteen patients with venous angiomas were treated between 1977 and 1987. In two cases, the venous angioma shared venous drainage with the associated AVMs. Cavernous malformations coexisted in two cases. Imaging follow-up was available in all but two cases. After treatment, complete obliteration of the venous angioma was observed in one patient, partial obliteration was observed in four, and no effect was found in four. In two patients where the AVMs used venous angiomas as venous drainage, only the AVMs were treated. Both

AVMs obliterated but the venous angiomas remained unchanged. Undue effects of radiation occurred in four patients; one patient had focal edema, and three had radionecrosis. Extirpation of the radionecrotic tissue 6 months after radiosurgery was necessary in one case. Literature elucidating the natural history of the venous angiomas became available only after the treatments included in the study were completed and was reviewed when we were drawing our conclusions [137]. We quoted Garner and associates [138], highlighting the statement that ‘‘surgical resection of venous anomalies is rarely indicated.’’ This fact was far from universally accepted at the time when these patients were treated.

Vein of Galen Malformations We have treated a series of nine Vein of Galen malformations (VGMs) including eight children aged 4–14 years and one adult with GKS [139]. Obliteration was achieved in four patients (three following a single treatment and one after two treatments) and partial obliteration in another three. Another patient with a significant initial response underwent repeat GKS but refused to undergo follow-up angiography after the second procedure in spite of the fact that the MRI could no longer visualize the malformation. In one case GKS and multiple embolization sessions had no effect on the size of the VGM. One patient experienced a transient neurological deficit. Another patient had evidence of a radiation-induced change on MRI, but the change was clinically silent.

Carotid-Cavernous Fistulas The ideal treatment of a spontaneous carotidcavernous fistula would consist of obliteration of the fistula with maintenance of the patency of the carotid artery. This goal is not achieved in practice by many of the current methods.


We have treated eight patients with carotidcavernous fistulas with Gamma Knife. The ages of these patients ranged from 27 to 70 years. Follow-up is available on six, and five of the fistulas were cured. No patient experienced any undue side effects. Since these patients usually have acute complaints that require immediate intervention, endovascular procedures are the first choice. If these procedures do not occlude the fistula, radiosurgery should be used.

Arterial Aneurysms Leksell insisted to try the GKS in arterial aneurysm. A 61-year-old lady who sustained a subarachnoid hemorrhage from a left posterior communicating artery aneurysm was treated with Gamma Knife [140]. The patient refused microsurgery, and the aneurysm was treated with two isocenters and received a prescription dose of 25 Gy. Over the next 11 months, there was a progressive decrease in size and finally obliteration of the aneurysm. The posterior communicating artery adjacent to the aneurysm also progressively narrowed in caliber and ultimately was obliterated without a neurological deficit. The patient refused a vertebral angiography, hence we cannot know whether the aneurysm would fill or not from posterior circulation. An additional 15 cases of arterial aneurysms were treated by Forster at Karolinska institute. All except one died of a hemorrhage a few weeks to months after the Gamma Knife treatment. For cases of perinidal and intranidal aneurysms associated with AVMs, Gamma Knife quite often occluded these lesions following obliteration of the AVM nidus. However, these aneurysms may increase the risk of hemorrhage during the latency period from the time of radiosurgical procedure to the resulting obliteration of the aneurysms. Therefore, they should be embolized before radiosurgery. If embolization is not feasible, radiosurgery can often obliterate these small arterial aneurysms.

66

Functional Disorders At the time when Leksell introduced radiosurgery, the term was synonymous with functional neurosurgery. Gammathalamotomies were used not only for tremor but also for intractable pain [141]. The Gasserian ganglion was irradiated for trigeminal neuralgia and gamma-capsulotomies were performed to interrupt fronto-limbic connections in the treatment of intractable anxiety and obsessive-compulsive disorders [142,143]. The introduction of better drugs, the emergence of non-ablative methods, the lack of imaging to provide the precision required by functional neurosurgery, and the fact that the placement of the lesion with Gamma Knife could not be corroborated by physiological methods, caused a decline in the use of Gamma Knife for functional disorders. However, recent developments in imaging techniques have led to a reassessment of the possibilities of functional radiosurgery.

Cancer Pain Gammathalamotomy for intractable pain was one of the first procedures performed with the Gamma Knife. Steiner et al. reported the outcome in 52 patients with terminal cancer treated with Gamma Knife for pain control [141]. Since CT scan or MRI was at that time not available, pneumoencephalography was used to target the thalamic centranum medianum (CM-Pf complex). Three by five and three by seven millimeter collimators were used. Lesions occurred in 21 of 36 patients that had postmortem examination. No lesion was observed with a maximum dose less than 140 Gy. The most effective lesions were more medially located near the wall of the third ventricle. Better results were observed for face and arm pain. Pain relief of variable degree was obtained in 26 patients and the pain relief was only temporary lasting not longer than 2–4 days. In 5 of 8 patients, a relatively satisfactory pain

1071

1072

66


relief lasted until the patient’s death 13, 10, 7, 4, and 1 months after the procedure, respectively. In three cases, the pain relief lasted for 9, 6, and 13 months, respectively and then returned with preoperative intensity. Hayashi et al. reported pain alleviation in patients with cancer pain and post-stroke thalamic pain following radiation of hypophysis with Gamma knife [144]. Young et al. obtained pain relief in chronic intractable pain using gammathalamotomy [145].

Trigeminal Neuralgia During his neurosurgical carrier, Leksell has been obsessed with the idea to alleviate pain. Twenty years before building the Gamma Knife, he used orthovoltage stereotactic technique to treat patients with trigeminal neuralgia (TN) and achieved long term relief of symptoms. From 1996–2003, we treated 151 cases of TN with Gamma Knife at Lars Leksell Center [13]. Radiosurgery was performed once in 136 patients, twice in 14 patients, and three times in 1 patient. One hundred twenty-two patients had typical TN, three with atypical TN, four with multiple sclerosisassociated TN, and seven with TN and a history of a cavernous sinus tumor. In each case, the chosen radiosurgical target was located 2–4 mm anterior to the entry of the trigeminal nerve into the pons. The maximum radiation doses ranged from 50 to 90 Gy. The mean time to relief of pain was 24 days (range 1–180 days). Forty-seven, Forty-five, and thirty-four percent of patients were pain free without medication at the 1-, 2-, and 3-year follow-ups, respectively. Ninety, seventy-seven, and seventy percent of patients experienced some improvement in pain at the 1-, 2-, and 3-year follow-ups, respectively. Twelve patients (9%) suffered the onset of new facial numbness after treatment. Although less effective than microvascular decompression, GKS remains a reasonable treatment option for those unwilling or unable to undergo more invasive surgical approaches.

Movement Disorders Purely out of historical interest, we mention that between 1968 and 1970, Leksell used the prototype of Gamma Knife for the production of thalamic lesions in five cases of tremor. The target was indirectly determined by using derived coordinates relative to anterior and posterior commissures as visualized by pneumoencephalography. The results were unsatisfactory. Following the introduction of stereotactic MRI, two patients with Parkinsonian tremor were treated [146]. Hirai and colleagues clarified the position, anatomic organization, and physiologic significance of the thalamus as it pertains to tremor, rigidity, and dyskinesia [147]. The correlations between neuroanatomic and electrophysiological findings in the human ventrolateral thalamic nuclei (e.g., VLa, VLp, VPLa, and VPLc) are now better understood. For GKS, the difficulty arises in identifying the VLp and VLa nuclei in the human thalamus purely by imaging. Rand observed improvement of tremor in four of seven patients with nucleus ventralis lateralis (NVL) lesion and improvement of rigidity in two (personal communication). In 4 of 8 patients treated with gamma-pallidotomy, the symptoms improved significantly. Two of three patients treated with an NVL lesion for intention tremor showed improvement. Duma et al. produced thalamic lesions in 34 patients and observed improved tremor in 63% of the patients [148]. Young et al. obtained similar results [149]. Ohye (personal communications) has used a single four millimeter isocenter and 130 Gy for gamma-thalamotomy in 27 patients with Parkinson’s disease. Tremor and/or rigidity improved in 85% of patients. Hirai (personal communications) has treated 14 patients with GKS for involuntary movement disorders. Of these 14 patients, 8 had tremor dominant Parkinson’s disease, 4 had rigidity and dyskinesia-dominant Parkinson’s disease, and 2 had essential tremors. Hirai’s target points were the VLp nucleus for


control of tremor and the VLa nucleus for control of rigidity and dyskinesia. Thirteen out of fourteen patients had subsequent improvement in symptoms. In nine of these patients, symptomatic improvement occurred by 50–80% in the patients’ Unified Parkinson’s Disease Rating Scale for tremor, rigidity and dyskinesia scores. Young et al. reported improvements in Unified Parkinson’s Disease Rating Scale tremor and rigidity scores in 74 out of 102 patients (73%). Of the patients with essential tremor, 88.2% became tremor-free [149]. An important change in the surgical management of movement disorders was the introduction of deep brain stimulation. Deep brain stimulation achieves amelioration of symptoms without a destructive lesion and has supplanted destructive lesions as the surgical procedure of choice.

Psychosurgery For obsessive-compulsive disorders Leksell used his open stereotactic system to target the frontolimbic connections in both anterior internal capsules and coined the term capsulotomy [150]. With the advent of Gamma Knife, he used it for psychosurgery instead of the open stereotactic tool. Mindus reported the series of patients with obsessive-compulsive disorders treated at Karolinska Institute; one group of 24 patients being treated with capsulotomy via a conventional thermocoagulation technique and followed for 1 year, another group of seven patients being treated by Gamma Knife and followed for 7 years. The clinical effects of these treatments were evaluated subjectively by two independent observers and were also rated on the Comprehensive Psychopathological Rating Scale. Ratings were performed 10 days before and 2, 6, and 12 months after surgery. The effects on the personality were evaluated by the Karolinska Scales of Personality. These scales have been developed to measure traits related to frontal lobe dysfunction and to reflect

66

different dimensions of anxiety proneness. At the 12-month follow-up, statistically and clinically significant improvement was noted in all assessments of symptomatic and psychosocial function. Freedom from symptoms or considerable improvement was noted in 79% of patients, and none were worse after the operation. A number of patients were preoperatively unable to work or function socially. Postoperatively, these patients could return to their previous occupation and to a normal social function. The results of gammacapsulotomy were found to be comparable to those of capsulotomy performed by the thermocoagulation technique.

Epilepsy Seizure was the presenting symptom in 59 of the 247 patients with AVMs of the brain treated by the senior author with Gamma Knife between 1970 and 1984. The treatment resulted in significant reduction of frequency or total content of seizures in 52 of these patients. Eleven were successfully taken off anticonvulsant medication. In three patients the seizure disorder stopped before the obliteration of AVMs. These observations prompted the idea of testing focal irradiation as a treatment modality for focal epilepsy. At University of Virginia, basic science research was done on changes in neuroexcitability after irradiation. The hippocampal slices from rats treated with the Gamma Knife were found to have a higher seizure threshold than those of controls when placed in solutions of varying concentrations of penicillin. This effect was lost at high concentrations (Henson and colleagues, personal communication, 2001). Using single doses of either 20 or 40 Gy to the hippocampus in a rat model of chronic spontaneous limbic epilepsy, a reduction in both the frequency and duration of spontaneous seizures was observed [151]. Histological evaluation of the targeted region revealed no signs of necrosis,

1073

1074

66


and hippocampal slice recordings revealed intact synaptically driven neuronal firing. Biochemical analysis of changes in rats brains after GKS, performed by Re´gis and colleagues, showed changes in the concentrations of excitatory and inhibitory amino acids (particularly gammaaminobutyric acid) [152]. Epilepsy has been treated with radiosurgery at many centers, but there have been few published long-term results. Barcia-Salorio and colleagues treated 11 patients with idiopathic epilepsy [153]. Complete relief from seizures was obtained in four patients, and significant reduction in seizure activity was seen in five patients. Re´gis and colleagues reported a case of mesial temporal lobe epilepsy treated with Gamma Knife [154]. They used 25 Gy given to the 50% isodose line. The patient was seizure-free after the treatment. Further results of 25 patients with medically intractable mesial temporal lobe epilepsy showed that of the 16 patients with more than 2 years follow-up, 13 were seizure-free and 2 were improved. There were three cases of nonsymptomatic visual field deficits. There was no mortality associated with the treatment [155]. The potential of a less invasive, nondestructive therapy to treat epilepsy prompted the creation of prospective European and a National Institutes of Health (NIH) sponsored multicenter studies of GKS for temporal lobe epilepsy. In the European study, three centers enrolled 21 patients with mesial temporal lobe epilepsy. The anterior parahippocampal cortex, the basal and lateral portions of the amygdala, and the anterior hippocampus were targeted, and patients received a mean dose of 24 Gy. At 2 years postradiosurgery, 65% of the patients were seizurefree. However, nine patients developed visual field deficits, and five suffered transient side effects including depression, headache, nausea, vomiting, and imbalance [155,156]. The NIH sponsored study randomized 40 patients into two dosage groups and monitored several clinical and imaging characteristics over 3 years

following radiosurgical treatment. These evaluation points include the effects on seizure frequency and severity, MR imaging, MR spectroscopy, and neuropsychological outcomes. Of four patients in the NIH study treated at Lars Leksell Center, two received a high dose (24 Gy) and became seizure free. Two cases received low dose (20 Gy). One became seizure free with auras; one had significant decrease of seizure attacks. All patients had headache and two had exacerbation of auras 9–12 months following Gamma Knife treatment. All patients had significant radiation induced changes on MRI at this period of time. One of the patients treated at another center as part of the NIH study experienced a serious adverse event that included persistent headache, visual changes, and cerebral edema, and these consequences necessitated a standard anterior temporal resection. Gamma Knife’s long-term effectiveness for epilepsy needs to be demonstrated. Also, it is unclear what underlying mechanisms are responsible for amelioration of seizures following radiosurgery. Some have suggested a ‘‘neuromodulation’’ phenomenon following GKS with accompanying glial cell reduction, stem cell migration, neuronal plasticity and sprouting, and biochemical changes [157]. Rigorous scientific studies evaluating the cellular and subcellular mechanisms responsible for improvements in epilepsy after GKS are thus far lacking.

Undue Effects The radiosurgical procedure is not associated with any immediate or short-term side effects per se. Patients sometime experience nausea. Rarely, seizures occurred in the posttreatment period, usually in patients with supratentorial lesions and who already had a history of seizure disorders. We recommend patients with seizure history maintain pre- and post-operative antiepileptic medications.


Radiation Induced Changes The imaging of radiation-induced changes is characterized by a bright signal on T2-weighted images on MRI (> Figure 66-19). In cases where this is associated with contrast enhancement on T1-weighted MRI, it presumably represents radiation-induced injury with an associated breakdown of the blood brain barrier. Guo reported that the radiation induced changes can be observed in 47% of cases following GKS for AVMs [158]. The onset of these changes occurred 3–15 months after the treatment in a majority of cases (92%) and more than 26 months after treatment in 8%. Progressive resolution of the radiation-induced effects is the usual course. The resolution is observed 1–17 months (mean 5 months) after the detection of changes. The clinical manifestations included headache, symptoms of raised intracranial pressure, and focal neurological deficits. In a small percentage of patients, this is associated with focal damage to neural tissue. Neurological deficits were still present at the time of the last follow-up in 3% of our patients. The nature of radiation-induced changes remains to be elucidated. These changes presumably represent a whole gamut of pathological processes, ranging from gliosis to true necrosis. It is important to emphasize that the signal changes on MRI associated with clinical deterioration are too frequently interpreted as radionecrosis despite the fact that the changes are usually transitory.

Delayed Cyst Formation A delayed cyst is defined as a collection of fluid at the site of the treated AVMs. A fluid cavity corresponding to previous hematoma or encephalomalacia should not be considered as a complication of radiosurgery. Up to date, small series of delayed cyst formation following GKS were

66

published by several authors [159,160]. Our large series of AVMs patients treated with GKS with long term follow-up provided more pertinent information about incidence and timing of cyst occurrence as well as its clinical symptoms [161]. In a total of 1,203 patients with long term MRI follow-up, 20 cysts were identified; 10 developed between 10 and 23 years, 9 between 5 and 10 years, and 1 in less than 5 years following the treatment (> Figures 66-20 and > 66-21). The incidence of cyst formation in the entire patient population was 1.6%, and 3.6% in those undergoing follow-up examination for more than 5 years. Six patients were symptomatic, including three with seizures and three with new neurological deficits. Two patients underwent craniotomy and drainage of the symptomatic cyst. In another patient, a cystoperitoneal shunt was implanted. Cyst wall specimens were obtained in two cases showing no evidence of neoplasia in either case. There are a number of hypothesis concerning the pathogenesis of the delayed cyst formation [160]. Certainly, hypotheses are important; they are like fishermen’s nets ‘‘only he who casts will catch’’ (Novalis). Nevertheless, without critical scrutiny, they are simply speculations. The validity of the causal agents in delayed cyst formation remains to be proven.

Radiosurgery Induced Neoplasia Kahan et al. [162] defined the criteria for a tumor to be considered as a result of irradiation: (1) the tumor must occur in the irradiation field; (2) it cannot be present prior to irradiation; (3) any primary tumor must differ histologically from the induced tumor; and (4) there must be no genetic predisposition for occurrence of a secondary malignancy or tumor progression. Based on the literature, the incidence of radiosurgery-induced neoplasia ranges between zero and three per 200,000 patients [49]. However, the

1075

1076

66


. Figure 66-19 Onset and resolution of radiation-induced changes of brain tissue surrounding an AVM nidus. Radiation induced changes appeared 6 months following GKS of a left basal ganglion AVM with a prescription dose of 20 Gy (a and b). These changes showed progressive regression (c) and a complete disappearance at 2 years following the onset (d). Angiography documented total obliteration of the AVM (not shown)

true incidence is likely to be higher because few of these 200,000 patients treated with radiosurgery were followed over a long period. Rowe presented the results of a study in which he cross-referenced patients treated with radiosurgery in Sheffield against England’s national mortality and cancer

databases [163]. With a group of 4,896 patients and more than 30,000 patient-years of data, 2 patients were found to have new malignant brain tumors. However, data covering at least 10 years was only available in 1,048 of these patients.


66

. Figure 66-20 Cyst formation following GKS for an AVM. The right temporal AVM visualized on the lateral carotid arteriogram (a) was cured as shown on a control angiogram obtained 2 years following GKS (b). The development of headache and personality changes prompted a MRI examination 7 years after GKS (c). The cyst was surgically decompressed (d)

We reviewed 1,333 patients with AVMs treated with Gamma Knife and followed with sequential MRI. A subset of 288 patients in this group underwent neuroimaging and participated in clinical follow up for at least 10 years [164]. In two cases, radiosurgically induced neoplasia were identified (> Figure 66-22). Each of the patients was found to have an incidental, uniformly enhancing, dura-based mass lesion; one precisely at the site of previous nidus, another one 5 cm from previous AVM. These lesions displayed

the imaging characteristics of a meningioma. Patients have been asymptomatic and refused any treatment. From our series, if we conservatively estimate that radiosurgery-induced lesions would be evident within a 10-year time interval, our incidence of radiosurgery-induced neoplasia is 2 in 2,880 person-years or 69 in 100,000 person-years. Thus, there is a 0.7% chance that a radiation-induced tumor may develop within 10 years following GKS. This is less than the 1.3%

1077

1078

66


. Figure 66-21 In a total of 1,203 AVM patients with long term MRI follow-up, 20 cysts were identified; one between 15 and 23 years, nine between 10 and 23 years, nine between 5 and 10 years, and one in less than 5 years following the GKS

risk over the first 10 years or 1.9% over 20 years detailed by Brada et al. in radiotherapy series [165]. However, our results encompass a follow-up period of only 10 years. It is our contention that radiation-induced neoplasia must be considered in broad terms when evaluating patients who have undergone GKS. The radiation passes through the head along as many as 201 (or 192 in Perfexion) different trajectories and distant areas of the brain are exposed to low doses of radiation. Therefore, even though the risk of radiosurgery-induced secondary tumor is low, it must be weighed in the treatment of pediatric patients and in patients with benign tumors and a long life expectancy.

Conclusion The legitimacy of radiosurgery is no longer in discussion. When Madjid Samii commenting a report on radiosurgery for vestibular schwannoma states ‘‘the study is a milestone in the treatment of vestibular schwannoma and in contemporary neurosurgery’’ [87] and when Leonard Malis writes that ‘‘the procedure has forced reluctant neurosurgeons to consider major changes in classic thinking about the proper care of many illnesses, including vascular malformations,

cavernous sinus meningiomas and acoustic neuromas’’ [88], it is time for neurosurgeons using radiosurgical tools to correct exaggerations as well as overstatements and to temper undue euphoria and unrealistic expectations. The Gamma Knife is a radiosurgical tool. Therefore, the ideal would be that only neurosurgeons who are able to remove a vestibular schwannoma or a meningioma of difficult approach, or who can extirpate a difficult AVM would have a Gamma Knife. They are the ones who would be able to make an unbiased decision on when to use the Gamma Knife instead of other neurosurgical instruments for best treatment of the given patient. Gamma Knife surgery should be used for AVMs located in the basal ganglia, brainstem and other locations of difficult approach. Microsurgery should not be easily discarded for vestibular schwannoma or skull base meningioma if the neurosurgeon is confident that he or she can perform it without damaging the patient’s quality of life. The radiosurgical and microsurgical tools are complementary and should be combined if required. When the ionizing focused beams are applied to treat lesions in children, the risk of secondary tumors should be kept in mind. To be sure, the rare risk of secondary tumor is of minor concern in adults aged 50 years or more. However a rate of 1.2% within 10 years and 1.9% within 20 years in children is not negligible. Therefore, it has to be taken into consideration at the decision making and radiosurgery should be used only in lack of other alternatives. In the past four decades, improved results with radiosurgery parallel a development in neuroimaging. This may result in more accurate definition of thalamic nuclei, in case research would win new insight in the function of the nuclei. The improved anatomic and physiological information could open new windows of opportunity for functional diseases. In the heated atmosphere of today’s ‘‘turf wars,’’ it should be remembered that the


66

. Figure 66-22 Antero-posterior and lateral views of vertebral angiograms demonstrate a right temporal AVM before GKS (a and b). Two years after GKS the nidus obliterated completely (c and d). Axial and coronal contrast-enhanced T1-weighted MR images (e and f) obtained 10 years postradiosurgery show a meningioma adjacent to the superior surface of tentorium. It is located in the area where the previous AVM was situated

1079

1080

66


radiosurgical tool can be used not only for brain lesions - currently indications are evolving for tumors in different organs. When it is used for a brain lesion, it is a neurosurgical tool and should be used by a neurosurgeon. When Walter Dandy used a pediatric cytoscope to penetrate the ventricles, he was performing a neurosurgical and not a urological intervention. If the Gamma Knife is used for a uveal melanoma, the ophthalmologist should be responsible for the procedure, and when the narrow beams are used for targeting the liver or pancreas it is not logical for the neurosurgeon to be responsible for the procedure. Leksell’s idea of radiosurgery additionally to novelty featured other ingredients characteristic of creativity, among them the ability to stimulate the minds of others for work of excellence. Over the past four decades, neurosurgeons, medical physicists, and radiation oncologists improved the Gamma Knife and adjusted the linear accelerators for radiosurgery. Currently, there are a number of sophisticated tools using focused narrow ionizing beams to treat lesions not only in the brain but also in the whole body. The litmus test of these medical tools is whether they will significantly improve the outcomes of the techniques. An in-depth investigation of this problem is overdue. The findings will suggest the line of research where spending brain and money would be most cost-effective. Imaging, software planning, pharmacology of radiation sensitizers and protective drugs for the normal tissue, ‘‘second factors’’ to enhance the effect of radiation on AVMs, and last but not least, to find a better physical agent as a source to supplant the ionizing beams of current radiosurgery are a few lines of investigation that may be considered. Following the observation of Pierre and Jacque Curie in 1880 that sound energy resulted when electric energy was applied to the surface of a quartz crystal, experiments with ultrasound energy culminated in clinical applications. Leksell, considering heat as the best physical

agent to destroy intracranial neuronal or pathological structures, by the 1950s had already investigated the possibility of using ultrasound in neurosurgery. He performed capsulotomy for obsessive-compulsive disorder using ultrasound, but erratic results and the need for craniotomy made him opt for ionizing beams as the physical agent for intracranial lesions. An integrated focused ultrasound-magnetic resonance imaging system with closed loop control of energy delivery and online tumor control has been recently developed and tested. The heat energy generated at the focus of the ultrasound beams successfully destroyed target tissue. The problem of craniotomy was addressed by Hynynen et al. [166], who were able to use ultrasound phased arrays to achieve delivery of focused ultrasound through the intact skull. Nevertheless, there remain many problems that require solutions. A detailed discussion of these challenges is beyond the scope of this conclusion, however they include heat buildup in bony structures and the protection of cranial nerves in close relation to tumors of the skull base, clinoid, parasellar space, auditory canal, foramen magnum, and foramen jugularis, which cannot tolerate the significant heat induced in the bone by the ultrasound. As long as these problems exist, ultrasound will not supplant the ionizing beams and the two physical agents will coexist and complement each other in the management of brain lesion with focused beams. One disgraceful aspect in radiosurgery and medicine in general is advertising. The widely touted statements of ‘‘noninvasive radiosurgery’’ and radiosurgery being ‘‘first choice of management for all neurosurgical indications’’ do not bear scrutiny. Ambition – while a most efficient and necessary quality for achieving work of excellence occasionally when temptation exists may result in acts of dishonesty. Like publication of inflated results, incorrect information of the patient or instead of considering the best option for a


given patient – performing radiosurgery always when the patient is referred specifically for radiosurgery. Until one is able to harness ambition and to quote Charles Drake, ‘‘be honest so that it hurts’’; until one is able to critically review possibilities and limitations of one’s activity; and until one avoids ‘‘business thinking’’; radiosurgery risks under-achievement and patients will be at risks.

References 1. Clarke RH, Horsley V. One method of investigating the deep ganglia and tracts of the central nervous system (cerebellum). Br Med J 1906;2:1799-800. 2. Spiegel E, Wycis H, Marks M, Lee A. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 3. Leksell L. Brain fragments in radiosurgery. In: Steiner L, editor. Radiosurgery: baseline and trends. New York: Raven Press; 1992. p. 623-92. 4. Hopewell JW, Wright EA. The nature of latent cerebral irradiation damage and its modification by hypertension. Br J Radiol 1970;43:161-7. 5. Jagannathan J, Sheehan JP, Pouratian N, Laws ER, Steiner L, Vance ML. Gamma Knife surgery for Cushing’s disease. J Neurosurg 2007;106:980-7. 6. Leber KA, Bergloff J, Langmann G, Mokry M, Schrottner O, Pendl G. Radiation sensitivity of visual and oculomotor pathways. Stereotact Funct Neurosurg 1995;64 Suppl 1:233-8. 7. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathways and cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neurosurg 1998;88:43-50. 8. Chen JC, Giannotta SL, Yu C, Petrovich Z, Levy ML, Apuzzo ML. Radiosurgical management of benign cavernous sinus tumors: dose profiles and acute complications. Neurosurgery 2001;48:1022-30. 9. Lim YL, Leem W, Kim TS, Rhee BA, Kim GK. Four years’ experiences in the treatment of pituitary adenomas with gamma knife radiosurgery. Stereotact Funct Neurosurg 1998;70 Suppl 1:95-109. 10. Sheehan JM, Vance ML, Sheehan JP, Ellegala DB, Laws ER Jr. Radiosurgery for Cushing’s disease after failed transsphenoidal surgery. J Neurosurg 2000;93:738-42. 11. Witt TC, Kondziolka D, Flickinger JC, Lunsford LD. Gamma knife radiosurgery for pituitary tumors. In: Lunsford LD, Kondziolka D, Flickinger J, editors. Gamma knife brain surgery. Progress in neurological surgery, vol 14. Basel: Karger; 1998. p. 114-27.

66

12. Noren G, Greitz D, Hirsch A, Lax I. Gamma knife surgery in acoustic tumours. Acta Neurochir Suppl (Wien) 1993;58:104-107. 13. Sheehan J, Pan HC, Stroila M, Steiner L. Gamma knife surgery for trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg 2005;102:434-441. 14. Tishler RB, Loeffler JS, Lunsford LD, Duma C, Alexander E III, Kooy HM, et al. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993;27:215-21. 15. Kihlstrom L, Lindquist C, Adler J. Histological studies of gamma knife lesions in normal and hypercholesterolemic rabbits. New York: Raven Press; 1992. 16. Nilsson A, Wennerstrand J, Leksell D, Backlund EO. Stereotactic gamma irradiation of basilar artery in cat. Preliminary experiences. Acta Radiol Oncol Radiat Phys Biol 1978;17:150-60. 17. Kamiryo T, Lopes MB, Berr SS, Lee KS, Kassell NF, Steiner L. Occlusion of the anterior cerebral artery after Gamma Knife irradiation in a rat. Acta Neurochir (Wien) 1996;138:983-90. 18. Cushing H, Bailey P. Tumors arising from the blood vessels of the brain. Springfield, IL: Charles C Thomas; 1928. 19. Andersson B, Larsson B, Leksell L, Mair W, Rexed B, Sourander P, et al. Histopathology of late local radiolesions in the goat brain. Acta Radiol Ther Phys Biol 1970;9:385-94. 20. Schneider BF, Eberhard DA, Steiner LE. Histopathology of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 1997;87:352-7. 21. Snell JW, Sheehan J, Stroila M, Steiner L. Assessment of imaging studies used with radiosurgery: a volumetric algorithm and an estimation of its error. Technical note. J Neurosurg 2006;104:157-62. 22. Laws ER Jr, Ebersold M, Piepgras DG. The results of transsphenoidal surgery in specific clinical entities. In: Laws ER Jr, Randall R, Kern EB, editors. Management of pituitary adenomas and related lesions with emphasis on transsphenoidal microsurgery. New York: AppletonCentury-Crofts; 1982. p. 277-305. 23. Laws ER Jr, Fode NC, Redmond MJ. Transsphenoidal surgery following unsuccessful prior therapy. An assessment of benefits and risks in 158 patients. J Neurosurg 1985;63:823-9. 24. Laws ER Jr, Thapar K. Recurrent pituitary adenomas. In: Landolt AM, Vance ML, Reilly P, editors. Pituitary adenomas. Edinburgh: Churchill-Livingtone; 1996. p. 385-94. 25. Laws ER Jr, Vance ML. Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 1999;10:327-36. 26. Laws ER, Sheehan JP. Pituitary surgery: a modern approach. Basel; New York: Karger; 2006. 27. Laws ER, Vance ML, Thapar K. Pituitary surgery for the management of acromegaly. Horm Res 2000;53 3:71-5.

1081

1082

66


28. Ciric I, Ragin A, Baumgartner C, Pierce D. Complications of transsphenoidal surgery: results of a national survey, review of the literature, and personal experience. Neurosurgery 1997;40:225-36. 29. Landolt AM, Haller D, Lomax N, Scheib S, Schubiger O, Siegfried J, et al. Octreotide may act as a radioprotective agent in acromegaly. J Clin Endocrinol Metab 2000;85:1287-9. 30. Landolt AM, Lomax N. Gamma knife radiosurgery for prolactinomas. J Neurosurg 2000;93 Suppl 3:14-18. 31. Mingione V, Yen CP, Vance ML, Steiner M, Sheehan J, Laws ER, et al. Gamma surgery in the treatment of nonsecretory pituitary macroadenoma. J Neurosurg 2006;104:876-83. 32. Kim MS, Lee SI, Sim JH. Gamma Knife radiosurgery for functioning pituitary microadenoma. Stereotact Funct Neurosurg 1999;72 Suppl 1:119-24. 33. Kim SH, Huh R, Chang JW, Park YG, Chung SS. Gamma Knife radiosurgery for functioning pituitary adenomas. Stereotact Funct Neurosurg 1999;72 Suppl 1:101-10. 34. Pouratian N, Sheehan J, Jagannathan J, Laws ER Jr, Steiner L, Vance ML. Gamma knife radiosurgery for medically and surgically refractory prolactinomas. Neurosurgery 2006;59:255-66. 35. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing disease. A report of 216 cases. Ann Intern Med 1988;109:487-93. 36. Nieman LK. Medical therapy of Cushing’s disease. Pituitary 2002;5:77-82. 37. Arnaldi G, Angeli A, Atkinson AB, Bertagna X, Cavagnini F, Chrousos GP, et al. Diagnosis and complications of Cushing’s syndrome: a consensus statement. J Clin Endocrinol Metab 2003;88:5593-602. 38. Chu JW, Matthias DF, Belanoff J, Schatzberg A, Hoffman AR, Feldman D. Successful long-term treatment of refractory Cushing’s disease with high-dose mifepristone (RU 486). J Clin Endocrinol Metab 2001;86:3568-73. 39. Izawa M, Hayashi M, Nakaya K, Satoh H, Ochiai T, Hori T, et al. Gamma knife radiosurgery for pituitary adenomas. J Neurosurg 2000;93 Suppl 3:19-22. 40. Jackson IM, Noren G. Gamma knife radiosurgery for pituitary tumours. Baillieres Best Pract Res Clin Endocrinol Metab 1999;13:461-9. 41. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the treatment of Cushing disease: long-term results. J Neurosurg 2002;97:422-8. 42. Morange-Ramos I, Regis J, Dufour H, Andrieu JM, Grisoli F, Jaquet P, et al. Gamma-knife surgery for secreting pituitary adenomas. Acta Neurochir (Wien) 1998;140:437-43. 43. Petrovich Z, Yu C, Giannotta SL, Zee CS, Apuzzo ML. Gamma knife radiosurgery for pituitary adenoma: early results. Neurosurgery 2003;53:51-9.

44. Mahmoud-Ahmed AS, Suh JH. Radiation therapy for Cushing’s disease: a review. Pituitary 2002;5:175-80. 45. Rahn T, Thoren M, Hall K, Backlund EO. Stereotactic radiosurgery in Cushing’s syndrome: acute radiation effects. Surg Neurol 1980;14:85-92. 46. Hayashi M, Taira T, Ochiai T, Chernov M, Takasu Y, Izawa M, et al. Gamma knife surgery of the pituitary: new treatment for thalamic pain syndrome. J Neurosurg 2005;102 38-41. 47. Hoybye C, Grenback E, Rahn T, Degerblad M, Thoren M, Hulting AL. Adrenocorticotropic hormoneproducing pituitary tumors: 12- to 22-year follow-up after treatment with stereotactic radiosurgery. Neurosurgery 2001;49:284-91; discussion 282–291. 48. Nagesser SK, van Seters AP, Kievit J, Hermans J, Krans HM, van de Velde CJ. Long-term results of total adrenalectomy for Cushing’s disease. World J Surg 2000;24:108-13. 49. Ganz JC. Gamma knife radiosurgery and its possible relationship to malignancy: a review. J Neurosurg 2002;97:644-52. 50. Ganz JC, Backlund EO, Thorsen FA. The effects of Gamma Knife surgery of pituitary adenomas on tumor growth and endocrinopathies. Stereotact Funct Neurosurg 1993;61 Suppl 1:30-7. 51. Levy RP, Fabrikant JI, Frankel KA, Phillips MH, Lyman JT, Lawrence JH, et al. Heavy-chargedparticle radiosurgery of the pituitary gland: clinical results of 840 patients. Stereotact Funct Neurosurg 1991;57:22-35. 52. Mauermann WJ, Sheehan JP, Chernavvsky DR, Laws ER, Steiner L, Vance ML. Gamma Knife surgery for adrenocorticotropic hormone-producing pituitary adenomas after bilateral adrenalectomy. J Neurosurg 2007;106:988-93. 53. Pollock BE, Young WF Jr. Stereotactic radiosurgery for patients with ACTH-producing pituitary adenomas after prior adrenalectomy. Int J Radiat Oncol Biol Phys 2002;54:839-41. 54. Wolffenbuttel BH, Kitz K, Beuls EM. Beneficial gammaknife radiosurgery in a patient with Nelson’s syndrome. Clin Neurol Neurosurg 1998;100:60-3. 55. Giustina A, Barkan A, Casanueva FF, Cavagnini F, Frohman L, Ho K, et al. Criteria for cure of acromegaly: a consensus statement. J Clin Endocrinol Metab 2000;85:526-9. 56. Vance ML. Endocrinological evaluation of acromegaly. J Neurosurg 1998;89:499-500. 57. Jezkova J, Marek J, Hana V, Krsek M, Weiss V, Vladyka V, et al. Gamma knife radiosurgery for acromegaly – longterm experience. Clin Endocrinol (Oxf) 2006;64:588-95. 58. Pollock BE, Jacob JT, Brown PD, Nippoldt TB. Radiosurgery of growth hormone-producing pituitary adenomas: factors associated with biochemical remission. J Neurosurg 2007;106:833-8.


59. Post KD, Habas JE. Comparison of long term results between prolactin secreting adenomas and ACTH secreting adenomas. Can J Neurol Sci 1990;17:74-7. 60. Yildiz F, Zorlu F, Erbas T, Atahan L. Radiotherapy in the management of giant pituitary adenomas. Radiother Oncol 1999;52:233-7. 61. Witt TC. Stereotactic radiosurgery for pituitary tumors. Neurosurg Focus 2003;14:e10. 62. Pan L, Zhang N, Wang EM, Wang BJ, Dai JZ, Cai PW. Gamma knife radiosurgery as a primary treatment for prolactinomas. J Neurosurg 2000;93 Suppl 3:10-13. 63. Molitch ME. Pathologic hyperprolactinemia. Endocrinol Metab Clin North Am 1992;21:877-901. 64. Ohmori K, Collins J, Fukushima T. Craniopharyngiomas in children. Pediatr Neurosurg 2007;43:265-78. 65. Hetelekidis S, Barnes PD, Tao ML, Fischer EG, Schneider L, Scott RM, et al. 20-year experience in childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1993;27:189-95. 66. Rajan B, Ashley S, Gorman C, Jose CC, Horwich A, Bloom HJ, et al. Craniopharyngioma – a long-term results following limited surgery and radiotherapy. Radiother Oncol 1993;26:1-10. 67. Leksell L, Liden K. A therapeutic trial with radioactive isotopes in cystic brain tumors. London: H. M. Stationary Office; 1951. 68. Backlund EO, Axelsson B, Bergstrand CG, Eriksson AL, Noren G, Ribbesjo E, et al. Treatment of craniopharyngiomas – the stereotactic approach in a ten to twentythree years’ perspective. I. Surgical, radiological and ophthalmological aspects. Acta Neurochir (Wien) 1989;99:11-19. 69. Bond WH, Richards D, Turner E. Experiences with radioactive gold in the treatment of craniopharyngioma. J Neurol Neurosurg Psychiatry 1965;28:30-8. 70. Guevara JA, Bunge HJ, Heinrich JJ, Weller G, Villasante A, Chinela AB. Cystic craniopharyngioma treated by 90yttrium silicate colloid. Acta Neurochir Suppl (Wien) 1988;42:109-12. 71. Julow J, Lanyi F, Hajda M, Szeifert G, Simkovics M, Toth S, et al. Further experiences in the treatment of cystic craniopharyngeomas with yttrium 90 silicate colloid. Acta Neurochir Suppl (Wien) 1988;42:113-19. 72. Kobayashi T, Kageyama N, Ohara K. Internal irradiation for cystic craniopharyngioma. J Neurosurg 1981; 55:896-903. 73. Leksell L, Backlund EO, Johansson L. Treatment of craniopharyngiomas. Acta Chir Scand 1967; 133:345-50. 74. Lindgren E, Westberg G. Radioactive bismuth phosphate for the treatment of craniopharyngioma. Acta Radiol Ther Phys Biol 1964;2:113-20. 75. Munari C, Landre E, Musolino A, Turak B, Habert MO, Chodkiewicz JP. Long term results of stereotactic

76.

77.

78.

79.

80.

81.

82.

83.

84. 85. 86.

87. 88. 89.

90. 91.

92.

66

endocavitary beta irradiation of craniopharyngioma cysts. J Neurosurg Sci 1989;33:99-105. Strauss L, Sturm V, Georgi P, Schlegel W, Ostertag H, Clorius JH, et al. Radioisotope therapy of cystic craniopharyngomas. Int J Radiat Oncol Biol Phys 1982;8:1581-5. Sturm V, Scheer KE, Schlegel W, Strauss L, Penzholz H, Georgi P. Intracavitary contact radiation of cystic craniopharyngiomas by stereotaxic Y-90 application. Strahlentherapie (Sonderb) 1981;76:113-18. Backlund EO. Colloidal radioisotopes as part of a multimodality treatment of craniopharyngiomas. J Neurosurg Sci 1989;33:95-7. Coffey RJ, Lunsford LD. The role of stereotactic techniques in the management of craniopharyngiomas. Neurosurg Clin N Am 1990;1:161-72. Duma CM, Lunsford LD, Kondziolka D, Harsh GRt, Flickinger JC. Stereotactic radiosurgery of cavernous sinus meningiomas as an addition or alternative to microsurgery. Neurosurgery 1993;32:699-704. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for tentorial meningiomas. Acta Neurochir (Wien) 1998;140:315-20. Steiner L, Lindquist C, Steiner M. Meningiomas and Gamma Knife radiosurgery. In: Al-Mefty, editor. Meningiomas. New York: Raven Press; 1991. p. 263-72. Henschen F. Zur Histologie und Pathogenese der Kleinhirnbruchenwinkeltumoren. Arch Psychiatr 1916; 56:21-122. Leksell L. A note on the treatment of acoustic tumours. Acta Chir Scand 1971;137:763-5. Prasad D, Steiner M, Steiner L. Gamma surgery for vestibular schwannoma. J Neurosurg 2000;92:745-59. Slattery WH III, Brackmann DE. Results of surgery following stereotactic irradiation for acoustic neuromas. Am J Otol 1995;16:315-19. Samii M, Matthies C. Gamma surgery for vestibular schwannoma. J Neurosurg 2000;92:892-6. Malis L. Gamma surgery for vestibular schwannoma. J Neurosurg 2000;92:894-6. Sheehan J, Yen CP, Arkha Y, Schlesinger D, Steiner L. Gamma knife surgery for trigeminal schwannoma. J Neurosurg 2007;106:839-45. Rades D, Fehlauer F. Treatment options for central neurocytoma. Neurology 2002;59:1268-70. Kim DG, Chi JG, Park SH, Chang KH, Lee SH, Jung HW, et al. Intraventricular neurocytoma: clinicopathological analysis of seven cases. J Neurosurg 1992; 76:759-65. Yasargil MG, von Ammon K, von Deimling A, Valavanis A, Wichmann W, Wiestler OD. Central neurocytoma: histopathological variants and therapeutic approaches. J Neurosurg 1992;76:32-7.

1083

1084

66


93. Yen CP, Sheehan J, Patterson G, Steiner L. Gamma knife surgery for neurocytoma. J Neurosurg 2007;107:7-12. 94. Szeifert GT PD, Kamiryo T. The role of gamma knife in the therapy of intracranial astrocytomas. In Second Annual Meeting of the Japanese society for Brain Tumor Surgery. Nagasaki, Japan, 1997. 95. Szeifert GT, Prasad D, Kamyrio T, Steiner M, Steiner LE. The role of the Gamma Knife in the management of cerebral astrocytomas. Prog Neurol Surg 2007;20:150-63. 96. Yen CP, Sheehan J, Steiner M, Patterson G, Steiner L. Gamma knife surgery for focal brainstem gliomas. J Neurosurg 2007;106:8-17. 97. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin LS. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002;50:41-6. 98. Payne BR, Prasad D, Szeifert G, Steiner M, Steiner L. Gamma surgery for intracranial metastases from renal cell carcinoma. J Neurosurg 2000;92:760-5. 99. Wronski M, Maor MH, Davis BJ, Sawaya R, Levin VA. External radiation of brain metastases from renal carcinoma: a retrospective study of 119 patients from the M. D. Anderson Cancer Center. Int J Radiat Oncol Biol Phys 1997;37:753-9. 100. Mingione V, Oliveira M, Prasad D, Steiner M, Steiner L. Gamma surgery for melanoma metastases in the brain. J Neurosurg 2002;96:544-51. 101. Goyal S, Prasad D, Harrell F Jr, Matsumoto J, Rich T, Steiner L. Gamma knife surgery for the treatment of intracranial metastases from breast cancer. J Neurosurg 2005;103:218-23. 102. Yen CP, Sheehan J, Patterson G, Steiner L. Gamma knife surgery for metastatic brainstem tumors. J Neurosurg 2006;105:213-19. 103. Backlund EO, Rahn T, Sarby B. Treatment of pinealomas by stereotaxic radiation surgery. Acta Radiol Ther Phys Biol 1974;13:368-76. 104. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for chordoma and chondrosarcoma: further experiences. Int J Radiat Oncol Biol Phys 1998;41:387-92. 105. Marchini G, Gerosa M, Piovan E, Pasoli A, Babighian S, Rigotti M, et al. Gamma Knife stereotactic radiosurgery for uveal melanoma: clinical results after 2 years. Stereotact Funct Neurosurg 1996;66 Suppl 1:208-13. 106. Zehetmayer M, Menapace R, Kitz K, Ertl A, Strenn K, Ruhswurm I. Stereotactic irradiation of uveal melanoma with the Leksell gamma unit. Front Radiat Ther Oncol 1997;30:47-55. 107. Krayenbiihl H. Discussion eds rapports sur ules angiomes supratentoriels. Bruxelles: Congres International de Neurochirurgie; 1957.

108. McKissock W, Hankinson J. The surgical treatment of supratentorial angiomas. Premier Congress International de Neurochirurgie, Rapports et Discussion Bruxelles; 1957. 109. Olivecrona H, Ladenheim J. Congenital arteriovenous aneurysms of the carotid and vertebral systems. Berin: Springer; 1957. 110. Wegemann HJ. Das Schicksal operierter Patienten mit arteriovenosem Aneurysma des Gehrins. Zurich: Juris Druck Verlag; 1969. 111. Johnson R. Radiotherapy of cerebral angiomas with a note on some problems in diagnosis. Berlin: Springer-Verlag; 1965. 112. Lindqvist M, Steiner L, Blomgren H, Arndt J, Berggren BM. Stereotactic radiation therapy of intracranial arteriovenous malformations. Acta Radiol Suppl 1986;369:610-13. 113. Steiner L, Leksell L, Greitz T, Forster DM, Backlund EO. Stereotaxic radiosurgery for cerebral arteriovenous malformations. Report of a case. Acta Chir Scand 1972;138:459-64. 114. Marks MP, Lane B, Steinberg GK, Snipes GJ. Intranidal aneurysms in cerebral arteriovenous malformations: evaluation and endovascular treatment. Radiology 1992;183:355-60. 115. Steiner L, Lindquist C, Steiner M. Radiosurgery. Adv Tech Stand Neurosurg 1992;19:19-102. 116. Steiner L, Greitz T, Leksell L. Radiosurgery in intracranial arteriovenous malformation. In sixth International Congress of Neurological Surgeons. Amsterdam: Excerpta Medica; 1977. 117. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997;40: 425-30. 118. Steiner L, Lindquist C, Adler JR, Torner JC, Alves W, Steiner M. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992;77:1-8. 119. Steiner L, Lindquist C, Adler JR, Torner JC, Alves W, Steiner M. Outcome of radiosurgery for cerebral AVM. J Neurosurg 1992;77:823. 120. Karlsson B, Lindquist C, Steiner L. Effect of Gamma Knife surgery on the risk of rupture prior to AVM obliteration. Minim Invasive Neurosurg 1996;39:21-7. 121. Guo WY, Karlsson B, Ericson K, Lindqvist M. Even the smallest remnant of an AVM constitutes a risk of further bleeding. Case report. Acta Neurochir (Wien) 1993;121:212-15. 122. Shin M, Kawahara N, Maruyama K, Tago M, Ueki K, Kirino T. Risk of hemorrhage from an arteriovenous malformation confirmed to have been obliterated on angiography after stereotactic radiosurgery. J Neurosurg 2005;102:842-6.


123. Yamamoto M, Jimbo M, Kobayashi M, Toyoda C, Ide M, Tanaka N, et al. Long-term results of radiosurgery for arteriovenous malformation: neurodiagnostic imaging and histological studies of angiographically confirmed nidus obliteration. Surg Neurol 1992;37:219-30. 124. Yen CP, Varady P, Sheehan J, Steiner M, Steiner L. Subtotal obliteration of cerebral arteriovenous malformations after gamma knife surgery. J Neurosurg 2007;106:361-9. 125. Sirin S, Kondziolka D, Niranjan A, Flickinger JC, Maitz AH, Lunsford LD. Prospective staged volume radiosurgery for large arteriovenous malformations: indications and outcomes in otherwise untreatable patients. Neurosurgery 2006;58:17-27. 126. Pan DH, Guo WY, Chung WY, Shiau CY, Chang YC, Wang LW. Gamma knife radiosurgery as a single treatment modality for large cerebral arteriovenous malformations. J Neurosurg 2000;93 Suppl 3:113-19. 127. Miyawaki L, Dowd C, Wara W, Goldsmith B, Albright N, Gutin P, et al. Five year results of LINAC radiosurgery for arteriovenous malformations: outcome for large AVMS. Int J Radiat Oncol Biol Phys 1999;44:1089-106. 128. Del Curling O Jr, Kelly DL Jr, Elster AD, Craven TE. An analysis of the natural history of cavernous angiomas. J Neurosurg 1991;75:702-8. 129. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cerebral cavernous malformations. J Neurosurg 1995;83:820-4. 130. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg 1991;75:709-14. 131. Karlsson B, Kihlstrom L, Lindquist C, Ericson K, Steiner L. Radiosurgery for cavernous malformations. J Neurosurg 1998;88:293-7. 132. Hasegawa T, McInerney J, Kondziolka D, Lee JY, Flickinger JC, Lunsford LD. Long-term results after stereotactic radiosurgery for patients with cavernous malformations. Neurosurgery 2002;50:1190-7. 133. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction of hemorrhage risk after stereotactic radiosurgery for cavernous malformations. J Neurosurg 1995;83:825-31. 134. Moriarity JL, Wetzel M, Clatterbuck RE, Javedan S, Sheppard JM, Hoenig-Rigamonti K, et al. The natural history of cavernous malformations: a prospective study of 68 patients. Neurosurgery 1999;44:1166-71. 135. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link MJ. Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000; 93:987-91. 136. Lindquist C, Guo WY, Karlsson B, Steiner L. Radiosurgery for venous angiomas. J Neurosurg 1993; 78:531-6.

66

137. Stein BM, Sisti MB, Mohr JP, Pile-Spellman J. Radiosurgery and venous malformations. J Neurosurg 1994;80:175-7. 138. Garner TB, Del Curling O Jr, Kelly DL Jr, Laster DW. The natural history of intracranial venous angiomas. J Neurosurg 1991;75:715-22. 139. Payne BR, Prasad D, Steiner M, Bunge H, Steiner L. Gamma surgery for vein of Galen malformations. J Neurosurg 2000;93:229-36. 140. Steiner L. Radiosurgery in cerebral arteriovenous malformations. In: Flamm E, editor. Cerebrovascular surgery, vol 4. New York: Springer-Verlag; 1985. p. 1161-215. 141. Steiner L, Forster D, Leksell L, Meyerson BA, Boethius J. Gammathalamotomy in intractable pain. Acta Neurochir (Wien) 1980;52:173-84. 142. Mindus P. Capsulotomy in anxiety disorders, a multidiscipline study. In: Neurosurgery. Stockholm: Karolinska Institute; 1991. 143. Rylander G. Stereotactic radiosurgery in anxiety and obsessive-compulsive neurosis. Amsterdam: Elsevier/ North-Holland Biomedical Press; 1979. 144. Hayashi M, Taira T, Chernov M, Izawa M, Liscak R, Yu CP, et al. Role of pituitary radiosurgery for the management of intractable pain and potential future applications. Stereotact Funct Neurosurg 2003;81:75-83. 145. Young RF, Jacques DS, Rand RW, Copcutt BC, Vermeulen SS, Posewitz AE. Technique of stereotactic medial thalamotomy with the Leksell Gamma Knife for treatment of chronic pain. Neurol Res 1995;17:59-65. 146. Lindquist C, Steiner L, Hindmarsh T. Gamma Knife thalamotomy for tremor: report of two cases. In: Steiner L, editor. Radiosurgery. Baseline and trends. New York: Raven Press; 1992. 147. Hirai T, Ohye C, Nagaseki Y, Matsumura M. Cytometric analysis of the thalamic ventralis intermedius nucleus in humans. J Neurophysiol 1989;61:478-87. 148. Duma CM, Jacques DB, Kopyov OV, Mark RJ, Copcutt B, Farokhi HK. Gamma knife radiosurgery for thalamotomy in parkinsonian tremor: a five-year experience. J Neurosurg 1998;88:1044-9. 149. Young RF, Vermeulen SS, Grimm P, Posewitz AE, Jacques DB, Rand RW, et al. Gamma Knife thalamotomy for the treatment of persistent pain. Stereotact Funct Neurosurg 1995;64 Suppl 1:172-81. 150. Bingley T, Leksel L, Meyerson BA. Long term results of stereotactic anterior capsulotomy in chronic obsessivecompulsive neurosis. In: Sweet W, editor. Neurosurgical treatment in psychiatry, pain and epilepsy. Baltimore: University Park Press; 1977. p. 287-99. 151. Chen ZF, Kamiryo T, Henson SL, Yamamoto H, Bertram EH, Schottler F, et al. Anticonvulsant effects of gamma surgery in a model of chronic spontaneous limbic epilepsy in rats. J Neurosurg 2001;94:270-80.

1085

1086

66


152. Regis J, Kerkerian-Legoff L, Rey M, Vial M, Porcheron D, Nieoullon A, et al. First biochemical evidence of differential functional effects following Gamma Knife surgery. Stereotact Funct Neurosurg 1996;66 Suppl 1:29-38. 153. Barcia-Salorio JL, Barcia JA, Hernandez G, Lopez-Gomez L. Radiosurgery of epilepsy. Long-term results. Acta Neurochir Suppl 1994;62:111-13. 154. Regis J, Peragui JC, Rey M, Samson Y, Levrier O, Porcheron D, et al. First selective amygdalohippocampal radiosurgery for ‘mesial temporal lobe epilepsy’. Stereotact Funct Neurosurg 1995;64 Suppl 1:193-201. 155. Regis J, Bartolomei F, Rey M, Genton P, Dravet C, Semah F, et al. Gamma knife surgery for mesial temporal lobe epilepsy. Epilepsia 1999;40:1551-6. 156. Regis J, Bartolomei F, Rey M, Hayashi M, Chauvel P, Peragut JC. Gamma knife surgery for mesial temporal lobe epilepsy. J Neurosurg 2000;93 Suppl 3:141-6. 157. Regis J, Bartolomei F, Hayashi M, Chauvel P. Gamma Knife surgery, a neuromodulation therapy in epilepsy surgery. Acta Neurochir Suppl 2002;84:37-47. 158. Guo WY. Radiological aspects of gamma knife radiosurgery for arteriovenous malformations and other nontumoural disorders of the brain. Acta Radiol Suppl 1993;388:1-34. 159. Kihlstrom L, Guo WY, Karlsson B, Lindquist C, Lindqvist M. Magnetic resonance imaging of obliterated

160.

161.

162.

163.

164.

165.

166.

arteriovenous malformations up to 23 years after radiosurgery. J Neurosurg 1997;86:589-93. Yamamoto M, Ide M, Jimbo M, Hamazaki M, Ban S. Late cyst convolution after gamma knife radiosurgery for cerebral arteriovenous malformations. Stereotact Funct Neurosurg 1998;70 Suppl 1:166-78. Pan HC, Sheehan J, Stroila M, Steiner M, Steiner L. Late cyst formation following gamma knife surgery of arteriovenous malformations. J Neurosurg 2005;102 Suppl:124-7. Cahan WG, Woodard HQ, Higinbotham NL, Stewart FW, Coley BL. Sarcoma arising in irradiated bone: report of eleven cases. 1948. Cancer 1998;82:8-34. Rowe J, Grainger A, Walton L, Silcocks P, Radatz M, Kemeny A. Risk of malignancy after gamma knife stereotactic radiosurgery. Neurosurgery 2007;60:60-65. Sheehan J, Yen CP, Steiner L. Gamma knife surgeryinduced meningioma. Report of two cases and review of the literature. J Neurosurg 2006;105:325-9. Brada M, Ford D, Ashley S, Bliss JM, Crowley S, Mason M, et al. Risk of second brain tumour after conservative surgery and radiotherapy for pituitary adenoma. BMJ 1992;304:1343-6. White J, Clement GT, Hynynen K. Transcranial ultrasound focus reconstruction with phase and amplitude correction. IEEE Trans Ultrason Ferroelectr Freq Control 2005;52:1518-22.

65 Gamma Knife: Clinical Experience A. Niranjan . L. D. Lunsford . J. C. Flickinger . J. Novotny . J. Bhatnagar . D. Kondziolka

Historical Review Professor Lars Leksell selected Cobalt-60 as the ideal photon radiation source for radiosurgery after investigating protons and cross fired photons form early generation linear accelerator [1,2]. The first Gamma Knife (179 Co-60 sources) created a discoid-shaped lesion suitable for movement disorder and intractable pain surgery. Clinical experience with the Gamma Knife began in 1967 with the treatment of a patient with craniopharyngioma. Lunsford and colleagues introduced the first clinical 201-source Gamma Knife unit (model U) to North America (the fifth gamma unit worldwide) which was installed in August 1987 at University of Pittsburgh Medical Center. To eliminate challenging reloading issues associated with model U design, the Gamma Knife was redesigned with sources arranged in a circular (O-ring) configuration. The second generation unit (Model B) was installed at the University of Pittsburgh in 1996. Later the robotic automated positioning system (APS) transformed this unit into the third generation technology model C in March 2000. In January 2005 the fourth generation Leksell Gamma Knife model 4-C was installed. The model 4-C was equipped with hardware and software enhancements designed to improve workflow and provide integrated imaging capabilities, especially image fusion. A completely redesigned version of the Gamma knife was introduced in year 2007. This unit is called the ‘‘Perfexion’’ model (> Figure 65-1). It allows irradiation of wide range of anatomic targets, and increases efficiency and patient flow. The dose

#


profile and dose delivery are improved. The beam size (4, 8, and 16 mm) is changed internally by robotic devices by moving radiation sources within a central collimator body. Sectors of different beam sizes can be mixed (> Figure 65-2). Changes in the dose profile are achieved by sector blocking of the Cobalt-60 sources. The treatable volume is greatly expanded to include targets anywhere in the cranial vault and upper neck. It uses robotics extensively, and has made the positioning of the patient even easier. The Perfexion Unit is especially valuable in the treatment of multiple brain metastases, since patients do not need to be re-positioned, and the risk of collision for lateral, inferior or posterior lesions is resolved using the expanded aperture of the gamma knife. Initial clinical trials with LGK Perfexion were completed at Marseille, France where the first clinical Perfexion was placed. Additional units are now operational in the UK, USA, and Canada. LGK Perfexion became operational at our institution on 28 September 2007. Between 1987 and 2007 8826 patients underwent gamma knife radiosurgery at our center (> Table 65-1).

Clinical Experience Vascular Malformation Radiosurgery Arteriovenous Malformations In the first 20 years of experience (1987–2007) in Pittsburgh, 1,164 patients with AVMs underwent single or multiple staged radiosurgery

1008

65

Gamma knife: clinical experience

. Figure 65-1 LGK Perfexion: A completely redesigned version of the Gamma knife which allows irradiation of wide range of anatomic targets, and increases efficiency and patient flow

. Figure 65-2 Collimator body of LGK Perfexion. The beam size (4 mm, 8 mm, and 16 mm) is changed internally by robotic devices by moving radiation sources within a central collimator body. Sectors of different beam sizes can be mixed. Changes in the dose profile are achieved by sector blocking of the Cobalt-60 sources

. Table 65-1 Indications treated with Gamma Knife radiosurgery at University of Pittsburgh (1987–2007) Brain disorder Vascular Disorders Benign Tumors Glial neoplasms Metastatic Tumor Functional Targets Miscellaneous Total

Number of patients treated 1329 2812 682 2587 869 347 8626

procedures (> Table 65-2). The goals of AVM radiosurgery are to achieve complete AVM obliteration, to improve symptoms, and to preserve existing neurological function. The chief benefit of radiosurgery is to eliminate the threat of spontaneous intracranial hemorrhage by gradual obliteration of the AVM nidus over 2–3 years [3,4]. Obliteration is a process resulting from endothelial proliferation within the AVM blood


. Table 65-2 Gamma Knife radiosurgery for Vascular Brain Disorders Brain disorder

Indications

Vascular Disorders

AVM Cavernous Malformation A V Fistula

Total

Number of patients treated 1164 132 34 1329

vessel walls, supplemented by myofibroblast proliferation. This leads to contraction and eventual obliteration of the AVM blood vessel lumens. We evaluated 906 patients who were eligible for 3 year follow-up. The median nidus volume was 3.4 cc (range, 0.065–57.7 cc) and the median margin dose was 20 Gy (range, 13–32). A single procedure was performed in 865 (95.5%) patients. Prospective volume-staged radiosurgery was performed in 41 (4.5%) patients. Repeat radiosurgery for incomplete nidus obliteration after 3 years was needed in 113 (12.5%) patients. At a median follow-up of 38 months (1–204) complete nidus obliteration was achieved in 78% (angiographic confirmation in 67%, and MRI in 33%) (> Figure 65-3). In addition 20.8% of patients had achieved partial nidus obliteration. A total of 38 hemorrhages (4.1%) occurred after radiosurgery. Seizure control improved in 51% of those who presented with seizures. Adverse radiation effects included new neurological deficits in 24 patients (2.6%) and peri-AVM MRI T2 signal increase in 108 patients (12%). Long-term complications included cyst formation or encephalomalacia in 16 patients (1.7%). No radiation induced tumors were detected.

Risk of Hemorrhage after AVM Radiosurgery In our previous reports we analyzed the risk of hemorrhage during the latency interval from radiosurgery until complete AVM obliteration and

65

studied the clinical and angiographic outcomes of 312 patients who had a mean follow-up of 47 months [5]. The actuarial hemorrhage rate from a patent AVM (before complete obliteration) was 4.8% per year during the first 2 years after radiosurgery and totaled 5.0% per year for the third to fifth years after radiosurgery. Multivariate analysis of clinical and angiographic factors correlated the presence of an unsecured proximal aneurysm with an increased risk of postradiosurgical hemorrhage. Aneurysms, immediately proximal (flow related) to the AVM, usually close as the AVM obliterates. No AVM hemorrhages were observed after radiosurgery in seven patients with intranidal aneurysms. We recommend that aneurysms more than one arterial branch division proximal to the AVM be secured by endovascular or microsurgical approaches prior to (if the aneurysm bled) or shortly after radiosurgery. Inoue et al. identified a single draining vein, deep drainage, AVMs with a varix, AVMs with venous obstruction, high-flow (shunt- and mixed-type) AVMs, and large AVMs with a volume of more than 10 cc as risk factors for hemorrhage [6]. No patient in our study suffered a hemorrhage after angiography had confirmed complete obliteration (n = 140) or suffered from an early draining vein without residual nidus (n = 19). In this study no clear hemorrhage reduction benefit was conferred on patients who had incomplete nidus obliteration in early ( Table 65-3). Long-term results have established radiosurgery as an important minimally invasive alternative to microsurgery. Tumor Growth Control

Long-term results of Gamma Knife radiosurgery for vestibular schwannomas have been documented [22,29–32]. Recent reports suggest tumor control rates of 93–100% after radiosurgery [22,29–48]. Kondziolka et al. studied 5- to 10-year outcomes in 162 vestibular schwannoma patients who had radiosurgery at the University of Pittsburgh [42]. In this study a long-term 98% tumor control rate was reported (> Figure 65-5). Sixty-two percent of tumors became smaller, 33% remained unchanged, and 6% became slightly larger. Some tumors initially enlarged 1–2 mm during the first 6–12 months after radiosurgery as they lost their central contrast enhancement. Such tumors generally regressed in volume compared to their pre-radiosurgery size. Only 2% of patients required tumor resection after radiosurgery. Noreń, in his 28-year experience with vestibular schwannoma radiosurgery, reported a 95% long-term tumor control rate. Litvack et al. reported a 98% tumor control rate at a mean follow-up of 31 months after radiosurgery using a 12 Gy margin dose [49]. Niranjan et al. analyzed the outcome of intracanalicular tumor radiosurgery performed at the University of Pittsburgh [50]. All patients (100%) had imaging-documented tumor growth control. Flickinger et al. performed an outcome analysis of acoustic neuroma patients treated between August 1992 and August 1997 at the University of Pittsburgh. The actuarial 5-year clinical tumor control rate (no requirement for surgical intervention) was 99.4 0.6% (> Figure 65-3) [32,34]. The long-term (10–15 year) outcome of benign tumor radiosurgery also has been evaluated. In

65

a study which included 157 patients with vestibular schwannomas, the median follow-up for the patients still living at the time of the study (n = 136) was 10.2 years. Serial imaging studies after radiosurgery (n = 157) showed a decrease in tumor size in 114 patients (73%), (> Figure 65-4 and 65-5) no change in 40 patients (25.5%), and an increase in three patients who later had resection (1.9%) [31]. No patient developed a radiation-associated malignant or benign tumor (defined as a histologically confirmed and distinct neoplasm arising in the initial radiation field after at least two years have passed). Hearing Preservation

Pre-radiosurgery hearing can now be preserved in 60–70% of patients, (> Figure 65-6) with higher preservation rates found for smaller tumors. In a long-term (5–10 year follow-up) study conducted at the University of Pittsburgh, 51% of patients had no change in hearing ability [34,42]. All patients (100%) who were treated with a margin dose of 14 Gy or less maintained a serviceable level of hearing after intracanalicular tumor radiosurgery [50]. Among patients treated after 1992, the 5-year actuarial rates of hearing level preservation and speech preservation were 75.2% and 89.2%, respectively, for patients (n = 89) treated with a 13 Gy tumor margin dose (> Figure 65-7a, b). The 5-year actuarial rates of hearing level preservation and speech preservation were 68.8 and 86.3%, respectively, for patients (n = 103) treated with >14 Gy as the tumor margin dose [32]. Unlike microsurgery, immediate hearing loss is uncommon after radiosurgery. If hearing impairment is noted, it occurs gradually over 6–24 months. Early hearing loss after radiosurgery (within 3 months) is rare and may result from cranial nerve edema or demyelination. Facial Nerve and Trigeminal Nerve Preservation

Facial and trigeminal nerve function can now be preserved in the majority of patients (>95%). In the early experience at University of Pittsburgh normal facial function was preserved in 79% of patients after 5 years and normal trigeminal nerve

1013

. Figure 65-4 Conformal gamma knife radiosurgery dose plan for acoustic neuroma. SPGR contrast enhanced MRI showing conformal dose plan in axial (a), coronal (b) and sagittal (c) plane. A margin dose of 12.5 Gy was prescribed to 50% isodose line (Yellow line marked with white arrow in a). The isodose lines are projected on 3D T2 weighted images (c, d, and e). The cochlea which is seen in 3D T2 weighted (single arrow in d) images receives less than 5 Gy (20%) (Green line marked by double arrows) of central dose

1014

65 Gamma knife: clinical experience


function was preserved in 73%. These facial and trigeminal nerve preservation rates reflected the higher tumor margin dose of 18–20 Gy used during the CT based planning era before 1991. In a later study using MR based dose planning, a 13 Gy tumor margin dose was associated with 0% risk of new facial weakness and 3.1% risk of facial numbness (5-year actuarial rates). A margin dose of >14 Gy was associated with a 2.5% risk of new onset facial weakness and a 3.9% risk of facial

. Table 65-3 Gamma Knife radiosurgery for Benign Brain Neoplasms Brain disorder

Indications

Benign Tumors

Vestibular Schwannoma Meningioma Pituitary Adenoma Non-Vestibular Schwannoma

Total

Number of patients treated 1290 1170 266 86 2812

65

numbness (5-year actuarial rates) [32]. Similar 10-year facial and trigeminal neuropathy rates have been documented [12]. None of the patients who had radiosurgery for intracanalicular tumors developed new facial or trigeminal neuropathies. Neurofibromatosis 2

Patients with vestibular schwannomas associated with neurofibromatosis 2 represent a special challenge because of the risk of complete deafness. Unlike the solitary sporadic tumors that tend to displace the cochlear nerve, tumors associated with NF2 tend to form nodular clusters that engulf or even infiltrate the cochlear nerve. Complete resection may not always be possible. Radiosurgery has been performed for patients with NF2. Subach et al. studied our first 40 patients (with 45 tumors) who were treated with radiosurgery for NF2. Serviceable hearing was preserved in 6 of 14 patients (43%), and this rate improved to 67% after modifications made to the technique in 1992. The tumor control rate was 98% [51] only one patient showed imaging documented growth. Normal facial nerve function and

. Figure 65-5 Axial contrast enhanced MRI showing of a 49 year old man showing left sided acoustic tumor at radiosurgery (a). Long-term (11 years) MRI follow-up shows significant tumor shrinkage (b)

1015

1016

65


. Figure 65-6 Axial contrast enhanced T1 weighted MR image (a and b) showing left frontal and right parietal brain metastases with significant surrounding edema. Three month follow-up MRI (right) showing tumor shrinkage and significant reduction in surrounding edema (c and d)

trigeminal nerve function was preserved in 81 and 94% of patients, respectively. In two recent series, [52,53] serviceable hearing was preserved in only 30% [52] and 40% [53] of cases, respectively. The tumor control rate was respectively 71% [52] and 79%. [53] Mathieu et al. updated outcomes of our NF2 series in 2007 [54]. The tumor control rate was 87.5%. The rate of serviceable hearing preservation using current technique was 52.6%. It now appears that preservation of serviceable hearing in patients with NF2 is an attainable goal using gamma knife radiosurgery. Early radiosurgery when the hearing level is still excellent may become an appropriate strategy in the future. At present we

generally delay radiosurgery in NF2 patients until we see hearing deterioration or tumor growth.

Meningioma The optimal treatment for meningioma when feasible is complete resection of tumor with its dural base. However, when meningiomas are attached to skull base cranial nerves or vascular structures, complete resection may not be possible. Recurrence rates are higher for meningiomas in critical locations where only subtotal resections are possible due to limited access and involve-


65

. Figure 65-7 (a) Axial contrast enhanced T1 weighted MR image shows a recurrent skull base at radiosurgery. (b) Long-term (12 years) follow-up MRI showing tumor regression

ment of critical structures. Radiosurgery offers an attractive option for patients with residual or recurrent meningioma as well as for patients in whom complete resection of tumor is considered attainable but only with unacceptable morbidity. Meningiomas attached to or within major venous sinuses, can be successfully treated by radiosurgery. Tumor regression may occur slowly over several years after radiosurgery. Radiosurgery provides long-term tumor control associated with high rate of neurological preservation and patient satisfaction [55]. Surgical excision is the preferred first line approach for most symptomatic convexity, anterior fossa, or lateral sphenoid ridge meningiomas. Even for these radiosurgery can be offered as a first management approach unless the tumor needs debulking because of mass effect. Malignant meningiomas require multimodality management that includes resection, radiosurgery, and radiation therapy. Gamma-knife radiosurgery (GKRS) has proven an effective strategy for many patients with recurrent meningiomas [56]. In general, the biological nature of the meningioma is the main factor that determines how effectively radiosurgery will control tumor growth. Recently, stereotactic radiosurgery has been performed for an increasing number of patients with smallto moderate-size meningiomas as an alternative

to surgical excision [55,57–67]. Gamma-knife radiosurgery is used as a first-line treatment and/or postoperative adjuvant therapy for suitable patients with meningiomas of the skull base, posterior fossa or cavernous sinus region [68]. Small, sharply demarcated tumors are the best candidates for radiosurgery. Gamma-knife procedures can be performed even after surgery and radiation therapy have failed [69]. Several centers have reported tumor control rates between 84 and 98% after meningioma radiosurgery with a range of follow-up intervals [55,57,59–61,64,65]. Significantly, most prior studies included patients with atypical or malignant tumors who had lower control rates after radiosurgery [63,66]. At our center 1,170 meningioma patients have undergone radiosurgery. Long-term follow-up (up to 20 years) now confirms the high tumor control rate and low morbidity of radiosurgery [70]. At many centers radiosurgery has become the preferred treatment for patients with small- to moderate-size meningiomas without symptomatic mass effect. Results of meningioma radiosurgery supports the concept that radiosurgery should be considered as primary management option for patients with tumors involving the skull base, where a Simpson Grade 1 resection often cannot be accomplished with acceptable risk [71–73].

1017

1018

65


The morbidity of radiosurgery for meningiomas of the cavernous sinus and petroclival regions has been analyzed separately and is considered overall to be 15 Gy). The risk for hypopituitarism after stereotactic radiosurgery thus becomes a primary function of the anatomy of the tumor and the dose prescribed. For recurrent tumors primarily where the pituitary stalk (and even at times the residual pituitary gland) is separate from the tumor, is easily visualized, when the stalk and gland are excluded from higher dose, the risk of hypopituitarism is reduced. For adenomas that cannot be visually separated from the normal gland, particularly if they extend upward to involve or compress the pituitary stalk, the risk is predominantly related to the dose necessary to effectively achieve all outcome goals for the functional status of the tumor (higher for secretory than nonsecretory adenomas). Gamma-Knife radiosurgery is superior to radiation therapy because there is a faster response and fewer adverse radiation effects. Response to radiosurgery is best with ACTH producing tumors, followed by GH producing tumors, prolactinoma having the poorest response usually because they have failed prior medical management due to their invasive nature. Hypopituitarism can be expected to occur in up to 30% within 4–5 years, but can be avoided by minimizing radiation to pituitary stalk and hypothalamus. Somatostatin analogs and dopamine agonists may have a radioprotective effect [82,89]. Although the radioprotective effect of these drugs was not confirmed in subsequent studies [78–80] it is advisable to stop these drugs prior to radiosurgery. Short-acting form of somatostatin analogs can be given until 2 weeks prior to GK. Long-acting somatostatin analogs should be discontinued as much as 4 months prior to GK. Dopamine agonists should be discontinued 2 months prior to radiosurgery.

1019

1020

65


After radiosurgery, once hormone levels are normal on medical therapy, somatostatin analogs should be stopped for 4 months each year to assess for biochemical cure. Similarly, dopamine agonists should be stopped for 2 months. A panel of tests to detect hypopituitarism should be done at 6 month intervals for the first 5 years and then yearly.

Craniopharyngioma Multi-modality therapy is often necessary for such patients because of the development of refractory cystic components of their tumors. Radiosurgery is usually part of a multi-modality management when prior therapies have failed [90,91]. Sixty-eight patients have undergone gamma knife radiosurgery as part of a primary or adjuvant management strategy for craniopharyngioma. Long-term follow-up in our patient series was available in 29 patients. The median tumor volume was 0.4 (range, 0.12–6.36) cm3. One to nine isocenters of different beam diameters were used. The median dose to the tumor margin was 12.5 Gy (range, 9–20), and the maximum dose was 25 Gy (range, 21.8–40). The dose to the optic apparatus was limited to less than 8 Gy. Clinical and imaging follow-up data were obtained at a median of 24 months (range, 13–150) from radiosurgery. Overall, 14 of 29 tumors regressed or vanished, and ten remained stable after radiosurgery. Further tumor growth was noted in five patients, three of whom underwent surgical resection and one had repeat radiosurgery. Two additional patients needed management for cyst enlargement. One patient with prior visual defect had further vision deterioration 9 months after radiosurgery. No patient developed new-onset diabetes insipidus. We found that stereotactic radiosurgery was a reasonable option in selected patients with small recurrent or residual craniopharyngiomas. Adverse radiation risks related to adjacent cranial nerve structures or the development of new

extraocular movement deficits are rare, providing that the optic nerve and tract dose is kept lower than 8 Gy or less in a single procedure. In general, we prefer the use of multi-modality management including microsurgery, radiosurgery, and intracavitary radiation rather than stereotactic or fractionated radiation therapy. The goal has been to maintain endocrinological function whenever possible, reduce the risks of visual dysfunction, and subsequently control tumor growth. There are other reports that have similar results in the management of craniopharyngiomas using gamma knife [86,92,93].

Glial Tumors Malignant Gliomas

Malignant gliomas continue to represent one of the most serious challenges in neurosurgery. Radiation therapy has become the mainstay of the treatment. The observation that local control and median survival can be improved through the dose escalation is the basis for the application of radiosurgery to malignant gliomas. Radiosurgery is used for boost irradiation of patients with malignant glial tumors in addition to conventional wide-margin fractionated radiotherapy. It has been used mainly for patients with tumors Table 65-4). Early radiosurgery reports widely varied in the outcomes for malignant gliomas with a median survival for GBM patients ranging from 9.5 to 26 months. These variations could result from patient selection biases and other prognostic factors [94]. We performed a retrospective study to evaluate the result of radiosurgery on 64 GBM and 43 anaplastic astrocytoma patients. The median survival for the GBM was 16 months after radiosurgery and 26 months after diagnosis. The 2-year survival rate was 51%. For patients with anaplas-


. Table 65-4 Gamma Knife radiosurgery for Glial Neoplasms Number of patients treated

Indications Astrocytoma Pilocytic Fibrillary Anaplactic GBM

81 40 94 308

Mixed

overall survival. No acute Grade 3 or Grade 4 toxicity was encountered. There appears to be a significant survival advantage using radiosurgery boost in patients with malignant glioma, especially if appropriately used with surgery and other adjuvant therapies. However, a carefully designed prospective randomized trial is needed to reliably establish any survival benefit from a radiosurgical boost for patients with malignant gliomas. Lower Grade Gliomas

Astro-Ologo Anaplastic Astro-Ologo GBM-Oligo Undifferentiated cell type-oligo Ependymoma Meduloblastoma Total

65

9 31 7 11 64 23 682

tic astrocytomas median survival after radiosurgery was 21 months and after diagnosis was 32 months. The 2-year survival rate after diagnosis was 67%. Other centers have recently reported survival rates that seem significantly improved as compared to 9-month median survival and 10% 2-year actuarial rate reported for standard therapy. Nwokedi et al. compared survival between 33 patients treated with external beam radiation therapy (EBRT) alone (Group 1), and 31 patients managed with EBRT plus a gamma knife radiosurgery (GK-SRS) boost (Group 2) [95]. GK-SRS was administered to most patients within 6 weeks of the completion of EBRT. The median EBRT dose was 59.7 Gy (range, 28–70.2 Gy), and the median GK-SRS dose to the prescription volume was 17.1 Gy (range, 10–28 Gy). Both groups were comparable in age, Karnofsky performance status, extent of resection, and tumor volume. The median survival was significantly better in patients treated with EBRT plus GK-SRS (13 months in EBRT alone versus 25 months in EBRT plus GK-SRS). Age, Karnofsky performance status, and the addition of GK-SRS were all found to be significant predictors of

Low-grade gliomas have been treated with radiosurgery. Simonova et al. treated 68 low-grade gliomas patients using gamma knife surgery [96]. The median patient age was 17 years and median target volume was 4,200 mm3. The median marginal prescription dose was 25 Gy. Ninety-five percent of patients were treated in five daily stages. These authors reported 83% rate of partial or complete tumor regression with a median time to response of 18 months. In this series the progression-free survival was 92% at 3 years and 88% at 5 years. Moderate acute or late toxicity was noted in 5% of patients. Kida et al. treated 51 patients harboring lowgrade gliomas (12 Grade I astrocytomas, 39 Grade II astrocytomas) using gamma knife [97]. The mean margin dose was 12.5 Gy for Grade I and 15.7 Gy for Grade II tumors. In the mean followup of 27.6 months grade I astrocytomas had a response rate of 50% and a control rate of 91.7%. Grade II astrocytomas had a 46.2% response rate and an 87.2% control rate. Despite the favorable histological characteristics and prognosis of pilocytic astrocytomas, some patients may not be cured after microsurgery because of an adverse location, recurrence, or tumor progression. Radiosurgery is an effective alternative therapeutic approach for these patients. Hadjipanayis et al. reported outcome in 38 patients harboring unresectable pilocytic astrocytomas who were treated with radiosurgery [98] at our center. The median radiosurgical dose to the tumor margin was 15 Gy (range, 9.6–22.5 Gy). After radiosurgery, serial imaging demonstrated complete tumor resolution in

1021

1022

65


ten patients, reduced tumor volume in eight, stable tumor volume in seven, and delayed tumor progression in 12. Three patients died of local tumor progression. Stereotactic radiosurgery is a valuable adjunctive strategy in the management of recurrent or unresectable pilocytic astrocytomas especially small volume, sharp bordered tumors. Radiosurgery can play an important role in the treatment of low-grade astrocytomas, and complete cure of these tumors has been achieved in at least some of the cases [98]. The acute complications following radiosurgery are unusual and limited to exacerbation of existing symptoms. The most frequently seen delayed complication of radiosurgical boost is tumor swelling radiation reaction in the tumor or surrounding brain swelling. Symptoms are usually controllable by steroid therapy. The reported incidence of radiation necrosis ranges from 2 to 22%. Re-operation rates ranging from 21 to 33% have been reported after radiosurgery. Neither radiation necrosis nor reoperation is associated with diminished length of survival.

Brain Metastases Radiosurgery (SRS) as the sole initial management or as a boost before or after whole brain radiation therapy (WBRT) has emerged as a widely practiced treatment modality for brain metastases. The goal of radiosurgery without WBRT is to achieve brain control without the possible long term neurotoxic or cognitive side effects of WBRT [99]. The rationale for radiosurgery, when used as a boost after WBRT, is to achieve improved local brain tumor control. Radiosurgery boost improves survival in selected patients in whom the predominant problem is brain disease rather than extracranial disease. Radiosurgery is also used as salvage treatment for progressive intracranial disease after surgery or WBRT. Traditionally radio-insensitive histologies tend to be more responsive to SRS than to conventional fractionated radiation treat-

ment. In addition, SRS causes indirect vascular injury and subsequent sclerosis of blood vessels, and eventual compromise of the blood supply and circulation within the tumor [100]. At our institution 2,587 patients have undergone radiosurgery for brain metastases (> Table 65-5). Retrospective series have consistently revealed local control of the target lesions in the range of 80–85% or even higher with a very acceptable side effect profile[101–107]. Prospective randomized trials have demonstrated that the one-year local control rate of target lesions with radiosurgery is 73%, which increases to 82–89% with the addition of WBRT [108,109]. Several studies have reported excellent local control (70–80% at 1 year) following radiosurgery for brain metastases [110,111] (> Figure 65-6). Other published series of patients treated with SRS have demonstrated a risk of distant brain failure at 1 year, ranging from 43 to 57% [112–115]. In general, the risk of new metastasis in patients with solitary tumors is approximately 37% (crude), but the actuarial risk is 50% at 1 year [116,117]. The histologic features or tumor type may play a role, with melanoma being more likely to be associated with multiple metastases than some other tumor types [118]. Despite a relatively high risk of new metastases outside the radiosurgery volume in patients who have SRS alone, retrospective studies have not . Table 65-5 Gamma Knife radiosurgery for Metastatic Neoplasms Primary Tumors Breast Sarcoma Gestro-intestinal Kidney Lung Melanoma Nasopharynx Thyroid Others Unknown primary Total indications

Number of patients treated 463 17 132 204 1168 392 29 12 106 64 2587


confirmed a survival benefit to adjuvant WBRT [113,119,120]. Freedom from local progression in the brain at 1 year was significantly superior in patients who received both SRS and WBRT compared with SRS alone (28% vs. 69%), although the overall survival rate was not significantly different [114]. A retrospective, multi-institutional study in which patients were treated with SRS alone (n = 268) or SRS + WBRT (n = 301) also reported no significant difference in the overall survival rate [161]. Despite the higher rate of new lesions developing in patients treated with SRS alone, the overall survival appears to be equivalent to SRS + WBRT since salvage therapies are fairly effective and patients’ extracranial disease is frequently the cause of death [113]. Only 24% of patients managed initially with radiosurgery alone required salvage WBRT. Pirzkall et al. reported that there was no survival benefit for an overall group of 236 patients with adjuvant WBRT but these authors noted a trend toward improved survival in a subset of patients with no extracranial tumor (15.4 vs. 8.3 months, p = 0.08) [119]. Chidel et al. reported 78 patients managed initially with SRS alone and 57 patients treated with SRS and adjuvant WBRT [67]. Whole-brain radiation therapy did not improve the overall survival rate but was useful in preventing both the local progression and the development of new brain metastases (74% vs. 48%, p = 0.06). These retrospective studies suggest that WBRT will improve local and distant control in the brain, but do not clearly demonstrate a survival advantage [113]. A multicenter retrospective analysis was performed with 502 patients treated at ten institutions in which all of the patients were treated with WBRT and SRS. The patients were stratified by the recursive partitioning analysis and compared with similar patients from the RTOG database who had been treated with WBRT alone. [121]. The study revealed that patients with higher KPS, controlled primary tumor, absence of extracranial metastases and lower RPA class

65

had statistically superior survival. The addition of an SRS boost resulted in a median survival of 16.1, 10.3, and 8.7 months, respectively, for RPA classes I, II, and III. This is in comparison to 7.1, 4.2, and 2.3 months for similar RPA class patients from the RTOG database. This improvement in overall survival, stratified by RPA class with an SRS boost, was statistically significant [121]. In a recent study SRS alone was found to be as effective as resection plus WBRT in the treatment of one or two brain metastases for patients in RPA classes I and II [122]. Stereotactic radiosurgery is an effective treatment for patients with multiple brain metastases. A substantial amount of published literature now supports use of radiosurgery in the treatment of multiple brain metastases. Stereotactic radiosurgery offers a very high control rate with a low risk of serious side effects.

Pineal Region Tumors Management of pineal region tumors remains a significant challenge because of the anatomic complexity of the area and the presence of critical brain and vascular structures. Microsurgical techniques are often successful in obtaining a tissue diagnosis; however, the likelihood of curative resection remains low. There are only few published reports on radiosurgery for pineal tumors [123,124]. At the authors’ institution 29 patients with parenchymal pineal tumors were treated between 1989 and 2007. Outcome analysis on 14 patients with longterm follow-up showed that local tumor control was achieved in 13 patients while one died of tumor progressions despite chemotherapy and craniospinal irradiation prior to radiosurgery. Neuroimaging follow-up showed complete disappearance of tumor in three, decrease in tumor size in seven, no change in tumor size in three, and tumor growth in one patient.

1023

1024

65


Skull Base Tumors Radiosurgery is a primary and adjuvant management for tumors of skull base [81,125–130]. From September 1987 through December 2004, 238 patients with a variety of skull base tumors were treated with gamma knife radiosurgery at the University of Pittsburgh Medical Center. These tumors and their subsequent management are described below in more detail. Non-vestibular Schwannomas

Eighty-six patients underwent radiosurgery for non-Vestibular nerve sheath tumors defined by clinical examination, high-resolution intraoperative imaging, and in selected cases prior surgery. Our results of trigeminal schwannoma have been recently published [127]. The records of 23 patients were reviewed with a median follow-up of 40 months. Twenty of twenty-three patients (91%) had tumor growth control, with regression noted in 15 and no further tumor growth in five. Patients who had subsequent tumor enlargement underwent a second radiosurgical procedure. Twelve of twenty-three trigeminal nerve sheath tumor patients (52%) reported symtomatic improvement. Nine (39%) had no change in their symptoms. Only two patients noted new neurological complaints such as facial weakness (one patient) or worsening of the pre-radiosurgical facial numbness (one additional patient). Of interest, trigeminal nerve sheath tumors have a much higher likelihood of developing transient, but occasionally impressive short-term swelling of their tumor in the first year after radiosurgery. This is quite distinct from those patients who have undergone acoustic tumor radiosurgery. In the majority of trigeminal neuroma patients, transient swelling is followed by delayed shrinkage, often of profound degree. Therefore, it is critical that patients and referring doctors do not despair during this transient tumor enlargement phase identified by imaging and sometimes associated with temporary neurological symptoms. Most such symptoms

will resolve as the tumors regress during the next 3–6 months. Radiosurgery using the gamma knife proved to be an effective management strategy for those patients who had undergone both primary as well as adjuvant (post microsurgery) radiosurgery [128]. Three patients underwent gamma knife radiosurgery for facial schwannomas, all identified at the time of prior microsurgery and associated with recurrence or subtotal prior resection. Tumors of the ninth and tenth cranial nerve pose special challenges. Twenty-six patients with jugular bulb schwannomas underwent radiosurgery for between August 1987 and September 2004. Most such patients present with symptoms related to imbalance, incoordination, dysphasia, or hearing loss. A total of 12 patients had previously undergone gross total resection with tumor recurrence, and four had undergone prior partial resection. Results to date show a high likelihood of long-term tumor growth control for such tumors. In an earlier report including 17 patients, we reported a tumor control rate of 94% (eight decreased and eight was stable in size) after Jugular foramen schwannoma radiosurgery [131]. Zhang et al. reported 96% (26/27) tumor growth control with a follow-up period of 38.7 months [130]. In the series of non vestibular schwannomas Pollock et al., reported 96% (22/23) tumor growth control after Gamma Knife radiosurgery [81]. Glomus Tumors

Radiosurgery using the gamma knife has been performed in 17 patients in a 20-year interval. The sparse number of patients (of 8,600 who had gamma knife radiosurgery) is accounted for by the tendency of such tumors to extend well below the skull base. When surgical resection is not feasible, we consider staged radiosurgery technologies such as LINAC based radiosurgery for the extracranial component and the gamma knife for the intracranial portion. Some patients also have undergone elective embolization for shrinkage of their tumor or subtotal microsurgical resection. Only one patient in our series had a glomus tympanicum


tumor. The new Perfexion model will facilitate extracranial radiosurgery to the level of C 4. Gamma knife radiosurgery appears to have a long-term high tumor control rate of glomus tumors, paralleling the benefit provided by fractionated radiation therapy. However, the gamma knife provides a superior biological effective tumor dose, with better dose sparing of the adjacent brainstem and cranial nerve structures. Pollock et al. in a series including 42 patients reported 98% tumor control after glomus jugulare radiosurgery at a mean follow-up of 44-month [132]. Neurological improvement or stability was observed in the majority of patients in published series. Centers using LINAC-based radiosurery continue to support radiosurgery as an effective and safe method of treatment for glomus jugulare tumors that results in low rates of morbidity. Hemangiomas

Radiosurgery for hemangioma was performed in eight patients. Hemangiomas of neurosurgical interest are histologically benign vascular epithelial cell origin tumors that most often occur in the orbit or cavernous sinus or both. These patients tend to present with ocular symptoms or signs such as orbital pain, ophthalmoplegia, proptosis or impaired visual acuity. They can, in fact, be diagnosed by their characteristic imaging appearance by MRI. Since they may hemorrhage dramatically at the time of attempted removal, it is prudent for surgeons considering biopsy or resection of such tumors to get the appropriate imaging in advance. Asymptomatic lesions do not require intervention but are often approached surgically in pursuit of a diagnosis. Symptomatic lesions require treatment. Options include en bloc resection, embolization, or radiation. Radiosurgery is a better option. In our relatively limited experience, some patients have had incomplete resection because of excessive blood loss, and one patient had undergone unsuccessful embolization. We recently reported the outcomes of four patients treated with radiosurgery with tumor doses ranging from 14–19 Gy at the margin

65

[133]. All patients had symptomatic improvement, and all had shown a dramatic reduction in the overall volume of their tumor. One patient had persistent diplopia. In our early experience, stereotactic radiosurgery proved to be a very effective management strategy, which avoided potentially serious complications associated with skull base microsurgery or embolization. The other reports including 3–5 cases in each, also achieved reduction in tumor volume after radiosurgery [123,134,135]. Hemangioblastoma

Over a 20-year interval 44 patients with intracranial hemangioblastomas, usually in conjunction with the syndrome of von Hippel-Lindau disease, have been treated by radiosurgery at our center. We studied outcome in 28 patients with 30 hemangioblastomas who had one year or longer follow-up. The mean patient age was 48 years (range, 28–83). The median tumor volume was 5.5 cc (range, 0.26–16.6). A median dose of 16 Gy (range, 11–20) was prescribed to tumor margins. Clinical and neuroimaging follow-up was obtained for all patients between 12–156 months (mean 41 months) after radiosurgery. Local tumor control was achieved in 28 of 30 tumors. The mean volume of lesions that were controlled by radiosurgery was 5 cc whereas median volume of tumors that failed radiosurgery was 11.5 cc. The lesions that were controlled by radiosurgery had received a median tumor margin dose of 16.4 Gy (range, 11–20) compared to 13.5 Gy (range, 13–14) prescribed to tumors that ultimately failed radiosurgery. The margin dose (16 Gy or more) was a significant predictor of tumor control after radiosurgery. At the last assessment, 20 patients (71%) were alive and eight (29%) had died. The mean survival after radiosurgery was 6.7 years. Early experience from several centers indicated that radiosurgery could lead to tumor control or regression [136,137] (> Figure 65-7). For the most part, we have treated tumors with documented tumor growth, which are usually solid and

1025

1026

65


almost exclusively located in the posterior fossa, cerebellum, and brainstem. Such tumors are generally treated when they have shown evidence of objective growth and neurological symptoms develop. Prophylactic radiosurgery for hemangioblastomas in the case of von Hippel-Lindau disease (VHL) is not performed unless tumor growth or new symptoms are documented. For those patients with cystic hemangioblastomas, we have less optimism related to the overall role of radiosurgery at least as a single option. However, surgical removal of a large cystic component of a tumor producing mass effect symptoms is usually appropriate followed by radiosurgery for any residual solid component. In selected cases stereotactic aspiration of the cyst followed by subsequent radiosurgery is feasible. Repeat radiosurgery may be required over many years when other tumors show additional growth [138]. Chordoma and Chondrosarcoma

We continue to regard these tumors as difficult tumors to manage. Almost invariably they require multi-modality management over the course of many years. These invasive tumors provide a management challenge because of their critical location and their tendency to aggressively recur locally despite multi-modality treatment. Radiosurgery has been used both as a primary and adjuvant management strategy [126,139]. During our 20-year experience, 28 patients with chordoma and 19 patients with chondrosarcomas have undergone management with radiosurgery (> Table 65-6). We recently analyzed outcome of 33 eligible patients (chordoma 14 and chondrosarcomas 19). Five-year actuarial tumor control rates for chondrosarcoma and Chordomas after single procedure were 78 and 57% respectively [140] (> Figure 65-7). Our recent analysis indicated both the promise and difficulties of multi-modality management for chordomas and chondrosarcomas of the skull base. We found that chondrosarcomas generally respond well to radiosurgery. At present, radiosurgery for chordomas is best considered as

an adjuvant management except for very small biopsy proven chordomas wherein high dose, highly conformal and highly selective radiosurgery can be given. The addition of radiosurgery after surgical resection and radiation therapy or proton beam radiation has shown to increase both local tumor control and survival. For tumors without compression of the brainstem or local mass effect, SRS is an alternative option to surgical resection. In our experience, none of our patients developed adverse radiation effects. Radiosurgery appears to be a safe and effective management for small volume tumors, but over the course of many years, especially from 5 to 10 years after initial surgery and radiosurgery, recurrence rates continue to increase. In such cases, repeat radiosurgery or perhaps fractionated radiation therapy and repeat radiosurgery should be considered. Most such patients require one or more microsurgical approaches . Table 65-6 Gamma Knife radiosurgery for miscellaneous Neoplasms Indications Pineal Region Tumor Craniopharyngioma Hemangioblastoma Chondrosarcoma Chordoma Hemangiopericytoma Glomus Tumor Hemabngioma Myoepithelioma Rhabdomyosarcoma Esthesioneuroblastoma Choroid Plexux papilloma Colloid cust Hamartoma Lymphoma Neurofibrosarcoma Dysembryoplastic Neuroepithelial tumors (DNET) Fibrohistiocytoma Invasive skullbase tumors Others Total indications

Number of patients treated 29 68 44 19 28 35 19 8 1 3 4 11 1 6 11 1 2 5 31 21 347


for tumor cytoreduction. More recently, we have embarked on the usage of endoscopic transsphenoidal resection followed by radiosurgery. Invasive Skull Base Cancers

After combined otolaryngological and neurosurgical procedures, we have used adjuvant radiosurgery for invasive skull base cancers (31 patients). Sixteen patients had adenocarcinomas, 14 had squamous cell carcinomas, and one patient had a metastatic neuroendocrine tumor. In such cases, radiosurgery has been used as an adjuvant or in combination with external beam fractionated radiation therapy. Many reports have documented the role of radiosurgery as salvage procedure for malignant tumors involving the skull base [139,141,142].

Radiosurgery for Functional Brain Disorders Trigeminal Neuralgia Radiosurgery At our institution 775 patients have undergone radiosurgery for management of trigeminal neuralgia. Our last detailed review studied 220 consecutive radiosurgery procedures for typical trigeminal neuralgia, all performed between 1992 and 1998 [143]. All 220 patients had trigeminal neuralgia that was idiopathic, long standing, and refractory to medication therapy. Most of the patients had a long history of medical treatment with the median symptom duration of 96 months (range, 3–564 months). Pain was predominantly distributed in the V2 and V3 distributions of the trigeminal nerve (29.5%), followed by V2 alone (22.3%) and V3 alone (13.2%). Prior surgery was performed in 135 patients (61.4%), including microvascular decompression, glycerol rhizotomy, radiofrequency rhizotomy, balloon microcompression, peripheral neurectomy, or ethanol injections. In the remaining 85 patients (38.6%), radiosurgery was the first surgical pro-

65

cedure performed. The median central dose at trigeminal nerve was 80 Gy. The pain relief after radiosurgery was graded into four categories: excellent, good, fair, and poor. Complete pain relief without the use of any medication was defined as an excellent outcome. Patients with complete pain relief with some medication were considered as good outcomes. Patients with partial pain relief (more than 50% pain relief) were considered to have a fair outcome [57]. No pain relief or less than 50% pain relief were considered as poor. Of the 220 patients, 47 (25.1%) required further additional surgical procedures because of insufficient pain control. These patients were considered as treatment failures (poor outcome), and the results after the additional procedure were excluded from this analysis. Most of the patients started experiencing pain relief within 6 months of radiosurgery (median, 2 months). At the initial followup assessment performed at 6 months, excellent results were obtained in 105 patients (47.7%), and excellent plus good results were found in 139 patients (63.2%). More than 50% pain relief (excellent, good, or fair) was noted in 181 patients (82.3%). At the last follow-up evaluation, 88 patients (40%) had excellent outcomes, 121 patients (55.9%) had excellent plus good outcomes, and 152 were fair or better (69.1%). Thirty patients (13.6%) had recurrence of pain after the initial achievement of pain relief (25 patients after complete relief, five patients after more than 50% relief) between 2 and 58 months after radiosurgery. Recurrences occurred at a mean of 15.4 months from irradiation. The median time to achieving complete pain relief (good or excellent) was also 2 months (2.05.1). Complete pain relief (good or excellent) was achieved in 64.93.2% of the patients at 6 months, 70.33.16% by 1 year, and in 75.43.49% of patients by 33 months. Complete pain relief (excellent or good) was achieved and maintained in 63.63.3% of patients at 1 year, 59.23.5% of patients at 2 years, and 56.63.8% of patients at 3 years. A history of no prior surgery

1027

1028

65


was the only factor significantly associated (p = 0.01) with achieving and maintaining complete pain relief. No patient sustained an early complication after any radiosurgery procedure. Seventeen patients (7.7%) developed increased facial paresthesia and/or facial numbness that lasted more than 6 months. Others have noted a dry eye, without significant facial numbness. No patient developed a mastication deficit after radiosurgery or noted problems in facial motor function. One patient (0.4%) developed deafferentation pain after radiosurgery. The low incidence of complications is the greatest advantage of stereotactic radiosurgery compared to all other surgical options. Radiosurgery can be repeated if pain returns after initial relief. We advocate repeat radiosurgery only if complete pain relief had been achieved with subsequent recurrence [144]. We advocate a maximum dose of 60–70 Gy at a second procedure.

Movement Disorder Radiosurgery Stereotactic radiosurgery is an option to manage movement disorders for patients who are high risk for surgery and anesthesia. Gamma Knife is the preferred radiosurgical tool for treatment of movement disorders. We have treated 80 movement disorders patients with radiosurgery (> Table 65-7). Gamma knife thalamotomy

The VIM nucleus of the thalamus is the target for a tremor patient. We have previously reported our experience with the treatment of essential tremor (ET) and Parkinson’s disease (PD) tremor using gamma Knife radiosurgery. Niranjan et al. evaluated 11 patients managed with gamma knife thalamotomy for essential and Multiple sclerosis (MS) related tremor [145]. All patients noted improvement in action tremor. Six of eight ET patients had complete tremor arrest, and the violent action tremor in all three patients

. Table 65-7 Gamma Knife radiosurgery for Functional Brain Disorders Number of patients treated

Indication Trigeminal neuralgia Cancer Pain Movement Disorder

OCD Total

775

Essential tremor Parkinson’s Disease Others

2 40 30 10 3 869

with MS was improved. One patient developed transient arm weakness. Duma et al. treated 42 patients with tremor from PD or ET with VIM thalamotomy using Gamma Knife. Median time of onset of improvement was 2 months [146]. No change in tremor occurred in four gamma knife thalamotomies (8.6%), ‘‘mild’’ improvement was seen in four (8.6%), ‘‘good’’ improvement was seen in 13 (28%), and ‘‘excellent’’ improvement in 13 (28%). In 12 thalamotomies (26%), the tremor was eliminated completely. One patient, after bilateral treatment, suffered a mild acute dysarthria 1 week after GK thalamotomy. Ohye, et al. reported 36 Gamma thalamotomies in 31 patients. Maximum dose was 150 Gy in the first six cases, which was subsequently reduced to 130 Gy [6,147]. In two patients undergoing repeat procedures the dose was decreased to 120 Gy. In all cases except one, a single 4-mm isocenter was used. In their 15 cases with more than 2 years follow-up, a clinically good result was seen in 87%, with no noticeable side effects. In a more recent report, these authors have compared the results of 51 patients who had thalamotomy after reloading of gamma Knife with that of previous patients. Young, et al., in a large series of patients reviewed their use of GK thalamotomy for the treatment of tremor [148]. Their series included 102 patients


with parkinsonian tremor, 52 patients with essential tremor, and four patients with tremor of other etiology. The single 4-mm collimator was used with doses varying from 110–160 Gy. At a median follow up of 52.5 months (11–93 months) 76% were tremor free, and 12% were ‘‘nearly free of tremor.’’ Thus there was failure in 12%. In 52 patients who had radiosurgery for disabling ET 92% were completely at 1 year and 88% were completely free after 4 years. Gamma Knife Pallidotomy

Duma et al. performed Gamma Knife pallidotomy on 18 patients with medically recalcitrant and disabling symptoms of PD. Only six patients (33%) showed improvement in rigidity and dyskinesia. Three patients (17%) were unchanged, and nine patients (50%) worsened. Of the six patients with improvement, two exhibited visual field deficits. Overall, four (22%) patients had a visual field deficit, three patients had speech and/or swallowing difficulties, three had worsening of their gait, and one had numbness in the contralateral hemibody. Nine patients (50%) had one or more complications related to treatment. Friedman, et al. had similar experience [149]. They described their results in four patients using Gamma Knife pallidotomy in advanced disease. No patient improved in a significant manner within the followup interval of 18 months. One patient experienced an improvement in his dyskinesia, but also became transiently psychotic and demented. The other three patients suffered no adverse effects. We do not recommend this procedure because of the high rate of morbidity noted. Radiosurgery for Epilepsy

Radiosurgery is an emerging therapeutic approach for the treatment of medically intractable seizures. Most experience of the treatment of medically refractory epilepsy is with gamma-knife radiosurgery for lesional epilepsy associated with arteriovenous malformations, cavernomas, and tumors. There have been some studies of the treatment of epilepsy associated with hypothalamic hamarto-

65

mas and mesial-temporal sclerosis as described below. There is evidence that gamma-knife radiosurgery improves seizure in patients affected by medically refractory epilepsy associated with hypothalamic hamartomas [150–153]. Overall, the results of these studies suggest that radiosurgery is a safe and effective option for the treatment of seizures associated with hypothalamic hamartomas. Marginal doses of 17 Gy or higher seem to be required when using gamma-knife radiosurgery. Regis et al. reported outcome of radiosurgery as a treatment for mesial-temporal-lobe epilepsy in a long-term multicentre trial [154]. Patients were selected for gamma-knife radiosurgery by use of the same criteria as for microsurgical amygdalohippocampectomy, including the presence of hippocampal sclerosis and the absence of space-occupying lesions. The target volume was approximately 7 cm3. The target included the head and body of the hippocampus, anterior part of the parahippocampal gyrus, and the basolateral region of the amygdaloid complex (sparing the upper and mesial part). A margin dose of 25 Gy was delivered to the 50% isodoseline. Analysis of seizure control after a 2-year follow-up showed a reduction of the median number of seizures from 6.2 to 0.3 per month. This study suggests that radiosurgery can be effective for medically refractory mesial-temporallobe epilepsy. The exact mechanism of seizure abolition after radiosurgery (or conventional irradiation) is not well understood. Depending on the dose and the target volume, radiosurgery can induce necrosis and consequent destruction of the epileptic focus and its pathways of spread. Alternatively, suppression of epileptic activity by a neuromodulatory effect at non-necrotising doses has been proposed as a possible mechanism of action [155]. In patients with seizures associated with arteriovenous malformations, seizure outcome seems to be independent of nidus occlusion [97,156,157], suggesting that radiosurgery can reduce seizure frequency through

1029

1030

65


an intrinsic effect on the epileptogenic cortex surrounding the nidus. Radiosurgery of the rat hippocampus induces neuromodulatory effects such as reduction of the cholinergic and excitatory aminoacid concentrations with preservation of the GABAergic system, suggesting that it might be possible to modify an epileptogenic cortex sufficiently for it to become non-epileptic, while preserving its functional role [155]. The experimental evidence suggests that, at least in rats, it is possible to achieve seizure control with subnecrotic doses of radiation [158–160].

11.

12.

13.

14.

References 1. Ammar A. Lars Leksell’s vision–radiosurgery. Acta Neurochir Suppl 1994;62:1-4. 2. Lunsford LD. Lars Leksell: notes at the side of a raconteur. Stereotact Funct Neurosurg 1996;67:153-68. 3. Liscak R, Vladyka V, Simonova G, Urgosik D, Novotny J, Jr, Janouskova L, Vymazal J. Arteriovenous malformations after Leksell gamma knife radiosurgery: rate of obliteration and complications. Neurosurgery 2007;60:1005-14; discussion 1015-6. 4. Pollock BE, Gorman DA, Coffey RJ. Patient outcomes after arteriovenous malformation radiosurgical management: results based on a 5- to 14-year follow-up study. Neurosurgery 2003;52:1291-96; discussion 1296-7. 5. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ, Kondziolka D. Hemorrhage risk after stereotactic radiosurgery of cerebral arteriovenous malformations. Neurosurgery 1996;38:652-9; discussion 659-61. 6. Inoue HK, Ohye C. Hemorrhage risks and obliteration rates of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 2002;97:474-6. 7. Karlsson B, Lax I, Soderman M. Risk for hemorrhage during the 2-year latency period following gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 2001;49:1045-51. 8. Maruyama K, Kawahara N, Shin M, Tago M, Kishimoto J, Kurita H, Kawamoto S, Morita A, Kirino T. The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 2005;352:146-53. 9. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. An analysis of the dose-response for arteriovenous malformation radiosurgery and other factors affecting obliteration. Radiother Oncol 2002;63:347-54. 10. Pollock BE, Kondziolka D, Flickinger JC, Patel AK, Bissonette DJ, Lunsford LD. Magnetic resonance imaging: an accurate method to evaluate arteriovenous mal-

15.

16.

17.

18.

19.

20.

21.

22.

formations after stereotactic radiosurgery. J Neurosurg 1996;85:1044-9. Maesawa S, Flickinger JC, Kondziolka D, Lunsford LD. Repeated radiosurgery for incompletely obliterated arteriovenous malformations. J Neurosurg 2000;92:961-70. Chopra R, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC. Long-term follow-up of acoustic schwannoma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2007;68:845-51. Pan DH, Guo WY, Chung WY, Shiau CY, Chang YC, Wang LW. Gamma knife radiosurgery as a single treatment modality for large cerebral arteriovenous malformations. J Neurosurg 2000;93 Suppl 3:113-9. Miyawaki L, Dowd C, Wara W, Goldsmith B, Albright N, Gutin P, Halbach V, Hieshima G, Higashida R, Lulu B, Pitts L, Schell M, Smith V, Weaver K, Wilson C, Larson D. Five year results of LINAC radiosurgery for arteriovenous malformations: outcome for large AVMS. Int J Radiat Oncol Biol Phys 1999;44:1089-106. Pollock BE, Kline RW, Stafford SL, Foote RL, Schomberg PJ. The rationale and technique of stagedvolume arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 2000;48:817-24. Sirin S, Kondziolka D, Niranjan A, Flickinger JC, Maitz AH, Lunsford LD. Prospective staged volume radiosurgery for large arteriovenous malformations: indications and outcomes in otherwise untreatable patients. Neurosurgery 2006;58:17-27; discussion 17-27. Raza SM, Jabbour S, Thai QA, Pradilla G, Kleinberg LR, Wharam M, Rigamonti D. Repeat stereotactic radiosurgery for high-grade and large intracranial arteriovenous malformations. Surg Neurol 2007;68:24-34; discussion 34. Flickinger JC, Kondziolka D, Lunsford LD, Kassam A, Phuong LK, Liscak R, Pollock B. Development of a model to predict permanent symptomatic postradiosurgery injury for arteriovenous malformation patients. Arteriovenous Malformation Radiosurgery Study Group. Int J Radiat Oncol Biol Phys 2000;46:1143-8. Hara M, Nakamura M, Shiokawa Y, Sawa H, Sato E, Koyasu H, Saito I. Delayed cyst formation after radiosurgery for cerebral arteriovenous malformation: two case reports. Minim Invasive Neurosurg 1998;41: 40-5. Izawa M, Hayashi M, Chernov M, Nakaya K, Ochiai T, Murata N, Takasu Y, Kubo O, Hori T, Takakura K. Long-term complications after gamma knife surgery for arteriovenous malformations. J Neurosurg 2005;102 Suppl:34-7. Pan HC, Sheehan J, Stroila M, Steiner M, Steiner L. Late cyst formation following gamma knife surgery of arteriovenous malformations. J Neurosurg 2005;102 Suppl:124-7. Koebbe CJ, Singhal D, Sheehan J, Flickinger JC, Horowitz M, Kondziolka D, Lunsford LD. Radiosurgery for dural arteriovenous fistulas. Surg Neurol 2005;64:392-8; discussion 398-399.


23. Linskey ME, Johnstone PA, O’Leary M, Goetsch S. Radiation exposure of normal temporal bone structures during stereotactically guided gamma knife surgery for vestibular schwannomas.[see comment]. J Neurosurg 2003;98:800-6. 24. Pan DH, Chung WY, Guo WY, Wu HM, Liu KD, Shiau CY, Wang LW. Stereotactic radiosurgery for the treatment of dural arteriovenous fistulas involving the transverse-sigmoid sinus. J Neurosurg 2002;96: 823-9. 25. Pollock BE, Nichols DA, Garrity JA, Gorman DA, Stafford SL. Stereotactic radiosurgery and particulate embolization for cavernous sinus dural arteriovenous fistulae. Neurosurgery 1999;45:459-66; discussion 466-7. 26. Shin M, Kurita H, Tago M, Kirino T. Stereotactic radiosurgery for tentorial dural arteriovenous fistulae draining into the vein of Galen: report of two cases. Neurosurgery 2000;46:730-3; discussion 733-4. 27. Hasegawa T, McInerney J, Kondziolka D, Lee JY, Flickinger JC, Lunsford LD. Long-term results after stereotactic radiosurgery for patients with cavernous malformations. Neurosurgery 2002;50:1190-97; discussion 1197-8. 28. Liscak R, Vladyka V, Simonova G, Vymazal J, Novotny J, Jr. Gamma knife surgery of brain cavernous hemangiomas. J Neurosurg 2005;102 Suppl:207-13. 29. Chung WY, Liu KD, Shiau CY, Wu HM, Wang LW, Guo WY, Ho DM, Pan DH. Gamma knife surgery for vestibular schwannoma: 10-year experience of 195 cases. J Neurosurg 2005;102 Suppl:87-96. 30. Hasegawa T, Kida Y, Kobayashi T, Yoshimoto M, Mori Y, Yoshida J. Long-term outcomes in patients with vestibular schwannomas treated using gamma knife surgery: 10-year follow up. J Neurosurg 2005;102:10-6. 31. Kondziolka D, Nathoo N, Flickinger JC, Niranjan A, Maitz AH, Lunsford LD. Long-term results after radiosurgery for benign intracranial tumors.[see comment]. Neurosurgery 2003;53:815-21; discussion 821-2. 32. Maruyama K, Kondziolka D, Niranjan A, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004;100:407-13. 33. Delbrouck C, Hassid S, Massager N, Choufani G, David P, Devriendt D, Levivier M. Preservation of hearing in vestibular schwannomas treated by radiosurgery using Leksell Gamma Knife: preliminary report of a prospective Belgian clinical study. Acta Oto-RhinoLaryngologica Belgica 2003;57:197-204. 34. Flickinger JC, Kondziolka D, Niranjan A, Lunsford LD. Results of acoustic neuroma radiosurgery: an analysis of 5 years’ experience using current methods.[see comment]. J Neurosurg 2001;94:1-6. 35. Flickinger JC, Kondziolka D, Pollock BE, Lunsford LD. Evolution in technique for vestibular schwannoma radio-

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

65

surgery and effect on outcome. Int J Radiat Oncol Biol Phys 1996;36:275-80. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis PS. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001;95:440-9. Harsh GR, Thornton AF, Chapman PH, Bussiere MR, Rabinov JD, Loeffler JS. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys 2002;54:35-44. Horstmann GA, Van Eck AT. Gamma knife model C with the automatic positioning system and its impact on the treatment of vestibular schwannomas. J Neurosurg 2002;97:450-5. Inoue HK. Low-dose radiosurgery for large vestibular schwannomas: long-term results of functional preservation. J Neurosurg 2005;102 Suppl:111-3. Ishihara H, Saito K, Nishizaki T, Kajiwara K, Nomura S, Yoshikawa K, Harada K, Suzuki M. CyberKnife radiosurgery for vestibular schwannoma. Minim Invasive Neurosurg 2004;47:290-3. Kondziolka D, Lunsford LD, Flickinger JC. Gamma knife radiosurgery for vestibular schwannomas. Neurosurg Clin N Am 2000;11:651-8. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas.[see comment]. N Engl J Med 1998;339:1426-33. Linskey ME. Stereotactic radiosurgery versus stereotactic radiotherapy for patients with vestibular schwannoma: a Leksell Gamma Knife Society 2000 debate. J Neurosurg 2000;93 Suppl 3:90-5. Linskey ME, Johnstone PA. Radiation tolerance of normal temporal bone structures: implications for gamma knife stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003;57:196-200. Linskey ME, Lunsford LD, Flickinger JC. Tumor control after stereotactic radiosurgery in neurofibromatosis patients with bilateral acoustic tumors. Neurosurgery 1992;31:829-38; discussion 838-9. Meijer OW, Wolbers JG, Vandertop WP, Slotman BJ. Stereotactische bestraling van het vestibulair schwannoom (acusticusneurinoom). Nederlands Tijdschrift voor Geneeskunde 2000;144:2088-93. Petit JH, Hudes RS, Chen TT, Eisenberg HM, Simard JM, Chin LS. Reduced-dose radiosurgery for vestibular schwannomas. Neurosurgery 2001;49:1299-306; discussion 1306-7. Shamisa A, Bance M, Nag S, Tator C, Wong S, Noren G, Guha A. Glioblastoma multiforme occurring in a patient treated with gamma knife surgery. Case report and review of the literature. J Neurosurg 2001;94:816-21. Litvack ZN, Noren G, Chougule PB, Zheng Z. Preservation of functional hearing after gamma knife surgery for vestibular schwannoma. Neurosurg Focus 2003;14:e3.

1031

1032

65


50. Niranjan A, Lunsford LD, Flickinger JC, Maitz A, Kondziolka D. Dose reduction improves hearing preservation rates after intracanalicular acoustic tumor radiosurgery. Neurosurgery 1999;45:753-62; discussion 762-5. 51. Subach BR, Kondziolka D, Lunsford LD, Bissonette DJ, Flickinger JC, Maitz AH. Stereotactic radiosurgery in the management of acoustic neuromas associated with neurofibromatosis type 2.[see comment]. J Neurosurg 1999;90:815-22. 52. Roche PH, Robitail S, Thomassin JM, Pellet W, Regis J. Radiochirurgie gamma knife des schwannomes vestibulaires associes a une neurofibromatose de type 2. Neuro Chir 2004;50:367-76. 53. Rowe J, Grainger A, Walton L, Silcocks P, Radatz M, Kemeny A. Risk of malignancy after gamma knife stereotactic radiosurgery. Neurosurgery 2007;60:60-5; discussion 65-6. 54. Mathieu D, Kondziolka D, Flickinger JC, Niranjan A, Williamson R, Martin JJ, Lunsford LD. Stereotactic radiosurgery for vestibular schwannomas in patients with neurofibromatosis type 2: an analysis of tumor control, complications, and hearing preservation rates. Neurosurgery 2007;60:460-8; discussion 468-70. 55. Kondziolka D, Levy EI, Niranjan A, Flickinger JC, Lunsford LD. Long-term outcomes after meningioma radiosurgery: physician and patient perspectives. J Neurosurg 1999;91:44-50. 56. Pollock BE, Stafford SL, Link MJ. Gamma knife radiosurgery for skull base meningiomas. Neurosurg Clin N Am 2000;11:659-66. 57. Chang SD, Adler JR, Jr. Treatment of cranial base meningiomas with linear accelerator radiosurgery. Neurosurgery 1997;41:1019-25; discussion 1025-7. 58. Duma CM, Lunsford LD, Kondziolka D, Harsh GR, Flickinger JC. Stereotactic radiosurgery of cavernous sinus meningiomas as an addition or alternative to microsurgery. Neurosurgery 1993;32:699-704; discussion 704-5. 59. Hakim R, Alexander E, III, Loeffler JS, Shrieve DC, Wen P, Fallon MP, Stieg PE, Black PM. Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998;42:446-53; discussion 453-4. 60. Kondziolka D, Flickinger JC, Perez B. Judicious resection and/or radiosurgery for parasagittal meningiomas: outcomes from a multicenter review. Gamma Knife Meningioma Study Group. Neurosurgery 1998;43:405-13; discussion 413-4. 61. Liscak R, Simonova G, Vymazal J, Janouskova L, Vladyka V. Gamma knife radiosurgery of meningiomas in the cavernous sinus region. Acta Neurochir (Wien) 1999;141:473-80. 62. Morita A, Coffey RJ, Foote RL, Schiff D, Gorman D. Risk of injury to cranial nerves after gamma knife radiosurgery

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

for skull base meningiomas: experience in 88 patients. J Neurosurg 1999;90:42-9. Ojemann RG. Management of cranial and spinal meningiomas (honored guest presentation). Clin Neurosurg 1993;40:321-83. Roche PH, Regis J, Dufour H, Fournier HD, Delsanti C, Pellet W, Grisoli F, Peragut JC. Gamma knife radiosurgery in the management of cavernous sinus meningiomas. J Neurosurg 2000;93 Suppl 3:68-73. Shafron DH, Friedman WA, Buatti JM, Bova FJ, Mendenhall WM. Linac radiosurgery for benign meningiomas. Int J Radiat Oncol Biol Phys 1999;43:321-7. Stafford SL, Pollock BE, Foote RL, Link MJ, Gorman DA, Schomberg PJ, Leavitt JA. Meningioma radiosurgery: tumor control, outcomes, and complications among 190 consecutive patients. Neurosurgery 2001;49:1029-37; discussion 1037-8. Subach BR, Lunsford LD, Kondziolka D, Maitz AH, Flickinger JC. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery 1998;42: 437-43; discussion 443-5. Nicolato A, Foroni R, Alessandrini F, Bricolo A, Gerosa M. Radiosurgical treatment of cavernous sinus meningiomas: experience with 122 treated patients. Neurosurgery 2002;51:1153-59; discussion 1159-61. Aichholzer M, Bertalanffy A, Dietrich W, Roessler K, Pfisterer W, Ungersboeck K, Heimberger K, Kitz K. Gamma knife radiosurgery of skull base meningiomas. Acta Neurochir (Wien) 2000;142:647-52; discussion 652-3. Kondziolka D, Mathieu D, Lunsford LD, Martin JJ, Madhok R, Niranjan A, Flickinger JC. Radiosurgery As Definitive Management Of Meningiomas. Neurosurgery 2008;62(1):53-8. Couldwell WT, Fukushima T, Giannotta SL, Weiss MH. Petroclival meningiomas: surgical experience in 109 cases. J Neurosurg 1996;84:20-8. De Jesus O, Sekhar LN, Parikh HK, Wright DC, Wagner DP. Long-term follow-up of patients with meningiomas involving the cavernous sinus: recurrence, progression, and quality of life. Neurosurgery 1996;39:915-9; discussion 919-20. DeMonte F, Smith HK, al-Mefty O. Outcome of aggressive removal of cavernous sinus meningiomas. J Neurosurg 1994;81:245-51. Iwai Y, Yamanaka K, Ishiguro T. Gamma knife radiosurgery for the treatment of cavernous sinus meningiomas. Neurosurgery 2003;52:517-24; discussion 523-4. Kurita H, Sasaki T, Kawamoto S, Taniguchi M, Terahara A, Tago M, Kirino T. Role of radiosurgery in the management of cavernous sinus meningiomas. Acta Neurol Scand 1997;96:297-304. Pollock BE, Stafford SL, Utter A, Giannini C, Schreiner SA. Stereotactic radiosurgery provides equivalent tumor control to Simpson Grade 1 resection for patients with


small- to medium-size meningiomas. Int J Radiat Oncol Biol Phys 2003;55:1000-5. 77. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery for nonfunctioning pituitary adenoma. Neurosurg Focus 2003;14:e9. 78. Attanasio R, Epaminonda P, Motti E, Giugni E, Ventrella L, Cozzi R, Farabola M, Loli P, Beck-Peccoz P, Arosio M. Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab 2003;88:3105-12. 79. Castinetti F, Taieb D, Kuhn JM, Chanson P, Tamura M, Jaquet P, Conte-Devolx B, Regis J, Dufour H, Brue T. Outcome of gamma knife radiosurgery in 82 patients with acromegaly: correlation with initial hypersecretion. J Clin Endocrinol Metab 2005;90:4483-88. 80. Gutt B, Wowra B, Alexandrov R, Uhl E, Schaaf L, Stalla GK, Schopohl J. Gamma-knife surgery is effective in normalising plasma insulin-like growth factor I in patients with acromegaly. Exp Clin Endocrinol Diabetes 2005;113:219-24. 81. Pollock BE, Flickinger JC. A proposed radiosurgery-based grading system for arteriovenous malformations. J Neurosurg 2002;96:79-85. 82. Sheehan JP, Niranjan A, Sheehan JM, Jane JA, Jr, Laws ER, Kondziolka D, Flickinger J, Landolt AM, Loeffler JS, Lunsford LD. Stereotactic radiosurgery for pituitary adenomas: an intermediate review of its safety, efficacy, and role in the neurosurgical treatment armamentarium. J Neurosurg 2005;102:678-91. 83. Kondziolka D, Maitz AH, Niranjan A, Flickinger JC, Lunsford LD. An evaluation of the Model C gamma knife with automatic patient positioning. Neurosurgery 2002;50:429-31; discussion 431-2. 84. Laws ER, Jr, Vance ML. Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 1999;10:327-36. 85. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the treatment of Cushing disease: long-term results. J Neurosurg 2002;97:422-28. 86. Lippitz BE, Kraepelien T, Hautanen K, Ritzling M, Rahn T, Ulfarsson E, Boethius J. Gamma knife radiosurgery for patients with multiple cerebral metastases. Acta Neurochir Suppl 2004;91:79-87. 87. Sheehan JM, Vance ML, Sheehan JP, Ellegala DB, Laws ER Jr. Radiosurgery for Cushing’s disease after failed transsphenoidal surgery. J Neurosurg 2000;93:738-742. 88. Vladyka V, Liscak R, Novotny J, Jr, Marek J, Jezkova J. Radiation tolerance of functioning pituitary tissue in gamma knife surgery for pituitary adenomas. Neurosurgery 2003;52:309-16; discussion 316-7. 89. Pollock BE, Nippoldt TB, Stafford SL, Foote RL, Abboud CF. Results of stereotactic radiosurgery in patients with hormone-producing pituitary adenomas: factors associated with endocrine normalization. J Neurosurg 2002;97:525-30.

65

90. Chiou SM, Lunsford LD, Niranjan A, Kondziolka D, Flickinger JC. Stereotactic radiosurgery of residual or recurrent craniopharyngioma, after surgery, with or without radiation therapy. Neuro Oncol 2001;3:159-66. 91. Hasegawa T, Kondziolka D, Hadjipanayis CG, Lunsford LD. Management of cystic craniopharyngiomas with phosphorus-32 intracavitary irradiation. Neurosurgery 2004;54:813-20; discussion 820-2. 92. Amendola BE, Wolf AL, Coy SR, Amendola M, Bloch L. Gamma knife radiosurgery in the treatment of patients with single and multiple brain metastases from carcinoma of the breast. Cancer J 2000;6:88-92. 93. Chung WY, Pan DH, Shiau CY, Guo WY, Wang LW. Gamma knife radiosurgery for craniopharyngiomas. J Neurosurg 2000;93 Suppl 3:47-56. 94. McDermott MW, Cosgrove GR, Larson DA, Sneed PK, Gutin PH. Interstitial brachytherapy for intracranial metastases. Neurosurg Clin N Am 1996;7:485-95. 95. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin LS. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002;50:41-6; discussion 46-7. 96. Simonova G, Novotny J, Jr, Liscak R. Low-grade gliomas treated by fractionated gamma knife surgery. J Neurosurg 2005,102 Suppl:19-24. 97. Kida Y, Kobayashi T, Mori Y. Gamma knife radiosurgery for low-grade astrocytomas: results of long-term follow up. J Neurosurg 2000;93 Suppl 3:42-6. 98. Hadjipanayis CG, Levy EI, Niranjan A, Firlik AD, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for motor cortex region arteriovenous malformations. Neurosurgery 2001;48:70-6; discussion 76-7. 99. Chang EL, Wefel JS, Maor MH, Hassenbusch SJ, III, Mahajan A, Lang FF, Woo SY, Mathews LA, Allen PK, Shiu AS, Meyers CA. A pilot study of neurocognitive function in patients with one to three new brain metastases initially treated with stereotactic radiosurgery alone. Neurosurgery 2007;60:277-83; discussion 283-4. 100. Szeifert GT, Massager N, DeVriendt D, David P, De Smedt F, Rorive S, Salmon I, Brotchi J, Levivier M. Observations of intracranial neoplasms treated with gamma knife radiosurgery. J Neurosurg 2002;97:623-6. 101. Auchter RM, Lamond JP, Alexander E, Buatti JM, Chappell R, Friedman WA, Kinsella TJ, Levin AB, Noyes WR, Schultz CJ, Loeffler JS, Mehta MP. A multiinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastasis. Int J Radiat Oncol Biol Phys 1996;35:27-35. 102. Bhatnagar AK, Kondziolka D, Lunsford LD, Flickinger JC. Recursive partitioning analysis of prognostic factors for patients with four or more intracranial metastases treated with radiosurgery. Technol Cancer Res Treat 2007;6:153-60. 103. Chernov MF, Nakaya K, Izawa M, Hayashi M, Usuba Y, Kato K, Muragaki Y, Iseki H, Hori T, Takakura K.

1033

1034

65 104.

105.

106.

107.

108.

109.

110.

111.

112.


Outcome after radiosurgery for brain metastases in patients with low Karnofsky performance scale (KPS) scores. Int J Radiat Oncol Biol Phys 2007;67:1492-8. DiLuna ML, King JT, Jr, Knisely JP, Chiang VL. Prognostic factors for survival after stereotactic radiosurgery vary with the number of cerebral metastases. Cancer 2007;109:135-45. Flickinger JC, Kondziolka D, Lunsford LD, Coffey RJ, Goodman ML, Shaw EG, Hudgins WR, Weiner R, Harsh GR, Sneed PK, et al. A multi-institutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994;28:797-802. Hussain A, Brown PD, Stafford SL, Pollock BE. Stereotactic radiosurgery for brainstem metastases: Survival, tumor control, and patient outcomes. Int J Radiat Oncol Biol Phys 2007;67:521-4. Mathieu D, Kondziolka D, Cooper PB, Flickinger JC, Niranjan A, Agarwala S, Kirkwood J, Lunsford LD. Gamma knife radiosurgery in the management of malignant melanoma brain metastases. Neurosurgery 2007;60:471-81; discussion 481-2. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami L, Rotman M, Mehta MP, Curran WJ, Jr. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363:1665-72. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K, Kenjyo M, Oya N, Hirota S, Shioura H, Kunieda E, Inomata T, Hayakawa K, Katoh N, Kobashi G. Stereotactic radiosurgery plus whole-brain radiation therapy vs stereotactic radiosurgery alone for treatment of brain metastases: a randomized controlled trial. JAMA 2006;295:2483-91. Chidel MA, Suh JH, Reddy CA, Chao ST, Lundbeck MF, Barnett GH. Application of recursive partitioning analysis and evaluation of the use of whole brain radiation among patients treated with stereotactic radiosurgery for newly diagnosed brain metastases. Int J Radiat Oncol Biol Phys 2000;47:993-9. Nieder C, Nestle U, Motaref B, Walter K, Niewald M, Schnabel K. Prognostic factors in brain metastases: should patients be selected for aggressive treatment according to recursive partitioning analysis (RPA) classes? Int J Radiat Oncol Biol Phys 2000;46:297-302. Chitapanarux I, Goss B, Vongtama R, Frighetto L, De Salles A, Selch M, Duick M, Solberg T, Wallace R, Cabatan-Awang C, Ford J. Prospective study of stereotactic radiosurgery without whole brain radiotherapy in patients with four or less brain metastases: incidence of intracranial progression and salvage radiotherapy. J Neuro Oncol 2003;61:143-9.

113. Flickinger JC, Kondziolka D, Lunsford LD, Pollock BE, Yamamoto M, Gorman DA, Schomberg PJ, Sneed P, Larson D, Smith V, McDermott MW, Miyawaki L, Chilton J, Morantz RA, Young B, Jokura H, Liscak R. A multi-institutional analysis of complication outcomes after arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 1999;44:67-74. 114. Hasegawa T, Kondziolka D, Flickinger JC, Germanwala A, Lunsford LD. Brain metastases treated with radiosurgery alone: an alternative to whole brain radiotherapy? Neurosurgery 2003;52:1318-26; discussion 1326. 115. Lutterbach J, Cyron D, Henne K, Ostertag CB. Radiosurgery followed by planned observation in patients with one to three brain metastases. Neurosurgery 2003;52:1066-73; discussion 1073-4. 116. Lavine SD, Petrovich Z, Cohen-Gadol AA, Masri LS, Morton DL, O’Day SJ, Essner R, Zelman V, Yu C, Luxton G, Apuzzo ML. Gamma knife radiosurgery for metastatic melanoma: an analysis of survival, outcome, and complications. Neurosurgery 1999;44:59-64; discussion 64-6. 117. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, Kryscio RJ, Markesbery WR, Foon KA, Young B. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998;280:1485-9. 118. Posner JB. Brain metastases: 1995. A brief review. J Neurooncol 1996.27:287-93. 119. Pirzkall A, Debus J, Lohr F, Fuss M, Rhein B, Engenhart-Cabillic R, Wannenmacher M. Radiosurgery alone or in combination with whole-brain radiotherapy for brain metastases. J Clin Oncol 1998;16:3563-9. 120. Sneed PK, Suh JH, Goetsch SJ, Sanghavi SN, Chappell R, Buatti JM, Regine WF, Weltman E, King VJ, Breneman JC, Sperduto PW, Mehta MP. A multi-institutional review of radiosurgery alone vs. radiosurgery with whole brain radiotherapy as the initial management of brain metastases. Int J Radiat Oncol Biol Phys 2002;53:519-26. 121. Sanghavi SN, Miranpuri SS, Chappell R, Buatti JM, Sneed PK, Suh JH, Regine WF, Weltman E, King VJ, Goetsch SJ, Breneman JC, Sperduto PW, Scott C, Mabanta S, Mehta MP. Radiosurgery for patients with brain metastases: a multi-institutional analysis, stratified by the RTOG recursive partitioning analysis method. Int J Radiat Oncol Biol Phys 2001;51:426-34. 122. Rades D, Bohlen G, Pluemer A, Veninga T, Hanssens P, Dunst J, Schild SE. Stereotactic radiosurgery alone versus resection plus whole-brain radiotherapy for 1 or 2 brain metastases in recursive partitioning analysis class 1 and 2 patients. Cancer 2007;109(12):2515-21. 123. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic gamma radiosurgery of meningiomas. Surg Neurol 2001;55:325-31.


124. Reyns N, Blond S, Gauvrit JY, Touzet G, Coche B, Pruvo JP, Dhellemmes P. Role of radiosurgery in the management of cerebral arteriovenous malformations in the pediatric age group: data from a 100-patient series. Neurosurgery 2007;60:268-76; discussion 276. 125. Miller RC, Foote RL, Coffey RJ, Gorman DA, Earle JD, Schomberg PJ, Kline RW. The role of stereotactic radiosurgery in the treatment of malignant skull base tumors. Int J Radiat Oncol Biol Phys 1997;39:977-81. 126. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for tentorial meningiomas. Acta Neurochir (Wien) 1998;140: 315-20; discussion 320-1. 127. Nettel B, Niranjan A, Martin JJ, Koebbe CJ, Kondziolka D, Flickinger JC, Lunsford LD. Gamma knife radiosurgery for trigeminal schwannomas. Surg Neurol 2004;62:435-44; discussion 444-6. 128. Pan L, Wang EM, Zhang N, Zhou LF, Wang BJ, Dong YF, Dai JZ, Cai PW. Long-term results of Leksell gamma knife surgery for trigeminal schwannomas. J Neurosurg 2005;102 Suppl:220-4. 129. Saringer W, Khayal H, Ertl A, Schoeggl A, Kitz K. Efficiency of gamma knife radiosurgery in the treatment of glomus jugulare tumors. Minim Invasive Neurosurg 2001;44:141-6. 130. Wang LG, Guo Y, Zhang X, Song SJ, Xia JL, Fan FY, Shi M, Wei LC. Brain metastasis: experience of the Xi-Jing hospital. Stereotact Funct Neurosurg 2002;78:70-83. 131. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for anterior foramen magnum meningiomas. Surg Neurol 1999;51: 268-73. 132. Pollock BE. stereotactic radiosurgery in patients with glomus jugulare tumors. Neurosurg Focus 2004;17:E 10. 133. Levy EI, Niranjan A, Thompson TP, Scarrow AM, Kondziolka D, Flickinger JC, Lunsford LD. Radiosurgery for childhood intracranial arteriovenous malformations. Neurosurgery 2000;47:834-41; discussion 841-2. 134. Nakamura N, Shin M, Tago M, Terahara A, Kurita H, Nakagawa K, Ohtomo K. Gamma knife radiosurgery for cavernous hemangiomas in the cavernous sinus. Report of three cases. J Neurosurg 2002;97:477-80. 135. Peker S, Ozduman K, Kilic T, Pamir MN. Relief of hemifacial spasm after radiosurgery for intracanalicular vestibular schwannoma. Minim Invasive Neurosurg 2004;47:235-7. 136. Georg AE, Lunsford LD, Kondziolka D, Flickinger JC, Maitz A. Hemangioblastoma of the posterior fossa. The role of multimodality treatment. Arquivos de NeuroPsiquiatria 1997;55:278-86. 137. Pan L, Wang EM, Wang BJ, Zhou LF, Zhang N, Cai PW, Da JZ. Gamma knife radiosurgery for hemangioblastomas. Stereotactic Funct Neurosurg 1998;70 Suppl 1:179-86.

65

138. Jawahar A, Shaya M, Campbell P, Ampil F, Willis BK, Smith D, Nanda A. Role of stereotactic radiosurgery as a primary treatment option in the management of newly diagnosed multiple (3–6) intracranial metastases. Surg Neurol 2005;64:207-12. 139. Lunsford LD, Kondziolka D, Flickinger JC, Bissonette DJ, Jungreis CA, Maitz AH, Horton JA, Coffey RJ. Stereotactic radiosurgery for arteriovenous malformations of the brain. J Neurosurg 1991;75:512-24. 140. Martin JJ, Niranjan A, Kondziolka D, Flickinger JC, Lozanne KA, Lunsford LD. Radiosurgery for chordomas and chondrosarcomas of the skull base. J Neurosurg 2007;107:758-64. 141. Firlik KS, Kondziolka D, Lunsford LD, Janecka IP, Flickinger JC. Radiosurgery for recurrent cranial base cancer arising from the head and neck. Head Neck 1996;18:160-5; discussion 166. 142. Lee DJ, Westra WH, Staecker H, Long D, Niparko JK, Slattery WH, III. Clinical and histopathologic features of recurrent vestibular schwannoma (acoustic neuroma) after stereotactic radiosurgery. Otol Neurotol 2003;24:650-60; discussion 660. 143. Maesawa S, Salame C, Flickinger JC, Pirris S, Kondziolka D, Lunsford LD. Clinical outcomes after stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg 2001;94:14-20. 144. Hasegawa T, Kondziolka D, Spiro R, Flickinger JC, Lunsford LD. Repeat radiosurgery for refractory trigeminal neuralgia. Neurosurgery 2002;50:494-500; discussion 500-2. 145. Niranjan A, Kondziolka D, Baser S, Heyman R, Lunsford LD. Functional outcomes after gamma knife thalamotomy for essential tremor and MS-related tremor. Neurology 2000;55:443-6. 146. Duma CM. Movement Disorder Radiosurgery. In: Loftus CM, Batjer HH, editors. Techniques in Neurosurgery. Philadelphia: Williams and Wilkins, 2003, pp 181-90. 147. Ohye C, Shibazaki T, Ishihara J, Zhang J. Evaluation of gamma thalamotomy for parkinsonian and other tremors: survival of neurons adjacent to the thalamic lesion after gamma thalamotomy. J Neurosurg 2000;93 Suppl 3:120-7. 148. Regine WF, Huhn JL, Patchell RA, St Clair WH, Strottmann J, Meigooni A, Sanders M, Young AB. Risk of symptomatic brain tumor recurrence and neurologic deficit after radiosurgery alone in patients with newly diagnosed brain metastases: results and implications. Int J Radiat Oncol Biol Phys 2002;52:333-8. 149. Friedman DP, Goldman HW, Flanders AE, Gollomp SM, Curran WJ, Jr. Stereotactic radiosurgical pallidotomy and thalamotomy with the gamma knife: MR imaging findings with clinical correlation – preliminary experience. Radiology 1999;212:143-50.

1035

1036

65


150. Mathieu D, Kondziolka D, Niranjan A, Flickinger J, Lunsford LD. Gamma knife radiosurgery for refractory epilepsy caused by hypothalamic hamartomas. Stereotact Funct Neurosurg 2006;84:82-7. 151. Regis J, Hayashi M, Eupierre LP, Villeneuve N, Bartolomei F, Brue T, Chauvel P. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Acta Neurochir Suppl 2004;91:33-50. 152. Regis J, Scavarda D, Tamura M, Nagayi M, Villeneuve N, Bartolomei F, Brue T, Dafonseca D, Chauvel P. Epilepsy related to hypothalamic hamartomas: surgical management with special reference to gamma knife surgery. Childs Nerv Syst 2006;22:881-95. 153. Regis J, Scavarda D, Tamura M, Villeneuve N, Bartolomei F, Brue T, Morange I, Dafonseca D, Chauvel P. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Semin Pediatr Neurol 2007;14:73-9. 154. Regis J, Rey M, Bartolomei F, Vladyka V, Liscak R, Schrottner O, Pendl G. Gamma knife surgery in mesial temporal lobe epilepsy: a prospective multicenter study. Epilepsia 2004;45:504-15. 155. Regis J, Kerkerian-Legoff L, Rey M, Vial M, Porcheron D, Nieoullon A, Peragut JC. First biochemical evidence of differential functional effects following Gamma Knife

156.

157.

158.

159.

160.

surgery. Stereotact Funct Neurosurg 1996;66 Suppl 1:29-38. Eisenschenk S, Gilmore RL, Friedman WA, Henchey RA. The effect of LINAC stereotactic radiosurgery on epilepsy associated with arteriovenous malformations. Stereotact Funct Neurosurg 1998;71:51-61. Steiner L, Lindquist C, Adler JR, Torner JC, Alves W, Steiner M. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992;77:1-8. Liscak R, Vladyka V, Novotny J, Jr, Brozek G, Namestkova K, Mares V, Herynek V, Jirak D, Hajek M, Sykova E. Leksell gamma knife lesioning of the rat hippocampus: the relationship between radiation dose and functional and structural damage. J Neurosurg 2002;97:666-73. Maesawa S, Kondziolka D, Dixon CE, Balzer J, Fellows W, Lunsford LD. Subnecrotic stereotactic radiosurgery controlling epilepsy produced by kainic acid injection in rats. J Neurosurg 2000;93:1033-40. Mori Y, Kondziolka D, Balzer J, Fellows W, Flickinger JC, Lunsford LD, Thulborn KR. Effects of stereotactic radiosurgery on an animal model of hippocampal epilepsy. Neurosurgery 2000;46:157-65; discussion 165-8.

58 Gamma Knife: Technical Aspects D. J. Schlesinger . C. P. Yen . C. Lindquist . L. Steiner

Lars Leksell designed the Gamma Knife1 unit (> Figure 58-1) from the ground up to be a tool of the neurosurgeon. The physics and engineering choices inherent in the design of the unit is perhaps best considered from the perspective of a surgeon end-user. Of all of the radiosurgical tools, the Gamma Knife is the oldest and best established. Since its first conception by Lars Leksell and Bo¨rje Larsson, the system has evolved through several generations of improvements; however the basic principles of the instrument remain primarily unchanged. This chapter will describe the physical principles and method of operation of the Gamma Knife, paying particular attention to the new features and changes found in the recently released Leksell Gamma Knife1 Perfexion™.

Introduction and Overview of Basic Concepts The technique Lars Leksell coined as ‘‘stereotactic radiosurgery’’ in 1951 grew out of his prior work in stereotactic neurosurgery and the idea that such surgery could be made ‘‘bloodless’’ using cross-firing beams of ionizing radiation [1,2]. Leksell’s early experiments attempted to harness the superior dosimetric and biological effect of proton beams generated by a cyclotron [3]. However, because of the extreme expense and difficulty associated with proton beams, Leksell abandoned this approach in search of alternative energy sources. Finally, after many prototypes and many failures, Leksell settled on the gamma emissions of cobalt-60 as an energy source that would best fit his goal of a minimallyinvasive form of stereotactic neurosurgery. #


60

Co Radioactive Decay

The cobalt-60 decay series begins with the creation 59 of radioactive 60 27 Co from stable 27 Co by bombarding it with neutrons in a reactor. 60 27 Co is therefore neutron-heavy, and lies above the line of stability with a high neutron/proton ratio. To regain a stable configuration, 60 27 Co decays through b-decay (> Figure 58-2). The nucleus first emits (with >99% probability) a b particle (or negatron) with energy 0.32 MeV, and a neutrino, both created when a neutron in the nucleus spontaneously disintegrates. The recoiling nucleus then releases two gamma photons at 1.17 and 1.33 MeV respectively [4]. In a Gamma Knife unit, the b-particles are absorbed by the source itself. It is the high-energy gamma photons which create the clinical effect through indirect ionization of tissue molecules, OH-free radical formation, and resulting chemical reactions with vital cellular targets such as chromosomal DNA [5].

Superposition of Beams Depending on the model of the unit, a Gamma Knife contains an array of 201 (model B and C) or 192 (Perfexion) individual cobalt-60 sources aligned with a collimation system. The collimation system (described in more detail below) directs the individual beams of gamma radiation to a very precise focus point. While an individual beam has a relatively low dose rate and causes minimal biological effect, the superposition of all beams at the focus point have a much higher dose rate. This simultaneous cross-firing of beams is the heart of the Gamma Knife technique

898

58

Gamma knife: technical aspects

. Figure 58-1 The Gamma Knife Perfexion at the Lars Leksell Center for Gamma Surgery at the University of Virginia

. Figure 58-2 The creation and radioactive decay of 60Co

and creates an advantageous target to normal tissue biological effect. The Gamma Knife can very precisely target small areas of tissue with a high radiation dose. However, because the energy

is spread out among the individual beams, the Gamma Knife can also achieve a very large dose gradient outside of the target. Gamma Knife treatments are therefore quite heterogeneous in


terms of dose distribution inside and outside of the target – a very different dose characteristic than is found in radiotherapy. Gamma Knife surgery is a high-dose, single-fraction procedure. Tumor control and tissue-sparing is achieved via the steep dose gradient at the target periphery rather than utilizing the radiobiological differential between normal and target tissue as in fractional radiotherapy. This suggests that a radiobiological effect different from that considered by traditional radiation oncology may be active during a Gamma Knife treatment [6,7].

58

. Figure 58-3 The Leksell stereotactic G-Frame and the Leksell coordinate system. (Image courtesy of Elekta AB, Stockholm)

Stereotactic Localization The ability to direct an array of beams on a single point does no good by itself unless the intended target can be precisely located at the focus point. The Gamma Knife operates on the principle of stereotaxy to achieve a high level of precision in localization. As with classic stereotactic neurosurgery, a stereotactic frame is mounted to a patient’s head and defines a reference Cartesian coordinate system known as the Leksell Coordinate System (> Figure 58-3). The base ring of Leksell frame defines the inferior limit of the coordinate system, with the origin located at a point to the right lateral, superior, and posterior to the base ring [3,8]. For the Gamma Knife, the arc and probe used for standard stereotactic procedures has been replaced by the cross-firing configuration of gamma rays. Patients are imaged with the frame attached as well as an external fiducial system which is visible on the resultant imagery. These images are used by the treatment planning system to locate targets in terms of this coordinate space. Once the coordinates for a given target are known, the patient can be positioned in the unit in such a way that the target coincides with the radiation focus point. During treatment, the stereotactic frame is locked to the Gamma

Knife unit so there is a direct connection between the internal machine coordinate system and the frame-defined stereotactic coordinate system.

Technical Description of the Gamma Knife The Gamma Knife with its plastic covers removed resembles a large metal sphere. The vast majority of this 20 metric ton ball of cast iron and tungsten is shielding that exists to protect approximately 20 g of 60Co at the heart of the unit. A Gamma Knife system, regardless of model, consists of six primary components as summarized in > Table 58-1.

The Radiation Body The radiation body (> Figure 58-4) makes up the bulk of what one sees when looking at a Gamma Knife unit. The radiation body is a more or less

899

900

58


. Table 58-1 The primary components of the Gamma Knife Component

Purpose

Radiation body

Primary shielding for the 60Co sources Container for sources and primary collimation system. Container for source sectors in the Perfexion model Mechanism to connect the stereotactic frame to the unit A mechanism to accurately position the patient in the radiation field Allows user control of the Gamma Knife Create appropriate radiation distributions

Central body

Frame docking mechanism Stereotactic positioning system Control panel Treatment planning station(s)

. Figure 58-4 The radiation body of the Gamma Knife model C. The five ports on the side of the radiation body align with reach row of 60Co sources to facilitate loading and unloading. The round control panel on the back of the unit facing the reader is part of the bearing of the central body. During loading and unloading the central body can rotate so a particular set of sources aligns with the loading ports on the side

spherically shaped container that serves as bulk shielding for the 60Co sources and houses the central body. On the front of the radiation body is a steel door which allows passage of the proximal end of the treatment table into the unit. The door has manual controls for opening or closing the entrance door in the unlikely event of a complete loss of system power. For loading and unloading sources, the radiation body has a series of loading channels formed in the right side which align with each row of sources in the central body. source with a corresponding loading channel on the radiation unit.

Central Body and Collimation System The 60Co Sources Inside the radiation unit is the central body, which houses the 60Co sources and the primary collimators (and in the Perfexion model Gamma Knife also houses the collimators, source sectors, and sector drive rods, with the sector motors at the rear of the unit). The central body is mounted on a bearing that is mechanically fixed during clinical use, but can be released and rotated during loading and unloading operations to align each

The 60Co sources are located within the central body of the unit in specially constructed source bushings (> Figure 58-5). Each source is a series of 18 60Co pellets, held inside a series of three welded stainless-steel cylinders, which are themselves held within the steel bushing assembly. Each source has an activity of slightly more than 30 Ci for a total activity at loading of


58

. Figure 58-5 The 60Co source assembly, showing the outer two steel cylinders and the outer assembly

approximately 6,000 Ci, providing a dose rate of greater than 3 Gy/min at the focal point.

Collimation System The details of the structure and operation of the central body and collimation system are one of the biggest differences between the Gamma Knife Perfexion and earlier B, and C models. In the earlier models, the 201 60Co source assemblies are configured in five concentric rings within the hemispherical central body. The sources are at a constant 400 mm from the focal point. Precisely machined beam channels consisting of a precollimator and a tungsten primary collimator are matched to the source assemblies. Each beam aligns with the focus point at the center of the unit with a tolerance of Figure 58-6a).

Each helmet has beam channels corresponding to the beam channels in the inner radiation body. The channels are fitted with removable tungsten-alloy collimators with circular apertures that result in a 4, 8, 14, or 18 mm field. There is one collimator helmet per collimator size, and while all four helmets are identical and can in theory take any size collimator, in practice each helmet is used with uniform-size collimators. Individual collimators can be replaced with tungsten-alloy ‘‘plugs’’ in order to affect desirable changes in the resulting dose distribution. In the Perfexion, the central body and collimation system are somewhat more complex because the 60Co sources in the Perfexion are not stationary and the collimators are entirely internal to the radiation body; there are no external helmets. In the Perfexion, the 192 60Co sources are grouped into eight independent source sectors (> Figure 58-6b). Each source sector is housed in an aluminum frame which is attached to sector drive motors at the rear of the radiation body via linear graphite bushings (> Figure 58-7). There are 576 collimators machined into 12 cm-thick tungsten in five concentric rings to align with the

901

902

58


. Figure 58-6 (a) The 4 mm collimator helmet for the model C (b) The internal, sector-based collimator of the Perfexion (image courtesy of Elekta AB, Stockholm)

source assembly configurations. Each source has three available collimators (4, 8, and 16 mm) as well as two shielded positions; ‘‘blocked’’ and ‘‘home.’’ The blocked position is used when repositioning a patient between isocenters or when shielding a critical structure. The home position is a shielded position used when the unit is inactive. To achieve a particular collimation, the sector drive motors move the sources along their bushings to the correct position over the appropriate

collimator opening. The position of each sector is monitored by linear and rotational encoders which have a positioning repeatability of Figure 58-8) which attaches to the frame and acts as an interface with the treatment table. The PPS has several potential advantages over the older APS system. Because the entire treatment table moves from position to position, the relative positions of the patient’s head and neck do not change during the treatment, increasing patient comfort. The PPS is also a simpler design than the APS, which required two independent sliders to agree on position. Linear and rotational encoders ensure that the PPS is always at a known, correct position with a repeatability of Figure 58-9) contains functionality to allow an operator to import treatment plans, verify all critical treatment

information, commence treatment, monitor treatment status, pause or end treatments, and in an emergency stop treatment and safely remove a patient. The console consists of a control panel, a computer monitor to display and monitor


58

treatment status, a video display for patient surveillance, a microphone for patient communication, and a keyboard for entering pertinent information.

distributions, evaluate treatment plans, verify treatment parameters, and export completed plans to the treatment unit. > Table 58-2 summarizes this functionality.

Planning System

Basic Treatment Process

The treatment planning system for the Gamma Knife, known as Leksell GammaPlan1 (Elekta Instruments AB, Stockholm), is housed in a separate computer, or computers from the clinical unit. Historically, treatment planning was accomplished on a Unix workstation with a direct serial connection to the Gamma Knife console computer for exporting treatment plans. With the Perfexion the treatment planning system runs on a Linux/PC platform and a TCP/IP-based communication system (older models are being retrofitted with the new planning system). The new version allows networking of treatment planning systems, communication with multiple gamma knife units, and simplified communication with external data sources such as hospital PACs and neuroimaging units. GammaPlan includes the functionality required to import patient images, create dose

The basic process of the Gamma Knife procedure begins with placement of the Leksell stereotactic frame, proceeds to imaging and treatment planning, pre-treatment checks, and concludes with the treatment itself and the removal of the stereotactic frame (> Figure 58-10).

Frame Placement The technique of Gamma Knife surgery begins with the placement of the Leksell G-Frame on the patient’s head by the neurosurgeon. The details of the procedure vary by center. The majority of institutions place the frame in procedure rooms using local anesthesia. In our center we believe the frame is best placed in an OR setting using controlled sedation, local anesthesia, and giving strict attention to aseptic conditions.

. Table 58-2 Overview of the functionality found in Elekta GammaPlan for used with the Gamma Knife Feature

Purpose

DICOM image import Image registration

Imports image data from imaging consoles or from institutional PACs systems Registers images into Leksell stereotactic space based on fiducial system visible in images Registers non-stereotactic image studies into the coordinate system of existing stereotactic images Creates a blended display of registered image studies from the same or different modalities (e.g., MR and CT to show both soft tissue and bone) To define targets, critical structures, and other structures of interest To define the skull morphology for use in dosimetry and collision calculations To define one or more dose distributions To create metrics by which to evaluate a treatment plan. Tools include dose volume histograms, mean/min/max doses, and point dose sampling Exports treatment plans to the clinical unit Prints treatment plans to be used as a written directive. Plans include all relevant treatment data which allows verification before and during treatment

Image co-registration Image fusion Contouring tools Skull data Treatment planning tools Treatment evaluation tools Treatment export Printing functionality

905

906

58


. Figure 58-10 Flowchart of basic treatment process

The Leksell G-Frame [10] (> Figure 58-11) consists of a rectangular aluminum base ring to which four aluminum posts are attached. The frame is attached to the patient through the posts using titanium pins which are screwed through the skin to the outer table of the skull. A variety of post shapes and lengths are available

to allow optimal fitting of the frame. In addition, frames are available which use replaceable plastic inserts which electrically isolate the pins from the frame in order to prevent pin-site heating in high-strength MRI units. The quality of the frame placement is absolutely critical to the subsequent steps in of the gamma surgery. Proper frame placement is an acquired skill and necessitates a comprehensive understanding of neuroanatomy as viewed from outside of the head. Because the Leksell frame determines the relative location of stereotactic space, and because the Gamma Knife has a finite treatment volume, it is critical that the surgical target be placed as close as possible to the center of the coordinate space. For patients with lateral lesions this may mean shifting the frame off of the midline of the patient. For patients with multiple lesions scattered across both hemispheres of the brain, this may mean choosing to treat all lesions on one-side of the brain in one session and returning for a second session to treat the other lesions1. In other cases, such as lesions of the parasellar region, it may be advantageous to align the frame with the base ring parallel and below the level of the anterior optic pathways. The resulting orientation of the brain may help shift dose away from these critical and radiosensitive structures, although there is often a tradeoff with increased dose to the lenses of the eye. Incorrect frame placement can result in great difficulty in achieving optimal dose distributions around critical structures, and in the case of lateral or posterior lesions can even make treatment impossible without repositioning the frame and re-imaging the patient.

1

Note that this is not as critical an issue for the Perfexion model Gamma Knife as it has a significantly larger available treatment volume. For the Perfexion, the optimal frame placement is generally neutral, with the frame centered on the midline of the head.


58

. Figure 58-11 Leksell G-Frame as it would appear affixed to a patient’s head and attached to the unit via the frame adapter

Imaging The accuracy of a Gamma Knife treatment is ultimately dependent on the neurosurgeon’s ability to visualize the intended target. Thus, the technique would be impossible without the advent of technology that allows three-dimensional views of anatomical structures in the brain, including tomographic modalities such as MR and CT, as well as bi-plane angiography. Each modality has advantages and disadvantages in particular situations as discussed below. Regardless of modality, all imaging proceeds using a stereotactic fiducial system. This consists of a box which attaches securely to the Leksell frame during imaging. Fiducial markers built into the box are made to be visible in the resulting image set and allow the Gamma Knife treatment planning system to calculate positions in the images relative to Leksell space. Patient immobilization during imaging is achieved by taking advantage of the rigid Leksell frame. Adapters designed for each modality attach to both the frame and the patient table and

are sufficient to prevent unwanted movement during imaging.

MR In the majority of centers, MR is the most used treatment modality because of its superior visualization of soft tissue structures and solid tumors. Typical MR protocols include T1-weighted preand post-contrast (Gadolinium-enhanced) images through the entire volume of the head. Sequences may be a collection of 2D image slices, or a true 3D acquisition such as the MP-RAGE [11] or its successors. Specialized sequences such as constructive interference in steady state (CISS) [12–14] protocols may be used for circumstances such as visualization of the internal auditory canals, the cerebellopontine angle, and parasellar regions. Current research is also exploring the use of more exotic MR protocols, such as MR fiber tractography using diffusion-tensor imaging, to better visualize critical brains structures not easily identified on standard pulse sequences [15].

907

908

58


MR images should be specified to have a field of view large enough for the fiducial markers to be clearly visible. Sequences should have a thin (1–3 mm) slice thickness with no gap between slices to ensure complete coverage of the head, increase the probability of detection of small lesions [16,17], and maximize the accuracy of volumetric measurements [18] and dose volume histograms. The MRI fiducial system consists of a plastic box which attaches directly to the stereotactic frame (> Figure 58-12a and b). Machined into

the box are channels with a distinctive ‘‘N’’ shape. The channels can be filled with a cupric sulfate solution which will enhance in commonly used pulse sequences. When resampled as axial and coronal sections, the fiducial marks appear as a series of dots in the image. By calculating the relative distances between the dots the treatment planning system can determine the 3D location of any point with respect to stereotactic space. It should be noted that MRI is susceptible to a variety of linear and nonlinear geometric distortion, primarily from gradient field nonlinearity

. Figure 58-12 (a) The MR fiducial box used with the Gamma Knife. (b) The resulting fiducial indicators on an MRI


and magnetic field inhomogeneities, including from the patient and frame [19]. Thus, all MR protocols used for radiosurgery should be validated by a qualified physicist to ensure minimal distortion. In cases where a target is close to an airtissue interface or some other area of quickly varying magnetic susceptibility one may consider using CT to help detect any distortion. There exist a variety of commercially available imaging phantoms which can help physicists detect and quantify the amount of distortion for a pulse sequence under a particular set of conditions, and can help minimize any risk from MR distortion.

CT Because of its relatively poor soft tissue visualization, CT imaging is a less-commonly used imaging modality. However in cases where MR is contraindicated or in cases where bony anatomy may provide useful information CT imaging is still quite useful. Gamma Knife CT protocols include pre- and post-contrast imaging with thin (1–3 mm) slice thickness and no gap between slices. Techniques such as CT cisternography can be useful in situations such as trigeminal neuralgia where structures in the CSF must be detected but when MR is not an available option [20]. CT

58

does not suffer from distortion in the same way as MR however the pins attaching the stereotactic frame to the patient cause artifacts which can obscure lesions proximal to the pin sites.

XA For vascular lesions such as arteriovenous malformations, digital subtraction angiography remains the imaging modality of choice. Gadoliniumenhanced MRI is sometimes used to correlate the extent of the AVM nidus with angiography, and there is current research investigating the use of MRA for nidus definition [21]. As with the MR and CT, images are acquired using a fiducial system (> Figure 58-13a and b), however the DSA system is based on projections rather than tomographic information. Images from some DSA systems must be geometricallycorrected to account for the curvature of the imageintensifierscreenbeforeimportingtheimages into the treatment planning system [22].

Skull Measurements The Gamma Knife treatment planning system requires the depth of the target point in each

. Figure 58-13 (a) The fiducial box used for digital subtraction angiography. (b) The resulting fiducial markers on an AP projection. Similar marks are found on the lateral projections

909

910

58


. Figure 58-14 ‘‘Bubble’’ device used to measure the dimensions of the patient’s skull. Skull measurements are used to calculate beam attenuation through the head and detect possible collisions when positioning the patient

beam direction in order to calculate photon attenuation and avoid collisions between the patient and the unit. GammaPlan accomplishes this using a skull scaling instrument (> Figure 58-14) to collect sampled skull measurement data, which is then interpolated and used to construct the required skull and frame model. A member of the treatment team uses a simple ruler to measure the distance from the scaling ‘‘bubble’’ to the patient’s head and records this information in the treatment planning system.

tolerable for the patient. As treatment planning is such an integral part of the Gamma Knife technique, it is explained in more detail in later sections.

Treatment Following the creation, approval, and export of a valid treatment plan to the Gamma Knife unit, the irradiation portion of the Gamma surgery may commence. > Figure 58-15 illustrates the steps in the procedure.

Planning Treatment planning is the process of creating a dose distribution that conformally treats the intended target. Treatment planning is an iterative technique requiring detailed knowledge of neuroanatomy, neuroradiology, the biological effect of single-fraction radiosurgery, and the compromises required to create a treatment which will be effective and at the same time

Verification of Treatment Parameters The Gamma Knife console receives a treatment planning file from the treatment planning system after the plan is approved and exported. Before treatment commences, the operator should have a printout (written directive) of the treatment which lists the vital treatment parameters such


58

. Figure 58-15 A flowchart of a Gamma Knife surgery

as patient identification, treatment plan identification, isocenter locations, collimator sizes, isocenter dwell times, etc. The operator should verify that the information on the written directive matches the information displayed on the console and in the treatment planning system before treatment commences.

previously described. If using the trunnion system, it is vital to use the utmost care in assuring that the correct coordinates have been set and all screws have been tightened. In the APS and PPS systems, the correct angle should be set (sensors on the unit will trigger a warning if this is not the case).

Docking the Patient

Clearance Checks

This step involves attaching the patient’s stereotactic frame to the treatment unit using one of the methods (trunnion, APS, frame adapter)

Moving a patient so an isocenter with a location far from the center of the head is at the focus point can cause the opposite side of the head or

911

912

58


. Figure 58-16 The clearance check tool used with the Perfexion. The arm of the tool rotates around a volume equal to the treatment cavity. When the patient is positioned relative to the clearance tool, a full rotation of the arm of the tool will detect and frame or skull collisions

stereotactic frame to collide with the treatment unit. In most cases the treatment planning system warns the user of possible collisions based on the acquired skull measurement data, and when using the APS the system tests all questionable locations to ensure they will avoid collisions. In cases with many collision checks, this can add significantly to the total treatment time for a patient. In the Perfexion, clearance checks are an infrequent problem; however they still can and do occur. The Perfexion has a specially designed clearance check tool (> Figure 58-16) which simulates the achievable treatment volume and simplifies the process of checking for collisions. The Perfexion also makes use of a ‘‘frame cap’’ (> Figure 58-17) which conservatively models a human scalp and is of dimensions known by the treatment planning system. The frame cap is placed over the stereotactic frame and allows the system to eliminate the possibility of collisions in many instances, reducing the required number of skull and frame measurements. Finally, the Perfexion has a collision sensor in the form of

. Figure 58-17 The Perfexion’s frame cap fits over the patient’s head and frame. It is of a known geometry, and assists the system in detecting collisions

an aluminum cap covering the treatment cavity. Pressure exerted on this cap will trigger the system to retract the sources to their shielded home position to minimize radiation exposure and permit manual removal of the patient from the unit.


Treatment Execution After the patient is comfortably situated and all checks are complete, the operator may commence treatment. The console displays data to allow the operator to keep track of treatment progress, and video and two-way audio surveillance of the patient allows the operator to react to any problems (medical or technical) that may arise. Following treatment, the surgeon removes the stereotactic frame from the patient’s head (often in the Gamma Knife suite), and the patient is either discharged or delivered to a room for overnight observation.

Treatment Planning GammaPlan (Elekta AB, Stockholm), the Gamma Knife’s treatment planning system, is a proprietary system specific to the Gamma Knife, making it somewhat different from linac-based delivery systems which often make use of generalized planning systems such as ADAC/Pinnacle [3] (Phillips Healthcare, Andover, MA). However, the decision to develop a specialized planning system is driven in part by the unique physics and specialized indications of Gamma Knife surgery, and likely simplifies the overall treatment planning process.

General Philosophy GammaPlan achieves the dual goals of allowing plans to be created in a reasonably short period of time and maintaining the look and feel of ‘‘surgery.’’ Gamma Knife surgery proceeds immediately following frame placement as soon as an approved plan can be generated. This is unlike traditional radiation therapy where days or even weeks can elapse between simulation and treatment. Because the patient is literally waiting for treatment with a frame on his head, short planning

58

times are important. Designing the planning system to have a ‘‘surgical’’ feel allows the neurosurgeon to approach the target from the perspective where he is most comfortable; that of a surgery, albeit a surgery on computer.

Basic Planning Process GammaPlan is expressly a forward-planning system. This means that the surgeon is responsible for defining and optimizing the plan. This is in contrast to inverse planning systems (commonly used with IMRT systems), which begins with a series of user-defined dose-volume constraints and the computer uses one of a variety of mathematical optimization algorithms in attempt to find a solution that best matches those constraints.2 Gamma Knife treatment planning proceeds in a series of steps as illustrated in > Figure 58-18.

Image Registration Image registration is the procedure in which the treatment planning system matches the fiducials visible in each image slice to an internal model of the fiducial system and stereotactic space. The system will report calculated deviations from the internal model so the user can be informed as to whether there are any gross distortions or other problems in the images.

Defining Targets and Critical Structures Following image registration, the next step in treatment planning is to define targets and critical 2

A reader interested in inverse planning for the Gamma knife may find references [23–25] useful. These methods have met little commercial acceptance in the Gamma Knife community.

913

914

58


. Figure 58-18 Procedures for treatment planning

structures. Using tools built into the treatment planning system, targets are outlined on a sliceby-slice basis. Critical structures are defined in a similar manner and may be marked as avoidance structures. Gamma Knife radiosurgery has traditionally followed a surgical approach to defining targets with the target outline matching as closely as possible to the visible target on the imagery. This is a departure from the approach in radiotherapy, where a gross target volume (GTV) is often expanded in steps to create a clinical target

volume (CTV) and planning target volume (PTV) to account for subclinical disease and setup error during the treatment [26].

Dose Matrices/targets The Gamma Knife treatment planning system makes use of one or more dose matrices (termed ‘‘targets’’ in recent releases of GammaPlan). A dose matrix is a calculation matrix with a fixed number (31 31 31) of sampling points. The


58

. Figure 58-19 A typical dose distribution (yellow line) is a superposition of a number of individual isocenters, or shots (red circles)

dose matrix defines an a volume over which GammaPlan will calculate dose and display isodose curves and also provides a method for prescribing a dose for a group of isocenters. Because the dose matrix has a fixed number of sampling points, the size of the matrix should be kept to a reasonable minimum that allows good visualization of the dose distribution and minimizes sampling error.

Defining a Treatment Field The goal of Gamma Knife treatment planning is to devise a plan that achieves complete coverage of the target at the desired dose while minimizing

dose to normal tissue. The neurosurgeon creates the dose distribution by defining one or more isocenters (commonly known as ‘‘shots’’) at locations within the volume of the target so that the prescription isosurface conformally matches the target (> Figure 58-19). Each isocenter defines a location to which the Gamma Knife must move the patient so as to place the isocenter in the radiation focus of the unit. Multiple isocenters allow the surgeon to create dose distributions with irregular shapes. In the model B and C gamma knives, the surgeon has a choice of four sizes of isocenters corresponding to

3

Note that some institutions use a superposition method to achieve isocenters with intermediate sizes [27].

915

916

58


the four collimator helmet sizes3. In the Perfexion model Gamma Knife the surgeon can vary the collimator size of the isocenter by sector. The surgeon may also vary parameters such as gamma angle, the weight of the isocenter relative to all other isocenters, the prescription isodose line, and the prescription dose. Shielding

It is frequently the case that the lesions targeted for Gamma Knife surgery are present close to critical structures such as the optic pathways, cranial nerves, etc. In these cases, the neurosurgeon may choose to protect these structures through the application of beam-blocking patterns that minimize the contribution of beams which intersect those structures. The practical effect of shielding is to compress the dose gradient in a particular direction, and shift the dose to a different direction. In the model B and C models it is possible to block individual beams by replacing collimators in the helmets with solid ‘‘plugs.’’ The ‘‘plugging patterns’’ as they are commonly called, are generated using the planning software using a beam-intersection method and refined by the neurosurgeon (> Figure 58-20a) [28]. With the Perfexion, manual plugging of individual beams has been replaced by the ability to automatically shield any combination of collimator sectors, so all beams in a shielded sector are blocked. Any contoured structure may be defined as an avoidance structure. The system will then compute for any combination of isocenters the beams which intersect the avoidance structure(s). If the number of intersecting beams in any sector exceeds a given threshold, the sector is blocked (> Figure 58-20b). Because the Perfexion’s collimation system is fully automatic, the burden on the treatment team when shielding has been eliminated. However, overuse of shielding can result in extended treatment times due to the reduced dose rate as the number of unblocked sources decreases.

Plan Evaluation Gamma Knife treatment plans are evaluated based on coverage of the target, conformity of the treatment field to the target, dose to normal and critical tissue, and the time and effort involved in executing the treatment. Treatment coverage and conformity are evaluated visually, using dose volume histograms, and often using one of a number of proposed indicies of plan conformity [29–32]. The conformity index attempts to quantitatively measure the degree to which the target is covered at the level of the prescribed dose and the degree to which normal tissue is included in the treatment field. In a perfectly conformal plan, the target would be completely covered by the prescribed radiation dose and there would be no spillover of dose outside of the target. Conformity evaluates a treatment plan at a particular dose. It does not evaluate the dose falloff and the irradiation of normal tissue at low doses. To address this, Paddick et al have devised the Gradient Index as a metric to gauge the dose falloff of a treatment plan below the prescription isodose [33]. Wagner, et al. attempt to combine the conformity index and gradient index ideas into a single metric they term the Confomity/Gradient (CGI) Index [34]. Several authors have also devised dose-volume constraints similar to those used in traditional radiation therapy such as the 12 Gy volume as a predictor for radiation necrosis following gamma surgery [35].

Treatment Planning – Tricks of the Trade Gamma Knife treatment planning is as much an art as a science and one of the drawbacks of forward-planning is that the end result depends greatly on the skill and experience of the individual performing the dose planning. In this section


58

. Figure 58-20 (a) Plugging. Individual isocenters may replace collimators with ‘‘plugs’’ in order to protect critical structures, in this case the anterior optic pathways in a treatment plan for a pituitary lesion. (b) Plugging with the Perfexion. The optic nerves are defined as avoidance structures. The hatched sector in the sector configuration diagram represents a sector which is to be shielded. The resulting isodose distribution can spare the optic pathways a high dose

917

918

58


we present some ‘‘tricks of the trade’’ the reader may find useful: 1.

2.

3.

4.

5.

6.

Targets should be filled with larger shots near the center and smaller shots on the periphery in order to maximize dose falloff outside of the target. Some reports suggest that placing shots along the central axis of a target is a useful first approximation [24]. Shots which are located significantly outside the boundary of the target should be avoided. While this technique can in some cases lead to a highly conformal plan at the prescription isodose line, it generally leads to poor dose falloff and a larger volume of normal tissue contained within lower isodose regions. When placing a shot, it is difficult to visualize where the shot fits in a plan without information for all three orthogonal directions. Therefore, it is often useful to plan using a workspace view that includes axial, coronal, and sagittal cuts through the brain. While Gamma Knife treatments are traditionally prescribed at the 50% isodose line, this does not always lead to optimal conformity or dose gradient [33]. Careful consideration should be used when choosing the appropriate isodose line in a given situation. Shots which are clustered together can cause hot spots in the plan and make the isodose distribution contract. This can be used as an advantage in treatment planning. Monitoring the hot spots (for instance, by displaying the 95% isodose line) can make it possible to predict when the prescription isodose line will expand or contract. Placing a small shot directly on a hot spot can provide a ‘‘control’’ shot that can be used intentionally to assist in planning by modulating the weight of the ‘‘control.’’ When planning parasellar tumors, changing the gamma angle to a lower angle may help align the major axis of the treatment plan

7.

8.

9.

to be parallel to the anterior optic pathways. This can help lower the dose to these critical structures and minimize the need for shielding. Because dose matrices in GammaPlan have a fixed number of sampling points, the dose matrix size should be kept to a reasonable minimum. Small-sized shots with large dose matrix sizes can lead to significant calculation error due to inadequate sampling frequency. Combinations of very large and very small shots (such as 18 and 4 mm) should be avoided. The difference in scale between these sizes is such that the smaller shot will have little significant effect on the resulting dose distribution. The B and C models ship with 100 physical plugs, but GammaPlan permits up to 166 plugs in a treatment plan. Taking advantage of this makes it easier to generate complicated plugging patterns. However, this technique should be used with caution so a plan is not approved with more than the physically-present number of plugs.

Gamma Knife Dosimetry From a dosimetric perspective, the basic physics the Gamma Knife treatment planning system needs to account for are the inverse square law, photon attenuation through tissue, reduction in dose rate due to collimation, and dose falloff away from the central axis of the beam. The algorithm used in GammaPlan makes the assumption that the brain is a homogenous, water-equivalent volume of tissue and that there is no buildup region near the surface. This makes the algorithm relatively simple from a physics standpoint, however from a geometric perspective it is quite complex as the dose


contribution from each of the 60Co sources must be computed. In the earlier Gamma Knife models (models B, C, and 4C), much of the beam data used in the dose calculation model was acquired empirically. While the general algorithm remains the same for all gamma knife models, the design of the Perfexion makes the details of the algorithm somewhat more complex, and in this new machine both empirical and monte-carlo methods were used to create the dose calculation model. Some of these differences created by the new Perfexion will be discussed at the end of this section.

Single-beam Model In the B, C, and 4C models of the Gamma Knife, the sources are within an approximation equal in activity, have identical collimation (for a given collimator helmet), and have identical 400 mm source to focus distances. Therefore for dosimetric purposes, each of the 201 beams may be considered to be identical.

58

To characterize any individual beam, a single source and beam channel was constructed to the specification used with the actual Gamma Knife unit (> Figure 58-21). From this single beam channel, the radial dose distribution and percent depth doses for a single beam were measured for each collimator size. From the percent depth doses an attenuation coefficient can be derived for the brain as well as output factors for each collimator size normalized to the 18 mm collimator.

Superposition Algorithm From the characterization of an individual beam and the assumption that all 201 beams are identical, a superposition of beam data leads to the final dose calculation model. The only difference between beams is the path each beam takes through the head. Thus, the problem becomes one of finding the depth each beam must traverse to reach the focal point so that the contribution of each beam may be appropriately attenuated.

. Figure 58-21 The single-beam measurement system used to characterize an individual beam channel (image courtesy of Elekta AB, Stockholm)

919

920

58


This problem is solved by interpolating the entrance point of the beam with the earlier described skull measurements via an iterative search algorithm. Once the depth is known, the beam may be adjusted by a constant exponential tissue attenuation factor and the dose rate contribution of the beam may be computed. The sum over all such beams leads to the dose rate at the focus point. To create a three-dimensional dose distribution at locations away from the isocenter, the radial dose distribution of each beam is used. For any point, the radial distribution and the attenuation coefficient allows computation of the difference in dose from the central axis of each beam (> Figure 58-22).

Dose Normalization Dose distributions for the Gamma Knife are traditionally normalized to the maximum dose point in the treatment plan. In treatment plans with multiple targets, isodose curves can be displayed relative to the global maximum dose point or to the local maximum dose point for

. Figure 58-22 The geometry used to determine the dose contribution of a single beam. To calculate the dose contribution at any point P from Beami, the depth and off-axis distance from the central beam must be computed. Note: df 0 if P is downstream

each target. Note that in situations where the surgeon intends to prescribe different doses to each target or to prescribe to different isodose lines for each target it is important to be aware of whether one is looking at a local set of isodose curves or curves normalized to the global plan maximum.

Differences with the Gamma Knife Perfexion The Perfexion departs from the idea that every beam may be treated identically. Because the Perfexion’s collimator body is conical, not hemispheric, each row of sources has a different source to focus distance ranging from 374 to 433 mm. In addition, in some rows the sources are placed at an angle to the beam channels, resulting in asymmetric treatment fields (> Figure 58-23a and b). The design changes mean that parameters such as output factors, attenuation factors, and source to focus distances are no longer constant, but instead vary by both collimator size and beam channel location. In addition, because the sources are angled to the beam channels the off-axis distributions must be stored as a two-dimensional table rather than a one dimensional table. The dose calculation model for the Perfexion is therefore more complex than its predecessors; however the design allows more complex isocenter distributions to be created. One example is an isocenter which is approximately ‘‘square’’ in the axial plane (> Figure 58-24).

Commissioning and Quality Assurance A robust quality assurance program is vital to ensure that the Gamma Knife is functioning within all radiological and mechanical tolerances and that all electromechanical and computerized


58

. Figure 58-23 (a) A cutaway of the beam channels in the Perfexion demonstrating how the sources are angled relative to the beam channels. (b) The resulting asymmetric beam at the focal point makes dose computation for the Perfexion more complex than in previous models (images courtesy of Elekta AB, Stockholm)

systems are functioning properly. > Table 58-3 summarizes some of the primary aspects of the system that are a part of any well-designed quality assurance program for the Gamma Knife. The remainder of this section will highlight some of the tools that ship with the Gamma Knife to assist in the quality assurance process.

Dose Rate (output) Tests Output tests for the Gamma Knife are conducted with a chamber/electrometer setup within an 80-mm spherical polystyrene phantom (> Figure 58-25) that ships with the Gamma Knife unit. The unit is calibrated to the largest

921

922

58


. Figure 58-24 A ‘‘square’’ dose distribution created using a composite shot with the Gamma Knife Perfexion

available collimator size (18 mm for the B, C, and 4C; 16 mm for the Perfexion). Field-sized dependent output factors are a standard quality assurance item in linearaccelerator systems; however for the Gamma Knife this is difficult due to the focal spot size. In the earlier models it is possible to verify the manufacturer-recommended output factors for the larger collimator sizes [36–38]. However, the small focal spot of the 4 mm collimator has been difficult to verify in a clinical setting. This becomes even more difficult with the Gamma Knife Perfexion as the output factors for individual beams are dependent on collimator position in the central body. However, unlike with the collimator jaws of a linac, the collimator openings of the Gamma Knife do not change size, so periodic output factor verification may be less of an issue.

Radiological Focus Point and Patient Positioning System Calibration In the Perfexion, the movements of the Patient Positioning System (PPS) and the position of the radiological focus point (RFP) are calibrated separately at installation. The movements of the PPS are verified for linearity using a laser interferometer. The mechanical alignment of the collimator beam channels are determined mechanically for the outer 4 mm collimators, and the relative locations of the remaining beam channels are known to a precise mechanical tolerance. Once the PPS and radiation body are mated at installation, it remains to calibrate the PPS versus the RFP. During installation and during interval quality assurance this calibration is accomplished

58


. Table 58-3 Components of a Gamma Knife QA program (note that individual centers may include more than appears on this list) System functionality Dose rate (output) Radial isocenter profile Focus point coincidence Patient positioning accuracy Timers Safety interlocks Treatment planning system Backup power supply

Model B, C, and 4C

Perfexion

Ionization chamber/ electrometer and water-equivalent phantom Film

Ionization chamber/ electrometer and waterequivalent phantom Film

Film

Film + diode tool

Trunnion centricity tool; APS QA tool Chamber/electrometer and independent timer Manual testing Independent dose calculation for a given test setup Power supply self test

Recommended tolerance

Frequency

3% of predicted dose

Monthly

1 mm of predicted profile at 50% isodose 0.4 mm

Yearly

Diode tool + clearance check tool

0.4 mm

Monthly

Chamber/electrometer and independent timer Manual testing Independent dose calculation for a given test setup Power supply self test

0.1 min or 10%

Monthly

No failures 1%

Daily Weekly

No failures

Monthly

Yearly

. Figure 58-25 Electrometer/ionization chamber and Elekta 16 cm diameter polystyrene phantom used for output tests

through the use of a specially-designed diode tool as well as radiochromic film such as GafChromic1 EBTor MD-55 (International Specialty Products, USA) films [39].

The diode tool consists of a diode mounted on a rigid metal frame engineered to dock to the PPS. The diode tool scans along a predefined volumetric space and searches for the maximum

923

924

58


. Figure 58-26 (a) The precision diode tool used to test radiological accuracy. (b) The film holder used for verifying beam profiles and positional accuracy

dose point as well as the penumbra regions of the profile for the 4 mm collimator. It compares these measurements against the stored installation/calibration information and reports an error if it is out of range with an accuracy of 0.4 mm and a repeatability of Figure 58-26a). The film tool is a special adapter which can hold a piece of radiochromic film. A pinhole is punched at the expected location of the

RFP. After irradiation, the exposed relative dose distribution can be analyzed and an RMS error determined (> Figure 58-26b).

Isocenter Profiles Isocenter profiles for the Gamma Knife are measured using the same film holder as in the previous tests; however with the film no pinprick


is required. The film dose distributions can be measured using optical density techniques and compared to the distributions predicted by the treatment planning system.

Future QA Radiosurgery in general may benefit for the ability to capture a true three-dimensional dose distribution for use in quality assurance. One promising area of research uses polymer-gel dosimetry in which a phantom filled with gel is irradiated with a specified dose distribution [40–44]. The gel polymerizes in proportion to the absorbed dose, and can be scanned with MRI or optical CT to obtain a 3D dose distribution. Questions regarding the accuracy and resolution of this technique have so far prevented its widespread adoption.

Summary of Features Specific to the Gamma Knife Perfexion Treatment Volume The elimination of the external collimation helmets required for the previous models opens up a much larger potential treatment volume in the Perfexion Gamma Knife. > Table 58-4 shows the differences in reachable range on the Perfexion

. Table 58-4 The treatment ranges of the trunnion system, model C APS, and Perfexion PPS Dimensions

Perfexion

Trunnion

APS

X

160 mm

100 mm

Y

180 mm

150 mm

Z

260 mm (distance from focus point to inner collimator surface)

125 mm

82 mm 120 mm 153 mm

58

versus the Trunnion and APS systems of the models B and C. This means that many cases which could not be optimally reached in one treatment or that required changes of gamma angle during the treatment in order to avoid collisions will now in most cases be easily treated in the new unit. > Figure 58-27 illustrates an example of a patient with multiple meningiomas widely distributed in the brain. In previous Gamma Knife units multiple treatment sessions would be required to reach all of the targets. With the Perfexion unit, these targets can be reached in a single session with no collision checks and no gamma angle changes, greatly reducing the treatment burden for the patient. In the infrequent situation where a potential collision is detected with the Perfexion, the previously described clearance tool that ships with the Gamma Knife reduces the effort required to control whether or not a collision will occur.

. Figure 58-27 Patient with multiple meningiomas distributed in both cerebral hemispheres. All lesions were successfully treated in a single Gamma Knife session on the Perfexion with no collision checks or angle changes

925

926

58


Composite Shots and Automated Shielding With the sector-based collimation system, any individual sector can be configured as any of the three available collimator sizes or to a blocked position. This means it is now possible to create composite isocenters which simultaneously make use of different collimator sizes. The promise of composite shots is that each isocenter can be more carefully tailored to match the shape of the target. The potential drawback is that the automation afforded by the Perfexion will make it tempting to increase the use of smaller collimators and shielding even in cases where larger collimators would suffice. The consequential increase in beam time will partly offset the gain in efficiency.

Automated Shielding With the Perfexion unit, the manual plugging of the older units becomes a fully automated process. In addition, the Perfexion make it possible to shield more (168 of 192) of the sources than was possible in the previous Gamma Knife models (100 of 201). This lowers the burden to the use of shielding and makes possible isodose distributions that were not conventionally possible with previous models. However, the granularity to which beam channels can be blocked has been reduced, as it is now possible to block only at the granularity of a sector. The ability to create annular plug patterns has been eliminated, which may limit the ability to shield critical structures superior to the target. Finally, the increased use of shielding may lengthen beam-on treatment time as there are fewer beams available to deliver the prescribed dose.

Reduced Shuttle Time The Perfexion significantly reduces the positioning time required between shots. In the Gamma

Knife models with no APS system, every isocenter requires stopping the treatment, manually setting coordinates, and restarting the treatment; a timeconsuming process. The APS automatically moves the patient from location to location; however in this system the patient is withdrawn from the focus of the Gamma Knife (the defocus position) while the coordinates are changing. The defocus/ reposition/refocus procedure takes 30 s to complete. Thus, with the APS system, there is significant time overhead for every isocenter added. The Perfexion’s PPS moves the patient at speeds up to 10 times (7 mm/s vs. 0.7 mm/s) faster than the APS. No defocusing of the patient is required as the sources move to a blocked position during repositioning. The result is that a much larger proportion of the Gamma Knife treatment is beam-on time with the Perfexion, reducing the overall treatment time for the patient.

Improved Patient Comfort The Gamma Knife Perfexion purports to significantly increase patient comfort. We have previously described the redesigned docking mechanism for the patient’s head and stereotactic frame and its implications for patient comfort. To further patient comfort, the Perfexion ships with a much thicker mattress and a treatment table that better supports the upper back and shoulders.

Lower Extracranial Dose The internal shielding and the beam channel directions of the Perfexion greatly reduce the extracranial dose patients will receive. Lindquist, et al. report doses on anthropomorphic phantoms which are up to 10 times lower than those reported for the B and C models [45]. The manufacturer claims an average dose of 4 mSv/h at a distance of 60 cm from the side surfaces of the unit. This is important for the safety of the device and the shielding required when designing treatment suites.


58

Future Possibilities

References

The elimination of the external collimation helmets opens up a much larger potential treatment volume in the Perfexion model. This means that many cases which could not be optimally reached in one treatment or that required awkward positional changes during the treatment will now in most cases be easily treated in the new unit. In addition, the Perfexion has the potential ability to treat indications down into the cervical spine (as low as 26 cm caudal from the vertex of the cranium [45], opening up the possibility of treating head and neck carcinomas and other pathologies. While this will require changes in fixation techniques and dose calculation algorithms [46], the new unit promises to significantly expand the pool of potential indications for the Gamma Knife.

1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102(4):316-19. 2. Lindquist C, Kihlstrom L. Department of Neurosurgery, Karolinska Institute: 60 years. Neurosurgery 1996;39(5): 1016-21. 3. Wu A, Lindner G, Maitz AH, et al. Physics of gamma knife approach on convergent beams in stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1990;18(4):941-9. 4. Khan FM. The physics of radiation therapy. Baltimore: Williams and Wilkins; 1984. 5. Hall EJ. Radiobiology for the radiologist. 3rd ed. London: J. B. Lippincott; 1988. 6. Buatti JM, Friedman WA, Meeks SL, Bova FJ. The radiobiology of radiosurgery and stereotactic radiotherapy. Med Dosim;1998;23(3):201-7. 7. Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993;25(2): 381-5. 8. Leksell GammaPlan 8.0 online reference manual. 1003197 Rev. 01 ed. Stockholm: Elekta Instrument AB; 2006. 9. Leksell gamma knife perfexion: system description. Art no 1002703. Stockholm: Elekta, AB; 2006. 10. Leksell L, Lindquist C, Adler JR, Leksell D, Jernberg B, Steiner L. A new fixation device for the Leksell stereotaxic system. Technical note. J Neurosurg 1987;66(4): 626-9. 11. Mugler JP III, Brookeman JR. Three-dimensional magnetization-prepared rapid gradient-echo imaging (3D MP RAGE). Magn Reson Med 1990;15(1):152-7. 12. Casselman JW, Kuhweide R, Deimling M, Ampe W, Dehaene I, Meeus L. Constructive interference in steady state-3DFT MR imaging of the inner ear and cerebellopontine angle. AJNR Am J Neuroradiol 1993;14(1): 47-57. 13. Held P, Fellner C, Fellner F, Seitz J, Strutz J. MRI of inner ear anatomy using 3D MP-RAGE and 3D CISS sequences. Br J Radiol 1997;70(833):465-72. 14. Stuckey SL, Harris AJ, Mannolini SM. Detection of acoustic schwannoma: use of constructive interference in the steady state three-dimensional MR. AJNR Am J Neuroradiol 1996;17(7):1219-25. 15. Hlatky R, Jackson EF, Weinberg JS, McCutcheon IE. Intraoperative neuronavigation using diffusion tensor MR tractography for the resection of a deep tumor adjacent to the corticospinal tract. Stereotact Funct Neurosurg 2005;83(5-6):228-32. 16. Litt AW, Kondo N, Bannon KR, Kricheff II. Role of slice thickness in MR imaging of the internal auditory canal. J Comput Assist Tomogr 1990;14(5):717-20. 17. Johnson CD, Fletcher JG, MacCarty RL, et al. Effect of slice thickness and primary 2D versus 3D virtual dissection on colorectal lesion detection at CT colonography in

Conclusions Lars Leksell conceived ‘‘radiosurgery’’ as a minimally-invasive technique for neurosurgeons. It is only very recently that radiosurgery has become a viable extracranial technique, and it is possible only because of tremendous advances in computing power, imaging technology, and manufacturing precision. Leksell developed his device without these advances, and its basic design remains basically unchanged to this day. It is a tribute to the elegant design of the Gamma Knife and its specificity to the problems of brain surgery that the Gamma Knife remains the ‘‘gold standard’’ by which new devices are compared. Over the coming decades it will be interesting to observe what changes new technology will bring to the field of radiosurgery. The authors hope that the accrued experience of the Gamma Knife will help temper what is often a rush to push new technology as a panacea to every clinical problem and will help clarify the appropriate role of radiosurgery both inside and outside of the brain.

927

928

58 18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.


452 asymptomatic adults. AJR Am J Roentgenol 2007; 189(3):672-80. Snell JW, Sheehan J, Stroila M, Steiner L. Assessment of imaging studies used with radiosurgery: a volumetric algorithm and an estimation of its error. Technical note. J Neurosurg 2006;104(1):157-62. Sumanaweera TS, Adler JR Jr, Napel S, Glover GH. Characterization of spatial distortion in magnetic resonance imaging and its implications for stereotactic surgery. Neurosurgery 1994;35(4):696-703; discussion 694-703. Worthington C, Hutson K, Boulware R, et al. Computerized tomography cisternography of the trigeminal nerve for stereotactic radiosurgery. Case report. J Neurosurg 2000;93 Suppl 3:169-71. Bednarz G, Downes B, Werner-Wasik M, Rosenwasser RH. Combining stereotactic angiography and 3D time-offlight magnetic resonance angiography in treatment planning for arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 2000;46(5):1149-54. Soderman M, Picard C, Ericson K. An algorithm for correction of distortion in stereotaxic digital subtraction angiography. Neuroradiology 1998;40(5):277-82. Shepard DM, Ferris MC, Ove R, Ma L. Inverse treatment planning for Gamma Knife radiosurgery. Med Phys 2000;27(12):2748-56. Wu QJ, Chankong V, Jitprapaikulsarn S, et al. Real-time inverse planning for Gamma Knife radiosurgery. Med Phys 2003;30(11):2988-95. Zhang P, Wu J, Dean D, et al. Plug pattern optimization for gamma knife radiosurgery treatment planning. Int J Radiat Oncol Biol Phys 2003;55(2):420-7. Morgan-Fletcher SL. Prescribing, recording and reporting photon beam therapy (Supplement to ICRU Report 50), ICRU Report 62. Br J Radiol 2001; 74(879):294. Thorsen FA, Ganz JC. Dose planning with the Leksell Gamma Knife: the effect on dose volume of more than one shot at the same target point. Stereotact Funct Neurosurg 1993;61 Suppl 1:151-63. Schlesinger D, Snell J, Sheehan J. Shielding strategies for Gamma Knife surgery of pituitary adenomas. J Neurosurg 2006;105(7):241-8. Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. Technical note. J Neurosurg 2000;93 Suppl 3:219-22. Borden JA, Mahajan A, Tsai JS. A quality factor to compare the dosimetry of gamma knife radiosurgery and intensity-modulated radiation therapy quantitatively as a function of target volume and shape. Technical note. J Neurosurg 2000;93 Suppl 3:228-32. Lomax NJ, Scheib SG. Quantifying the degree of conformity in radiosurgery treatment planning. Int J Radiat Oncol Biol Phys 2003;55(5):1409-19.

32. Wu QR, Wessels BW, Einstein DB, Maciunas RJ, Kim EY, Kinsella TJ. Quality of coverage: conformity measures for stereotactic radiosurgery. J Appl Clin Med Phys 2003;4(4):374-81. 33. Paddick I, Lippitz B. A simple dose gradient measurement tool to complement the conformity index. J Neurosurg 2006;105(7):194-201. 34. Wagner TH, Bova FJ, Friedman WA, Buatti JM, Bouchet LG, Meeks SL. A simple and reliable index for scoring rival stereotactic radiosurgery plans. Int J Radiat Oncol Biol Phys 2003;57(4):1141-9. 35. Korytko T, Radivoyevitch T, Colussi V, et al. 12 Gy gamma knife radiosurgical volume is a predictor for radiation necrosis in non-AVM intracranial tumors. Int J Radiat Oncol Biol Phys 2006;64(2):419-24. 36. Ma L, Li XA, Yu CX. An efficient method of measuring the 4 mm helmet output factor for the Gamma knife. Phys Med Biol 2000;45(3):729-33. 37. Bilge H, Osen Z, Senkesen O, Kucucuk H, Cakir A, Sengoz M. Determination of output factors for the Leksell gamma knife using ion chamber, thermoluminescence detectors and films. J BUON 2006;11(2):223-7. 38. Cheung JY, Yu KN, Ho RT, Yu CP. Monte Carlo calculated output factors of a Leksell Gamma Knife unit. Phys Med Biol 1999;44(12):N247-N249. 39. Sanders M, Sayeg J, Coffey C, Patel P, Walsh J. Beam profile analysis using GafChromic films. Stereotact Funct Neurosurg 1993;61 Suppl 1:124-9. 40. Maryanski MJ, Ibbott GS, Eastman P, Schulz RJ, Gore JC. Radiation therapy dosimetry using magnetic resonance imaging of polymer gels. Med Phys 1996;23(5): 699-705. 41. Scheib S, Crescenti R, Vogelsanger W, et al. Application of normoxic polymer gels in 3D-dosimetry for radiosurgery. Z Med Phys 2006;16(3):180-7. 42. Watanabe Y, Akimitsu T, Hirokawa Y, Mooij RB, Perera GM. Evaluation of dose delivery accuracy of Gamma Knife by polymer gel dosimetry. J Appl Clin Med Phys 2005;6(3):133-42. 43. Karaiskos P, Petrokokkinos L, Tatsis E, et al. Dose verification of single shot gamma knife applications using VIPAR polymer gel and MRI. Phys Med Biol 2005; 50(6):1235-50. 44. Sandilos P, Tatsis E, Vlachos L, et al. Mechanical and dose delivery accuracy evaluation in radiosurgery using polymer gels. J Appl Clin Med Phys 2006;7(4):13-21. 45. Lindquist C, Paddick I. The Leksell Gamma Knife Perfexion and comparisons with its predecessors. Neurosurgery 2007;61 Suppl 3:130-40; discussion 131-140. 46. Solberg TD, Holly FE, De Salles AA, Wallace RE, Smathers JB. Implications of tissue heterogeneity for radiosurgery in head and neck tumors. Int J Radiat Oncol Biol Phys 1995;32(1):235-9.

75 Gamma Knife Radiosurgery: Technical Issues D. Kondziolka . A. Niranjan . J. Novotny . J. Bhatanagar . L. D. Lunsford

The Gamma Knife was developed by Lars Leksell and Borje Larsson, to achieve their goal of an efficient, precise, hospital-based stereotactic radiosurgery system [1]. Clinical work with the Gamma Knife began in 1967 and the first patient had a craniopharyngioma. The patient’s head was immobilized using a plaster-molded headpiece. Subsequently, gamma knife surgery was performed in patients with pituitary tumors, vestibular schwannomas, vascular malformations, and functional disorders such as intractable pain. In 1975, a series of surgical pioneers at the Karolinska Hospital, Stockholm began to utilize a new Gamma Knife, redesigned to create a more spheroidal dose-profile better suited for the treatment of intracranial tumors and vascular malformations. Units 3 and 4 were placed in Buenos Aires and Sheffield England in the early 1980s. Lunsford et al. introduced the first clinical 201-source Gamma Knife unit to North America (the fifth gamma unit worldwide) and the first patient was managed in August 1987 at the University of Pittsburgh Medical Center [2]. The encouraging results of radiosurgery for benign tumors and vascular malformations led to an exponential rise of radiosurgery cases and installations of radiosurgical units (> Table 75-1). In recent years metastatic brain tumors have become the most common indication of radiosurgery. Brain metastases now comprise 30–50% of radiosurgery cases at busy centers.

#


The Evolution of Gamma Knife Technology Models A, B, and C There have been numerous changes to the Gamma Knife since the original 1967 design. In the first models (Model U or A) 201 Cobalt sources were arranged in a hemispheric configuration. These units presented challenging Cobalt-60 loading and reloading issues. To facilitate reloading, the unit was redesigned so that sources were arranged in a circular configuration (Model B, C, and 4C) (> Figure 75-1). Gamma Knife radiosurgery usually involves the use of single or multiple isocenters of different beam diameters to achieve a treatment plan that conforms to the 3-dimensional volume of the target. The total number of isocenters may vary depending upon the size, shape, and location of the target. Each isocenter has a set of three Cartesian (X, Y, Z) stereotactic coordinates corresponding to its location in three-dimensional space as defined using a rigidly fixed stereotactic frame. When multiple isocenters are used, the stereotactic coordinates will need to be set individually. In 1999, the Model C Gamma Knife was introduced and first installed in the United States at the University of Pittsburgh Medical Center in March 2000 [3]. This technology combined dose planning advances with robotic engineering. The unit incorporated an automatic positioning system

1226

75

Gamma knife radiosurgery: technical issues

. Table 75-1 Brain Disorders treated World-wide Using Gamma Knife Radiosurgery by December 2006

Brain disorder

Indications

Vascular disorders

AVM

Benign tumors

Malignant tumors

Functional targets

Ocular disorders

Aneurysm Cavernous malformation Other vascular Vestibular schwannoma Trigeminal schwannoma Other schwannoma Meningioma Pituitary adenoma Pineal region tumor Craniopharyngioma Hemangioblastoma Hemangiopericytoma Chordoma Glomus tumor Other benign tumors Glial tumors (grade I–II) Glial tumors (grade III–IV) Metastatic tumor Chondrosarcoma Nasopharyngeal carcinoma Other malignant tumors Trigeminal neuralgia Parkinson’s disease Pain Epilepsy Obsessive compulsive disorder Other functional targets Uveal melanoma Glaucoma Other ocular disorders Total indications

Number of patients treated 48,407 270 1,887 3,777 36,843 2,312 1,005 49,558 31,901 3,150 3,397 1,656 946 1,619 1,107 3,490 2,169 23,610 14,1210 520 1,277

(APS) with submillimetric accuracy, used to move the frame to each coordinate. This technology obviates the need to manually adjust each set of coordinates in a multiple-isocenter plan. The robot eliminates the time spent removing the patient from the helmet, setting the new coordinates for each isocenter and repositioning the patient in the helmet. This has significantly reduced the total time spent to complete the procedure and also increased accuracy and safety [4–8]. The other features of the Model C unit include an integral helmet changer, dedicated helmet installation trolleys, and color-coded collimators. In 2005 the fourth generation Leksell Gamma Knife model 4C was introduced. The first unit was installed at the University of Pittsburgh in January of 2005. The model 4C is equipped with enhancements designed to improve workflow and provide integrated imaging capabilities. The imaging enhancements available with the Leksell GammaPlan, offers image fusion capability. These images can also be exported to a CD-ROM, so the referring physician can receive pre- or post-operative images for reference and follow-up. The planning information can be viewed on both sides of the treatment couch. The helmet changer and robotic Automatic Positioning System are faster and reduce total treatment time.

5,609 25,198 1,309 566 2,243 140 867 1,354 210 65 397,672

LGK Perfexion The newest iteration of Gamma Knife technology is the PERFEXION unit. Beginning in 2002, an invited group of neurosurgeons, radiation oncologists and medical physicists was asked by the manufacturer to define specifications for a new Leksell Gamma Knife system. The group agreed on five critical features for a new system: (1) best dosimetry performance, (2) unlimited cranial reach, (3) best radiation protection for patient and stuff, (4) full automation of the treatment process,


75

. Figure 75-1 Schematic diagram of the Leksell Gamma Knife 4C

(5) patient and staff comfort, and (6) similar radiation dose profiling as prior units. The new unit was first installed in 2006, and was first used at our center in September 2007 (> Figure 75-2). The radiation unit was redesigned. A total of 60 192 Co sources were arranged in a cylindrical configuration in five concentric rings. This differs substantially from the previous hemispherical arrangements and results in different source to focus distances for each ring varying from 374 to 433 mm. The primary and secondary collimators have been replaced by a single large 120 mm thick tungsten collimator array ring (> Figure 75-3). Consequently no collimator helmets are needed for the PEREFXION system. Three collimators are available for the PERFEXION system. The 4 and 8 mm collimators remain, and a new 16 mm collimator

replaces the prior 14 and 18 mm collimators. The tungsten collimator array is subdivided into eight identical but independent sectors, each containing 72 collimators (24 collimators for 4 mm, 24 collimators for 8 mm, 24 collimators for 16 mm). The collimator size for each sector is changed automatically by moving 24 sources over the selected collimator set. Each sector with 24 sources can be moved independently into five different positions: (1) sector in home position when system is standby, (2) 4 mm collimator, (3) 8 mm collimator, (4) 16 mm collimator, and (5) sector off position (defined as the position between the 4 and 8 mm collimators providing blocking of all 24 beams for that sector) (> Figure 75-3). Sector movement is performed by servo-controlled motors with linear scales located at the rear of the radiation unit.

1227

1228

75


. Figure 75-2 Leksell Gamma Knife PERFEXION

The radiation cavity has been increased by more than 300% compared to previous models. However, due to an improved collimation system (120 mm tungsten ring), the average distance from source to focus is very close to previous models. This results in similar output for the prior 18 mm and new 16 mm administrations. The increase in the volume of the radiation cavity to more than three times allows for a greater mechanical treatment range in X/Y/Z. It is (160/180/220 mm) for the PERFEXION system compared to (100/120/ 165 mm) for other gamma knife models. This provides virtually unlimited cranial reach, so crucial in the care of patients with multiple brain metastases [9]. To date, we have seen no problems related to potential helmet collisions. The Automatic Positioning System (APS) used in the C units was replaced by the Patient Positioning System (PPS). Rather than just the head, the whole couch moves into pre-selected stereotactic coordinates. This provides better patient comfort and allows complete of the majority of radiosurgeries in one single run. Docking of the patient into the PPS is done by means of an

adaptor that attaches to the standard stereotactic Leksell G frame with three clips. The adapter is then directly docked to the PPS (> Figure 75-4). The patient can be attached in three different positions, with gamma angles of 70, 90 or 110 reflecting neck flexion or extension. The gamma angle is the only treatment parameter that requires manual set up. The PPS has repeatability better than 0.05 mm. The redesigned hardware of the PERFEXION unit has had significant impact on the planning software Leksell GammaPlan PFX (LGP PFX), a new version of the LGP running on a PC platform with the Linux operating system. There are in principle three possible approaches in the treatment planning: (1) use of classical combinations of 4, 8, and 16 mm isocenters (shots), (2) use of composite shots containing combinations of 4, 8, 16 mm or blocked sectors, and (3) dynamic shaping using blocked selected sectors to protect volumes defined as critical structures (> Figure 75-5). The most revolutionary change in the treatment planning is the ability to generate a single isocenter composed of different beam diameters. Such


75

. Figure 75-3 Diagrams of the PERFEXION Gamma Knife radiation unit and collimator system. (a) Cross section of the Leksell Gamma Knife PERFEXION radiation unit. (b) Each sector holds 24 60Co sources and can be moved independently on other sectors to the desired position to define collimator size or block groups of beams. (c) Sector position in 4 mm collimator. (d) Sector position in 8 mm collimator. (e) Sector position in 16 mm collimator

a composite shot design allows an optimized dose distribution shape for each individual shot. The setup of any sectors, combinations of different collimators, or blocking takes only minimal time (a few seconds done automatically). The new PERFEXION system provides further improvements in patient and staff radiation shielding. The sectors are always in the off position (blocked) during patient transportation in the treatment position, transition into new stereotactic coordinates, pause or emergency interrupt. These results in significantly (about 5–10 times) lower extracranial irradiation to the patient compared to modelsBandC.Ourpreliminarycomparisonstudy shows that for a patient with ten brain lesion, the total time saved is about 1.5–2.0 h compared to other systems. The new Leksell Gamma Knife PERFEXION provides excellent dosimetry

performance, unlimitedcranialreach, enhancedradiationprotectionforpatientandstaff,fullautomation of the treatment process and better patient and staff comfort compared to previous models (> Table 75-2). Thus, the PERFEXION unit provides the potential to increase the spectrum of treatable indications including multiple brain metastasis, access to the upper cervical spine, and other pathologies of the head and neck. The use for the lower to mid-cervical spine will require the development of a new fixation device.

The Radiosurgery Procedure In the following sections we discuss the basic technical steps of gamma knife radiosurgery using LGK 4C and PERFEXION.

1229

1230

75


. Figure 75-4 Leksell stereotactic frame is docked by means of frame adapter in the Leksell Gamma Knife PERFEXION

. Figure 75-5 Example of the treatment plan with composite shots for the Leksell Gamma Knife PERFEXION. Multiple 4 mm collimators were used to design dose plan for a vestibular schwannomas. Sector blocking was used in one shot to achieve high conformity and sharp dose fall

Daily Quality Assurance Gamma Knife quality assurance testing is performed by an authorized medical physicist every morning. The purpose of Daily Quality Assurance is to assure proper system function in standard treatment conditions plus verify all safety and

emergency functions. The medical physicist ensures that all system tests required by Nuclear Regulatory Commission (NRC) regulations are performed and functioning properly. These tests include the permanently mounted radiation monitor inside the treatment room and it’s remote indicator, hand held radiation monitor, patient


75

. Table 75-2 Technical Specifications of Different Leksell Gamma Knife Units Comparing parameter Accuracy Radiological accuracy Positioning accuracy Positioning repeatability Radiation safety Room shielding required around back wall, 180 Room shielding required around front wall, 180 Body dose to patient lower than other devices Treatment planning Mechanical treatment range X/Y/Z Shape of accessible volume Effective target dose rate Composite shot Dynamic shaping Workflow Typical treatment time Approximate patient set up time Stereotactic coordinates set up time Collimator size set up time Collimator blocking set up time Composite collimator set up time

Leksell Gamma Knife PERFEXION

Leksell Gamma Knife 4C

Leksell Gamma Knife B

Figure 59-5). Spiegelmann et al. [18,19] have reported their experience. They reviewed the methods and results of linear accelerator (LINAC) radiosurgery in 44 patients with acoustic neuromas who were treated between 1993 and 1997. Computerized tomography scanning was selected as the stereotactic imaging modality for target definition. A single, conformally shaped isocenter was used in the treatment of 40 patients; two or three isocenters were used in four patients who harbored very irregular tumors. The radiation dose directed to the tumor border was the only parameter that changed during the study period: in the first 24 patients who were treated the dose was 15–20 Gy, whereas in the last 20 patients the dose was reduced to 11–14 Gy. After a mean follow-up period of 32 months (range 12–60 . Figure 59-5 Four years post-treatment, the MRI scan shows the schwannoma to be much smaller

Linac radiosurgery

months), 98% of the tumors were controlled. The actuarial hearing preservation rate was 71%. New transient facial neuropathy developed in 24% of the patients and persisted to a mild degree in 8%. Radiation dose correlated significantly with the incidence of cranial neuropathy, particularly in large tumors (4 cm3). Several reports on smaller series of patients treated with linear accelerator-based radiosurgery for vestibular schwannomas have been published in recent years. Martens et al. reported on 14 patients with at least 1 year of follow-up after radiosurgery on the LINAC unit in the University Hospital in Ghent, Belgium [20]. A mean marginal dose of 19.4 Gy (range 16–20 Gy) was delivered to the 70% isodose line with a single isocenter. Mean follow-up duration was 19 months (range 12–24 months.). During this relatively short follow-up interval, 100% radiographic tumor control has been achieved (29% regressed, 71% stable, zero enlarged). Rates of delayed facial and trigeminal neuropathy were 21% and 14%, respectively and two of three facial nerve deficits resolved. Preoperative hearing was preserved 50% of the time. Valentino and Raimondi reported on 23 patients treated with LINAC radiosurgery in undergone previous surgery. Total radiation dose to the tumor margin ranged from 12 to 45 Gy (median 30 Gy) and was delivered in 1–5 sessions. One or two isocenters were used and mean duration of follow-up was 40 months (range 24–46 months.). Results using this less conventional method of multi-session radiosurgery were comparable to other radiosurgical techniques. Tumor control was achieved in 96% of patients (38% regressed, 58% stable, 4% enlarged), facial and trigeminal neuropathies each occurred at a rate of 4%, and ‘‘hearing was preserved at almost the same level as that prior to radiosurgery in all patients.’’ Ishihara [22] discussed 38 patients treated with Cyberknife radiosurgery. The treatment volumes of these two groups were 0.5–24.0 cm3

59

(mean 4.7 cm3), and 0.5–41.6 cm3 (mean 8.2 cm3). Target irradiation was administered in 1–3 fractions (mean 2.5 fractions). The total marginal radiation doses were 15.0–20.5 Gy (mean 17.0 Gy), and 11.9–20.1 Gy (mean 16.9 Gy), respectively. After a mean follow-up period of 31.9 months (range 12–59 months, median 27 months), 94% of the tumors were controlled. Only one patient in the group with non-serviceable hearing underwent additional surgical resection for a presumed increase in tumor size. The hearing preservation rate was 93%. Facial weakness did not develop in any of the patients in the serviceable hearing group. New trigeminal symptoms did not develop in any patients in either group. The use of LINAC radiosurgery for acoustics is briefly discussed in reports by Delaney[23] and Barcia [24]. As of December 2008, the University of Florida experience with vestibular schwannomas comprised 448 patients.

Meningiomas Meningiomas are the most common benign primary brain tumor, with an incidence of approximately 7/100,000 in the general population. Surgery has long been thought to be the treatment of choice for symptomatic lesions and is often curative. Many meningiomas, however, occur in locations where attempted surgical cure may be associated with morbidity or mortality, such as the cavernous sinus or petroclival region [25,26]. In addition, many of these tumors occur in the elderly, where the risks of general anesthesia and surgery are known to be increased. Hence there is interest in alternative treatments, including radiation therapy and radiosurgery, either as a primary or adjuvant approach. Simpson, in a classic paper, described the relationship between completeness of surgical resection and tumor recurrence [27]. A grade I resection, that is complete tumor removal with

935

936

59

Linac radiosurgery

excision of the tumor’s dural attachment and involved bone, has a 10% recurrence rate. A grade II resection, complete resection of the tumor and coagulation of its dural attachment, has up to a 20% recurrence rate. Grade III resection is complete tumor removal without dural resection or coagulation. Grade IV resection is subtotal and Grade V resection is simple decompression. Recurrence rates in the Grade IV and V groups basically reflect the natural history of the tumor, with high rates of recurrence over time. Unfortunately, some common meningioma locations, such as the cavernous sinus or petroclival region are not readily amenable to a complete dural resection or coagulation strategy because of location and the proximity of vital neural and vascular structures. In addition, relatively high complication rates have been described for meningioma surgery in some locations and in the elderly. Pollock and colleagues recently analyzed 198 patients with meningiomas less than 35 mm in diameter, treated with either surgical resection or gamma knife radiosurgery [28]. Tumor recurrence was more frequent in the surgical resection group (12% vs. 2%). No statistically significant difference was detected in the 3 and 7 year actuarial progression-free survival rate between patients with Simpson Grade 1 resections and those who underwent radiosurgery. Progression Complications were lower in the radiosurgery group. Multiple linear accelerator radiosurgical series have been published [29–34]. Hakim and colleagues described the largest such series, and the only one of the group to report actuarial statistics [35]. One hundred twenty-seven patients with one hundred fifty-five meningiomas were treated. Actuarial tumor control for patients with benign tumors was 89.3% at 5 years. Six patients (4.7%) had permanent radiation induced complications. The University of Florida report on linear accelerator radiosurgery treatment of meningiomas is the largest yet published [36]. Two hundred ten patients were treated from May 1989 through December 2001.

All patients had follow-up for a minimum of 2 years and no patients were lost to follow-up. Actuarial local control for benign tumors was 100% at 1 and 2 years, and 96% at 5 years (see > Figures 59-6 and > 59-7). Actuarial local control for atypical tumors was 100% at 1 year, 92% at 2 years, and 77% at 5 years. Actual control for malignant tumors was 100% at 1 and 2 years, but only 19% at 5 years. Permanent radiation induced complications occurred in 3.8%, all of which involved malignant tumors. These tumor control and treatment morbidity rates compare well with all other published series. We found that reliance on imaging characteristics rather than surgical pathology did not yield a high incidence of missed diagnoses. During the time interval of this study, only two patients were treated as presumed meningiomas and later found to have other diagnoses. One had a dural-based metastasis which was surgically excised when it enlarged. The other had a hemangiopericytoma of the lateral cavernous sinus which was surgically excised when it enlarged.

. Figure 59-6 MRI scan shows right cavernous sinus meningioma. The patient presented with a sixth nerve paresis

Linac radiosurgery

. Figure 59-7 Three years later, the meningioma is barely visible. The sixth nerve paresis completely resolved. We believe that radiosurgery is the treatment of choice for many cavernous sinus meningiomas

Radiosurgery for Malignant Tumors Malignant tumors are radiobiologically more amenable to fractionated radiotherapy than benign lesions. Malignancies tend to infiltrate surrounding brain, resulting in poorly definable tumor margins. A priori, these two traits of cerebral malignancies would seem to make radiosurgery an unattractive treatment option. Nevertheless, SRS has proven to be a useful weapon in the armamentarium against malignant brain tumors. The most common applications of SRS to malignant tumors are the treatment of cerebral metastases and the delivery of an adjuvant focal radiation ‘‘boost’’ to malignant gliomas.

Cerebral Metastases Metastatic brain tumors are up to ten times more common than primary brain tumors with an annual incidence of between 80,000 and

59

150,000 new cases each year [37]. Fifteen to forty percent of cancer patients will be diagnosed with a brain metastasis during the course of their illness. Once a brain metastasis has been diagnosed, the median life expectancy is less than 1 year. However, in many patients, aggressive treatment of metastatic disease has been shown to restore neurologic function and prevent further neurological manifestations. Debate exists concerning the optimum treatment for metastatic brain disease. In autopsy series, brain metastases occur in up to 50% of cancer patients [38]. Approximately 30–40% present with a solitary metastasis. Brain metastases frequently cause debilitating symptoms which can seriously impact the patient’s quality of life. With no treatment or steroid therapy alone, survival is very limited (1–2 months). Whole brain radiotherapy extends median survival, but the duration of survival is typically low (3–4 months). Several randomized trials have suggested that, when possible, surgery followed by WBRT is superior to WBRT alone. Patchell et al. reported a randomized clinical trial involving 46 patients with a single metastasis and well-controlled systemic disease [39]. They found a significant improvement in survival (40 vs. 15 weeks) and local recurrences in the CNS (20% vs. 52%) for patients in the surgery plus whole brain radiotherapy arm of the study. Likewise, Noordijk et al. randomized 66 patients and found a significant survival advantage (10 vs. 6 months) for the combination therapy arm [40]. In contrast, Mintz et al. studied a group of 84 patients and did not show an advantage of surgery plus radiotherapy over radiotherapy alone [41]. It has been suggested that the inclusion of a higher percentage of patients with active systemic disease and lower performance scores did not allow the benefit of improved local control to affect survival in this series. Haines points out that survival and quality of life are the most important outcomes measures in evaluating a clinical treatment for cancer [42]. Surrogate

937

938

59

Linac radiosurgery

endpoints, like local control, are inherently unreliable, especially when the definition of local control is changed. This applies to a comparison of radiosurgery with surgery for brain metastasis. In surgical series, local control means no visible tumor on follow-up scans. In radiosurgical series, local control means no growth (or sometime minimal growth) on follow-up scans. These endpoints are unlikely to be equivalent. In addition, comparison of current results to historical controls is fraught with hazard to selection bias. This issue led to erroneous conclusions about the efficacy of brachytherapy for malignant gliomas and to overly optimistic reports regarding the efficacy of intra-arterial chemotherapy. Of equal import is the difficulty and variability of reporting standards for local control. Few series provide actuarial local control. They simply provide a ‘‘raw’’ number at an arbitrary point in time. Less commonly appreciated is the difficulty in documenting local control. Many of these patients die away from the medical center where radiosurgery was performed. It is frequently impossible to determine from family or local physician telephone interview, whether the proximate cause of death was loss of local control, new intracranial disease (loss of regional control), or systemic disease. Most radiosurgical series have assumed that, unless an MRI was performed documenting local loss of local control prior to death, local control was maintained. This assumption may lead to a systematic overestimation of local control rates. Sturm [43–45], Black [46,47], and Adler [48–50] published early reports on linear accelerator radiosurgery for brain metastases. Alexander [47] reported on 248 patients. Median tumor volume was 3 cc and median tumor dose was 15 Gy. Median survival was 9.4 months. Actuarial local control was 85% at 1 year and 65% at 2 years. Auchter et al. reported a multiinstitutional study of 122 patients [51]. Actuarial 1 and 2 year survivals were 53% and 30% respectively. Local control was 86%. Many other LINAC series have been reported [45,52–59].

As radiosurgery has emerged as a treatment option, clinicians have attempted to define prognostic factors, which may help to define patient populations most likely to benefit from radiosurgical treatment [60–62]. Multiple factors have been discerned from retrospective analysis and include: Karnofsky performance scale score, status of systemic disease, histology, number of metastases, volume of metastases, time interval between the diagnosis of the primary lesion and the metastatic lesion, pattern of enhancement [63,64], the RTOG recursive partitioning categories [65], and radiation dose. The University of Florida published their early experience with radiosurgery for brain metastases [66] in 2004. Three hundred eighty three patients were treated. Median survival was 9 months. Melanoma histology and increasing number of metastases predicted poorer survival. Increasing age, somewhat surprisingly, slightly improved survival, possibly because younger patients tended to have more radioresistant histologies. Actuarial local control was 75% (> Figures 59-8 and > 59-9). Increasing dose provided better control and eloquent location was also associated with better control (possibly because eloquent tumors tended to be discovered at a smaller size). Regional control was poorer in melanoma or breast patients and in those with synchronous presentation of brain metastasis and primary tumor. In this retrospective analysis, whole brain radiotherapy did not improve regional control. Andrews et al. [67] reported the phase III results of RTOG protocol 9508, the first prospective, randomized trial of WBRT with or without SRS for the treatment of metastatic brain disease to reach full accrual. Three hundred thirtythree patients with 1–3 newly diagnosed brain metastases, KPS 70, and no history of WBRT were randomly allocated either WBRT alone (167 patients) or WBRT followed by an SRS boost (164 patients). Analysis revealed a significant survival advantage for patients with a single brain

Linac radiosurgery

59

. Figure 59-8 The patient with known breast carcinoma presented with a symptomatic pontine lesions. She was treated with radiosurgery (15 Gy to the 80% isodose line)

. Figure 59-9 Three years later, the site of the lesion was barely visible

metastasis treated with WBRT plus SRS rather than WBRT alone (median survival 6.5 months vs. 4.9 months), but failed to reveal a similar advantage for patients with 2–3 brain metastases.

However, all patients in the WBRT plus SRS group were significantly more likely to have a stable or improved KPS six months post-treatment, and were less likely to be dependent on corticosteroids, than those treated with WBRT alone. Multivariate analysis revealed RPA Class I and non-small cell lung primary to be significantly associated with improved survival. Analysis also revealed significantly better 1-year local control of lesions treated with WBRT plus SRS, compared to those treated with WBRT alone (82% vs. 71%). The authors concluded that the addition of SRS to WBRT should be standard treatment for patients with a single brain metastasis and considered for patients with 2–3 brain metastases. Aoyama et al. [68] recently reported results of the first prospective, multi-institutional, randomized trial of SRS with or without WBRT for the treatment of metastatic brain disease. One hundred thirty-two patients with 1–4 brain metastases 3 cm in diameter and KPS 70 were randomly allocated either SRS plus WBRT (65 patients) or SRS alone (67 patients).

939

940

59

Linac radiosurgery

Adjuvant WBRT did not affect survival; patients who received both SRS and WBRT had a median survival of 7.5 months and a 1-year actuarial survival rate of 38.5%, compared to 8.0 months and 28.4% for those treated with SRS alone. However, patients who did not initially receive WBRT underwent significantly more salvage procedures. Multivariate analysis revealed age Figures 59-9 and > 59-10). If the patient’s clinical status changes, she/he is followed more closely at clinically appropriate intervals. Each patient is scheduled to undergo cerebral angiography at three years post-radiosurgery, and a

Linac radiosurgery

. Figure 59-10 Pretreatment angiogram shows a left parietal AVM. It was treated with radiosurgery (17.5 Gy to the 70% line using three isocenters)

definitive assessment of the success or failure of treatment is made based on the results of angiography. If no flow is observed through the AVM nidus, the patient is pronounced cured and is discharged from follow-up. If the AVM nidus is incompletely obliterated, appropriate further therapy (most commonly repeat radiosurgery on the day of angiography) is prescribed, and the treatment/follow-up cycle is repeated.

The University of Florida Experience From 18 May 1988 to 6 November 2007, 606 patients with arteriovenous malformations were treated at the University of Florida. The mean age was 40 (4–78 years). The median treatment volume was 6 cc (0.2 – 45.3 cc). Many patients early in the series were treated with single isocenters but in recent years an effort has been made to produce highly conformal plans by employing multiple isocenters. The median radiation dose to the periphery of the AVM was

59

1750 cGy and the mean follow-up duration was 31 months. Angio or MRI cure rates vary according to AVM size: 10 cc = 35%. With a second treatment, cure rates substantially increase, even for larger AVMs: 10 cc = 78%. Ellis et al. [88] performed a detailed analysis of treatment failures in our series in 1998. He found that 26% of the failures were due to targeting error, at least in part. Statistical predictors of failure were increasing AVM size, decreasing treatment dose, and increasing Spetzler-Martin score. Of particular interest were the ‘‘cutpoints’’ which were identified. There was a dramatic increase in cure rates when the peripheral dose was raised to a least 15 Gy. There was a dramatic decrease in cure rate when AVM size exceeded 10 cc (size D). In a more recent analysis, a study was undertaken to determine which factors were statistically predictive of radiographic and clinical outcomes in the radiosurgical treatment of arteriovenous malformations [96]. The computerized dosimetry and clinical data on 269 patients were reviewed. The AVM nidus was hand contoured on successive enhanced CT slices through the nidus, to allow detailed determination of nidus volume, target miss, normal brain treated, dose conformality and dose gradient. In addition, a number of patient and treatment factors, including Spetzler-Martin score, presenting symptoms, dose, number of isocenters, radiographic outcome, and clinical outcome were subjected to multivariate analysis. None of the analyzed factors were predictive of permanent radiation induced complications or of hemorrhage after radiosurgery in this study. Eloquent AVM location and 12 Gy volume correlated with the occurence of transient radiation induced complications. Better conformality correlated with a reduced incidence of transient complications. Lower SpetzlerMartin scores, higher doses, and steeper dose gradients correlated with radiographic success.

943

944

59

Linac radiosurgery

When AVMs are not cured, current practice frequently involves a ‘‘retreatment,’’ usually 3 years after the original treatment. We reviewed the cases of 52 patients who underwent repeat radiosurgery for residual AVM at our institution between December 1991 and June 1998 [97]. In each case, residual arteriovenous shunting persisted beyond 36 months after the initial treatment. The mean interval between the first and second treatments was 41 months. Each AVM nidus was measured at the time of original treatment and again at the time of retreatment, and dosimetric parameters of the two treatments were compared. After retreatment, patients were followed, and their outcomes evaluated, according to our standard post-AVM radiosurgery protocol. Definitive endpoints included angiographic cure, radiosurgical failure (documented persistence of AVM flow 3 years after retreatment), and death. The mean original lesion volume was 13.8 cc and the mean volume at retreatment was 4.7 cc, for an average volume reduction of 66% after the initial ‘‘failed’’ treatment. As pointed out above, retreatment when studied in our entire patient population has substantially increased overall cure rate, with minimal morbidity. Zipfel and Friedman [98] sought to determine which morphological features of arteriovenous malformations (AVMs) are statistically predictive of preradiosurgical hemorrhage, postradiosurgical hemorrhage, and neuroimagingdefined failure of radiosurgical treatment. Archived CT dosimetry and available angiographic and clinical data for 268 patients in whom AVMs were treated with linear accelerator radiosurgery were retrospectively reviewed (> Figure 59-11). Many of the morphological features of AVMs, including location, volume, compact or diffuse nidus, neovascularity, ease of nidus identification, number of feeding arteries, location (deep or superficial) of feeding arteries, number of draining veins, deep or superficial venous drainage, venous stenoses,

. Figure 59-11 Two years later, the angiogram is normal

venous ectasias, and the presence of intranidal aneurysms, were analyzed. In addition, a number of patient and treatment factors, including patient age, presenting symptoms, radiation dose, repeated treatment, and radiological outcome, were subjected to multivariate analyses. A larger AVM volume (odds ratio [OR] 0.349; p = 0.004) was associated with a decreased rate of pretreatment hemorrhage, whereas periventricular location (OR 6.358; p = 0.000) was associated with an increased rate of pretreatment hemorrhage. None of the analyzed factors was predictive of hemorrhage following radiosurgery. A higher radiosurgical dose was strongly correlated with neuroimaging-defined success (OR 3.743; p = 0.006), whereas a diffuse nidus structure (OR 0.246; p = 0.008) and associated neovascularity (OR 0.428; p = 0.048) were each associated with a lower neuroimaging-defined cure rate. A strong correlation between CT scanning and angiography was noted for both nidus structure (p = 0.000; Fisher exact test) and neovascularity (p = 0.002; Fisher exact test). Clinical follow-up information is available on 472 patients. Twelve (2.5%) have sustained permanent radiation induced neurological deficits.

Linac radiosurgery

59

. Table 59-1 LINAC AVM radiosurgery series with patient numbers greater than 20 Series Betti [99] Colombo [100] Engenhart [101] Friedman et al. [96] Schlienger [102] Young [103]

#Patients

Angiographic success rate

286 180 212 554 169 66

82% 80% 72% 74% 64% 50%

Fifteen (3%) have sustained transient deficits. Most importantly, 41 have had post-treatment hemorrhages (eight fatal). One must remember that the major drawback of radiosurgery for arteriovenous malformations is the continued risk of hemorrhage during the latent period (typically 2 years). The results of other LINAC radiosurgery series are summarized in > Table 59-1.

1 patient 2.2% 4.3% 1.8%

11.

12.

13. 14.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-9. 2. Betti OO, Derechinsky VE. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurol Chir Suppl 1984;33:385-90. 3. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985;16:154-60. 4. Hartmann GH, Schlegel W, Sturm V, et al. Cerebral radiation surgery using moving field irradiation at a linear accelerator facility. Int J Radiat Oncol Biol Phys 1985;11:1185-92. 5. McGinley PH, Butker EK, Crocker IR, Landry JC. A patient rotator for stereotactic radiosurgery. Phys Med Biol 1990;35:649-57. 6. Podgorsak EB, Olivier A, Pla M, Lefebvre PY, Hazel J. Dynamic stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1988;14:115-26. 7. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988; 22:454-64. 8. Friedman WA, Bova FJ. The University of Florida radiosurgery system. Surg Neurol 1989;32:334-42. 9. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med 1998;339(20):1426-33. 10. Samii M, Matthies C. Management of 1,000 vestibular schwannomas (acoustic neuromas): surgical management

Permanent radiation complications

15.

16.

17.

18.

19.

20.

21.

22.

23.

and results with an emphasis on complications and how to avoid them. Neurosurgery 1997;40:11-23. Samii M, Matthies C. Management of 1,000 vestibular schwannomas (acoustic neuromas): the facial nerve – preservation and restitution of function. Neurosurgery 1997;40:684-95. Samii M, Matthies C. Management of 1,000 vestibular schwannomas (acoustic neuromas): hearing function in 1,000 tumor resections. Neurosurgery 1997;40:248-62. Leksell L. A note on the treatment of acoustic tumors. Acta Chir Scand 1971;137:763-5. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis PS. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001; 95(3):440-9. Friedman WA, Bradshaw P, Myers A, Bova FJ. Linear accelerator radiosurgery for vestibular schwannomas. J Neurosurg 2006;105(5):657-61. Beegle RD, Friedman WA, Bova FJ. Effect of treatment plan quality on outcomes after radiosurgery for vestibular schwannoma. J Neurosurg 2007;107(5):913-6. Vachhani JA, Friedman WA. Radiosurgery in patients with bilateral vestibular schwannomas. Stereotact Funct Neurosurg 2007;85(6):273-8. Spiegelmann R, Gofman J, Alezra D, Pfeffer R. Radiosurgery for acoustic neurinomas (vestibular schwannomas). Isr Med Assoc J 1999;1(1):8-13. Spiegelmann R, Lidar Z, Gofman J, Alezra D, Hadani M, Pfeffer R. Linear accelerator radiosurgery for vestibular schwannoma. J Neurosurg 2001;94(1):7-13. Martens F, Verbeke L, Piessens M, Van Vyve M. Stereotactic radiosurgery of vestibular schwannomas with a linear accelerator. Acta Neurochir Suppl 1994; 62:88-92. Valentino V, Raimondi AJ. Tumour response and morphological changes of acoustic neurinomas after radiosurgery. Acta Neurochir 1995;133:157-63. Ishihara H, Saito K, Nishizaki T, et al. CyberKnife radiosurgery for vestibular schwannoma. Minim Invasive Neurosurg 2004;47(5):290-3. Delaney G, Matheson J, Smee R. Stereotactic radiosurgery: an alternative approach to the management of acoustic neuromas. Med J Austral 1992;156:440.

945

946

59

Linac radiosurgery

24. Barcia Salorio JL, Hernandez G, Ciudad J, Bordes V, Broseta J. Stereotactic radiosurgery in acoustic neurinoma. Acta Neurochir Suppl 1984;33:373-6. 25. Sekhar LN, Jannetta PJ, Burkhart LE, Janosky JE. Meningiomas involving the clivus: a six-year experience with 41 patients. Neurosurgery 1990;27:764-81. 26. Sekhar LN, Altschuler EM. Meningiomas of the cavernous sinus. In: Al-Mefty O, editor. Meningiomas. New York: Raven Press; 1991. p. 445-60. 27. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957;20:22-39. 28. O’Neill BP, Iturria NJ, Link MJ, Pollock BE, Ballman KV, O’Fallon JR. A comparison of surgical resection and stereotactic radiosurgery in the treatment of solitary brain metastases. Int J Radiat Oncol Biol Phys 2003; 55:1169-76. 29. Valentino V, Schinaia G, Raimondi AJ. The results of radiosurgical management of 72 middle fossa meningiomas. Acta Neurochir 1993;122:60-70. 30. Villavicencio AT, Black PM, Shrieve DC, Fallon MP, Alexander E, Loeffler JS. Linac radiosurgery for skull base meningiomas. Acta Neurochir (Wien) 2001; 143(11):1141-52. 31. Engenhart R, Kimmig BN, Hover KH, et al. Stereotactic single high dose radiation therapy of benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 1990; 19:1021-6. 32. Spiegelmann R, Nissim O, Menhel J, Alezra D, Pfeffer MR. Linear accelerator radiosurgery for meningiomas in and around the cavernous sinus. Neurosurgery 2002;51 (6):1373-79; discussion 1379–80. 33. Chuang CC, Chang CN, Tsang NM, et al. Linear accelerator-based radiosurgery in the management of skull base meningiomas. J Neurooncol 2004;66(1–2): 241-9. 34. Deinsberger R, Tidstrand J, Sabitzer H, Lanner G. LINAC radiosurgery in skull base meningiomas. Minim Invasive Neurosurg 2004;47(6):333-8. 35. Hakim R, Alexander E, III, Loeffler JS, et al. Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998;42:446-54. 36. Friedman WA, Murad G, Bradshaw P, et al. Linear accelerator radiosurgery for meningiomas. J Neurosurg 2005;103:206-9. 37. Lohr F, Pirzkall A, Hof H, Fleckenstein K, Debus J. Semin Surg Oncol 2001;20:50-6. 38. DeAngelis LM. Brain tumors. N Engl J Med 2001; 344:114-23. 39. 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. 40. 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;29:711-7.

41. Mintz AH, Kestle J, Rathbone MP, et al. A randomized trial to assess the efficacy of surgery in addition to radiotherapy in patients with a single cerebral metastasis. Cancer 1996;78(7):1470-6. 42. Haines SJ. Moving targets and ghosts of the past: outcome measurement in brain tumour therapy. J Clin Neurosci 2002;9:109-12. 43. Sturm V, Kober B, Hover KH, et al. Stereotactic percutaneous single dose irradiation of brain metastases with a linear accelerator. Int J Radiat Oncol Biol Phys 1987;13:279-82. 44. Sturm V, Kimmig B, Engenhardt R, et al. Radiosurgical treatment of cerebral metastases. J Stereotactic Funct Neurosurg 1991;57:7-10. 45. Voges J, Treuer H, Erdmann J, et al. LINAC radiosurgery in brain metastases. Acta Neurochir Suppl 1994; 62:72-6. 46. Black PM. Solitary brain metastases: radiation, resection, or radiosurgery? Chest 1993;103:367S-9S. 47. Alexander E, Moriarty TM, Davis RB, et al. Stereotactic radiosurgery for the definitive, noninvasive treatment of brain metastases. J Nat Cancer Inst 1995;87:34-40. 48. Adler JR, Cox RS, Kaplan I, Martin DP. Stereotactic radiosurgical treatment of brain metastases. J Neurosurg 1992;76:444-9. 49. Fuller BG, Kaplan ID, Adler J, Cox RS, Bagshaw MA. Stereotaxic radiosurgery for brain metastases: the importance of adjuvant whole brain irradiation. Int J Radiat Oncol Biol Phys 1992;23:413-8. 50. Joseph J, Adler JR, Cox RS, Hancock SL. Linear accelerator-based stereotaxic radisourgery for brain metastases: the influence of number of lesions on survival. J Clin Oncol 1996;14:1085-92. 51. Auchter RM, Lamond JP, Alexander E, et al. A multiinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastasis. Int J Radiat Oncol Biol Phys 1996;35:27-35. 52. Becker G, Jeremic B, Engel C, et al. Radiosurgery for brain metastases: the Tuebingen experience. Radiother Oncol 2002;62:233-7. 53. Breneman JC, Warnick RE, Albright RE, et al. Stereotactic radiosurgery for the treatment of brain metastases. Cancer 1997;79:551-7. 54. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Treatment selection factors for stereotactic radiosurgery of intracranial metastases. Int J Radiat Oncol Biol Phys 1995;32:1161-6. 55. Caron J-L, Souhami L, Podgordak EB. Dynamic stereotactic radiosurgery in the palliative treatment of cerebral metastatic tumors. J Neuro Oncol 1992;12:173-9. 56. Gutin PH, Wilson CB. Radiosurgery for malignant brain tumors. J Clin Oncol 1990;8:571-3. 57. Mehta MP, Rozental JM, Levin AB, et al. Defining the role of radiosurgery in the management of brain metastases. Int J Radiat Oncol Biol Phys 1992;24:619-25. 58. Mehta M, Noyes W, Craig B, et al. A cost-effectiveness and cost-utility analysis of radiosurgery vs. resection for

Linac radiosurgery

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

single-brain metastases. Int J Radiat Oncol Biol Phys 1997;39:445-54. Valentino V. The results of radiosurgical management of 139 single cerebral metastases. Acta Neurochir Suppl 1995;63:95-100. Cho KH, Hall WA, Gerbi BJ, Higgins PD, Bohen M, Clark HB. Patient selection criteria for the treatment of brain metastases with stereotactic radiosurgery. J Neuro Oncol 1998;40:73-86. Fernandez-Vicioso E, Suh JH, Kupelian PA, Sohn JW, Barnett GH. Analysis of prognostic factors for patients with single brain metastasis treated with stereotactic radiosurgery. Rad Onc Invest 1997;5:31-7. Maor MH, Dubey P, Tucker SL, et al. Stereotactic radiosurgery for brain metastases: results and prognostic factors. Int J Cancer 2000;90:157-62. Goodman KA, Sneed PK, McDermott MW, et al. Relationship between pattern of enhancement and local control of brain metastases after radiosurgery. Int J Radiat Oncol Biol Phys 2001;50:139-46. Shiau C-Y, Sneed PK, Shu H-KG, et al. Radiosurgery for brain metastases: relationship of dose and pattern of enhancement to local control. Int J Radiat Oncol Biol Phys 1997;37:375-83. Gaspar L, Scott C, Rotman M, et al. Recursive partitioning analysis of prognostic factors in three radiation therapy oncology group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997;37:745-51. Ulm AJ, Friedman WA, Bova FJ, Bradshaw P, Amdur RJ, Mendenhall WM. Linear accelerator radiosurgery in the treatment of brain metastases. Neurosurgery 2004; 55:1076-85. Andrews DW, Scott CB, Sperduto PW, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363(9422):1665-72. Aoyama H, Tago M, Kato N, et al. Neurocognitive function of patients with brain metastasis who received either whole brain radiotherapy plus stereotactic radiosurgery or radiosurgery alone. Int J Radiat Oncol Biol Phys 2007;68(5):1388-95. Swinson BM, Friedman WA. Linear accelerator stereotactic radiosurgery for metastatic brain tumors: 17 years of experience at the University of Florida. Neurosurgery 2008;62(5):1018-31; discussion 1031-32. Bernstein M, Laperriere N, Glen J, Leung P, Thomason C, Landon AE. Brachytherapy for recurrent malignant astrocytoma. Int J Radiat Oncol Biol Phys 1994; 30(5):1213-7. Chang CN, Chen WC, Wei KC, et al. High-dose-rate stereotactic brachytherapy for patients with newly diagnosed glioblastoma multiformes. J Neuro Oncol 2003; 61(1):45-55. Prados MD, Gutin PH, Phillips TL, et al. Interstitial brachytherapy for newly diagnosed patients with

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

59

malignant gliomas: the UCSF experience. Int J Radiat Oncol Biol Phys 1992;24(4):593-7. Shrieve DC, Alexander E, III, Wen PY, et al. Comparison of stereotactic radiosurgery and brachytherapy in the treatment of recurrent glioblastoma multiforme. Neurosurgery 1995;36(2):275-82; discussion 82–4. Hall WA, Djalilian HR, Sperduto PW, et al. Stereotactic radiosurgery for recurrent malignant gliomas. J Clin Oncol 1995;13(7):1642-8. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radiosurgery for high-grade gliomas: the University of Florida experience. Int J Radiat Oncol Biol Phys 1995; 32(1):205-10. Masciopinto JE, Levin AB, Mehta MP, Rhode BS. Stereotactic radiosurgery for glioblastoma: a final report of 31 patients. J Neurosurg 1995;82(4):530-5. Regine WF, Patchell RA, Strottmann JM, Meigooni A, Sanders M, Young AB. Preliminary report of a phase I study of combined fractionated stereotactic radiosurgery and conventional external beam radiation therapy for unfavorable gliomas. Int J Radiat Oncol Biol Phys 2000;48(2):421-6. Prisco FE, Weltman E, de Hanriot RM, Brandt RA. Radiosurgical boost for primary high-grade gliomas. J Neuro Oncol 2002;57(2):151-60. Gannett D, Stea B, Lulu B, Adair T, Verdi C, Hamilton A. Stereotactic radiosurgery as an adjunct to surgery and external beam radiotherapy in the treatment of patients with malignant gliomas. Int J Radiat Oncol Biol Phys 1995;33(2):461-8. Roberge D, Souhami L. Stereotactic radiosurgery in the management of intracranial gliomas. Technol Cancer Res Treat 2003;2(2):117-25. Curran WJ, Jr., Scott CB, Horton J, et al. Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology group malignant glioma trials. J Natl Cancer Inst 1993;85(9):704-10. Sarkaria JN, Mehta MP, Loeffler JS, et al. Radiosurgery in the initial management of malignant gliomas: survival comparison with the RTOG recursive partitioning analysis. Int J Radiat Oncol Biol Phys 1995;32(4): 931-41. Kondziolka D, Flickinger JC, Bissonette DJ, Bozik M, Lunsford LD. Survival benefit of stereotactic radiosurgery for patients with malignant glial neoplasms. Neurosurgery 1997;41(4):776-83; discussion 83–5. Ulm AJ, III, Friedman WA, Bradshaw P, Foote KD, Bova FJ. Radiosurgery in the treatment of malignant gliomas: the University of Florida experience. Neurosurgery 2005;57(3):512-7; discussion 517. Irish WD, Macdonald DR, Cairncross JG. Measuring bias in uncontrolled brain tumor trials – to randomize or not to randomize? Can J Neurol Sci 1997; 24(4): 307-12. Curran WJ, Jr., Scott CB, Weinstein AS, et al. Survival comparison of radiosurgery-eligible and -ineligible

947

948

59

Linac radiosurgery

malignant glioma patients treated with hyperfractionated radiation therapy and carmustine: a report of Radiation Therapy Oncology Group 83–02. J Clin Oncol 1993; 11(5):857-62. 87. Souhami L, Seiferheld W, Brachman D, et al. Randomized comparison of stereotactic radiosurgery followed by conventional radiotherapy with carmustine to conventional radiotherapy with carmustine for patients with glioblastoma multiforme: report of Radiation Therapy Oncology Group 93–05 protocol. Int J Radiat Oncol Biol Phys 2004;60(3):853-60. 88. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti JM. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998;89(1):104-10. 89. Pollock BE, Flickinger JC, Lunsford LD, Maitz A, Kondziolka D. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998; 42(6):1239-44; discussion 44–7. 90. Bova FJ, Friedman WA. Stereotactic angiography: an inadequate database for radiosurgery? Int J Radiat Oncol Biol Phys 1991;20:891-5. 91. Blatt DL, Friedman WA, Bova FJ. Modifications in radiosurgical treatment planning of arteriovenous malformations based on CT imaging. Neurosurgery 1993;33:588-96. 92. Spiegelmann R, Friedman WA, Bova FJ. Limitations of angiographic target localization in planning radiosurgical treatment. Neurosurgery 1992;30:619-24. 93. Flickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-response analysis of arteriovenous malformation obliteration after radiosurgery. Int J Radiat Oncol Biol Phys 1996;36:873-9. 94. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997;40(3):425-30; discussion 30–1. 95. Pollock BE, Kondziolka D, Lunsford LD, Bissonette D, Flickinger JC. Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery 1996;38(2): 318-24. 96. Friedman WA, Bova FJ, Bollampally S, Bradshaw P. Analysis of factors predictive of success or complications in arteriovenous malformation radiosurgery. Neurosurgery 2003;52:296-308. 97. Foote KD, Friedman WA, Ellis TL, Bova FJ, Buatti JB, Meeks SL. Salvage retreatment after failure of radiosurgery in patients with arteriovenous malformations. J Neurosurg 2003;98:337-41. 98. Zipfel GJ, Bradshaw P, Bova FJ, Friedman WA. Do the morphological characteristics of arteriovenous malformations affect the results of radiosurgery? J Neurosurg 2004;101(3):393-401. 99. Betti OO, Munari C, Rosler R. Stereotactic radiosurgery with the linear accelerator: treatment of arteriovenous malformations. Neurosurgery 1989;24:311-21.

100. Colombo F, Pozza F, Chierego G, Casentini L, De Luca G, Francescon P. Linear accelerator radiosurgery of cerebral arteriovenous malformations: an update. Neurosurgery 1994;34:14-21. 101. Engenhart R, Wowra B, Debus J, et al. The role of high-dose single fraction irradiation in small and large intracranial arteriovenous malformations. Int J Radiat Oncol Biol Phys 1994;30:521-9. 102. Schlienger M, Atlan D, Lefkopoulos D, et al. LINAC radiosurgery for cerebral arteriovenous malformations. Int J Radiat Oncol Biol Phys 2000;46:1135-42. 103. Young C, Summerfield R, Schwartz M, O’Brien P, Ramani R. Radiosurgery for arteriovenous malformations: the University of Toronto experience. Can J Neurol Sci 1997;24:99-105.

67 Linac Radiosurgery: Technical Aspects F. J. Bova . W. A. Friedman

Introduction Radiosurgery, a term coined in 1951 by the Swedish neurosurgeon Lars Leksell [1], was first practiced using an orthovoltage x-ray apparatus, then a particle accelerator and finally a Co-60 isotope unit. The majority of Leksell’s clinical work was carried out using the latter device, known as the Gamma Knife. With this dedicated tool Leksell was able to pioneer many radiosurgical procedures. By the mid-1980s only a handful of GammaKnife’s existed, limiting the application of this new technique. By the 1980s the linear accelerators had, for the most part, replaced Cobalt-60 units for routine radiation therapy. In the mid 1980s researchers began to adapt linacs as radiosurgical tools [2–4]. These developments had immediate effects on the field of radiosurgery. Once the successful application of the linac based radiosurgery paradigm, was demonstrated, hundreds of radiation oncologists and neurosurgeons found that their facilities could purchase the necessary hardware and software to upgrade their linac making them radiosurgery capable. Surprisingly, this upgrade could be accomplished for a tenth the cost of the Gamma Knife. In less then a decade the number of radiosurgical units worldwide went from fewer than half a dozen to hundreds. As radiation oncologists became more involved in the day-to-day use of radiosurgical techniques, they realized that stereotactic techniques had application beyond the traditional neurosurgical procedures pioneered by Leksell. These applications were both intra and extra cranial. Many of the new extra cranial applications required #


further innovative targeting technologies employing ultrasound, orthogonal x-ray and conebeam. ‘‘Radiosurgery’’ was initially defined as a ‘‘single-fraction’’ technique that uses ‘‘stereotactic principles’’ for targeting and treatment of ‘‘intracranial’’ lesions through the use of ‘‘multiple noncoplanar beams’’. Over the past decade the stereotactic techniques on which radiosurgery are founded have been applied to fractionated treatments, termed stereotactic radiotherapy. Currently the line between single fraction radiosurgery and a limited number of fraction, i.e., usually less then five, radiosurgery and fractionated stereotactic radiosurgery, FRS, is not particularly distinct. For this text the term radiosurgery will be used to designate the single fraction technique. In considering the accuracy and precision necessary in a radiosurgical system, it is important to distinguish between the addition of a margin to account for the inclusion of microscopic disease and the necessity of adding a margin because of system uncertainty. If the lack of system accuracy requires the clinician to prescribe a margin of normal tissue to all stereotactic targets to ensure that the identified target is included in the selected isodose volume, normal tissues will unnecessarily be exposed to radiation. For example, adding a margin of 2 mm to a 24 mm target increases the volume irradiated from 7.2 to 11.5 cc, an volume increase of 60%, all of which is normal tissue. Another aspect of the stereotactic treatment must be considered before one can discuss the requirements of a radiosurgical system. Most radiosurgical targets cannot be imaged through conventional radiotherapy techniques involving

1088

67

Linac radiosurgery: technical aspects

simulation and portal verification. The majority of radiosurgical targets can only be visualized during diagnostic examinations, such as angiography or from a three-dimensional data set assembled from computed tomography (CT) and magnetic resonance imaging (MRI). Recently functional imaging in the form of positron emission tomography, PET, and functional magnetic resonance imaging, f MRI, have begun to be used to help define target as well as normal critical structures. At the time of treatment, i.e., the alignment of the patient to the therapeutic x-ray beam, none of these imaging modalities are available, blinding the clinician and forcing the reliance upon the spatial integrity of the stereotactic targeting system. The quality assurance procedures necessary to guarantee system accuracy under these conditions make up a large part of each radiosurgical treatment. Also, the use of a stereotactic system for radiosurgery is more technically demanding than the use of the same system for other stereotactic techniques. In the case of a stereotactic biopsy the neurosurgeon can target a region and acquire a series of biopsies along a probe track or from multiple placements of the biopsy probe. In the case of deep brain stimulation the neurosurgeon has the stimulation to help verify probe placement. The radiosurgeon has none of these verification tools available. It is this absolute dependence on stereotactic targeting that places more stringent requirements on radiosurgical systems than on other stereotactic neurosurgical procedures. When deciding the accuracy and precision necessary for stereotactic applications, one is inevitably told that the most inaccurate portion of the procedure is the identification of the target. This often comes in the form of the question; ‘‘Since we don’t know precisely where the target begins or ends, why should we worry about system accuracy?’’ This line of logic has two serious flaws. The first is that all inaccuracies are cumulative, starting with diagnostic imaging, through treatment planning and then through treatment

delivery. The fact that the edge of the target volume cannot be accurately defined is little justification for not being able to precisely predict a prescribed dose distribution. The second is that avoidance of critical structures, such as the optic chiasm, is often as important as the inclusion of target tissues. Very often the adjacent critical structures, the location of which are exactly known, place constraints on the prescribed dose. Inaccurate targeting and treatment delivery can as easily include critical structures as it can exclude target tissues. Knowing the position of the highisodose volumes as well as the location of the dose gradient to within a pixel of its true position is an absolute necessity. Over the years many different immobilization techniques have been employed in radiation oncology in attempts to provide a high degree of alignment between diagnostic imaging, that allowed the target and normal tissues to be appreciated, and patient alignment with the therapeutic beam. Radiosurgery initially overcame this obstacle through the use of a rigidly fixed reference system in the form of a stereotactic frame. The first step of each stereotactic procedure is the application of the reference frame to the patient’s skull. Once applied, this frame remains rigidly attached to the patient’s skull throughout the procedure. Several frames have been developed for general stereotactic applications, and many of these have been adapted for radiosurgical use. Each of these systems has unique advantages and disadvantages as well as its own coordinate system. They all, however, have one common feature: once the frame is fixed to the patient’s skull, a rigid relationship between that patient’s intracranial anatomy and the frame’s coordinate system is established. Frameless approaches have been pioneered [5] each system substitutes a somewhat different object or structure for the rigid frame. All require the same assumption of a rigid alignment between the patient’s anatomy and the reference system.


All of the basic principles, a long with the advantages and disadvantages of a frame based approach, apply to these newer techniques. Initially all diagnostic information necessary for targeting and planning had to occur after ring placement. However, with the advent of image fusion techniques, non-stereotactic exams, exams carried out prior to the application of the headring, can be registered to the frame-based exam, providing significant flexibility in obtaining target images. The three most common diagnostic procedures used for stereotactic localization are angiography, CT, and MR with fMRI, PET and DTI are beginning to find application in the planning process. Angiography provides unique information concerning vascular structures, and CT and MRI provide information about both target tissues and normal anatomy and allow a full threedimensional model of the patient’s intracranial anatomy to be reconstructed. Stereotactic imaging systems can be separated into two categories. The first includes systems that produce images containing all the information necessary to compute the stereotactic coordinates of all tissues contained within that image. Embedded in each image is a description of the stereotactic coordinate system. The fiducial marker on each image allows the trajectory of the x-ray beam relative to the stereotactic coordinate system to be computed. In the case of CT the axial sections are not required to be absolutely parallel to the plane of the stereotactic reference system. The plane of the scan relative to the reference system can be computed with the information provided within each transaxial image. When an independent system is used, the stereotactic application can be designed so that all necessary quality assurance can be carried out on the images used in each procedure. This is usually accomplished through overdefined fiducial systems or by verifying the geometry of fixed fiducial markers.

67

The second type of stereotactic system, a dependent system, does require special knowledge of the imaging conditions. For biplanar imaging, either orthogonality or exact knowledge of the angle between image planes and the x-ray source to detector distance is required. When a dependent system is used, the quality assurance of the stereotactic procedure must be linked to the quality assurance of the diagnostic equipment at the time of image acquisition. Such quality assurance issues may include the precision of the indexing of the CT patient table, the orthogonality of the CT gantry to the axis of the stereotactic system, the alignment of the laser alignment system within the CT gantry, or the orthogonality of plane film images within a biplanar angiographic procedure. While it is feasible to use dependent reference systems, there appears to be little to gain and a great deal to lose in terms of potential for errors and increased quality assurance time. It is recommended that an independent system be used whenever possible. To bring a better focus to the discussion of stereotactic frames and coordinate systems as applied to radiosurgery and stereotactic radiotherapy, one system, the Brown–Robert–Wells (BRW) system [6], will be used throughout this text. The basic stereotactic system is shown in > Figure 67-1. It consists of a frame, angiographic localizer, CT localizer, and a method of fixing the frame to a linac. This system has three orthogonal axes: anteroposterior, lateral, and axial. The origin of the anteroposterior axis as well as of the lateral axis is defined at the center of the circular reference frame with the origin of the axial axis is defined to be 80 mm from the superior surface of the ring. The function of each localization system is to embed a set of fiducial marks in each image (> Figure 67-2). The goal of treatment planning, as is stated in almost every text written about the process, is to devise a plan that concentrates the dose to the target tissues while sparing all normal tissues. A perfect plan would therefore provide the

1089

1090

67


. Figure 67-1 BRW Headring, patient being CT scanned and patient mounted on stereotactic floorstand aligned for treatment

prescribed dose to the target tissues and no dose to any normal tissue, an impossible task. This would require a dose distribution with an infinitely steep dose gradient, that is, a dose gradient that instantaneously decreases from a target dose to zero dose, i.e., a step function. For a specific target volume the best plan is the plan which most highly conforms to the target volume while providing the minimum dose to all normal, nonetarget, tissues. The multiarc treatment technique used in linac radiosurgery is capable of producing conformal dose volumes that only allow a few tenths of a cc of normal tissue to be included in the prescription volume while providing a dose gradient that

decreases from a specified target dose to a subclinical dose, usually the target isodose level to half that isodose level, within approximately 3 mm. This distribution can be easily achieved using a treatment planning technique know as sphere packing for both circular and non-circular target volumes. The sphere packing technique allows the radiosurgeon to apply a set of non-coplanar beams, usually grouped into several non-coplanar arcs. These arcs effectively allow the planner to apply hundreds of small circular non-coplanar beams, all with unique entrance and exit trajectories to the target volume. The result of such a beam arrangement is a spherical distribution of


67

. Figure 67-2 BRW CT localizer, without headring, aligned with transaxial images showing fiducial rods in each image

dose with the high dose region being the volume where all beams overlap and a rapid decrease in dose as the beams diverge away from the target tissues. This approach represents a class solution that when applied with accompanying rules for sphere spacing and weighting, can produce a highly conformal plan with steep dose gradients at the edge of the treatment volume. The planning process for non-spherical target volumes involves the repeat application of these spheres of dose with the alignment of the outer target edge with the outer shell of the dose spheres, (> Figure 67-3). This technique provides a somewhat inhomogeneous dose inside the target volume. However, unlike routine fractionated radiation therapy, where the target volume contains a high percentage of normal cells, the radiosurgical target volume is nearly 100% target cell, containing almost no normal cells. It is this high concentration of target cells, small total treatment volume of doses about 10% of the prescription dose, as well as minimal dose to normal tissues, that allow sphere packing to provide excellent clinical results.

In many clinical situations it is desirable to provide a high dose gradient along the entire surface of the target volume. This is generally true because most intra cranial tissue has similar sensitivity to radiation. One exception are tissues involved in optical processing. There are clinical situation where it is desirable to provide a nonuniformly steep gradient. In such cases the aim of planning may be to provide the appropriate dose concentration over the target tissues while orienting the steepest possible dose gradient in the direction of the most important adjacent normal tissue structure. For example, when treating a target involving or near the pituitary, the most important adjacent normal tissue structure is usually the optic chiasm. Treating this target with a uniform dose gradient can produce a relatively high dose to the brainstem (> Figure 67-4). A simple modification of this plan can create a much steeper dose gradient in the direction of the optic chiasm that then results in a less steep dose gradient lateral to the target tissues. In this case these lateral tissues are less important than the tissues superior to the target,

1091

1092

67


. Figure 67-3 Non-spherical target planned with three isocenters, image showing the location of each of the three isocenters

. Figure 67-4 Two treatments plans for treating a pituitary tumor. The plan on the left uses a symmetric set of nine arcs and the plan on the right uses six arcs selected to maximize the dose gradient in the direction of the optic chiasm


and this plan modification provides significant clinical benefit. The requirements of a radiosurgical planning system center on providing information necessary to appreciate the target’s size and shape and the relative location of all normal tissues as well as providing the information necessary to optimize dose to the target tissues and dose avoidance to all normal tissues. The parameters available for plan optimization are collimator size, arcing angle, arc placement, isocenter position and in the case of multiple isocenters, relative isocenter weight [7].

Equipment Certification Linac Setup and Isocentric Testing The first procedure in preparing a linac for radiosurgery is to measure the isocentric accuracy of the gantry and patient support systems to allow the combined system’s isocentric accuracy to be evaluated. This procedure does not necessarily parallel the isocentric test performed during acceptance testing or yearly quality assurance testing. In most cases the accepted method of testing the accuracy of a linear accelerator is to individually test the isocentric accuracy of the gantry, collimator and patient support system. Although the customary procedure allows for the accuracy of each subsystem to be evaluated, it does not test the combined accuracy of the gantry and patient support system. If for example, the gantry’s isocentric accuracy is at its high point i.e., outer limit compared to the mean or nominal isocenter, when the gantry is at 0 and the beam is pointed straight down, and if the patient support system is also at one of its extremes at 0 , the combined accuracies of the total gantry plus patient support system will exceed the individual accuracies of either subsystem. It is therefore possible to have both rotational specifications to be within their tolerance

67

of plus or minus 1 mm and have the effective combined accuracy to approach twice the individually certified specification. While these issues are important for all non-coplanar based linac based treatments, they are especially important for older equipment, which is often underused and often relegated to the less frequent procedures such as radiosurgery.

Radiation Testing of Isocenter The testing of radiation isocenter is one of the most important quality assurance procedures in radiosurgery. This procedure provides the certification that the beam delivery system can remain focused on the target coordinate. This test should be executed so that it examines the beamtarget alignment at points throughout the stereotactic space and throughout the extent of beam gantry and patient support motion. For example, if during CT or angiography the stereotactically defined space includes coordinates of – 100 to 100 mm in each of the three orthogonal directions, anteroposterior, lateral and axial, the test should have test points at these extremes. Since more than 95% of the radiosurgical targets are confined to a space defined by a 100 100 100 mm cube, one can make an argument for concentrating on system tests using plus or minus 50-mm points as opposed to plus or minus 100-mm points. However, as will be discuss, pretreatment setup test should always test the accuracy of actual patient specific target settings. It is best to first systematically test each orthogonal axis separately and to then test combinations of axes. It is not sufficient to test the center of the target space or any other single pointing space and to consider that a validation of accuracy across all available stereotactic space. The treatment process requires that tissues throughout the available target space be localized and brought to the linac’s isocenter. The linac QA process must guarantee that errors such as the potential misalignment of axis with the linacs

1093

1094

67


target alignment system are fully evaluated and eliminated. The exact method of carrying out this test depends on the stereotactic system being used. These systems can be divided into basic types. The first division concerns the patient-isocenter alignment. There are systems that use auxiliary patient support systems, such as a floor stand, to position the patient at the system’s isocenter. These systems usually rely upon mechanical micrometer scales for aligning the target tissue to the linac’s isocenter. There are also systems that use the linac’s native patient support system, i.e., patient table support, and rely upon room laser or optical systems for alignment to the linac’s isocenter. The commercial floor stand systems provide very rigid and robust alignment, requiring little dayto-day adjustment as well as providing stability during patient rotations. The table mount systems often require daily certification of laser alignment or calibration of optical alignment systems prior to a radiosurgical procedure. Plus as the equipment ages, these systems can often require added realignment whenever patient table rotations, new arcing planes, are required. Because the stereotactic process begins with diagnostic imaging and ends with the alignment and application of a therapeutic beam, it is critical that institutional procedures exist to certify the entire localization-treatment chain. Many systems provide stereotactic phantoms with targets that are compatible with stereotactic CT image acquisition as well as with megavoltage x-ray beam alignment testing. Some such phantoms are rigid, providing a set of targets, set at known stereotactic coordinates (> Figure 67-5). Others systems provide variable phantoms, allowing for the testing at user defined coordinates. Each of these systems allow for the errors in localization and treatment to be separate measured separately. This is often helpful when trying to minimize system errors during instillation or in evaluating long term system drifts. If neither of these systems are available it is suggested that a commercial phantom system

. Figure 67-5 Absolute stereotactic target. This target is compatible with both the systems CT localizer as well as the linac stereotactic mounting systems. Allowing end-to-end system testing

capable of detecting sub millimeter alignment errors be employed prior to system certification. As previously mentioned the radiosurgery treatment procedure is fully reliant on the mechanical alignment of the stereotactic coordinate system. The testing of system alignment should therefore be a requirement prior to each and every radiosurgery patient treatment. One technique for such a test was initially designed by Ken Winston and Wendell Lutz and is commonly referred to as the Winston-Lutz test [3]. In this test a variable stereotactic phantom is set to the patient’s stereotactic coordinates. The phantom is then mounted onto the linac patient support system and the linac is fitted with the beam collimation system that will be used for treatment. Films are then taken at a variety of gantry patient support settings. These setting are usually selected to demonstrate the alignment of the target to the radiation beam through the range of movement that will be required to execute the patient’s radiosurgical treatment. Initially these alignments were recorded on x-ray sensitive film


and then developed. This has now been replaced by radiochromic films [8] that require no developing and by electronic portal imaging. The accuracy of the treatment can be measured by evaluating the accuracy at which the target is aligned with the center of the therapeutic beam. This test and the subsequent image analysis can usually be carried out in less then 10 min. In carrying out the above test it is recommended that the setting of the stereotactic phantom and the alignment setting of the linac be carried out by two different individuals to add a double blind verification of patient specific alignment parameters. If the test fails, which it will from time to time due to human errors in setting of stereotactic coordinates, the participants should again recheck their own work and repeat the test from scratch. It is important to provide independent verification of setup parameters during this highly precise single fraction therapy.

Stereotactic Imaging Radiosurgical procedures require that all targets be precisely imaged in stereotactic space. The three diagnostic examinations most commonly associated with stereotactic techniques are angiography, CT, and MRI. Because of the necessity to compute radiological paths during the doseplanning phase of the procedure and the need to display the estimated dosimetry on the appropriate anatomical structures, CT is a required database for each radiosurgical procedure. For vascular structures such as arteriovenous malformations, planar angiography is the gold standard for diagnoses. As will be discussed, angiography does have substantial limitations that can make precise three-dimensional planning impossible. MRI, while capable of providing a true threedimensional database, is susceptible to spatial nonuniformities. It is best to use an MRI data set only in combination with and registered to

67

a spatially precise CT data set. Each of these imaging modalities requires complete evaluation prior to use in a radiosurgical procedure.

Plane Film Angiography Angiography remains the gold standard for the identification of an arteriovenous malformation. The angiographic examination allows the clinician to follow the flow of contrast through the malformation. Through rapid sequential imaging the flow of contrast can be followed starting with the arterial phase and progressing through the shunting nidus and into the venous return. Through the 1990s plane film angiography using a sequence of images that freeze in time delivered this detailed progression of the contrast. Typically these images are taken in 0.125–0.25 s intervals. Modern systems depend upon electronic imaging, initially image intensifier technology and more recently flat solid state imaging systems, all of which substantially increased the rate of image acquisition. Throughout this technological progression images were first spatially flat, film systems, then warped when image intensifiers were introduced and now again spatially flat with the introduction of solid-state imagers. It is important to understand the technology being used and the QA and potential correction algorithms required to accurately compute stereotactic coordinates from angiographic views. The implication of orthogonal image sequencing is that when two views that best detail the nidus are chosen, they are guaranteed to be taken at different times. Because of the rapid progression of the contrast in most lesions, this shift in time results in the anteroposterior and the lateral images documenting different physical structures. This can result in a substantial discrepancy between the two orthogonal views. For three-dimensional objects, two orthogonal views cannot always provide an accurate threedimensional description. Such an object is shown

1095

1096

67


in > Figure 67-6. The concave nature of this nidus cannot be appreciated through the two orthogonal views. While it has been suggested that the use of several other noncoplanar views can aid in the reconstruction of nonspherical targets, these solutions solve only a small subset of cases. These tend to be thin, elongated structures that lend themselves to simple orthogonal correlations along their length. In general the nidus of an arteriovenous does not lend itself to such simple deconvolution. This time shift along with the inability to precisely appreciate overlapping structures results I a skew error during data evaluation [9]. . Figure 67-6 Three potential target shapes. While each target can be projected onto an orthogonal plan the resultant orthogonal images cannot be reconstructed to describe the original structure

To relate the nidus to stereotactic space, a set of fiducial markers is attached to the head ring. These markers provide reference points on each angiographic view. From these markers the exact stereotactic coordinate of the center of the spherical object can be obtained. Such a set of films is shown in > Figure 67-7. A true three-dimensional database is required to solve this problem. A contrast-enhanced CT scan can demonstrate the true three-dimensional nature of a nidus. Early CT angiographic sequences required many compromises to avoid CT x-ray tube overheating. It was common for a 1990s circa scan of the head to require 30 min. Modern multi slice scanners have little trouble providing a scan of the entire brain in less than 10 s. These rapid scan sequences coupled with maximum intensity image reconstructions have allowed CT angiography, CTA, to provide high quality true three dimensional information on the shape and location of a nidus. These scan sequences are compatible with routine stereotactic CT image acquisition.

CT Localization Computed tomography (CT) is the most reliable imaging modality for radiosurgery treatment planning. CT numbers correspond directly to electron density. This is important, because the knowledge of electron density within tissue is necessary for correctly calculating the x-ray beam attenuation characteristics. CT also provides an accurate spatial database that introduces little image distortion, and provides an accurate representation of the patient’s external contour and internal anatomy. It is best to obtain stereotactic CT images using a scanner independent fiducial localizer, such as the Brown-RobertsWells (BRW) compatible CT localizer. This localizer attaches to the stereotactic ring using three tooling balls providing a precise and repeatable


67

. Figure 67-7 Anterior and lateral angiographic view showing the nidus of an AVM along with projections of fiducial markers, allowing exact computation of target within the headings stereotactic space

fit. Since the geometry of the localizer is known relative to the head ring, stereotactic coordinates may be accurately calculated based on the CT scan of the localizer. Analyzing the nine rods provides the plane relative to the headring as well as a map of the image pixels to their ring coordinates. The characteristic N shape allows for a scanner independent calculation of stereotactic coordinates. In other words, the x,y,z coordinates of any point in space can be mathematically determined relative to the head ring rather than relying on the CT coordinates. This method provides more accurate spatial localization, and minimizes the CT scanner quality assurance requirements [10]. If a phantom base is part of the radiosurgical system, it can be used as a primary reference in CT image evaluation. While some phantoms are not configured in the proper orientation for insertion into a CT scanner, this can usually be solved by simply modifying the base of the phantom so that it can be properly mounted onto the CT patient support system. As with the evaluation of linac isocentric accuracy, the full stereotactic space should be evaluated. A table similar to > Table 67-1 should be used to determine the test points.

. Table 67-1 Positions for validating the alignment of stereotactic space and treatment delivery space Anteriorposterior mm

Lateral mm

Axial mm

CSV CSV

CSV CSV

CSV

CSV CSV

CSV

Extreme right setting Extreme left setting CSV

CSV

CSV

Extreme anterior space Extreme posterior space

Extreme right space Extreme left space

CSV Extreme anterior setting Extreme posterior setting CSV CSV

CSV Extreme superior setting Extreme inferior setting Extreme superior space Extreme inferior space

CSV – Center of Stereotactic Volume. For the BRW Systems this would be AP = 0.0, Lat = 0.0, Axial = 0.0

When using a three-dimensional database such as CT or MRI, the user should consider the size of the individual pixels during data acquisition. If, for example, a scan diameter of 35 cm is used and a matrix of 256 256 pixels is chosen,

1097

1098

67


the in-plane pixel dimension is 1.37 1.37 mm. If for the same scan diameter a 512 512 matrix isused,theindividualpixeldimensiondecreasesbya factor of four to 0.68 0.68 mm. This change doubles the resolution of the in-plane image. Similarly, if one uses thick slices during image acquisition, the out-of-plane resolution is degraded. This point can be seen by looking at the coronal reconstruction of an axially acquired data set. The coronal image shown in > Figure 67-8 has been reformatted from a transaxial data set that contained 0.68 0.68 1 mm pixels. > Figure 67-8 also demonstrates the same reformatting, except this time the data set was sampled from transaxial data

with pixel dimensions of 0.68 0.68 5 mm. As can be seen, the vertical resolution of the second image has been significantly degraded.

Magnetic Resonance Imaging Localization If an MRI-compatible stereotactic system is available, it can be tested in the same manner as the CT system. MRI has capabilities not available on CT, and MRI is susceptible to image perturbations not found in stereotactic CT scans. For example, MRI allows image acquisition in planes

. Figure 67-8 Object and two sets of 3D renderings and sagittal reconstructions. The left set are from 1 mm thick CT scans and the ones on the right are from 5 mm thick CT scans


other than transaxial. Sagittal as well as coronal images are routinely available. Also, many different MRI sequences can be used. Depending on the target tissues, T1 or T2 weighted images may be required to provide target normal tissue contrast. Each of these parameters can introduce image nonlinearities that must be evaluated. It is also important to appreciate image perturbations induced by the patient being scanned. It is often difficult to appreciate the distortions that MR scanning can introduce if direct MR scanning is used. Most MR fields of view are most uniform towards the center of the scan and less uniform towards the outer edges. When one examines the stereotactic localizers one realizes that the fiducials that are used to map imaged pixels to ring coordinates lie towards the outer edge, the region that is most likely to have spatial non-uniformities. It should also be realized that susceptibility artifacts occur in regions where large local changes in MR response are present. The fiducial markers usually have high signal values and sit in air, which has an exceedingly low MR signal. These two factors combine to produce unreliable MR stereotactic coordinate mapping. It is therefore recommended that stereotactic MRI never be used as a stereotactic database without the correlation to an identical stereotactic CT database. As a replacement to directly obtained stereotactic MRI, many systems have adopted image fusion. In this technique a nonstereotactic MRI is obtained prior to the application of the stereotactic head ring. A stereotactic CT scan is obtained after ring placement. Through one of several techniques the nonstereotactic MRI is rotated, translated, and sized to match the CT scan pixel for pixel. Once this has been accomplished, a new stereotactic MRI database can be prepared for target localization and treatment planning. > Figure 67-9 shows a split-screen view of such a matched data set. On the right side of each view is the MRI data and on the left side is the CT data. As can be seen, the internal anatomy has

67

been matched. In > Figure 67-10 a set of split views is shown at the target level.

Treatment Plan Optimization As with other treatment optimization systems, each radiosurgery planning system is based on a specific dose model. Unlike radiotherapy planning systems, which must accommodate large radiation portals, typically 5 5 cm–40 40 cm, the radiosurgical planning system’s range is approximately 4 cm down to 0.5 cm in diameter. These small fields present some special problems as well as some unique opportunities for radiosurgical beam models. The limitations on field size enable the radiosurgical systems to use relatively simple dose models. For circular fields a model that predicts field shape according to the product of tissue phantom ratios and off-axis ratios is usually sufficient. For irregular fields this model can be expanded to account for the irregular fields and field edge effects. The small fields used allow for the models to remain relatively simple as compared with those of normal external beam planning systems.

Photon Radiosurgery Paradigm Radiosurgery requires large dose deposition within the target volume and a steep dose gradient resulting in very little dose delivered to normal tissue. Clearly, the single static photon beam fails to meet these criteria. Any target that is not located at the exact depth of dmax (the depth at which a dose of radiation reaches maximum intensity) has a tissue–maximum ratio of less than 1, and the point of dmax will receive a higher dose than will the target. Since beam intensity only decreases at approximately 5–6% per centimeter up and down the beam path from isocenter, tissues near the target will receive nearly the full target dose.

1099

1100

67


. Figure 67-9 Inspection windows for judging proper alignment of CT, left and bottom, and MR, right and top, scans

The basic photon radiosurgery paradigm relies on the use of multiple tightly focused beams that use unique entry paths and converge on a point of interest. This concept is illustrated in > Figure 67-11, in which 16-mm diameter 6-MV x-ray beams converge in the center of a 17.5-cm-diameter spherical phantom. For a static beam the target (isocenter) dose is 64% of the maximum dose. With two beams, the target dose is 2 (64%) = 128% of the dose at either dmax. After normalizing to the point of maximum dose, the dose at entry dmax for either beam is only 78% of the maximum dose. Similarly, the four beams shown in > Figure 67-11c result in a peak dose of 4 (64%) = 256%, and the eight

beams in > Figure 67-11d result in 512% at the isocenter. Normalizing these to the point of maximum dose demonstrates the reduction in relative dose delivered to the entry dmax for each beam. Adding more beams that coincide only at the point of interest results in a dose distribution that is highly peaked at their point of intersection and that has a steep dose gradient outside of the target volume. One may logically surmise that the optimal dose gradient results from an infinite number of beams that irradiate the sphere from all possible entry angles (4p irradiation). Because of the physical constraints of the patient’s geometry in relation to the dose delivery systems, 4p irradiation is


67

. Figure 67-10 Inspection windows for judging alignment of Ct and MR scans taken through target region

impossible to achieve. All photon radiosurgical paradigms are attempts to irradiate isotopically typically over half of a sphere, or 2p irradiation. Linacs use multiple noncoplanar arcs focused on the target. > Figure 67-12 demonstrates commonly used 2p irradiation schemata. Almost all of the published literature detailing clinical results for radiosurgical treatments have relied on circular collimators that resultant in dose distributions are nominally spherical for the target isodose lines. For most typical plans between 5 and 11 non-coplanar arc are arranged, generally covering the available 2p. These beam arrangements allow for many hundreds of beams to be applied to the target. Each having unique

entrance and exit pathways, only overlapping over the target tissues. When fewer beams are used then less geometric beam concentration is available, each beam must supply a higher percentage of the total target dose along its entrance and exit pathway, resulting in a decrease in dose concentration and higher doses to tissues along the various beam paths. The high geometric concentration of dose not only provides a high dose concentration but the rapid divergence of beam paths provides a very steep dose decrease as one moves away from the edge of the area of geometric overlap. Typically this decrease in dose, referred to as the dose gradient, allows for the dose to drop from

1101

1102

67


. Figure 67-11 Intensity profiles of one beam, two beams, four beams, eight beams and 36 beams demonstrating the increase in concentration of dose with increasing geometric convergence of beams at the target point and divergence away from the target point

. Figure 67-12 Set of nine arcs evenly spaced

the target or prescription dose to half that dose in approximately 3–4 mm in all directions. In > Figure 67-13 a profile of the radiation dose intensity across the spherical dose distribution. From this curve one can appreciate the concept behind single isocenter radiosurgical doses being prescribed to the 80% isodose line. As mentioned earlier the radiosurgical target volume usually has a very high concentration of

tumor cells. The aim of the therapy is to provide a dose sufficient to eliminate the cells within the target. While the target volume contains almost 100% target cells the non-target volume, i.e., normal tissue volume, contains almost no target cells. Just as one aim of radiosurgery is to provide cell death to all target cells, a parallel and equally important goal is to minimize the dose to normal tissues. To achieve both goals it is advantageous


. Figure 67-13 A plot of the intensity of a single isocenter using five evenly spaced arcs. The choice of a target isodose that coincides with the steep dose fall off, i.e., the 80% of maximum dose, reduces the volume of the tissue receiving greater then one half the prescribed tumor dose

to prescribe to a point along the surface of the dose distribution where the dose is decreasing very rapidly. If such a prescription point is placed at the edge of the target volume then the shell of normal tissue exposed to the high prescription dose will be minimized. This principle is demonstrated in > Figure 67-14, which contains the above mentioned dose cross plot. If, as is shown in > Figure 67-14, the 95% isodose point is chosen as the prescription point then the dose would decrease from 95 to 47.5%, i.e., half of the prescription dose, in 5.7 mm. If, as in > Figure 67-14 the 90% isodose point was placed at the edge of the target volume then it would require 4.5 mm for the dose to decrease from the prescription level, 90%, to half that intensity, 40%. If we repeat this exercise for the 80% isodose level we see the distance is 3.8 mm and the 70% isodose level we see that 3.9 mm is required. It can therefore be seen that when the 80% isodose point is chosen for the prescription isodose level the thickness of the shell between target dose and half target dose is minimized. When non-spherical targets are encountered then the single isocenter treatment technique is

67

expanded to utilize not a single large sphere to cover the target while including normal tissues but rather the packing of the target volume with multiple spheres covering or packing the volume, with spherical distributions. In general the spheres are arranged such that the outside of the target volume is aligned with the outer edge of various spheres. This allows the high dose gradient to be aligned with the target-normal tissue interface. This provides prescription dose to the target while minimizing the dose to normal tissues. There is only one caveat when using multiple isocenters and that is the shift in prescription isodose level from 80 to 70% of the maximum dose in the target volume. This shift is the result of very small hot spots that result from spheres being pack next to one another. Since these small hot spots are within the target volume and as we have mentioned all target cells are to receive doses that result in cell death no negative clinical results have ever been reported due to these hotspots. For many spherical targets a single isocenter is sufficient to provide 100% target coverage with exceedingly little normal tissue contained within the prescription isodose shell. The next level of complexity is for targets that are elongated. Many acoustic neurinoma have relative circular portions that are adjacent to the brain step and a section that extends down the internal auditory canal. This shape can easily be covered by two isocenters. > Figure 67-15 shows such a target and prescription isodose. In this case the first, a 14 mm isocenter covering the more medial spherical portion of the target volume and a 10 mm centered towards the lateral extent of the target. The combination of these two spherical isocenters results in a very conformal treatment. As mentioned earlier the technique of sphere packing allows the edge of the sphere, which has a very steep dose gradient to be aligned with the edge of the target volume. In this case the first isocenter is placed with this steep dose

1103

1104

67


. Figure 67-14 A set of graphs demonstrating that selecting the prescription isodose at the 80% of max dose provides a more rapid fall off then selecting the 95, 90 or 70% of the maximum dose points

edge providing the steep gradient in the direction of the brain stem. The region where the twoisocenter plans present the previously mentioned hot spot is along the line that joins the two isocenters. As mentioned this is always within the target volume and has not been shown to result in any treatment complications. Examining this distribution further one finds that the target volume is 0.8 cc and the prescription isodose volume is 0.9 cc providing only 0.1 cc of normal tissue within the prescription treatment volume. One can also determine that the volume of the isodose shell that encompasses one-half the treatment isodose value, in this case the prescription was 1250 cGy to the 70% isodose shell

so the one-half treatment shell would be the 35% isodose shell which would deliver 625 cGy, is only.95 cc. This shell is on average 4.1 mm from the treatment isodose shell. This means that the on average the dose decreases from 1250 to 625 cGy in less then 4.1 mm. This same principle can be applied to significantly more complex targets. > Figure 67-16 demonstrated the application of multiple isocenters to cover a complex target with adjacent critical structures. The important parameters in developing the above plan are the selection of the correct collimators, the correct placement of the isocenters and the relative intensities of the isocenters. The planning system should


have tools to allow these optimization parameters to be readily manipulated and optimized. > Figure 67-17 shows a planning system screen with the tools required for isocenter placement, spacing and weighting.

. Figure 67-15 Isodose distribution for the treatment of a nonspherical acoustic neurinoma. Demonstrating the typical degree of conformality and dose gradient obtainable using two correctly placed, spaced and weighted isocenters

67

The above examples are can be planned on either a fully merged CT-MR dataset or preplanned on a non-stereotactic MR prior to ring placement. This function allows the clinician to obtain the MR targeting scan days before the planned therapy and to plan the treatment at their leisure. On the day of treatment the stereotactic headring can be rigidly fixed to the patient, the preplan transferred to the fused dataset and rebalanced. This process can easily be automated, resulting in a planning session on the day of therapy that is simply an inspection of the previously-approved plan. Using this process the isodose coverage at the time of preplanning is identical to the coverage after plan transfer and rebalancing. The described arcing delivery paradigm is one method of providing the requisite number of noncoplanar beams that converge precisely at a target point and rapidly diverge while traveling through non tissue tissues. As previously mentioned Leksell [11] had developed a dedicated radiation delivery unit that achieved almost identical spherical dose distributions, with accompanying steep dose gradients. The most widely commercialized design consisted of 201 non-coplanar pencil radiation

. Figure 67-16 Multi-isocenter circular cone plan demonstrating the ability of the technique to not only provide a conformal plan but to also maintain a steep dose gradient

1105

1106

67


. Figure 67-17 Screen showing the tools used to produce the plan in figure 15. Isocenter placing, weighting and spacing tools are interactively used to optimize the dose distribution

beams. While 201 beam provided almost identical dose distributions as the five non-coplanar arcs spread out over nearly 2p it has been demonstrated that an equivalent distribution can be delivered from a few as 15 beam spread across the same geometry [12]. Alternate approaches for providing the needed high target conformality as well as providing normal tissue sparing have used a technique that modulates the intensity of the radiation beam, IMRT. One of the early explanations of intensity modulation suggested that this could be achieved through the previously described sphere packing technique [13], however, iterative dose

optimizations have been used to implement a majority of the commercial intensity modulation algorithms [14]. The use of IMRT has found wide clinical acceptance in larger field radiation dose delivery. While IMRT has shown to provide equivalent conformal target coverage, it has also been shown to provide overall shallower dose gradients [15,16] when compared to multiple isocenter circular beam techniques. This may in part be due to the tradeoff of fewer beams being utilized. The fewer beams are in part due to the computation time required for each fiend, the QA required per IMRT beam or the simple extrapolation of an adequate number of larger fields to a


small field technique. The previously reference work of St John [12] indicates that utilizing fewer then 15 non-coplanar beam paths will yield this lower dose gradient and higher doses to normal tissues. A solution to the initial IMRT problem posed by Barth [13] has been demonstrated by St John. This planning and delivery technique provides the radiosurgeon with the conformality and gradient benefits of multiple isocenter planning, combined with the speed of single isocenter delivery. When this approach is combined with the automated multi isocenter planning, as demonstrated by Wagner [17] plans as demonstrated in > Figure 67-16 can be derived in under 1 min of cpu planning time and delivered in under 20 min. Aside from the general desire to reduce dose to normal tissue the 12 Gy dose volume has been shown to correlate with permanent radiation induced complications [18–22]. With the upper dose for many radiosurgical targets being in the 20 to 22.5 Gy range it is critical to provide a very steep dose gradient, one that decreases the target dose to approximately one half the target dose value over a short distance. With the critical parameters being conformality and dose gradient a simple metric for scoring radiosurgical plans has been suggested [23]. This metric is comprised of two scores, one for inclusion of normal tissue in the prescription isodose shell and a second that measures the average dose gradient between the prescription isodose shell and the shell that encloses the one-half the prescription isodose value. This is the gradient previously discussed and shown in > Figure 67-15. When multiple sets of rival stereotactic radiosurgery plans were ranked with respect to this single score index, the resulting plan rankings closely matched the plan rankings according to biologic indices (calculated nontarget brain normal tissue complication probabilities). When applied to radiosurgery treatment outcomes have been shown to have sensitivities

67

to not only conformality and dose gradients but also the time required to deliver the treatment. It has been demonstrated that when treatment times extend beyond 15–20 min that sub lethal damage repair begins to decrease the overall effect of the prescription dose [24,25]. For multiple isocenter planning and treatment delivery this is usually not a problem. Although a multiple isocenter plan may require over an hour to fully deliver each individual isocenter seldom requires more then 20 min of dose delivery time. So although the total treatment delivery to the acoustic neurinoma shown in > Figure 67-15 may require 25 min the cells covered by each of the two isocenters receive their full dose in less then 15 min. Once delivered moving to the next isocenter then treats the next subsection of the target, again in less than 20 min. This may not be the case if complex modulations are required or if extended alignments are required between non-coplanar beam sets.

Treatment The final treatment plan may have several isocenters each requiring a set of arcs and one or more collimators. As previously discussed the setup of the treatment unit and setting of the patient specific target coordinates is best carried and tested using a double-blinded phantom test. It is also very helpful to have a checklist that walks the treatment team through the entire procedure from first step to last. > Figure 67-18 shows several pages of the checklist for the above detailed two-isocenter treatment. It is suggested that at a minimum of three treatment staff be involved in the setup and treatment process: Two staff to setup of the linac and phantom for initial testing plus a third team member whose sole purpose is not to setup the patient or the linac but to observe and provide a blinded check of all settings and procedures.

1107

1108

67


. Figure 67-18 Printout of single isocenter plan. This printout is used to guide the QA during treatment delivery. Each treatment team member checks that the treatment plan is precisely followed

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-19. 2. Colombo F, Benedetti A, Pozza F, et al. Stereotactic radiosurgery utilizing a linear accelerator. Appl Neurophysiol 1985;48:133-45. 3. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988;22:454-64. 4. Friedman WA, Bova FJ. The University of Florida radiosurgery system. Surg Neurol 1989;32:334-42.

5. Bova FJ, Meeks SL, Friedman WA, Buatti JM. Optic-guided stereotactic radiosurgery. Med Dosim 1998;23:221-8. 6. Brown RS. A computerized tomography–computer graphics approach to stereotaxic localization. J Neurosurg 1979;50:715-20. 7. Friedman WA, Buatti JM, Bova FJ, et al. LINAC Radiosurgery – a practical guide. Berlin: Springer; 1998. 8. Mack A, Czempiel H, Kreiner HJ, Du¨rr G, Wowra B. Quality assurance in stereotactic space. A system test for verifying the accuracy of aim in radiosurgery. Med Phys 2002;29(4):561-8. 9. Siddon RL, Barth NH. Stereotaxic localization of intracranial targets. Int J Radiat Oncol Biol Phys 1987; 13:1241-6. 10. Huh SN. Incorporation of magnetic resonance imaging and digital angiography in the application of stereotactic radiosurgery. Gainesville: University of Florida; 1994 (dissertation). 11. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two cases of intractable pain. Acta Chir Scand 1968;134(8):585-95. 12. St John TJ, Wagner TH, Bova FJ, Friedman WA, Meeks SL. A geometrically based method of step and shoot stereotactic radiosurgery with a miniature multileaf collimator. Phys Med Biol 2005;50(14):3263-76. 13. Barth, NH. An inverse problem in radiation therapy. Int J Radiat Oncol Biol Phys 1990;18(2):425-31. 14. Webb S. Advances in treatment with intensity-modulated conformal radiotherapy. Tumori 1998;84(2):112-26. 15. Nakamura JL, Pirzkall A, Carol MP, Xia P, Smith V, Wara WM, Petti PL, Verhey LJ, Sneed PK. Comparison of intensity-modulated radiosurgery with gamma knife radiosurgery for challenging skull base lesions. Int J Radiat Oncol Biol Phys 2003;55(1):99-109. 16. Sankaranarayanan V, Ganesan S, Oommen S, Padmanaban TK, Stumpf J, Ayyangar KM. Study on dosimetric parameters for stereotactic radiosurgery and intensity-modulated radiotherapy. Med Dosim 2003;28 (2):85-90. 17. Wagner TH, Yi T, Meeks SL, Bova FJ, Brechner BL, Chen Y, Buatti JM, Friedman WA, Foote KD, Bouchet LG. A geometrically based method for automated radiosurgery planning. Int J Radiat Oncol Biol Phys 2000;48(5):1599-611. 18. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. Analysis of neurological sequelae from radiosurgery of arteriovenous malformations: how location affects outcome. Int J Radiat Oncol Biol Phys 1998;40(2):273-8. 19. Flickinger JC, Kondziolka D, Lunsford LD, Pollock BE, Yamamoto M, Gorman DA, Schomberg PJ, Sneed P, Larson D, Smith V, McDermott MW, Miyawaki L, Chilton J, Morantz RA, Young B, Jokura H, Liscak R. A multi-institutional analysis of complication outcomes after arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 1999;44(1):67-74.


20. Flickinger JC, Kondziolka D, Lunsford LD, Kassam A, Phuong LK, Liscak R, Pollock B, Arteriovenous Malformation Radiosurgery Study Group. Development of a model to predict permanent symptomatic postradiosurgery injury for arteriovenous malformation patients. Int J Radiat Oncol Biol Phys 2000;46(5):1143-8. 21. Korytko T, Radivoyevitch T, Colussi V, Wessels BW, Pillai K, Maciunas RJ. Einstein DB. 12 Gy gamma knife radiosurgical volume is a predictor for radiation necrosis in non-AVM intracranial tumors. Int J Radiat Oncol Biol Phys 2006;64(2):419-24. 22. Friedman WA, Bova FJ, Bollampally S, Bradshaw P. Analysis of factors predictive of success or complications in arteriovenous malformation radiosurgery. Neurosurgery 2003;52(2):296-307; discussion 307–8.

67

23. Wagner TH, Bova FJ, Friedman WA, Buatti JM, Bouchet LG, Meeks SL. A simple and reliable index for scoring rival stereotactic radiosurgery plans. Int J Radiat Oncol Biol Phys 2003;57(4):1141-9. 24. Shibamoto Y, Ito M, Sugie C, Ogino H, Hara M. Recovery from sublethal damage during intermittent exposures in cultured tumor cells: implications for dose modification in radiosurgery and IMRT. Int J Radiat Oncol Biol Phys 2004;59(5):1484-90. 25. Tomita N, Shibamoto Y, Ito M, Ogino H, Sugie C, Ayakawa S, Iwata H. Biological effect of intermittent radiation exposure in vivo: recovery from sublethal damage versus reoxygenation. Radiother Oncol 2008;86(3):369-74.

1109

57 Overview of Radiosurgery Technology M. Schulder

Introduction Stereotactic radiosurgery (SRS) is still often referred to as a new or ‘‘high tech’’ treatment. And yet it is 57 years since Lars Leksell invented the concept and coined the term radiosurgery [1]. After experimenting with orthovoltage irradiation and considering heavy particle use, Leksell invented the Gamma Knife (GK) in 1967 [2]. In the early 1980s, Betti and Derechinsky, and Colombo, et al. adapted linear accelerators (Linacs) to deliver photon based SRS [3–6], thereby ushering in the second generation of SRS devices. Linacs offered a promising alternative form of external beam radiotherapy to the gamma knife [7–9]. Improvements in imaging and especially in computer power led to ever-increasing sophistication, speed, and clinical use of SRS. And yet, what had not changed for nearly 50 years was the concept that delivering SRS required the use of isocenters and either a Linac or a GK to deliver a conformal treatment with a steep dose gradient. This notion has been termed ‘‘sphere packing’’ or the use of multiple spherical isocenters created by circular collimators for irradiating irregularly shaped lesions [8]. SRS, invented as a single-session technique of minimally invasive neurosurgery utilizing ionizing radiation, was understood to be completely different from standard, fractionated methods of radiation therapy. In the late 1990s this paradigm began to change. Radiation oncologists came to appreciate the value of stereotactic localization,

#


now available without the use of rigid frames, and neurosurgeons came to understand the potential benefits of fractionation [10–12]. Though each treatment was distinct, to some degree the line between SRS and stereotactic radiotherapy (SRT) became blurred. Studies confirmed that fractionated dose delivery of SRT offers the combined benefits of stereotactic targeting and minimal radiobiologic harm to normal tissue [13,14]. SRS treatment plans still should be based on a set of quantitative parameters. Dosimetry indices for SRS planning include the coverage index (the isodose surface that covers the entire target volume as a fraction of the prescription isodose); the homogeneity index (the ratio of the maximum dose in the target volume to the prescription isodose); and the conformity index (the ratio of the prescription isodose volume to the target volume) the last two of which should be less than two [15,16]. The conceptual changes noted above, together with great leaps in the technology of radiation delivery, have transformed the radiosurgical landscape. Rapid innovations in Linacbased devices, and to some extent in charged particle SRS technology, are redefining the field [17]. The implementation of multiaxial robotic technology in certain Linac machines has greatly refined the integrative capabilities of SRS devices in all aspects of planning and treatment [3,7,18]. This chapter focuses mainly on the newer, commercially available radiotherapy applications of Linac and GK, the most common forms of SRS worldwide.

868

57

Overview of radiosurgery technology

Linear Accelerator SRS Linacs produce a radiation beam with a magnetron and gantry. Accelerated electrons generated from the magnetron are injected into an evacuated copper tube. There, they collide with a heavy metal target (typically tungsten) to produce photons. The scattered photons are subsequently focused through a portal to create a directed, or collimated, beam of gamma radiation. The X-rays produced by modern Linac electron guns usually have higher energies (4–6 MeV) than those generated through isotope decay [19]. The Linac is mounted to a gantry that rotates through an arc around the patient. This frame-gantry geometry of Linacs is suitable for treating lesions throughout the body. The circular gantry movement lends itself to the development of Linac based SRS, wherein the use of multiple noncoplanar arcs rotated around an isocenter deposit a high dose of radiation at that isocenter in essentially an additive fashion. Treatment volume could be altered by changing collimator diameter. Adjusting the spherical shape of a typical unmodified treatment would be done by adding isocenters, and to a lesser extent by adjusting beam length and weight. Recent technological advances in the dosimetry and physics of SRS have led to a new generation of treatment planning software that has simplified and increased the versatility of Linacbased SRS. The newest generation of Linac-based SRS devices includes X-Knife (Radionics, Burlington, MA), Novalis (Helmstetten, Germany), Trilogy (Varian Medical Systems, Palo Alto, CA), and CyberKnife (CK) (Accuray, Sunnyvale, CA). Compared to the GK, Linac sources offer a wider array of treatment options for both traditional SRS and fractionated stereotactic radiotherapy treatment [17]. Other technical and electronic innovations have further enhanced the accuracy and precision of the newest versions of Linac radiosurgical devices. Nonetheless, the mechanical complexity of these devices demands that strict

quality control measures be implemented and routinely followed [17]. The finely collimated X-ray beams from Linac sources can be effectively targeted to both intracranial and extracranial lesions. Replicable data suggest that planning and treatment of Linac SRS is comparable overall to those of the Gamma Knife [17,20]. The physical characteristics of Linac devices render them capable of irradiating extracranial lesions such as tumors of the spine, prostate and lung [17,21]. Treatment of each isocenter can usually be performed in less than 10 min on a Linac-based SRS device [19]. Linac systems can operate using conventional circular arcs, conformal SRS based on multiple static-shaped beams, dynamic arc SRS employing micromultileaf collimator (MMLC) field shaping, and intensity modulated radiotherapy (IMRT). While all of these techniques can be implemented, MMLCs lie at the core of modern Linac-based applications for SRS [19].

Micromulitleaf Collimators and Inverse Planning Multileaf collimation for beam shaping and delivery originated in Japan in the 1960s [19]. By the 1990s, medical physicists and clinicians had learned to shape focused radiation to conform to irregular concave surfaces while decreasing the dose reaching normal tissues. This approach involved modulating the intensity of the radiation beam within the field of radiation, and it signaled the advent of IMRT. The technique was later refined with the use of a computercontrolled, flexible high resolution beam shaping device called a multileaf collimator (MLC). MLCs are used to shape the radiation beam during IMRT. The micro-multileaf collimator (MMLC) was eventually designed for smaller intracranial lesions that could not be treated with beams transmitted through collimators with wider leaf widths [3,19].


SRS treatment planning evolved as an iterative, forward planning process. Using a computer model based on stereotactic computed tomography (CT) and/or magnetic resonance imaging (MRI), clinicians and physicists place isocenters in the target and examine such features as conformality, dose gradient, and dosing of adjacent areas. This process can be very time consuming when planning SRS using circular collimators for patients whose lesions have complex shapes and, in particular, are close to such critical structures as the optic chiasm or brainstem. Inverse treatment planning evolved as a numerical optimization strategy for finding optimizing beam directions during SRS and IMRT procedures. In 1996 Carol et al. described their experience with the Linac-based Peacock system, designed to deliver highly conformal radiation using inverse planning and MMLC [22]. The Peacock system was used for fully fractionated radiation therapy (RT), and required implantation of skull screws for repeat fixation. While not truly ‘‘radiosurgical,’’ this invention set the stage for what has become an increasingly common and powerful concept in SRS: the use of the MMLC. The MMLC contains many small vanes that change position as the Linac rotates. Consequently, the need for ‘‘sphere packing’’ is avoided, and even complex shapes can be treated without the need to reposition the patient. Only target volume, the risk for critical volumes, and other clinical considerations need limit the use of SRS. In essence, by using an MMLC and an inverse planning algorithm, clinicians can shape the dose such that it has conformality for the target lesion. When used with either an IMRT or nonisocentric beam, the inverse planning algorithm spares, or ‘‘wraps around,’’ around healthy tissue [21]. Yet, planning these treatments in a ‘‘traditional,’’ forward manner is a formidable task. By taking advantage of modern computer speed and power, however, the treatment team can define key parameters such as the target and the dose

57

tolerance of adjacent structures. Technologically sophisticated Linac-based SRS systems, including some with computer-controlled robotic functions for MMLC performance, generate a plan based on these constraints. The physicians and physicists can adjust the plan by changing various parameters of the treatment plan, depending upon the core features and options available through the particular SRS system software. This section will discuss several systems that bring new technology to bear on SRS, each from a different perspective.

Adapted Linac: Radionics XKnife Description The Radionics XKnife (Integra Radionics, Burlington MA) was the first dedicated Linac for SRS, developed in the 1980s [9] and made commercially available shortly thereafter. It is a 3-dimensional (3-D) treatment planning software and hardware system meant to adapt a conventional Linac for SRS and SRT. Before the advent of the MMLC, radiosurgical plans and delivery were realized using multiple, non-coplanar arcs. The Linac rotates around a fixed isocenter and target shaping was done by changing couch angle and to a lesser extent by adjusting arc length and weighting. This system was used with excellent results [23]. However, as treatment technology evolved, the limitations of the couch mount system spurred certain technological advances. In particular, conformal treatments of nonspherical targets were difficult to achieve with the couch mount, as isocenter changes were time-consuming and required multiple steps to verify. This was an obstacle to overcome, especially for patients needing SRS for benign tumors or functional disorders. The current XKnife version features tighter tolerances for isocentric rotation and fixed primary and collimation units that minimize

869

870

57


gantry sag. The basic principle of a rotating gantry that irradiates X-rays to the patient lying on a couch during the procedure has remained unchanged. However, the system has evolved with major innovations such as treatment planning algorithms, fractionated SRT treatment protocols, and a relocatable frame. The XKnife is compatible with most Linacs and requires only simple mounting and dismounting. Components of the latest XKnife include stereotactic localizers and immobilizers, SRS collimators, a MMLC, and instrumentation for enhanced quality assurance. The precision and accuracy of the XKnife treatments depend totally on the rotational accuracy of the treatment couch and gantry [24], both of which are computer-controlled by the XKnife software. The device contains advanced algorithms for the localization, planning, setup, and delivery of radiation therapy without the need for separate software. In addition to treatment planning and SRS, conformal MMLC uses are included.

controlled by the XKnife Version 4 IMRT software. Using equipment already on the linear accelerator, the XKnife Version 4 IMRT customizes dose delivery for treating brain tumors and vascular malformations as well as head, neck, spine, and body tumors. With RT real-time planning, the XKnife Version 4 is equipped to deliver collimated beam radiation therapy in frameless stereotaxy [25]. Cranial accessories for the XKnife include the standard BRW head ring, the GTC relocatable head ring, and the TLC pediatric head ring. These devices are used for precise fixation, localization, and repositioning of targets in the head and neck. Target sites in the spine and body are immobilized and localized for fractionated SRT using a vacuum-based stereotactic system. The Radionics HDRT is a comprehensive platform that manages patient repositioning and internal anatomical shifts during extracranial frameless stereotactic treatments. The HDRT integrates Radionics stereotactic localizers, XKnife IMRT, Linac-based CT imaging, and Radionics stereotactic ImageFusion [25].

Performance Features During SRS or SRT, the XKnife MMLC provides greater conformality than do circular collimators or blocks. This hybrid device contains 62 leaves that can be opened to various widths along the rotational arc of the gantry, as dictated by the treatment plan. The uniform leaf width of 4 mm at the isocenter is appropriate for finely shaping small and large field patterns. Through computerized control of the motorized leaves of the MMLC, the clinician can maximize the delivery of dose to the target volume with minimal radiation to nearby normal tissue. The MMLC has a large field size of 10 cm 12 cm for treating irregular tumor volumes, including lesions in the brain, head and neck, and spine. Other advanced optimization techniques providing increased dose conformality are also

Evidence-based Studies on Quality and Efficacy of the XKnife Urie et al. compared the Radionics circular collimators with the MMLC XKnife treatment planning system to determine any differences in tissue maximum ratios (TMRs). Beam data, including TMRs, output factors, penumbrae, and isodose distributions of these fields were measured. The mechanical accuracy of the MMLC compared well with that of the circular collimators, while also producing a smaller leakage dose. Dose distributions were basically equivalent for circular collimators and for simulated circular fields (total, 360) of four arcs of the MMLC. The resulting TMRs were the same as those of circular collimators of equivalent size.


However, since the MMLC is bulkier than the circular collimators, mechanical collisions may occur [26]. Recent clinical data support the technical efficiency and safety of the Radionics SRS system. Hillard et al. conducted a retrospective study of 10 patients undergoing multiple SRS treatments with the XKnife for multiple brain lesions. Each patient received at least two treatments delivered to at least three isocenters with a minimum follow-up period of 6 months. The average of the maximum doses directed to a point within the whole brain and or a critical brain structure varied widely, ranging from 159 cGy to the left optic nerve to 2,402 cGy within the brain. The authors found no complications that could be attributed to the SRS, except for one patient who developed seizures linked with radiation necrosis. Hillard et al. concluded that multiple SRS treatments at the cumulative doses delivered in this study were safe for patients with multiple brain lesions [27]. Plowman and Doughty compared dose gradient ‘‘fall-off ’’ at the margin of the target (the distance between isodoses) between the Gamma Knife and XKnife. Both techniques produced similar values for isocenter treatment volumes up to 1.5 cm diameter, but these results diverged when planning treatments for larger and more complex targets. While more frequent multiple isocenter ‘‘shots’’ of the gamma unit produced greater conformity in complex target volumes, it did so at the cost of steeper internal dose gradients. For instance, in the treatment of an acoustic neuroma (vestibular schwannoma), the XKnife generated a considerably smaller internal dose gradient (11%) than in the GK (100%) (28). That said, the clinical significance of this dose heterogeneity is uncertain. The authors suggested that this may have contributed to improved hearing preservation in patients with acoustic neuroma who have been treated with XKnife SRT [28]. Other evidence also supports

57

the safety and efficacy of the XKnife for acoustic neuromas. McClelland evaluated 20 patients who received fractionated SRT (54 Gy in 1.8 Gy daily fractions) via the XKnife 4.0 3-D planning system for the treatment of acoustic neuromas. Local tumor control occurred in each patient and there were no complications All nine patients who had preserved hearing prior to treatment had sustained hearing preservation at the last follow-up. Four patients experienced decreased hearing after treatment [29]. > Figure 57‐1 shows a conformal plan for single session SRS with the XKnife. Experience with the XKnife indicates that adapting a regular Linac with inverse planning SRS software and an MMLC allows for the full gamut of single session and fractionated stereotactic treatments to be delivered with safety and efficacy. This may indeed be an attractive option for centers looking to upgrade an older SRS system or to install a new one, in that they may not want to justify the installation of a dedicated radiosurgical device.

Dedicated Isocentric Linac for SRS: Novalis Tx System Description The Novalis SRS system was among the first devices in a new generation of 6-MeV electron Linacs that emerged during the mid 1990s [30]. The Novalis Tx system, a joint venture of a stereotactic neurosurgical vendor (Munich) and a Linac manufacturer (Varian, Palo Alto, CA) utilizes a dedicated Linac, MMLC for cone-based beam shaping, and a treatment planning system, which are all integrated through an information management system. Designed for rapid and high dose delivery, the Novalis Tx is used to treat lesions that range widely in size and shape. As in the case of Linac-based radiotherapy

871

872

57


. Figure 57‐1 XKnife treatment plan for a 54-year-old man with a left acoustic neuroma. The 1,250 cGy prescription dose is conformal, and the plan has a craniocaudal orientation to minimize irradiation of the brainstem

devices in general, this system is designed for both single, high-dose sessions in SRS and multiple, smaller dose sessions in SRT [31].

Performance Features The Novalis Tx can generate radiation fields from a minimum of 3 mm 3 mm to a maximum of 98 mm 98 mm in size. It can rapidly deliver multiple beam energies of up to 18 MeV for targeting deep-seated lesions. The dynamic, computer-controlled, fine ‘‘beam shaping’’ capabilities of the MMLC are able to precisely match the contour of the target lesion. Full doses of radiation can be delivered to tumors with varying complicated shapes while sparing irradiation to nearby healthy tissue or critical structures such as the brainstem or spinal cord [31]. After the path to frameless SRS was blazed with the Cyberknife (see below), the Novalis system was adapted to allow for SRS in one or more

fractions without the use of a stereotactic frame [31]. The ExacTrac X-Ray 6D imaging system employs infrared tracking to monitor patient movement during treatment. Similar tothe CK registration system, this imaging technology consists of two X-ray units recessed into the Linac room floor and two flat-panel silicon detectors mounted to the ceiling (> Figure 57‐2). Fluoroscopic images are registered to digitally reconstructed radiographs, based on pre-treatment CT scans. Any malalignment is corrected by movement of the couch until precise registration is confirmed, after which treatment is resumed. The adaptive respiratory gating module, another feature of the ExacTrac system, tracks the patient’s breathing in real time. The respiratory gating software activates the Linac beam only when the lesion is at the isocenter position for radiation delivery. Regardless of patient body movement, the Novalis robotics-driven systems are able to direct a maximum, effective dose of radiation to quickly penetrate the entire target


. Figure 57‐2 Patient being prepared for Novalis SRS

anatomical site from multiple angles. This process allows for accurate spinal SRS to be delivered without the need for fiducial marker insertion. While the Novalis Tx does not constitute a robotics-based Linac SRS device in a strict sense, its robotic functions are pivotal to performing complex set-ups during both stereotactic (SRS and SRT) and IMRT treatments of cranial, head and neck, or spine indications.

Evidence-based Studies on Quality and Efficacy of Novalis SRS Devices In five patients receiving IMRT treatment for either recurrent pituitary tumors or meningiomas, the Novalis-based pretreatment technique helped to streamline and optimize the treatment plan before fitting patients with a head ring [32]. Wurm et al. reported that the Novalis image-guided noninvasive SRS was accurate and viable for treating intracranial

57

benign and malignant lesions [31] (> Figure 57‐3). For 2-mm treatment planning CT slice thickness, the hidden target test (phantom measurements) for the ExacTrac robotics module revealed an overall system accuracy of 1.04+/0.47 mm for the Novalis device after stereoscopic X-ray verification and correction. The overall average translational setup error was 0.31+/0.26 mm and the rotation error was 0.26+/0.23 . These values are comparable with verification and correction of the Cyberknife [31,33,34] Pedroso et al. evaluated Novalis shaped beam and IMRT delivered as a single dose (SRS, hypofractionated radiotherapy (HSRT), or a fractionated regimen (FSRT) in eight patients with chordoma. Treatment resulted in local tumor control in all cases, although followup was only about 26 months overall. Clinical assessments were based on volumes and dose range for optic apparatus and brainstem. In the FSRT group, two lesions disappeared, one diminished in volume, and the other was unchanged. The authors suggest that the biological effects and treatment outcome of Linac-based radiation are equivalent, whether delivered from a gantry or a robot [35]. Other studies also confirm the accuracy and versatility of the BrainLab Novalis system for both frame-based and frameless image-guided, cranial and extracranial SRS, including functional treatments [36]. In framed SRS, a Novalis Linac device attached to a 4-mm collimator was safe and effective for treatment of trigeminal neuralgia in 32 patients treated during a 12-month interval. Clinical outcomes compared well with those of the Gamma unit [37]. In a study of the geometrical accuracy of Novalis SRS, Rahimian, et al. showed that positioning errors of the Novalis system were less than 1 mm in all axes [38]. They stressed the need for careful quality assurance (QA) procedures to ensure that this level of accuracy was reached, especially before treating patients with the high and critically placed doses needed for the relief of trigeminal neuralgia. Novalis-based stereotactic irradiation has also been used to treat spinal or paraspinal lesions

873

874

57


. Figure 57‐3 Novalis treatment plan beam’s eye view for SRS in a patient with acoustic neuroma, showing planned position of MMLC vanes as gantry rotates around isocenter

such as Ewing sarcoma, metastatic carcinoma and adenocarcinoma, hemangioblastoma, vascular malformations, schwannoma, meningioma, and chordoma [39]. The Novalis system is a powerful dedicated Linac for SRS and SRT that can also be used for IMRT and other conformal treatments in a standard fractionation regimen. Users may enjoy this flexibility but as in any system must be diligent in maintaining strict levels of QA to ensure safe and effective treatments.

Varian Trilogy Image-guided Radiotherapy Platforms Description The Trilogy device (also produced by Varian) epitomizes the incorporation of stereotaxis into an overall RT planning system. Trilogy IGRS consists of a Linac, two gantry/emission heads, beam

collimators, robotic motorized arms mounted on the imaging device, a multi-modal volumetric imaging system, and a remote control operated treatment couch [3,15] (> Figure 57‐4). Setup error and organ motion during treatment are tracked via image-guided motion management techniques for immobilization and positioning. High quality digital image data generated through intraoperative 2D imaging techniques of radiography and fluoroscopy and 3D conebeam CT are used to precisely locate target lesions with submillimetric accuracy. The Trilogy imaging management system also interfaces with CT/PET scanners [15,31,40].

Performance Features Setup and monitoring of the patient and target position are performed in real time for each treatment beam during a Trilogy-based procedure.


. Figure 57‐4 Trilogy gantry and integrated digital imaging system (courtesy of Varian Medical Systems)

An optically guided frameless device is used for immobilization and relocalization. This consists of an optical fiducial array that is attached to a customized bite block of the patient. The Trilogy optical imaging system then determines the patient’s position and motion by tracking the optical markers in real time throughout treatment. Optical markers may be active infrared light-emitting diodes (IRLEDs) or passive reflective spheres [15]. The efficiency of Trilogy is aided by the OnBoard Imager (OBI), a gantry-mounted imaging system controlled by robotic technology and integrated software control. It consists of a highperformance, low energy kilovoltage (kV) X-ray imaging source, a large-area flat-panel amorphous silicon (aSi) digital X-ray detector, and two robotic arms that independently position the kV source and OBI at right angles to the treatment beam. High resolution X-ray images produced during treatment via the OBI are on par with those used for treatment planning [3,15]. To maximize image-guided performance, Varian has coupled the OBI with the Trilogy accelerator for fluoroscopic, radiographic, and volumetric cone-beam CT imaging and delivery within a single machine. CT images, produced

57

from very fine slices, provide key reference images [15]. To some extent this parallels the technology developed for Tomotherapy (see below), and points to the increasing incorporation of high quality digital imaging in radiation delivery systems. The Trilogy system allows for tracking and stereotactic treatment of extracranial targets without the need for fiducial marker insertion. Optimal immobilization is achieved using both physical techniques and software methods. A complementary or alternative approach utilizes Trilogy Respiratory Gating. This system electronically monitors the beam, controlling beam activity in response to the patient’s own respiratory waveform. Trilogy Respiratory Gating uses an infrared camera that tracks a passive marker block placed on the patient’s chest or abdomen [3,15]. Any changes detected in the patient’s normal breathing pattern trigger a gating mechanism that turns the radiation beam on or off. The beam does not go on unless the lesion falls within the planned treatment target area [3,15]. Besides SRS, Trilogy offers several external beam therapies, including 3-Dimensional conformal radiation therapy (3D CRT), IMRT, IGRT and ‘‘Dynamic Adaptive Radiotherapy’’ (DART). Minute lesions can be stereotactically targeted, due to the tight isocentric alignments on three axes rendered in part by the robotic components. The isocentric accuracy of the beam on the gantry and two collimator axes is 0.5 mm or less. The isocenter radius of the couch rotational axis is 0.75 mm or less [15]. Equivalent dose distributions are attained by creating as many beam directions as possible and by managing beam sizes. Geometrically optimized noncoplanar beam arrangements (the ‘‘bouquet of beams’’) and circular collimators or MLCs with narrow leaf sizes limit the size of beam apertures. The result is a high conformal dose distribution with rapid dose fall-off for healthy surrounding tissues in all three dimensions. The Trilogy system can operate either in dynamic

875

876

57


mode (‘‘sliding window’’) through an MMLC, or in segmental mode (‘‘step-and-shoot’’). The dynamic mode confers a high degree of spatial fidelity, leading to the most uniform dose in the target, the steepest dose gradients, and the lowest dose delivered to normal tissues [15].

Evidence-based Studies on Quality and Efficacy of Trilogy Radiotherapy Platforms The capability of the Trilogy to deliver high dose rates in a short time is a key feature of this system. Typical dose rates for Trilogy stereotactic mode (6 MV beam, up to 1,000 monitor units (MU /min dose rate, up to 6,000 MU/field total dose, up to a maximum field size of 15 cm 15 cm, and up to 60 MU/deg dose rate for arc-based treatments) range from 600 cGy/min to 1,000 cGy/min. Through its advanced capabilities, the Trilogy accelerator can also produce an 18-mV high-energy beam (40 cm 40 cm, up to 600 cGy/min at 100 cm). Compared to other Linacs, the Trilogy system delivers a stereotactic dose rate 20% higher at a rate of 1,000 MU per min. Fan et al. advise exercising cautious when using the Varian recommended beam data to perform Trilogy SRS [41]. Trilogy-based treatments have shown high dose conformity for various intracranial and extracranial conditions. In a case study, a single fraction 20 Gy of frameless Trilogy SRS was delivered to a single frontal lobe metastasis located in Broca’s area (0.6 cm3/1.0-cm diameter). A single isocenter with a 10-mm cone and five 100 arcs resulted in high dose conformality with sharp (3 mm) dose fall-off [15]. VanderSpek, et al. performed single fraction single isocenter Trilogy IMRS for multiple brain metastases in 10 patients. The median prescribed dose was 16 (range, 14–18) Gy and the median total planning target volume (PTV) was 35.0 cm3. Eight patients either showed an improvement in

their symptoms or did not experience complications with IMRS [42]. In a quantitative study, the automated quality assurance QA procedure for SRS on a Varian Trilogy Linac was found to be quick, precise and an efficient replacement for the conventional film-based quality assurance method [43].

Linac SRS: The Future Linacs will be the most oft-used tools used for SRS in coming years for a variety of reasons: (1) they are the most common devices for administering RT, hence can be adapted for stereotactic treatments: (2) as noted, Linacs are being created as inherently stereotactic devices for use in standard RT; (3) radiation oncologists control their purchase, implementation, and use in almost all cases. We are still at the beginning of this era of allinclusive Linac systems that provide imaging, inverse planning for IMRT and SRS, and routine stereotactic localization. Much work needs to be done to assess the different technical solutions that have been proposed. For example, in an assessment of SRS treatment plans involving dynamic conformal arcs for intracranial lesions, the shift from a 5-mm leaf size in the Novalis MLC to a 3-mm leaf size in the Varian Millennium MMLC was linked with improved target conformity and increased normal tissue sparing. However, ongoing studies are required to determine whether the parameter of size alone applies to all collimators, regardless of total leaf count, and whether smaller leaf size necessarily translates into clinical improvement [44]. Other issues to assess will be the confirmation of frameless localization as a routine method for even high-dose single session SRS (as in for functional SRS for patients with trigeminal neuralgia or movement disorders). The benefits of fluoroscopic versus CT registration should be examined. The cost-effectiveness of dedicated SRS units will be compared with Linacs adapted for


SRS. Without question, Linac-based SRS will be an ever larger component of clinical neurosurgery and radiation oncology.

Tomotherapy Description Tomotherapy, which literally means ‘‘slice therapy,’’ is a novel platform for administering IMRT. It can deliver higher dose radiation therapy to target sites, including multiple sites simultaneously, in various regions of the body while still limiting the amount of radiation to healthy tissue. Also known as helical tomotherapy (HT), it seamlessly integrates into a single system (1) a customized inverse treatment planning system; (2) onboard, real time, online megavoltage CT image-guided patient positioning; and (3) a spiral pattern of IMRT treatment delivery. The HT device also includes a MLC and a quality assurance module individualized for each patient [21,45–48]. The HT prototype was developed in the late 1980s by the Radiotherapy Research Group, led by Mackie and Reckwerdt at the University of Wisconsin. It was based in part on principles of serial tomotherapy and a modulated multileaf collimator technology, both of which characterized the abovementioned Peacock system [49]. The current HT version, the Tomotherapy Hi-Art Treatment System, is an intensitymodulated, rotational radiation therapy that contains a ring gantry geometry with fan-beam delivery [21,45,47]. HT utilizes a helical CT platform instead of the static treatment arm and couch platform found in traditional Linacbased SRS devices. In the HT unit, a miniaturized 6 MV Linac standing only 18 inches high is mounted to a CT gantry, such that the combined device is used for both imaging and treatment. The accelerator revolves around the patient as he moves slowly through the gantry ring [21,45,47].

57

Since 2003, HT has been used clinically as an alternative SRS technique to GK, CK, and other Linac-based SRS devices for the treatment of brain tumors [21,45,47]. For imaging, the HT unit generates a megavoltage CT (MVCT) 3-D scan that delineates the precise contours of each tumor just prior to irradiation. This is essential for verifying the location of the tumor and adjusting the patient’s position. Accurate patient setup is achieved by coregistering pretreatment daily MVCT scans with the initial planning CT scan [21,45,50,51]. Since daily MVCT localization can be attained with frameless stereotaxis, internal or external fiducial markers for tumor localization may not be needed [52]. In the Tomotherapy Hi-Art System, the Linac uniquely delivers thousands of minute radiation beamlets from every point along a spiral pattern as the photon energy forms multiple circles around the gantry ring. The beamlets are delivered from various angles, intersecting with multiple targets and, at the same time minimizing the amount of radiation that reaches neighboring tissue (> Figure 57‐5). These capabilities enable HT machines to irradiate lesions usually not targeted with conventional Linac SRT and SRS devices [21,45].

Performance Features Daily IMRT treatment with HT is continually modified through an innovative verification method called dose reconstruction [49,51]. By providing a comparison of CT radiation dose images with planned dose images, future treatments are modified to correct for errors in completed treatments. The intensity of the beams can be changed, if necessary, by adjusting the width of the MLC leaves. This process of daily correction during multi-treatment hypofractionation sessions is called adaptive radiation therapy [51]. During HT treatment, the beam

877

878

57


. Figure 57‐5 Illustration of helical tomotherapy concept. Radiation is delivered from 360 as the patient moves through the gantry (courtesy of TomoTherapy Inc.)

treatment team be highly skilled as both a system user and an expert in managing all technical aspects of verification QA [53].

Evidence-based Studies on Quality and Efficacy of Tomotherapy

path is totally shielded to reduce radiation ‘‘scatter’’ to the patient. In a technologically advanced fashion, the motion of the MV radiation source in the HT unit is synchronized with the movement of both the patient and the MLC. As the Linac moves in a circular pattern delivering the beamlets, the patient slowly advances in and out of the center of the gantry ring. Simultaneously, the shape of the radiation beam, and essentially the angle of the treatment, are modified by the pneumatically driven MLC, which has two sets of interlaced leaves resembling a zipper. As the treatment couch moves, the beamlets can be turned on or off by opening and closing the 64 leaves of the MLC [21,50]. The dynamic delivery of radiation in HT is complex, and carefully orchestrated by multi-level integrated, automated systems that control coordinated patterns of motion in the gantry, treatment couch, and multileaf collimator leaves. Yet, this complex scheme is partly hidden from the end-user. Therefore, it is imperative that the medical physicist on the

Tomotherapy has been used mainly as a means of delivering fractionated RT throughout the body in a way that is more precise than standard conformal treatments. As such, there is relatively little literature analyzing it as a tool for SRS. Yartsev conducted a small scale study (n = 12) comparing HT plans with other radiotherapy modalities in the treatment of small brain tumors. Of five methods evaluated, proton techniques (see below) exhibited the overall best dose distributions. HT resulted in higher target dose uniformity (average SD ¼ 1.3%), and showed a risk of irradiation to critical structures comparable to that of other Linac methods [54]. Shi, et al. found that HT treatment plans generate fairly uniform doses to target treatment sites in the head and neck, brain, lung, prostate, pelvis, and cranio-spinal axis when compared with Linacbased step-and-shoot IMRT planning for special treatment sites. Their findings also confirm that HT produces smaller integral doses to healthy structures when delivered to these target anatomical sites [55]. In eight patients treated for spine tumors (seven with metastatic disease), the Tomotherapy Hi-ART System produced accurate set-up and delivery without the need for fiducial markers. Acute and late toxicity was minimal, and four patients were still alive (median overall survival, 5.1 months) at the completion of the study [52]. The on-board MVCT image system increases the accuracy and precision of patient positioning over time during delivery of multiple fractions. Repeat CT imaging and re-planning is highly advantageous, especially for IMRT treatment of head and neck cancer. Daily setup corrections may be


needed due to weight changes and subsequent alteration in anatomy near the target lesion. Hansen, et al. reported that the median rate of decline of the clinical target volume was 1.7–1.8% per treatment day, accompanied by a shrinking volume that was often asymmetric in patients with head or neck cancer. In a case study, the modified radiation dose delivered to a patient had a sparing effect on spinal cord and trachea [45,56]. In a comparison of single fraction HT and Gamma Knife SRS used to treat single brain metastasis in 5 patients, these two techniques produced different dosimetric and conformal indices. Compared to Gamma Knife SRS, HT was linked with larger lower isodose line volumes and longer treatment times [57]. HT is not without other limitations. A single slice of MVCT is reconstructed from a 180 rotation in 5 s, but it contains motion artifacts from breathing. Yet, a full 360 cone-beam CT scan usually produced in 45–60 s is also constrained by motion artifacts. Future refinements to HT imaging technology eventually may resolve this problem [45,52,58]. Gutie´rrez, et al. evaluated the feasibility of using composite HT planning to deliver whole brain radiotherapy in 10 patients with metastatic brain tumors treated with an integrated boost of SRS. The planning used HT with the original CT scans and MRI-CT fusion-defined target and normal structure contours. Composite HT planning achieved favorable outcomes of hippocampal avoidance; homogeneous whole brain dose distribution equal to that of conventional whole brain radiotherapy; and radiosurgically equivalent dose distributions to individual metastases [59]. Welsh, et al. reported that HT treatment has yielded excellent results for reducing the risk of alopecia in patients undergoing standard whole brain radiotherapy (WBRT). By planning a scalp-sparing treatment, HT appeared to be superior in hair preservation when compared with both conventional modalities and other IMRT techniques used for WBRT. The superior

57

dose-rate of HT may offer advantages over conventional Linac-based IMRT, particularly as clinical targets become more complicated [60]. Documented clinical benefits of HT include the amount of the time required to create plans, overall tumor target conformality, capacity for conformally avoiding critical structures, and the time required to deliver each fraction [60]. Delivery of the conventional dose of 300 cGy with the HT method may take approximately 20 min for all steps, including the MV CT used for image-guidance and set-up. By contrast, complex IMRT delivery of a 100 cGy dose fraction takes an estimated 30 min with a Varian EX accelerator in dynamic delivery mode. A standard 300 cGy dose delivery from the Varian EX would, therefore, become intolerably long and impractical for clinical use [60,61]. Helical tomotherapy has been an important addition to the SRS/SRT landscape, and has spurred interest throughout the field in the combination of digital sectional imaging as an integral part of stereotactic irradiation. Clinically, its advantages will probably be felt mainly in conformal and tissue sparing treatments of patients with relatively large target volumes (> Figure 57-6).

Cyberknife: A Robotic Linac Description The Cyberknife (CK) (Accuray, Sunnyvale CA) is an innovative tool in SRS. Developed in the early 1990s by John Adler and collaborators at Stanford University, the CK was the first device to allow for frameless SRS for treating targets in the head as well as anywhere in the body (> Figure 57‐7). The latter was the impetus for development of the CK. After a fellowship in Stockholm in the early 1980s studying SRS with Lars Leksell, Adler sought an SRS technology that would not be limited to intracranial targets

879

880

57


. Figure 57-6 Tomotherapy treatment plan for a patient with brainstem glioma (courtesy Dr. Jed Pollack, Long Island Radiation Therapy, Lake Success, NY)

. Figure 57‐7 Patient about to undergo SRS with Cyberknife


(J Adler, personal communication). The CK employs an X-band Linac, small and light enough to be mounted on an industrial robot. SRS is delivered by a stop-and-shoot technique that avoids the need for isocentric planning. Instead, the target is ‘‘painted’’ with the prescription dose. Treatment planning is done with an inverse technique, after clinicians define the target, designate the dose and number of fractions, and set dose constraints for adjacent critical structures [3,62,63]. The CK can deliver both single-fraction and multifraction SRS with submillimetric accuracy. The CK frameless image-guidance system includes precisely calibrated, paired orthogonally placed (90-degree offset), diagnostic X-ray tubes that are rigidly attached to the ceiling. Real-time images of the anatomical treatment site are generated using two amorphous silicon X-ray screens, or detectors that produce high-resolution digital images. These are

57

registered to digitally reconstructed radiographs (DRRs) generated from the patient’s pretreatment CT scan [62,64,65] (> Figure 57‐8). CK was the first SRS system to use 3D digital data for enhanced complex treatment planning for SRS [3].

Performance Features of the CK The CK has gained increasing acceptance as a SRS tool, largely because of its high level of accuracy, rapid radiation dose fall-off and precision on a par with that of frame-based SRS systems [62,64–67]. Through an image-guided control loop connecting the imaging system and the robotic manipulator, the CK provides real-time tracking of both patient motion and the exact position of the lesion. An advanced algorithm controls the loop, which in turn prompts the tracking system to automatically

. Figure 57‐8 Screen view of the Cyberknife registration system. Mismatches between the new fluoroscopic images and the DRR from pretreatment CT (left) are corrected by automated table movement until accurate overlay is confirmed

881

882

57


re-position the robotic manipulator to locate and track the target, and to move the radiation beam into alignment with the target lesions. The miniaturized Linac in the CK is advantageous for rapidly positioning the robotic arm over a wide range of beam orientations [65]. Radiation beams can be aimed almost anywhere in space. In a typical treatment, the target is irradiated with 80–120 stereotactic, precisely cross-fired, non-isocentric radiation beams that produce complex overlapping arrangements. As noted above, patient positioning is corrected via co-registration of projection images that are taken with the silicon detectors positioned on either side of the patient’s anatomy. An algorithm directs the computer to correlate the pair of alignment radiographs of the target from the perspective of the two cameras with the patient’s treatment planning CT scan [68,69]. This mechanism allows for SRS to be delivered using only mask immobilization. Digital imaging and registration can be repeated after each beam, but in practice this would unnecessarily lengthen treatment times. Depending on the degree of patient movement, images can be repeated more or less often, but typically at least after 10 beams are delivered. Movements over several mm, which are unusual, may require manual repositioning by the radiation therapist to bring the couch back into position. When the CK algorithm redetermines the spatial coordinates of the target lesion, and then relays to the robot the information needed to correct the patient positioning, the patient himself is not moved. Instead, the CK image-based computer software accurately directs adjustment of the treatment couch [68]. Treatment plan optimization in the CK system is based on inverse planning algorithms, including a planned dose distribution analysis for selecting beam directions and weights of each node surrounding the target volume [70]. The CK usually is programmed to function with

a non-isocentric technique, using multiple overlapping isocenters to irradiate irregularly shaped lesions. In addition, the CK also has algorithms that can direct a beam to be delivered to a single isocenter for treating spherical targets [3]. During treatment, the robotic arm can be deployed to 1,200 positions for controlling beam direction [68,71]. In practice, between 80 and 120 beams typically are used to keep treatment time to about 30 min per session, on average. Yu, et al. compared treatment plans obtained with the CK with those of other SRS therapies, including GK, Linac multiple arcs, conformally shaped static fields, and IMRT. The evaluations were based on dosimetric indices such as dose-volume histograms and other standard radiosurgical parameters (target coverage, homogeneity index, and conformity index). While all of the tested treatment modalities provided the same level of full target coverage, the treatment plans for the CK yielded more flexibility for a given target size and shape. Conformal dosimetry with the CK and Gamma Knife was similar for lesions of limited volume, regardless of shape. Again, the ability of the CK to irradiate irregularly shaped targets is a function of its nonisocentric treatment option [68]. Giller, et al. suggested that the maximum error reported for the robotically controlled CK system is not just comparable with frame-based SRS systems, but may even be superior to many frameless and frame-based systems. They proposed that the CK may therefore be feasible for SRS treatment of malignant brain tumors in infants [72].

Evidence-based Studies on Quality and Efficacy of the CK The frameless technology of the CK renders it effective in treating not only patients with head and neck cancers and spinal lesions, but also


abdominal and thoracic tumors, pancreatic and liver cancer, and pelvic tumors such as prostate cancer and select rectal tumors [73–77]. Yet, it is in the realm of intracranial SRS that the CK initially made its mark as a technically feasible and practical treatment device for both malignant disease and functional disorders. In an early study, Chang, et al. evaluated the effects of CK in singlefraction SRS treatment (mean radiation dose, 18.1 Gy; range, 12–36 Gy) on 84 tumors (primarily metastatic carcinoma) in 72 patients. Their findings demonstrated that the overall accuracy of the frameless CK system is comparable to that of other Linac-based systems requiring invasive stereotactic frames for skeletal fixation [78]. By 2001, Chang and Adler reported that the clinical results for nearly 1,900 intracranial tumors and arteriovenous malformations (AVM) treated with the CK closely paralleled the outcomes of other SRS techniques [79]. Shimamoto et al. reviewed clinical outcomes of CK irradiation (dose range dose, 9 to 30 Gy) on 66 metastatic brain tumors in 41 patients. Tumors treated with at least 24 Gy were less likely to progress than lesions treated with 20 Gy (p = 0.0244; log-rank test). No severe side effects occurred. The complete response (CR) rate was significantly higher if the maximum dose to the tumor was at least 24 Gy (p = 0.0045) [80]. More recently, Collins et al. assessed CK treatment planning parameters for fractionated SRS of 82 skull base lesions in 80 patients. The plans showed excellent conformity, homogeneity, and percent target coverage for both simple and complex skull base tumors percent target coverage [70]. In a multicenter retrospective analysis of 95 patients treated for idiopathic trigeminal neuralgia, Villavicencio et al. identified optimal radiosurgical treatment parameters for the treatment of idiopathic trigeminal neuralgia. Of the 95 patients, 64 (67%) initially experienced excellent pain relief, with a median time to pain relief of 14 days. In contrast to previous reports, in this

57

study higher doses were used (median maximal dose, 78 Gy; range, 70–85.4 Gy) and longer segments of the trigeminal nerve were irradiated (median length, 6 mm; range, 5–12 mm), resulting in both better pain relief and a higher incidence of hypesthesia. Post-treatment numbness was linked with improved pain relief. At a mean follow-up of 2 years, 47 patients (50%) had experienced sustained pain relief, and were completely off pain medications at follow-up (mean 2 years). The overall complication rate was 18% [81]. A major impact of the CK has been on SRS for patients with skull base tumors. The ability to fractionate has allowed for treatments in 2–5 sessions to be given to patients with vestibular schwannomas [82] and parasellar tumors [83] (> Figure 57‐9). Results have shown tumor control of 95% or better with few complications. Hearing preservation was maintained in about 70% of patients, comparable to Gamma Knife and other SRS series [82]. Patients with parasellar lesions did not develop visual loss, except in one patient who had prior fractionated irradiation [83].

Spinal SRS with the Cyberknife The first description of spinal SRS was by Hamilton et al. [84]. This groundbreaking innovation did not in the end prove practical, due to the need to fix a registration marker to the patient’s spinous process. Besides the invasive aspect of the procedure, this meant the treatment would be done prone and most likely under general anesthesia. The CK was the first tool to make spinal SRS practical. This has included treatment of patients with metastatic tumors but also with intradural and intramedullary neoplams, and AVMs [79]. Gerszten, et al. analyzed the effects of single-fraction SRS delivered via the CK to cervical spine lesions in 115 consecutive patients. Both benign and metastatic tumors were treated,

883

884

57


. Figure 57-9 Cyberknife treatment plan for 34-year-old man with recurrent pituitary tumor in left cavernous sinus and nearing the left side of the optic chiasm. He received 2,500 cGy in 5 daily fractions

and lesion volume ranged from 0.3 to 232 ml (mean, 27.8 ml). Tumor radiation dose was held at 12–20 Gy to the 80% isodose line (mean 14 Gy). Skull bony landmarks were used to locate and track cervical spine lesions. Fiducial markers were positioned to pinpoint the lower spinal lesions. Patients showed rapid recovery and good response after a short treatment time in an outpatient setting. No acute radiation toxicity or new neurological deficits were observed during the follow-up period (3–24 months) [65]. Data from an anthropomorphic phantom study confirm the submillimetric accuracy of the CK in spinal SRS. Based on head and torso

phantoms, Yu et al., reported an average treatment delivery precision of 0.3 0.1 mm for CT slice thickness (mean, 0.7 0.3 mm; range, 0.625–1.5 mm). The tracking error for fiducial markers was less than 0.3 mm for radial translations up to 14 mm, and less than 0.7 mm for rotations up to 4.5 [68]. These values are generally comparable with those of previous phantom studies of the targeting of brain lesions using standard SRS head frames [68,85–89]. CK software now supports accurate tracking of spinal targets without the need for fiducial placement. The error of this system was found to be 0.6 mm, hence as accurate as that attained with fiducial implantation [90].


SRS treatment of spinal conditions using the CK can be completed in a single day, and may significantly improve local control of cancer of the spine [91]. In a prospective evaluation of 18 patients with malignant sacral lesions treated with single-fraction SRS via the CK, pain improved in all 13 patients who had symptoms before treatment. Followup imaging showed no tumor progression. No acute radiation toxicity or new neurological deficits were reported during the followup period (mean, 6 months). In contrast to conventional external-beam radiotherapy, the CK was able to effectively deliver large doses of radiation while sparing nearby radiosensitive structures, particularly the spinal cord and cauda equina. The volume of the cauda equina receiving more than 8 Gy ranged from 0 to 1 ml (mean, 0.1 ml) [92]. Similarly, SRS administered with the CK was tolerated in 15 patients receiving singlefraction radiosurgery for treatment of benign spinal lesions (12 cervical, one thoracic, and two lumbar). Dose plans for the CK did not produce any acute radiation-induced toxicity or new neurological deficits during the follow-up period. No tumor progression was evidenced in follow-up imaging (mean, 12 months) [66]. A recent study further demonstrates the shortterm clinical benefits of single fraction SRS in treating benign extramedullary spinal neoplasms. Gerszten et al. evaluated the effects of SRS (mean intratumoral dose, 2,164 Gy) on the treatment of benign intradural extramedullary spinal tumors, primarily located in the cervical region in 73 patients [120]. The cases included neurofibroma (25), schwannoma (35), and meningioma (13). SRS was used mainly as a primary treatment modality or as a therapy of postsurgical radiographic progression. The median follow-up period was 37 months. Longterm radiographic tumor control was seen in all cases, with long-term pain improvement achieved in 22 (73%) out of 30 cases. New symptoms associated with radiation-induced

57

spinal cord toxicity developed in three patients 5–13 months after treatment. Studies conducted over longer follow-up periods are critical to assessing whether SRS has long-term efficacy for benign spinal tumors [67].

Definition of SRS and the Cyberknife SRS was defined as a single session treatment by Leksell, and this concept went unchallenged for 50 years. The advent of the CK called this into question. The frameless technique, with accuracy shown to be equivalent to that achieved with a stereotactic frame, allowed SRS to be done in more than one fraction. Prior stereotactic fractionation, done with relocatable frames, had meant the use of standard fractionation regimens – typically 25–30 treatments – but with smaller volumes and less irradiation of surrounding tissues than with standard fixed field treatments. The CK began to be used to treat patients with lesions that were too large or critically located for single session SRS, in up to five sessions, depending on the perceived risk of hypofractionation [82,83]. There does not seem to be any logical reason to define SRS as necessarily confined to a single session. With this in mind a committee convened by the leadership of American organized neurosurgery redefined SRS as a stereotactically targeted, image guided treatment of 1–5 sessions [13]. This expanded definition of course applies to all devices used for SRS, including the other Linac based systems discussed above. This has allowed neurosurgeons to recommend for their patients the best individualized treatment, without concern that they will lose their involvement in patient care on the one hand, or prescribe a suboptimal single session treatment on the other.

885

886

57


Leksell Gamma Knife Perfexion Description The Leksell Gamma Knife Perfexion (Elekta, Stockholm, Sweden), an updated version of previous Gamma Knife (GK) units, features a new design and major improvements in performance capabilities. The Perfexion is a fully robotized GK that can treat patients with a wide variety of lesions in the skull base and neck. Like its predecessors, the Perfexion still uses a multi-isocenter treatment plan, but it incorporates inverse treatment planning and depends less on decreased forward treatment planning [3,19,50,93,94]. Compared to earlier models of the GK, the Perfexion is better able to treat target volumes of up to 300% greater than previously possible, while simultaneously sparing healthy tissue near eloquent areas [19,50,93,94]. Full automation of the dose-delivery process in the Perfexion has decreased manual input and errors. As a result of the smoother workflow of integrated components of the system, both treatment time and discomfort to patient have been significantly reduced [3,93].

Performance Features The radiation unit of the Perfexion has been reconfigured with a new beam geometry of 192 Cobalt-60 sources arranged in a cone section pattern. This modification translates into a higher dose rate for any given source activity. The redesigned cylindrical shape of the collimator helmet of the Perfexion allows anatomic structures in once inaccessible locations to be targeted for SRS treatment. The caudal reach of the Perfexion extends further than in the traditional helmet of the GK Model C unit, thus providing improved peripheral coverage. With the redesigned collimator

. Figure 57‐10 Graphic illustration of the single Perfexion collimator helmet. Different collimator diameters and blocking patterns can be automatically set without the need for manual helmet changes (courtesy of Elekta, Inc.)

helmet, upper cervical lesions as well as extracranial head and neck tumors can now be treated [19,93,94]. The Perfexion is the only instrument in the GK series to contain a single collimator within the housing unit (> Figure 57‐10). The collimator consists of a larger, single, 12-cm thick tungsten array arranged in a series of five concentric rings. Older models of the GK have multiple standard collimators with fixed apertures, but these require manual changing between shots requiring different collimator sizes. In the Perfexion, the multiple, exchangeable collimator helmets have been replaced with an integrated automated adjustable collimator system. The built-in modulating apertures of the new collimator system are automatically adjusted by internal controls, thereby eliminating the time-consuming process of manually switching helmets [19,50,93]. The longer axial diameter of the Perfexion collimator helmet allows peripherally located targets to be positioned at isocenter with less potential for collimator-frame collisions. The ability to automatically adjust the size of the collimator


openings is essential for protecting critical structures adjacent to the target from excessive irradiation. These performance-related design features enable the Perfexion to create complex shapes of isodose volumes through dynamic shaping of dose distribution [19,93,94]. The robotic-driven automatic positioning system of the Perfexion operates through the couch unit rather than the head frame. Repositioning the patient by moving the couch instead of the head frame is less disruptive for the patient. In addition, the robotic-controlled couch system facilitates a seamless transition between ‘‘shots’’ fired in a multi-isocenter treatment plan [19,93] (> Figure 57-11).

57

the Perfexion relative to the GK 4C. These findings, combined with other documented clinical benefits of the Perfexion, indicate that it meets the highest standards of an SRS instrument [94]. As a frame-based SRS device, the GK does not lend itself readily to fractionated treatments. However, the Perfexion unit has greatly increased the versatility of this instrument, which will remain an important way to deliver SRS for the foreseeable future.

Heavy Particle Radiotherapy Description

Evidence-based Studies on Quality and Efficacy of the Leksell Gamma Knife Perfexion Studies confirm that compared to previous models of the GK, the Perfexion has an increased ability to shape isodose volumes, even for single isocenter treatment plans. As noted, the new collimator arrangement and the automated patient positioning system have led to decreased patient transit times and less exposure during movement between isocenters. The larger radiation cavity of the Perfexion also permits more regions of the head and neck to be targeted for treatment [93]. In a blinded randomized prospective comparative study on various treatment related parameters, including QA, 29 patients were treated with the GK 4C and 30 patients with the GK Perfexion. Patients underwent treatment for various lesions, including multiple metastases. Treatment with the Perfexion unit was associated with greater radiation protection and with collision-free procedures. The time of the SRS treatment, time of clinician and physicist intervention on the machine, and time of the QA procedure were all reduced with

Charged particle, or heavy particle beams have been used since the 1950s to treat intracranial tumors, arteriovenous malformations (AVMs), and subfoveal neovascularization [95]. To some extent this reflects the fact that heavy particle therapy provided the only way, until the invention of the GK in the late 1960s, to deliver highly conformal irradiation with a steep dose gradient. Today, while there are still relatively few facilities offering this method, heavy particle irradiation, particularly with protons and carbons, is gaining momentum as an efficacious treatment. Unlike radiation from gamma or photon sources that decays exponentially, there is no exit dose of proton energy deposition beyond the target volume. When proton beam energy is delivered to the tissue, it first enters the area as a slowly rising dose followed by a rapid rise to a maximum (the Bragg peak), and then a fall to near zero [96]. In practice, an overly sharp radiation deposition is insufficient to treat a volume such as a tumor. Therefore heavy particle beams need to be shaped and flattened with a variety of filters to create a spread-out Bragg peak (SOBP) [97] (> Figure 57‐12). During treatment with heavy particles, beams are delivered via either a passive (scattering) or active (scanning) method. The technique can be selected

887

888

57


. Figure 57‐11 Gamma Knife Perfexion treatment plan for patient with a right vestibular schwannoma. Note the high number of shots, the different collimator sizes, and the blocking patterns (courtesy of Dr. Jean Re´gis, University Hospital La Timone, Marseilles)

by changing the operating parameters of several pieces of equipment such as the accelerator and collimator. The Bragg peak coincides with the treatment volume, and is determined by the beam energy that can be altered by controlled by adjusting the accelerator. In pencil beam scanning, multiple beams of varying energies are superimposed on each other, thus causing the treatment volume to exhibit an SOPB [97]. Heavier ion beams produce better dose distributions precisely because of this higher biologic property: the ability to generate a higher energy transfer in the Bragg peak as compared with the entrance point of beam transport [98,99]. The physical characteristics of protons and other high energy, heavy particles make them attractive sources of radiation beams for therapeutic radiation. Charged-particle SRS is particularly advantageous for the conformal treatment of large and/or irregularly shaped lesions, as well as for minimizing radiation reaching eloquent brain structures located near the target site. As newgeneration intensity-modulated proton techniques

emerge, proton SRS may yield even more favorable outcomes for treating large target volumes while sparing critical structures such as the spinal cord, eyes, and brain [96,100]. Newer units also are smaller and have beams that exit from relatively compact gantries that can rotate around an isocenter (> Figure 57‐13).

Evidence-based Studies of Heavy Particle Radiotherapy Proton irradiation produces high dose isodose conformality, particularly for tumors with complex geometry, especially concavity [101]. In a clinical comparison assessment of proton-based and proton SRS for the treatment of AVMs, vestibular schwannomas, and pituitary adenomas, protonbased techniques achieved excellent outcomes. The results were at least on par with those of photon-based radiation therapies. Compared to photon irradiation, proton radiation beams produced superior conformality in dose distribution, particularly in larger sized lesions, and showed good sparing effects of normal tissue [96].


. Figure 57‐12 The spread-out Bragg peak 2000;31:6, with permission)

(Europhysics News

Proton beam SRS improved tumor control for vestibular schwannoma [102,103] and showed reasonable rates of facial and trigeminal nerve functional preservation [102]. Weber, et al. evaluated 88 patients with vestibular schwannomas (mean, 16 mm; range, 2.5–35 mm) who were treated with proton beam SRS that applied two to four convergent fixed beams of 160-MeV protons. Facial nerve function (House-Brackmann Grade 1) and trigeminal nerve function were normal in 79 patients (89.8%). Hearing was good or excellent (Gardner-Robertson (GR) Grade 1 in eight patients (9%) and serviceable in 13 patients (15%) at follow-up (median period, 38.7 months). Seven (33.3%) of the 21 patients%) with functional hearing (GR Grade 1 or 2) retained serviceable hearing ability (GR Grade 2). The actuarial 2- and 5-year tumor control rates were 95.3% (95% confidence interval (CI), and 93.6% (95% CI). The actuarial 5-year cumulative radiological reduction rate was 94.7% (95% CI,). Actuarial 5-year normal facial and trigeminal nerve function preservation rates were 91.1% (95% CI) and 89.4% (95% CI) [102]. Cumulative data indicate that fractionated proton beam radiotherapy improves tumor control in patients with skull base tumors [98,104] including adenoid cystic carcinoma of

57

. Figure 57‐13 A patient about to be treated with proton beam therapy using a compact gantry (Europhysics News 2000;31:6, with permission)

the skull base [105] and pediatric base of skull tumors [106]. Proton radiation therapy offers excellent prospects for durable tumor control in patients with low-grade chondrosarcomas or chordomas, and without causing undue complications in some series [98,107]. However, others have reported an alarming 50% risk of visual loss after proton treatment of clivus chordomas [108]. Similarly, in a comparison of conformal radiation methods, proton therapy was superior to conventional photon irradiation and photon IMRT for craniospinal axis irradiation and posterior fossa boost in a pediatric patient with medulloblastoma [109]. The dose to the cochlea was reduced from 101.2% of the prescribed posterior fossa boost dose to 33.4% with IMRT, and only 2.4% with protons. Dose to 50% of the heart volume was decreased from 72.2% for conventional X-rays to 29.5% for IMRT and 0.5% for protons. The considerably enhanced effect of protons on nontarget tissue sparing may be favorable for long-term toxicity,

889

890

57


particularly in regard to and endocrine and cardiac function [109]. Carbon ion radiotherapy has been proposed because these heavy ions can provide highly effective dose-localization in the body. Growing evidence reveals that carbon ion radiotherapy is effective in treating patients with tumors of the skull base and head and neck, as well as bone and soft-tissue sarcomas, prostate cancer, non-small-cell lung cancer, and hepatocellular carcinomas [99]. In studies conducted in Japan, carbon ion hypofractionated radiotherapy using larger doses per fraction resulted in local control rates for locally advanced tumors with minimal toxicity. Overall treatment times of carbon ion radiotherapy were less than conventional radiotherapy [110]. This treatment had a sparing effect on critical structures such as the spinal cord, eyes, and brain [100]. Higher doses of carbon ion radiotherapy administered in combination with X-ray radiotherapy and chemotherapy for malignant gliomas increased the survival rate in patients [111]. The physical and biological properties of carbon ion beams appear to have advantages over those of photons. Mizoe, et al., conducted phase I/II dose-escalation studies and clinical phase II studies evaluating the effects of carbon ion radiotherapy on 1,601 patients with various types of malignant tumors. All patients except those with malignant glioma were delivered carbon ion radiotherapy alone. The fraction number and overall treatment time was fixed for each tumor site, and administered to one field per day and 3 or 4 days per week. The total dose was escalated by 5 or 10% increments to guarantee safety. Carbon ion radiotherapy resulted in high local control rates for locally advanced tumors, including advanced head and neck tumors, and pathologically non-squamous cell type of tumors, producing minimized toxicity. Carbon ion radiotherapy has led to improved outcomes in several cancers, including chordoma and chondrosarcoma of

the skull base and cervical spine, locally advanced prostate carcinomas in high-risk patients, and post-operative pelvic recurrence of rectal cancer [112]. However, in this study, severe late complications of the recto-sigmoid colon and esophagus developed in patients receiving high dose levels of carbon ion radiotherapy for prostate, uterine cervix and esophageal cancer. These adverse effects disappeared when patients were administered safe dose levels that were established after improved irradiation methods were implemented. The results showed that carbon ion could be delivered with larger per fraction doses of hypofractionated radiotherapy and in a shorter treatment time compared to conventional radiotherapy [112].

Heavy Particle Therapy and SRS There are about 20 proton bean units in operation around the world, with a handful of carbon ion facilities. For some conditions, especially in patients with chordoma, there is an assumption by many clinicians that proton beam irradiation is the treatment of choice [113]. Compared to conventional fractionated radiotherapy 10-year survival rates are much better, on the order of 65% versus under 20%. On the other hand, the use of stereotactic photon SRT or SRS has shown much better results than with RT alone, with 82% progression free and overall survival of 82% at 8 years [114]. These results exemplify this debate: to what extent does the Bragg peak and high energy deposition of heavy particles confer an advantage over photon SRS? Or does the presumed benefit of protons and other heavy particles mainly reflect the historical fact that this was the only real way of delivering highly focused therapeutic irradiation, until relatively recently? This question is hardly academic. Commercial installations of proton bean units cost on the order of


$100 million; for heavier ions, with more complex cyclotron requirements, costs will be even higher. Medical centers around the US and elsewhere are considering implementation of heavy particle units. Can these centers, and national health systems around the world, sustain the costs involved? Academia and industry may eventually pave the way for practical proton-based SRS instrumentation. Caporaso, et al. and colleagues at Lawrence Livermore National Laboratory are designing a novel compact CT-guided proton therapy system that uses a dielectric wall accelerator. This device can be housed in a conventional Linac vault, and feasibility tests of an optimization system are underway. The completed proton therapy system will permit rapid, optimized CT-based IMPT delivery for the treatment of larger target volumes as well as for motion management in pediatric patients. It will be capable of generating high electric field gradients by using alternating insulators and conductors and short pulse times. This proton therapy device is based on emerging new technologies such as high gradient, vacuum insulators, solid dielectric materials, SiC photoconductive switches and compact proton sources [115]. Certain vendors, including Varian and Still River Systems Proton Therapy may succeed in creating practical proton beam devices that are less expensive – but still on the order of $15 million per. Perhaps most neurosurgeons interested in radiosurgery should focus their energies on continuing to improve techniques of photon and gamma SRS.

Conclusions Stereotactic radiosurgery was, for the first 45 years of its existence, the provenance of a few neurosurgeons (and allied radiation oncologists and medical physicists) who took a special interest in this then esoteric technique. As a result of

57

the success of the SRS pioneers, this has now become a standard part of stereotactic neurosurgery – so much so that neurosurgical training programs now require SRS to be taught to all residents. As neurosurgeons understood more of the history and radiobiology of radiation therapy, the benefits of fractionation led to the current definition of SRS. At the same time, the obvious advantages of the stereotactic technique became clear to manufacturers of Linacs and to radiation oncologists as a whole. It was inevitable that radiation therapy in general would become increasingly stereotactic – in other words, that SRS would become more like RT, and that RT in turn would become more like SRS. Newer SRS devices and the treatment planning processes used to implement SRS therapy are changing the practice of neurosurgery. While SRS will never totally replace conventional surgical neurology, the distinct advantages that each of the newer SRS modalities offers has broadened the range of options for intervention that are available to neurosurgeons. The fundamental goals for the safe and effective use of SRS remain unchanged: dose homogeneity, target coverage, healthy structure sparing, and dose conformality [21]. The challenge of SRS today lies in determining which competing device is best for delivering the optimal radiation doses for a given target. As the contents of this chapter reveal, the lines separating conventional SRS, hypofractionation, and SRT are thin. In addition, the clinician may have to decide whether the treatment should be delivered using either 3D conformal radiotherapy or IMRT [40]. The answer to the clinician’s dilemma may very well be that SRS can be delivered with equivalent safety and efficacy by all of the systems discussed in this chapter. The evolutionary landmarks in SRS over the past 50 years include advances in computerized, real-time imaging technology, inverse planning techniques, and the ability to localize and treat both intra- and extracranial targets with submillimetric precision [3]. Frameless SRS has led to

891

892

57


techniques for fractionation, reduced pain and discomfort to patients, and decreased requirements for anesthesia in pediatric patients [30]. The growing clinical use of IMRT for targets in the upper head-and-neck region is a direct consequence of accurate and reproducible thermoplastic mask–based stereotactic head-and-neck fixation [116]. The use of advanced digital imaging has made extracranial targets accessible to SRS [30]. As a result of new technologies in surgical robotics, some SRS devices are now equipped with enhanced computerized features, including robotic-assisted planning, robotic patient positioning, and robotic beam delivery. The OnBoard Imager of Varian Trilogy is integrated with robotic functions [3,30] just as the CK [62,79] and the updated GK Perfexion utilize robotics for pivotal functions [94]. Many of the limitations of earlier versions of the GK have also been overcome with the redesign of the Perfexion, including significant changes in the helmet and collimator [94]. In Trilogy and CK, imaging systems that rapidly and precisely calculate the target volume as well as patient positioning are fully integrated with complex automated programs that factor continuous patient movement into the treatment plan [3,21,79]. In some cases, the novelty of emerging platforms is the hybrid configuration of the instrumentation and software, as in the case of a Trilogy imaging system integrated with the CK for use in treating prostate cancer [21,76]. The clinical applications of SRS have greatly expanded since the concept was first described in 1951, and the number of published studies on clinical indications treatable with SRS has exponentially increased [30]. The advantage of one SRS system over another ultimately may depend on the planning target volume and its proximity to critical structures. Yet, the fact remains that comparisons among a broad array of specific SRS/SRT devices are lacking [117]. Evidence from clinical studies will assist in formulating a generally accepted method for evaluation of the

planning technique. As IMRT gains increased use for SRS and SRT treatment of irregularly shaped tumors, inverse planning techniques to determine dose optimization distributions have become more widely accepted [118]. Another important challenge is ensuring that research on the clinical implementation of SRS keeps pace with the commercial availability of varying technologies. Manufacturers, clinicians, and investigators must engage in discussion of the technological tools and applications of SRS instruments. The key current issues on the table include dosimetric algorithms, methods of identifying targets for treatment, and discussions of field sizes and immobilization. Only through continued clinical research can adequate data be generated to not only resolve the immediate issues at hand, but also to address new concerns that will surely arise [119]. The stereotactic method has moved out of the brain, to the spine, and completely out of the neurosurgical realm and into the rest of the body. Patients with tumors of the chest, pancreas, liver, prostate, and elsewhere are being offered SRS or SRT, a trend that will no doubt only increase. Neurosurgeons will confine their SRS treatments to patients with disorders of the nervous system and their colleagues in other surgical disciplines will learn the principles of SRS appropriate to their patients’ needs. Whether with newer Linac systems, dedicated systems like the Cyberknife, the Perfexion Gamma Knife, heavy particle irradiation, or the descendants of any of these devices, stereotactic radiosurgery will be an increasingly powerful tool in the hands of neurosurgeons. Leksell’s brainchild has become the father to one of the most important and exciting fields of medicine in the twenty-first century.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-9.


2. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psyc 1983;46:797-803. 3. Jayarao M, Chin LS. Robotics and its applications in stereotactic radiosurgery. Neurosurg Focus 2007;23(6):E6. 4. Betti O, Derechinsky V. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurochir Suppl 1984;33:385-90. 5. Betti OO, Derechinsky VE. Irradiacion hiperselectiva encefalica con acelerador lineal: Unidad Multihaz Jean Talairach. Prensa Me´dica Argentina 1983;70:102-7. 6. Colombo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C, Chierego G, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985;16:154-60. 7. Niranjan A, Gobbel GT, Kondziolka D, Lunsford LD. Heritage of radiosurgical research, current trends and future perspective. Prog Neurol Surg 2007;20:359-74. 8. Hoh DJ, Liu CY, Pagnini PG, Yu C, Wang MY, Apuzzo ML. Chained lightning, part I: Exploitation of energy and radiobiological principles for therapeutic purposes. Neurosurgery 2007;61(1):14-27. 9. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988;22(3):454-64. 10. Maciunas RJ. Computer-assisted neurosurgery. Clin Neurosurg 2006;53:267-71. 11. Linskey ME. The changing role of stereotaxis in surgical neuro-oncology. J Neurooncol 2004;69(1–3):35-54. 12. Housepian EM. Stereotactic surgery: the early years. Neurosurgery 2004;55(5):1210-4. 13. Barnett GH, Linskey ME, Adler JR, Cozzens JW, Friedman WA, Heilbrun MP, et al. American Association of Neurological Surgeons; Congress of Neurolofical Surgeons Washington Committee Stereotactic Radiosurgery Task Force. Stereotactic radiosurgery–an organized neurosurgery-sanctioned definition. J Neurosurg 2007;106(1):1-5. 14. Niranjan A, Gobbel G, Kondziolka D, Lunsford D. Radiosurgical research: what has it told us? what do we still need to know?. Tech Neurosurg 2003;9(3):242-50. 15. Huntzinger C, Friedman W, Bova F, Fox T, Bouchet L, Boeh L. Trilogy image-guided stereotactic radiosurgery. Med Dosim 2007;32(2):121-33. 16. Shaw E, Kline R, Gillin M. Radiation therapy oncology group: radiosurgery quality assurance guidelines. Int J Radiat Oncol Biol Phys 1993;27:1231-9. 17. Chen JC, Girvigian MR. Stereotactic radiosurgery: instrumentation and theoretical aspects – part 1. Permanente J 2005;9(4). 18. Schweikard A, Schlaefer A, Adler JR. Resampling: an optimization method for inverse planning in robotic radiosurgery. Med Phys 2006;33(11):4005-11. 19. Hoh DJ, Liu CY, Pagnini PG, Yu C, Wang MY, Apuzzo ML. Chained lightning, part I: Exploitation of energy and radiobiological principles for therapeutic purposes. Neurosurgery 2007;61(1):14-27.

57

20. Suh JH, Saxton JP. Conventional radiation therapy for skull base tumors: an overview. Neurosurg Clin N Am 2000;11(4):575-86. 21. Andrews DW, Bednarz G, Evans JJ, Downes B. A review of 3 current radiosurgery systems. Surg Neurol 2006;66 (6):559-64. 22. Carol M, Grant WH, III, Pavord D, Eddy P, Targovnik HS, Butler B, et al. Initial clinical experience with the Peacock intensity modulation of a 3-D conformal radiation therapy system. Stereotact Funct Neurosurg 1996;66 (1–3):30-4. 23. Alexander E III, Loeffler JS. Radiosurgery using a modified linear accelerator. Neurosurg Clin N Am 1992;3:167-90. 24. Gerbi BJ, Higgins PD, Cho KH, Hall WA. Linac-based stereotactic radiosurgery for treatment of trigeminal neuralgia. J Appl Clin Med Phys 2004;5(3):80-92. 25. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997;37 (3):679-88. 26. Urie MM, Lo YC, Litofsky S, FitzGerald TJ. Miniature multileaf collimator as an alternative to traditional circular collimators for stereotactic radiosurgery and stereotactic radiotherapy. Stereotact Funct Neurosurg 2001;76 (1):47-62. 27. Hillard VH, Shih LL, Chin S, Moorthy CR, Benzil DL. Safety of multiple stereotactic radiosurgery treatments for multiple brain lesions. J Neurooncol 2003;63(3):271-8. 28. Plowman PN Doughty D. Stereotactic radiosurgery, X: clinical isodosimetry of gamma knife versus linear accelerator X-knife for pituitary and acoustic tumours. Clin Oncol (R Coll Radiol) 1999;11:(5)321-9. 29. McClelland S III, Gerbi BJ, Higgins PD, Orner JB, Hall WA. Safety and efficacy of fractionated stereotactic radiotherapy for acoustic neuromas. J Neurooncol 2008;86 (2):191-4. 30. Niranjan A, Lunsford LD. Radiosurgery: where we were, are, and may be in the third millennium. Neurosurgery 2000;46(3):531-43. 31. Wurm RE, Erbel S, Schwenkert I, Gum F, Agaoglu D, Schild R, et al. Novalis frameless image-guided noninvasive radiosurgery: initial experience. Neurosurgery 2008;62(5 Suppl):A11-18. 32. Jensen RL, Wendland MM, Chern SS, Shrieve DC. Novalis intensity-modulated radiosurgery: methods for pretreatment planning. Neurosurgery 2008;62 (5 Suppl):A2-9; discussion A9-10. 33. IGRT Adaptive Gating, radiotherapy solutions. 34. Yan H, Yin FF, Kim JH. A phantom study on the positioning accuracy of the Novalis Body system. Med Phys 2003;30(12):3052-60. 35. Pedroso AG, De Salles AA, Tajik K, Golish R, Smith Z, Frighetto L, et al. Novalis Shaped Beam Radiosurgery of

893

894

57 36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46. 47.

48. 49.

50.


arteriovenous malformations. Neurosurg 2004;101 Suppl 3:425-34. Teh BS, Paulino AC, Lu HH, Chiu JK, Richardson S, Chiang S, et al. Versatility of the Novalis system to deliver image-guided stereotactic body radiation therapy (SBRT) for various anatomical sites. Technol Cancer Res Treat 2007;6(4):347-54. Chen JC, Girvigian M, Greathouse H, Miller M, Rahimian J. Treatment of trigeminal neuralgia with linear accelerator radiosurgery: initial results. J Neurosurg 2004;101 Suppl 3:346-50. Rahimian J, Chen JC, Rao AA, Girvigian MR, Miller MJ, Greathouse HE. Geometrical accuracy of the Novalis stereotactic radiosurgery system for trigeminal neuralgia. J Neurosurg 2004;101 Suppl 3:351-5. Frighetto L, De Salles A, Medin P, Selch M. Shaped beam stereotactic radiosurgery and radiotherapy for the brain and spine. Techniques in Neurosurgery. Radiosurgery 2003;9(3):204-17. Huntzinger C, Friedman W, Bova F, Fox T, Bouchet L, Boeh L. Trilogy image-guided stereotactic radiosurgery. Med Dosim 2007;32(2):121-33. Fan J, Paskalev K, Li J, Wang L, Chen L, Price R, et al. MO-D-AUD-07: determination of output Ffactors for stereotactic radiosurgery beams by Monte Carlo and measurements. Med Phys 34(6):2522. VanderSpek L, Wang J, Alksne J, Murphy K. Single fraction, single isocenter Intensity Modulated Radiosurgery (IMRS) for multiple Brain netastases: dosimetric and early clinical experience. Int J Radiat Oncol Biol Phys 2007;69(3):S265. Mamalui-Hunter M, Drzymala R, Willcut V, Santanam L, Low D. TH-D-AUD-08: Automated stereotactic radiosurgery quality assurance using a portal imager. Med Phys 2007;34(6):2642. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997;37 (3):679-88. Yartsev S, Kron T, Van Dyk J. Tomotherapy as a tool in image-guided radiation therapy (IGRT): current clinical experience and outcomes. Biomed Imaging Interv J 2007;3(1):e16. Mackie TR. History of tomotherapy? Phys Med Biol 2006;51(13):R427-53. Bauman G, Yartsev S, Coad T, Fisher B, Kron T. Helical tomotherapy for craniospinal radiation. Br J Radiol 2005; 78(930):548-52. Tomotherapy Product Information. Mackie R, Balog J, Ruchala K, Shepard D, Aldridge S, Fitchard E, et al. Tomotherapy. Semin Radiat Oncol 1999;9:108-17. Baisden JM, Benedict SH, Sheng K, Read PW, Larner JM. Helical TomoTherapy in the treatment of central

51.

52.

53.

54.

55.

56.

57.

58. 59.

60.

61.

62.

63.

64.

nervous system metastasis. Neurosurg Focus 2007; 22(3):E8. Welsh JS, Lock M, Harari PM, Tome´ WA, Fowler J, Mackie TR, et al. Clinical implementation of adaptive helical tomotherapy: a unique approach to image-guided intensity modulated radiotherapy. Technol Cancer Res Treat 2006;5(5):465-79. Kim B, Soisson ET, Duma C, Chen P, Hafer R, Cox C, et al. Image-guided helical Tomotherapy for treatment of spine tumors. Clin Neurol Neurosurg 2008;110 (4):357-62. Balog J, Soisson E. Helical tomotherapy quality assurance. Int J Radiat Oncol Biol Phys 2008;71(1 Suppl): S113-7. Yartsev S, Kron T, Cozzi L, Fogliata A, Bauman G. Tomotherapy planning of small brain tumours. Radiother Oncol 2005;74(1):49-52. Shi C, Penãgarıćano J, Papanikolaou N. Comparison of IMRT treatment plans between linac and helical tomotherapy based on integral dose and inhomogeneity index. Med Dosim 2008;33(3):215-21. Hansen EK, Bucci MK, Quivey JM, et al. Repeat CT imaging and replanning during the course of IMRT for head-and-neck cancerInt J Radiat Oncol Biol Phys 2006;64(2):355-62. Penãgarıćano JA, Yan Y, Shi C, Linskey ME, Ratanatharathorn V. Dosimetric comparison of Helical Tomotherapy and Gamma Knife Stereotactic Radiosurgery for single brain metastasis. Radiat Oncol 2006;1:26. Zhang T. Respiratory gating and 4-D tomotherapy. Med Phys 2004;31:3529. Gutie´rrez AN, Westerly DC, Tome´ WA, Jaradat HA, Mackie TR, Bentzen SM, et al. Whole brain radiotherapy with hippocampal avoidance and simultaneously integrated brain metastases boost: a planning study. Int J Radiat Oncol Biol Phys 2007;69(2):589-97. Welsh JS, Mehta MP, Mackie TR, Orton N, Jaradat H, Khuntia D, et al. Helical tomotherapy as a means of delivering scalp-sparing whole brain radiation therapy. Technol Cancer Res Treat 2005;4(6):661-2. Roberge D, Parker W, Niazi TM, Oliveres M. Treating the contents and not the container: Dosimetric study of hair-sparing Whole Brain Intensity Modulated Radiation Therapy. Technol Cancer Res Treat 2005;4:567-70. Adler JR,Murphy MJ, Chang SD, Hancock SL. Image-guided robotic radiosurgery. Neurosurgery 1999;44:1299-307. Gildenberg PL, Woo SY. Multimodality program involving stereotactic surgery in brain tumor management. Stereotact Funct Neurosurg 2000;75(2–3):147-52. Chang SD, Main W, Martin DP, Gibbs IC, Heilbrun MP. An analysis of the accuracy of the CyberKnife: a robotic frameless stereotactic radiosurgical system. Neurosurgery 2003;52:140-7.


65. Gerszten PC, Ozhasoglu C, Burton SA, Vogel W, Atkins B, Kalnicki S, Welch WC. Evaluation of CyberKnife frameless real-time image-guided stereotactic radiosurgery for spinal lesions. Stereotact Funct Neurosurg 2003;81(1–4):84-9. 66. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA, Kalnicki S, Welch WC. CyberKnife frameless singlefraction stereotactic radiosurgery for benign tumors of the spine. Neurosurg Focus 2003;14(5):e16. 67. Hara W, Soltys SG, Gibbs IC. CyberKnife robotic radiosurgery system for tumor treatment. Expert Rev Anticancer Ther 2007;7(11):1507-15. 68. Yu C, Gabor J, Apuzzo M, Zbigniew P. Dosimetric comparison of CyberKnife with other radiosurgical modalities for an ellipsoidal target. Neurosurgery 2003;53 (5):1155-63. 69. Romanelli P, Chang SD, Koong A, Adler John R. Extracranial radiosurgery using the CyberKnife. Techniques in neurosurgery. Radiosurgery 2003;9(3):226-31. 70. Collins SP, Coppa ND, Zhang Y, Collins BT, McRae DA, Jean WC. CyberKnife radiosurgery in the treatment of complex skull base tumors: analysis of treatment planning parameters. Radiat Oncol 2006;1:46. 71. Muacevic A, Wowra B. Cyberknife radiosurgery – overview. European neurological disease. Touch briefings; 2007. 72. Giller CA, Berger BD, Gilio JP, Delp JL, Gall KP, Weprin B, et al. Feasibility of radiosurgery for malignant brain tumors in infants by use of image-guided robotic radiosurgery: preliminary report. Neurosurgery 2004;55 (4):916-24. 73. Adler JR, Chang SD, Murphy MJ, Doty J, Geis P, Hancock SL. The Cyberknife: a frameless robotic system for radiosurgery. Stereotact Funct Neurosurg 1997;69:124-8. 74. Gibbs IC. Frameless image-guided intracranial and extracranial radiosurgery using the Cyberknife robotic system. Cancer Radiother 2006;10(5):283-7. 75. Hara W, Soltys SG, Gibbs IC. CyberKnife robotic radiosurgery system for tumor treatment. Expert Rev Anticancer Ther 2007;7(11):1507-15. 76. Pawlicki T, Cotrutz C, King C. Prostate cancer therapy with stereotactic body radiation therapy. Front Radiat Ther Oncol 2007;40:395-406. 77. Romanelli P, Schweikard A, Schlaefer A, Adler J. Computer aided robotic radiosurgery. Comput Aided Surg 2006;11(4):161-74. 78. Chang SD, Poen J, Hancock SL, Martin DP, Adler JR, Jr. Acute hearing loss following fractionated stereotactic radiosurgery for acoustic neuroma: report of two cases. J Neurosurg 1998;89:321-5. 79. Chang SD, Adler JR. Robotics and radiosurgerythe cyberknife. Stereotact Funct Neurosurg 2001;76 (3–4):204-8. 80. Shimamoto S, Inoue T, Shiomi H, Sumida I, Yamada Y, Tanaka E, et al. CyberKnife stereotactic irradiation

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

57

for metastatic brain tumors. Radiat Med 2002;20 (6):299-304. Villavicencio AT, Lim M, Burneikiene S, Romanelli P, Adler JR, McNeely L, et al. Cyberknife radiosurgery for trigeminal neuralgia treatment: a preliminary multicenter experience. Neurosurgery 2008;62(3): 647-55. Chang SD, Gibbs IC, Sakamoto GT, Lee E, Oyelese A, Adler JR Jr. Staged stereotactic irradiation for acoustic neuroma. Neurosurgery 2005;56(6):1254-61. Adler JR Jr, Gibbs IC, Puataweepong P, Chang SD. Visual field preservation after multisession cyberknife radiosurgery for perioptic lesions. Neurosurgery 2006;59 (2):244-54. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady JR. Preliminary clinical experience with linear acceleratorbased spinal stereotactic radiosurgery. Neurosurgery 1995;36(2):311-9. Yu C, Apuzzo MLJ, Zee CS, Petrovich Z. A phantom study of the geometric accuracy of computed tomographic and magnetic resonance imaging svereotactic localization with the Leksell stereotactic system. Neurosurgery 2001;48:1092-9. Drzymala RE, Mutic S. Stereotactic imaging quality assurance using an anthropomorphic phantom. Comput Aided Surg 1999;4:248-55. Maciunas RJ, Galloway RL Jr, Latimer JW. The application accuracy of stereotactic frames. Neurosurgery 1994;35(4):682-94. Serago CF, Lewin AA, Houdek PV, Gonzalez-Arias S, Hartmann GH, Abitbol AA, et al. Stereotactic target point verification of an X ray and CT localizer. Int J Radiat Oncol Biol Phys 1991;20:517-23. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys 1988;14:373-81. Ho AK, Fu D, Cotrutz C, Hancock SL, Chang SD, Gibbs IC, et al. A study of the accuracy of cyberknife spinal radiosurgery using skeletal structure tracking. Neurosurgery 2007;60(2 Suppl 1):147-56. Gerszten PC, Welch WC. Cyberknife radiosurgery for metastatic spine tumors. Neurosurg Clin N Am 2004;15 (4):491-501. Gerszten PC, Ozhasoglu C, Burton SA, Welch WC, Vogel WJ, Atkins BA, et al. CyberKnife frameless singlefraction stereotactic radiosurgery for tumors of the sacrum. Neurosurg Focus 2003;15(2):E7. Lindquist C, Paddick I. The Leksell Gamma Knife Perfexion and comparisons with its predecessors. Neurosurgery 2007;61(3 Suppl):130-40. Regis J, Tamura M, Guillot C, Muracciolle X, Nagaje M, Porcheron D. Radiosurgery. Radiosurgery of the head and neck with the world’s first fully robotized 192 Cobalt-60 source Leksell Gamma Knife Perfexion in clinical use. www.elekta.com/healthcare-internationalgamma-knife-surgery.php

895

896

57


95. Levy P, Schulte RWM, Slater JD, Miller DW, Slater JM. Stereotactic radiosurgery. The role of charged particles. Acta oncologica 1999;38(2):165-9. 96. Chen CC, Chapman P, Petit J, Loeffler J. Proton radiosurgery in neurosurgery. Neurosurg Focus 2007;23 (6):E5. 97. Suit H, Goldberg S, Niemierko A, Trofimov A, Adams J, Paganetti H, et al. Proton beams to replace photon beams in radical dose treatments. Acta Oncol 2003;42 (8):800-8. 98. Schulz-Ertner D, Tsujii H. Particle radiation therapy using proton and heavier ion beams. J Clin Oncol;2007;25(8):953-64. 99. Schulz-Ertner D, Karger CP, Feuerhake A, Nikoghosyan A, Combs SE, Ja¨kel O, et al. Effectiveness of carbon ion radiotherapy in the treatment of skull-base chordomas. Int J Radiat Oncol Biol Phys 2007;68(2):449-57. 100. Levin WP, Kooy H, Loeffler JS, DeLaney TF. Proton beam therapy. Br J Cancer 2005;93(8):849-54. 101. Hug EB. Protons versus photons: a status assessment at the beginning of the 21st Century. Radiother Oncol 2004;73 Suppl 2:S35-7. 102. Weber DC, Chan AW, Bussiere MR, Harsh GR, Ancukiewicz M, Barker FG II, et al. Proton beam radiosurgery for vestibular schwannoma: tumor control and cranial nerve toxicity. Neurosurgery 2003;53: 577-88. 103. Harsh GR IV, Thornton AF, Chapman PH, Bussiere MR, Rabinov JD, Loeffler JS. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys 2002;54:35-44. 104. Hug EB. Review of skull base chordomas: prognostic factors and long-term results of proton-beam radiotherapy. Neurosurg Focus 2001;10(3):E11. 105. Pommier P, Liebsch NJ, Deschler DG, Lin DT, McIntyre JF, Barker FG 2nd, et al. Proton beam radiation therapy for skull base adenoid cystic carcinoma. Arch Otolaryngol Head Neck Surg 2006;132(11):1242-9. 106. Hug EB, Sweeney RA, Nurre PM, Holloway KC, Slater JD, Munzenrider JE. Proton radiotherapy in management of pediatric base of skull tumors. Int J Radiat Oncol Biol Phys 2002;52(4):1017-24. 107. Hug E, Slater J. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. Neurosurg Clin North Am 2000;11:627-38. 108. Bowyer J, Natha S, Marsh I, Foy P. Visual complications of proton beam therapy for clival chordoma. Eye 2003;17 (3):318-23. 109. St Clair WH, Adams JA, Bues M, Fullerton BC, Shell S, Kooy HM, et al. Advantage of protons compared to conventional X-ray or IMRT in the treatment of a

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

pediatric patient with medulloblastoma. Int J Radiat Oncol Biol Phys 2004;58:727-34. Tsujii H, Mizoe J, Kamada T, Baba M, Tsuji H, Kato H, et al. Clinical results of carbon ion radiotherapy at NIRS. J Radiat Res (Tokyo) 2007;48 Suppl A:A1-A13. Mizoe JE, Tsujii H, Hasegawa A, Yanagi T, Takagi R, Kamada T, et al. Organizing Committee of the Central Nervous System Tumor Working Group. Phase I/II clinical trial of carbon ion radiotherapy for malignant gliomas: combined X-ray radiotherapy, chemotherapy, and carbon ion radiotherapy. Int J Radiat Oncol Biol Phys 2007; 69(2):390-6. Tsujii H, Mizoe JE, Kamada T, Baba M, Kato S, Kato H, et al. Overview of clinical experiences on carbon ion radiotherapy at NIRS. Radiother Oncol 2004;73 Suppl 2:S41-9. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999;175 Suppl 2:57-63. Debus J, Schulz-Ertner D, Schad L, Essig M, Rhein B, Thillmann CO, et al. Stereotactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base. Int J Radiat Oncol Biol Phys 2000;47 (3):591-6. Caporaso GJ, Mackie TR, Sampayan S, Chen YJ, Blackfield D, Harris J, et al. A compact linac for intensity modulated proton therapy based on a dielectric wall accelerator. Phys Med 2008;24(2):98-101. Dietmar G, Bogner J, Dieckmann K, Po¨tter R. Is maskbased stereotactic head-and-neck fixation as precise as stereotactic head fixation for precision radiotherapy? Int J Radiat Oncol Biol Phys 2006;66(4Suppl 1):S61-6. Grabenbauer GG, Ernst-Stecken A, Schneider F, Lambrecht U, Ganslandt O. Radiosurgery of functioning pituitary adenomas: comparison of different treatment techniques including dynamic and conformal arcs, shaped beams, and IMRT. Int J Radiat Oncol Biol Phys 2006;66(4 Suppl 1):S33-9. Grzadziel A, Grosu A, Kneschaurek P. Threedimensional conformal versus intensity-modulated radiotherapy dose planning in stereotactic radiotherapy: application of standard quality parameters for plan evaluation. Int J Radiat Oncol Biol Phys 2006;66 (4 Suppl 1):S87-94. Wurm R, Okunieff P. Intracranial and extracranial stereotactic radiosurgery and radiotherapy. Int J Radiat Oncol Biol Phys 2006;66(4 Suppl 1):S1-2. Gerszten PC, Burton SA, Ozhasoglu C, McCue KJ, Quinn AE. Radiosurgery for benign intradural spinal tumors. Neurosurgery 2008;62(4):887-95; discussion 895-6.

69 Proton Beam Radiosurgery: Clinical Experience H. A. Shih . P. H. Chapman . J. S. Loeffler

Introduction The clinical benefits of proton radiation lie within the ability to deliver radiation therapy with improved conformality. This translates into a reduction of potential treatment-related adverse effects. In some situations, the improved dose distribution also permits for higher doses to be delivered without increasing the normal tissue injury probability. Despite these advantages of proton radiotherapy, injury to normal tissues still occur and include serious morbidities such as cranial neuropathies or neurocognitive dysfunction. Such injuries are often delayed in onset and may arise months to years following treatment. Neurological toxicities are perhaps most devastating in patients with benign lesions as they can significantly compromise long-term quality of life. Potential adverse effects of treatment should be carefully weighed when considering any treatment by proton radiosurgery.

Benign CNS Lesions Arteriovenous Malformations When surgery or embolization is not an option for the management of arteriovenous malformations (AVMs), radiosurgery can be considered. When applied, the purpose of radiation therapy for AVMs is to completely obliterate the nidus. This eliminates the risk of intracranial hemorrhage that otherwise occur at rate of 2–4% per year [1]. Common indications for radiosurgery #


include surgical inaccessibility, expected high probability of surgical morbidity due to large AVM size or eloquent location, medically inoperable status, or patient refusal. When treating with radiosurgery, factors to be considered include AVM size, location, volume, venous drainage, and patient age [2,3]. Photon-based stereotactic radiosurgery can effectively obliterate relatively small lesions with low treatment-related toxicity and cure rates of approximately 80% [4–6]. The rate of obliteration is related to dose and increases with time. Re-irradiation of incompletely responding lesions can raise obliteration rates further without adding significant morbidity [7]. Proton radiosurgery can achieve the same success as photon-based techniques for treating small lesions, but the primary benefit of proton radiosurgery is best appreciated when treating larger unresectable AVMs. Proton radiosurgery has been utilized for this purpose often in situations when photon-based methods would be considered unsafe. Proton radiation enables safer dose escalation than photon-based therapy. Even at doses equivalent to photons, proton radiosurgery would be expected to have fewer long-term sequelae as a result of the spared excess normal tissue radiation exposure while maintaining equivalent obliteration rates. AVMs have been treated with proton radiosurgery at the Harvard Cyclotron Laboratory (HCL) since 1965 [8]. In the initial report of 75 patients followed for 2–16 years, symptomatic improvement was achieved in 75% of those who had presented with seizures or severe headaches. Of 62 patients with angiographic follow up, 20% showed complete

1132

69

Proton beam radiosurgery: clinical experience

AVM obliteration, 56% showed AVM size reduction of greater than 50%, and 13% showed no radiographic changes. Seifert et al. [9] reported a limited study of 63 patients referred to a proton treatment facility in the U.S. but who had both pre-radiation and post-radiation follow up locally. Patients were largely selected because of surgically inoperable AVMs or patient refusal of surgery. Among those patients with AVMs 3 cm in diameter, 76% of them had clinical improvement in symptoms, most commonly of seizures, headaches, or other neurological symptoms. Twelve percent of patients with small AVMs had progression of symptoms. As the size of AVM increased, the obliteration rate decreased while the rate of adverse effect increased (33 and 44.5%, respectively, for AVMs >6 cm diameter). Likewise, 86% of patients with SpetzlerMartin grade I or II AVMs had clinical improvement whereas only 54% of grade III and 24% of grade IV patients experienced improvement. The limitation of this study was the lack of treatment details of the proton radiation delivered. The iThemba LABS in the Republic of South Africa reported their proton radiosurgery experience with AVMs [10]. These investigators chose their radiation doses judiciously and frequently hypofractionated in two or four fractions for large volume lesions. At a median follow up of 62 months, they achieved a 67% obliteration rate for AVM treatment volumes of Figure 61-2).

Production of Protons for Therapy When a hydrogen atom is stripped off, using electric fields, its orbiting negatively charged electron, what remains is a proton. In a treatment facility, the protons are usually initially accelerated in a linear accelerator before being injected into a cyclotron or a synchrotron where they are accelerated to the required energy for treatment. A cyclotron will accelerate protons to the highest energy of that particular system. If protons of lower energies are required, an energy degrading process is used to accomplish that. A synchrotron, on the other hand, can produce protons of a specified energy in one pulse and change the energy on every pulse. The exiting narrow beam of high energy protons from either a cyclotron or a synchrotron can then be either spread out by various scattering devices (scattered protons), or deflected by special magnets to scan across the

958

61

Proton beam radiotherapy: technical and clinical aspects

. Figure 61-1 Dosimetric advantange of protons

vertical or horizontal dimensions of a tumor (scanning protons). Currently most treatment facilities use the scattered proton beams for therapy but increasing number of sites are investigating the use of scanning proton beams because of the potential of improved dose distributions over scattered proton beams.

Clinical Experience of Proton Therapy in Brain and Skull-base Tumors Meningioma Approximately 90–95% of intracranial meningiomas are benign, 5% are atypical and 3% malignant. Surgery is the primary treatment. However, incomplete resection can result in high recurrence rate over time (up to 55% at 10 years and 91% at 15 years) [1]. Tumors at certain locations such as

the cavernous sinus, petro-clival region or sphenoid wing are not generally suitable for aggressive surgery because of the significant morbidity associated with it. Several retrospective series have demonstrated a reduction of recurrence after incomplete surgery or good local control with radiation therapy [2–5]. Because benign meningiomas are usually well delineated on MRI, they are well suited to be irradiated by highly conformal radiation therapy techniques. Proton therapy is such a technique. When compared to other photon based conformal techniques, proton therapy could reduce the volume of normal brain receiving a low to moderate dose of radiation because of the physical characteristics of protons as described above. The reported 3–5 year control rate of recurrent, incompletely resected meningiomas(including a few cases of atypical and malignant histologies) with proton therapy(+ or a component of photon therapy) to doses of 50–74 CGE is 80–100% [6–9]. > Table 61-1 summarizes the result. Late


61

. Figure 61-2 Medulloblastoma

. Table 61-1 Results of proton therapy for meningiomas Investigators

Follow-up (months)

Local control

Survival

Noel et al.a Vernimmen et al. Wenkel et al. Weber et al.

37 (median) 40 (mean) 53 (median) 34

4-year: 87.5% +/ 12% 89% 93% 10-year: 77% 3-year: 91.7%

4-year: 88.9% +/ 11% – 5-year: 93% 3-year: 92.7%

Reference [6] [7] [8] [9]

a

Includes a few patients with atypical/malignant meningiomas

complication rates were reported as 10–25%. For atypical and malignant meningiomas specifically, one series from Boston reported 5-and 8-year control rates of 38%(atypical) and 52%(malignant), and 19%(atypical) and 17%(malignant) respectively [10]. The investigators noted significantly improved local control for proton therapy versus photon radiation therapy (80% vs. 17% at 5 years, p = 0.003).

Craniopharygioma Craniopharyngioma is curable with complete resection. However, in many cases a gross total resection is not possible without a significant morbidity because of the attachment of the tumor to critical structures such as the optic nerve/chiasm and hypothalamus. Radiation therapy after tumor cyst decompression and biopsy

959

960

61


has been reported to produce an equivalent curerate as gross total resection [11–13]. In one retrospective series, the neuron-cognitive outcome was reported to be better with radiation therapy than with surgery [14]. Proton therapy has been used in a small number of patients with reported 5-year and 10-year local control rates of 93 and 85% respectively [15,16]. The functional status of the living adult patients was reported as unaltered from the pre-radiotherapy status. One of five children showed some learning difficulties [15]. One of twelve patients in another series developed new-onset panhypopituitarism [16]. Craniopharyngioma is currently considered suitable for proton therapy.

Pituitary Adenoma Surgery is the primary treatment for pituitary adenoma. Residual tumor (macroscopic or microscopic) after surgery can be present if the tumor invades the cavernous sinus, has a large supraseller extension, or significantly erodes the bony floor of the sella. Post-operative radiation therapy in doses of 45–50.4 Gy has produced a 10-year relapse-free survival rate higher than 90% [17]. Currently one of the conformal radiation techniques is recommended. The experience with proton therapy is limited. One series with relatively short follow-up reported no progression of tumor in 44 out of 47 patients [18]. The only morbidity was hypopituitarism.

Low-grade Glioma Low-grade glioma is usually subjected to resection. If a gross total resection is achieved, adjuvant radiation therapy is generally not necessary. Even if the resection is incomplete, if the tumor is not in a eloquent area where a recurrence could cause significant functional deficit, radiation therapy can be postponed till evidence of recurrence. This

approach is more relevant in children because the delay in radiation therapy allows for the further development of the young brain. A small experience of conformal proton therapy for 27 children with progressive or recurrent low-grade astrocytoma has been reported [19]. The target doses were between 50.4 and 63 CGE (Cobalt Gray Equivalent). The mean follow-up period was 3.3 years (0.6–6.8 years). The local control and survival was respectively 87 and 93% for central tumors, 71 and 86% for hemispheric tumors and 60 and 60% for brainstem tumors. All children with local control were reported to have maintained their performance status. All children with optic pathway tumors maintained or improved their visual status. One child with Type 1 neurofibromatosis developed Moyamoya syndrome. Because proton therapy, when compared to photon therapy reduces the volume of normal brain receiving low to modest dose of radiation, it has at least a theoretical advantage in children [20].

Medulloblastoma Radiation therapy is a very important modality in any potentially curative treatment program for medulloblastoma. Because medulloblastoma has a propensity to spread in the leptomeningeal space, irradiation of the entire craniospinal axis is usually required. Additional radiation is directed to the tumor bed after craniospinal irradiation. Several dosimetric studies comparing protons with photons showed a clear advantage of proton therapy in its ability to reduce the radiation dose to the eyes, optic chiasm, hypothahamus-pituitary axis, cochleae,thyroid, lung, heart, liver and kidneys [21,22,23]. In addition, one study estimated that proton therapy for medulloblastoma could reduce the risk of second malignancy by ten folds when compared to photon therapy [24]. A recent costeffective evaluation concluded that proton therapy could be cost-effective and cost-saving compared with conventional photon radiation therapy in the


treatment of children with medulloblastoma [25]. Increasing number of children with medulloblastoma are currently being treated with proton therapy. Prospective studies to evaluate the late morbidity and quality of life of children treated with proton craniospinal irradiation are underway.

Chordoma/Chondrosarcoma Chordoma and chondrosarcoma at the base of skull usually present with therapeutic challenges. Although surgery is the mainstay of treatment, total resection is not always possible because of the adjacent vital structures. Post-operative radiation therapy for subtotally resected tumor is usually indicated. Conventional photon radiation therapy to moderately high doses of radiation (50–55Gy) has produced local recurrence rate of 80–100% [26,27]. Higher doses of radiation (66–80 CGE) delivered by proton therapy have produced 3-year and 5-year local control rates of 91.6–94%, and 75% respectively for chondrosarcoma. For chordoma the 3- to 4-year local control rates have been reported to be 53.8–87.3%, and the 3- to 5-year survival rates have been reported to be 80.5– 93.8% [28,29,30,31] (> Table 61-2). In pediatric patients, the local control rates for chordoma and chondrosarcoma have been reported to be 60 and 100% respectively [32]. Proton therapy has been accepted as the standard treatment for chordoma and chondrosarcoma. Although occasionally chordomas can metastasize to lymph

61

nodes, lungs or bones, or recur along the surgical pathway, the predominant failure pattern for chordomas of the base of skull is local, and savage therapy after a local failure is generally unsuccessful. Thus aggressive upfront treatment is justified. These high doses deliverable with proton therapy nonetheless have been associated with significant complications in a minority of patients. The rate of significant complications such as cranial nerve injury or brain necrosis has been reported to be 7–18% [30,33]. However endocrinopathy has been reported in a significant number of patients [34].Currently 70 CGE is deemed appropriate for chondrosarcoma. The best radiation dose for chordoma is still under study. There are ongoing two prospective randomized studies comparing 70 CGE against 78 CGE, and 72 CGE against 80 CGE.

Conclusions The physical characteristics of protons render them attractive particles for radiation therapy of many tumors in the brain and base of skull, especially in children. The new technologies of proton therapy such as scanning beam and intensitymodulated proton therapy, couple with new images which can more clearly delineate tumors, tumor subpopulations and specific functional areas of the brain will bring about exciting investigations that hopefully will significantly improve the therapeutic ratio of radiation therapy for future patients.

. Table 61-2 Results of proton therapy for chordomas Investigators

Follow-up (months)

Local control

Survival

Reference

Fagundes et al. Hug et al. Noel et al. Weber et al.

54 (median) 33 (mean) 31 (median) 29 (median)

71% 76% 4-year: 53.8% 3-year: 87.3%

– 79% 5-year: 80.5% 3-year: 93.8%

[28] [29] [30] [31]

961

962

61


References 1. Mirimanoff RO, Dosoretz D, Linggood R, et al. Meningioma: analysis of recurrence and progression following neurosurgical resection. J Neurosurg 1985;62:18-24. 2. 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. 3. Nutting C, Brada M, Brazil L, et al. Radiotherapy in the treatment of benign meningiomas of the skull base. J Neurosurg 1999;90:823-7. 4. Pourel N, Auque J, Bracard S, et al. Efficacy of external fractionated radiation therapy in the treatment of meningiomas: a 20-year experience. Radiother Oncol 2001;61:65-70. 5. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation therapy (IMRT) for meningiomas. Int J Radiat Oncol Biol Phys 2002;53:1265-70. 6. Noel G, Habrand JL, Mammar H, et al. Highly conformal therapy using proton component in management of meningiomas. Preliminary experience of the Centre de Protontherapie d’Orsay. Srahlenther Onkol 2002;178 (9):480-5. 7. Vernimmen FJ, Harris JK, Wilson JA, et al. Stereotactic proton beam therapy of skull base meningiomas. Int J Radiat Oncol Biol Phys 2001;49(1):99-105. 8. Wenkel E, Thornton AF, Finkelstein D, et al. benigh meningioma: partially resected, biopsies, and recurrent intracranial tumors treated with combined proton and photon radiotherapy. Int J Radiat Oncol Biol Phys 2000;48(5):1363-70. 9. Weber DC, Lomax AJ, Rutz HP, et al. Spot-scanning proton radiation therapy for recurrent, residual or untreated intracranial meningiomas. Radiother Oncol 2004;71(3):251-8. 10. Hug EB, Devries A, Thornton AF, et al. Management of atypical and malignant meningiomas: role of high-dose, 3D-conformal radiation therapy. J Neurosurg 2000;48 (2):151-60. 11. Weiss M, Sutton L, Marcial V, et al. The role of radiation therapy in the management of childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1989;17:1313-21. 12. Minniti G, Saran F, Traish D, et al. Fractionated stereotactic conformal radiotherapy following conservative surgery in the control of craniopharyngiomas. Radiother Oncol 2007;82(1):90-5. 13. Combs SE, Thilmann C, Huber PE, et al. Achievement of long-term local control in patients with craniopharyngiomas using high precision stereotactic radiotherapy. Cancer 2007;109(11):2308-14. 14. Merchant TE, Kiehna EN, Kun LE, et al. Phase II trial of conformal radiation therapy for pediatric patients with craniopharygioma and correlation of surgical factors and radiation dosimetry with change in cognitive function. J Neurosurg 2006;104(2 Suppl):94-102.

15. Luu QT, Loredo LN, Archambeau JO, et al. Fractionated proton radiation treatment for pediatric craniopharyngioma: preliminary report. Cancer J 2006;12(2):155-9. 16. Fitzek MM, Linggood RM, Adams J, et al. Combined proton and photon irradiation for craniopharyngioma: long-term results of the early cohort of patients treated at Harvard Cyclotron Laboratory and Massachusetts General Hospital. Int J Radiat Oncol Biol Phys 2006;64 (5):1348-54. 17. Brada M, Rajan B, Traish D, et al. The long-term efficacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf) 1993;38:571-8. 18. Ronson BB, Schulte RW, Han KP, et al. Fractionated proton beam irradiation of pituitary adenomas. Int J Radiat Oncol Biol Phys 2006;64(2):425-34. 19. Hug EB, Muenter MW, Archambeau JO, et al. Conformal proton radiation therapy for pediatric low-grade astrocytomas. Strahlenther Onkol 2002;178(1):10-17. 20. Bolsi A, Fogliata A, Cozzi L. Radiotherapy of small intracranial tumours with different advanced techniques using photon and proton beams: a treatment planning study. Radiother Oncol 2003;68(1):1-14. 21. St Clair WH, Adams JA, Bues M, et al. Advantage of protons compared to conventional X-ray or IMRT in the treatment of a pediatric patient with medulloblastoma. Int J Radiat Oncol Biol Phys 2004;58(3): 727-34. 22. Lee CT, Bilton SD, Famiglietti RM, et al. Treatment planning with protons for pediatric retinoblastoma, medulloblastoma, and pelvic sarcoma: how do protons compare with other conformal techniques? Int J Radiat Oncol Biol Phys 2005;63(2):362-72. 23. Timmermann B, Lomax AJ, Nobile L, et al. Novel technique of craniospinal axis proton therapy with the spotscanning system: avoidance of patching multiple fields and optimized ventral dose distribution. Strahlenther Onkol 2007;183(12):685-8. 24. Miralbell R, Lomax A, Cella L, et al. Potential reduction of the incidence of radiation-induced second cancers by using proton beams in the treatment of pediatric tumors. Int J Radiat Oncol Biol Phys 2002;54(3):824-9. 25. Lundkvist J, Ekman M, Ericsson SR, et al. Costeffectiveness of proton radiation in the treatment of childhood medulloblastoma. Cancer 2005;103(4):793-801. 26. Catton C, O’Sullivan B, Bell R, et al. Chordoma: longterm follow-up after radical photon irradiation. Radiother Oncol 1996;41(1):67-72. 27. Colli B, Al-Mefty O. Chordomas of the craniocervical junction: follow-up review and prognostic factors. J Neurosurg 2001;95(6):933-43. 28. Fagundes MA, Hug EB, Liebsch NJ, et al. Radaition therapy for chordomas of the base of skull and cervical spine: patterns of failure and outcome after relapse. Int J Radiat Oncol Biol Phys 1995;33(3):579-84.


29. Hug EB, loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 1999;91(3):432-9. 30. Noel G, Feuvret L, Calugaru, et al. Chordomas of the base of skull and upper cervical spine. One hundred patients irradiated by a 3D conformal technique combining photon and proton beams. Acta Oncol 2005; 4497:700-8. 31. Weber DC, Rutz HP, Pedroni ES, et al. results of spotscanning proton radiation therapy for chordoma and chondrosarcoma of the skull base: the Paul Scherrer Institute experience. Int J Radiat Oncol Biol Phys 2005;63 (2):401-9.

61

32. Hug EB, Sweeney RA, Nurre PM, et al. Proton radiotherapy in management of pediatric base of skull tumors. Int J Radiat Oncol Biol Phys 2002;52(4):1017-24. 33. Santoni R, Liebsch N, Finkelstein DM, et al. Temporal lobe (TL) damage following surgery and high-dose photon and proton irradiation in 96 patients affected by chordomas and chondrosarcomas of the base of skull. Int J Radiat Oncol Biol Phys 1998;41(1):59-68. 34. Pai HH, Thornton A, Katznelson L, et al. Hypothalamic/ pituitary function following high-dose conformal radiotherapy to the base of skull: demonstration of a dose-effect relationship using dose-volume histogram analysis. Int J Radiat Oncol Biol Phys 2001;49(4):1079-92.

963

Stereotactic Radiosurgery

56 Radiobiology of Stereotactic Radiosurgery D. C. Shrieve . J. S. Loeffler

Radiobiology is the study of the effects of radiation on biological systems. Ionizing radiation is known to cause biological effects due to a cascade of events occurring in rapid sequence following the irradiation. An understanding of the biophysical and radiobiological principles governing effects following the administration of therapeutic radiation is essential to the safe and efficacious practice of radiotherapy and radiosurgery. This chapter will discuss these principles in the context of what is commonly referred to as stereotactic radiosurgery, or single fraction radiotherapy. The radiobiology of single fraction treatment is firmly rooted in the radiobiology of fractionated radiotherapy.

Types of Ionizing Radiation Gamma rays and X-rays. Gamma rays and X-rays are electromagnetic radiation with energies ranging from 100 to 2 billion electron volts (eV). X-rays are produced when electrons transition from a higher to lower energy level, usually in the outer shell of heavy atoms and are thus produced outside the nucleus. X-rays may be products of radioactive decay (electron capture) or may be produced from X-ray tubes or linear accelerators, which accelerate electrons onto a heavy metal target producing both a continuous spectrum of photon energies called bremsstrahlung and monoenergetic characteristic X-rays. Gamma rays are photons emitted by radioactive nuclei and have a much narrower range of energies than X-rays, 10 keV to 10 MeV. Once produced, gamma rays and X-rays are #


indistinguishable. 60Co, commercially produced from 59Co, is the most common source of gamma rays used in radiotherapy and undergoes betadecay with a half-life of 5.27 years. It is the subsequent gamma ray emission that makes 60Co applicable to radiotherapy and, more commonly, to stereotactic radiosurgery. 60Co decay results in emission of two discreet gamma energies, 1.17 and 1.33 MeV, giving an effective average energy of 1.21 MeV. Protons. The use of protons in radiotherapy and radiosurgery is based on the physical properties of these particles and the related characteristics of dose deposition in irradiated tissues [1]. Dose deposition is characterized by the Bragg peak. Quantitatively, the entrance dose for particle beams is relatively low compared to photons. An unaltered proton beam will deposit more than 50% of its energy over a narrow 2- to 3-cm path at a depth in water that depends on the beam energy. The beam may be altered to spread the Bragg peak to conform to the thickness and depth of the volume to be treated. However, the entrance dose is significantly increased in this case (> Figure 56-1). The biological effectiveness of X-rays, gamma rays and protons are roughly equivalent and each is considered to be low LET (linear energy transfer), sparsely ionizing radiation. Of note, protons have only a slightly higher radiobiological effectiveness (RBE) than 60Co and megavoltage X-rays. In practice this small difference is accounted for by calculating dose for protons in cobalt Gray equivalents (CGE), whether for single or multiple fractions. Proton radiotherapy and radiosurgery are fully discussed elsewhere in this volume.

854

56

Radiobiology of stereotactic radiosurgery

. Figure 56-1 Schematic representation of depth dose curves for a 160 MeV proton beam. Both unmodulated and spread-out Bragg Peak curves are shown. A 10 MVp X-ray curve is shown for comparison

Basic Principles of Radiobiology

Mammalian Cell Survival Curves

Direct Versus Indirect Effects of Radiation

Cell survival following single doses of ionizing radiation is a probability function of absorbed dose measured in the unit Gray (Gy)*. Typical mammalian cell survival curves obtained following single-dose irradiation in culture have a characteristic shape (> Figure 56-2): a low-dose ‘‘shoulder’’ region is followed by a more steeply sloped, or continuously bending, portion at higher doses. Measurement of radiation doseresponse of cultured human tumor cells has largely been limited to malignant tumor cell types. Studies on such cell lines have shown that the apparent radiosensitivity depends heavily on culture conditions and the assay used to assess cell survival [3–5]. The shoulder region is interpreted as accumulation of sublethal damage at low doses with lethality resulting

When cells are irradiated with low LET radiation, the majority of photon interactions are with water molecules, producing a fast electron and an ionized water molecule through Compton Scattering. These fast electrons interact with other water molecules through further ionizing events. The resulting positively charged water molecules dissociate into H+ ions and OH free hydroxyl radicals with an extremely short half-life (1010 s). Hydroxyl radicals are highly reactive and have sufficient energy to break chemical bonds in nearby (within 2 nm) molecules. This indirect effect of radiation, through the free radical intermediary, produces about 70% of radiation damage. The direct effect results from direct interaction of fast electrons with biologically important molecules (DNA) [2].

*1 J/kg


. Figure 56-2 Curve for mammalian cell survival as a function of single dose of radiation (red line) given as a single fraction. The a/b is 10 Gy, a dose at which the contributions to cell killing by single events (aD, dashed blue line) and the interaction of sublethal events (bD2) are equal

from the interaction of two or more such sublethal events [2,6,7]. It may be considered that DNA is the target molecule for cell killing by ionizing radiation and that a double strand break in the DNA is necessary and sufficient to cause cell death (defined as loss of ability to divide). Double strand breaks may be effectively produced by a single particle track or by the interaction of two single strand breaks caused by separate particle tracks and occurring closely in space and time (> Figure 56-3). Single strand breaks alone may be repaired and therefore represent sublethal damage. Such a model is described by the linear-quadratic formula 2

SF ¼ eðaDþbDÞ

where SF is surviving fraction and D is dose of radiation in Gy [8], a is the coefficient related to single event cell killing and b the coefficient related to cell killing through the interaction of sublethal events. a/b is the ratio of the relative contributions of these two components to overall cell kill. a/b is the single dose at which overall cell killing is equally due to these two components (> Figure 56-2). aD ¼ bD2 or D ¼ a=b

56

. Figure 56-3 Schematic representation of double-strand break production by single events (aD) or interaction of events (bD2)

Most mammalian cell survival curves are well fit to the linear-quadratic model [3,4]. Cell survival following a single dose of radiation in vitro reflects the intrinsic radiosensitivity of a particular cell type to a particular type of radiation [3]. Cell types and tissues may vary in the a/b, resulting in slightly different shaped response curves (> Figure 56-4a). The a component varies little from tissue to tissue and variation in a/b is largely due to variation in the amount of b-type damage. A cell or tissue demonstrating a low a/b will have a relative abundance of b-type damage compared to those demonstrating a high a/b.

Radiobiology of Fractionated Radiotherapy A spectrum of fractionation schedules are used to treat intracranial disease, ranging from single fraction radiosurgery to fully fractionated courses of radiotherapy involving a series of 30 or more daily treatments. For fractionated radiotherapy each dose (fraction) produces similar biological effects, given sufficient interfraction interval (> Figure 56-4b). The linear quadratic formula for fractionated doses becomes 2 n SF ¼ eadbd

855

856

56


where d is the dose per fraction and n is the total number of fractions. A basic principle of radiobiology and radiotherapy is that dose fractionation ‘‘spares’’ virtually all cell and tissue types. ‘‘Sparing’’ in this context means that, for a given total dose, there will always be less molecular damage and a lower

level of biological effect associated with multiple fractions compared to a single dose. As the number of fractions increases, the total dose (n x d) required to achieve a certain level of biological effect also increases (> Figure 56-5). The magnitude of the sparing effect of dose fractionation varies, however, and depends on a/b.

. Figure 56-4 Comparison of single-dose effect curves (a) and fractionated dose-effect curves (B) for low (blue) and high (red) a/b tissues. The small advantage seen in the low dose region sparing low a/b tissues (A) is amplified through dose fractionation (b)

. Figure 56-5 The effect of dose fractionation on the biological effectiveness of x radiation for low a/b (blue lines) versus high a/b (red lines) tissues. Isoeffect curves show the increase in total dose required to maintain biological effectiveness with increasing number of fractions


56

The biologically effective dose (BED) is represented by d BED Gya=b ¼ nd 1þ a=b

context of single dose radiosurgery, a time factor is not likely to be important and has not, therefore, been included.

Where BED is expressed in Gya/b to indicate that it should be used only to compare effects in tissues with the same a/b, n is the number of fractions of dose d and nd is, therefore, the total dose (D). BED can be expressed as BED Gya=b ¼ D F

Biologically Effective Dose of Single Dose Radiosurgery

where F is a ‘‘fractionation factor’’ d F ¼ 1þ a=b F increases with increasing dose/fraction d but the effect is greatest for lower a/b and may be negligible for very high a/b, since as a/b increases F approaches 1. The linear-quadratic formulation is a means of estimating the effects of dose fractionation. Other factors, such as a rapid doubling time may be accounted for by additional terms [9]. In the

The biological effectiveness of single dose radiosurgery increases with dose much more precipitously than does the BED of fractionated radiotherapy. This is due to the unavoidable increase in both total dose and dose per fraction in the BED equation, which for radiosurgery becomes D2 BED Gya=b ¼ Dþ a=b So while a doubling in fractionated radiotherapy dose will exactly double the BED (assuming dose per fraction is not changed), doubling a radiosurgery dose will always more than double the BED by an additional increment dependant on a/b (> Figure 56-6)

. Figure 56-6 Biologically Effective Dose as a function of total dose for radiosurgery (solid line) and fractionated radiotherapy (2 Gy per fraction)

857

858

56


D2 : a=b Various doses of radiosurgery may be compared to the dose of fractionated radiotherapy required to achieve an equivalent BED [10]. This comparison may be made only for tissues having the same a/b. A separation is again noticed between curves for tissues with different a/b (> Figure 56-7).

Tumor and Normal Tissue Radiobiology Radiobiology of CNS tumors. Tumors treated with radiosurgery range from highly malignant brain metastases to ‘‘benign’’ neoplasms, such as vestibular neuroma and meningioma and vascular malformations. Less differentiated, rapidly proliferating tumors, such as brain metastases are generally radiosensitive and radioresponsive. These are often referred to as ‘‘early responding’’ tissues and exhibit a dose response with a high a/b of about 10. More differentiated, slowly proliferating targets and normal CNS are considered to be ‘‘late responding’’ and exhibit a low a/b of 5 or less.

. Figure 56-7 Isoeffective doses for radiosurgery and fractionated radiotherapy (2 Gy per fraction) for low a/b tissues (blue line) and high a/b tissues (red line)

The probability of tumor control (TCP) is a function of the likelihood of inactivating all tumor cells in a given tumor following Poisson statistics. If it is assumed that every tumor cell must be killed to control a tumor, the probability of tumor control (TCP) is given by TPC ¼ eSFN where SF is the surviving fraction and N the total number of cells in the tumor. SF N is then the average number of viable cells remaining in a tumor following a certain treatment. TPC is the probability of no cells remaining viable under these conditions. TPC is a function of total dose and dose per fraction (BED), N (tumor bulk) and radiosensitivity of the tumor. This model leads to a sigmoid dose response curve for TCP (> Figure 56-8). The shape of this curve, indicating no probability of tumor control at low doses, a high probability at high doses and a steep rise in TPC through an intermediate range of doses is characteristic of all radiation dose response data. It should be appreciated that radiosensitivity is not equivalent to radioresponsiveness. Some CNS tumors are very radioresponsive but inevitably recur (e.g., CNS lymphoma), while others may show little or no radiographic evidence of response but are well controlled by modest radiation doses (e.g., meningioma, acoustic neuroma). Benign CNS tumors do not typically grow rapidly and do not shrink rapidly following radiotherapy or radiosurgery. Lack of growth should, therefore, be the goal of such therapy. Benign tumors of the CNS are slowly proliferating, relatively differentiated and have a delayed response to radiotherapy or radiosurgery. Estimates of a/b for meningioma and acoustic neuromas are 2.3–4 Gy [11]. AVMs are extremely slow growing and exhibit a delayed response to radiosurgery. The a/b for AVMs has been estimated to be 2.2 Gy [12]. Brain metastases have relatively rapid rates of growth, are typically derived from highly malignant,


. Figure 56-8 Curves schematically comparing the probability of tumor control (TCP) with the probability of a normal tissue complication (NTCP). (a) The curves are positioned relatively close to one another. Normal tissue complications may be avoided only by minimizing the dose to the critical normal structure. Such a situation may occur when a normal structure such as optic nerve lies adjacent to a benign tumor being treated with single dose radiosurgery. (b) Dose fractionation separates the TCP and NTCP curves allowing for a higher probability of tumor control without significant risk of normal tissue complication. The ‘‘Uncomplicated Cure’’ curve is TCP-NTCP

undifferentiated primary cancers and respond quickly to radiosurgery, consistent with their having a high a/b of approximately 10 Gy.

Factors Affecting the Radiosensitivity of Tumors Physical factors. The relative biological effectiveness (RBE) of a particular beam of radiation depends upon the type of radiation emitted

56

(photons, protons, heavy particles, neutrons) and the energy of the emitted radiation. In general, the RBE is related to the linear energy transfer (LET) of the beam. LET is a measure of the amount of energy deposited per length along the particle track, usually expressed in keV/mm. Photons and protons are considered to be sparsely ionizing or low LET radiation with BED very nearly equal to 1 (reference radiation for RBE is 250 keV X-rays). Neutrons and heavy particles are considered densely ionizing and high LET radiation. In general LET increases with increasing mass or charge and decreases with increasing energy (velocity). LET varies from 0.2 keV/mM for Co60 to more than 1,000 keV/mM for high energy Fe ions (cosmic radiation). RBE reaches a peak at an LET of about 100 keV/ mM and can be as high as about 10 for some endpoints. Another physical factor affecting radiosensitivity is dose rate. Below a dose rate of about 1 Gy per minute cells can repair sublethal damage more quickly than it is produced. This leads to increased survival for a given dose compared to higher dose rates. The dose rate effect is most important between dose rates of about 1 cGy per minute and 100 cGy (1 Gy) per minute. Therapeutic radiotherapy is usually delivered at up to 10 Gy per minute. It has been suggested that overall treatment time for radiosurgery may also be important, especially for multi-isocenter methods, such as gamma knife, when treatments may be delivered over an hour or more. Biological factors. The major biological factor affecting radiosensitivity is position in the cell cycle. In general cells are most sensitive to radiation in the post replication phases of the cell cycle (G2 and mitosis). Cells are most resistant to radiation during replication (S phase). An intermediate radiosensitivity is found in cells that are in the pre-replicative phase (G1 or G0). The magnitude of the difference seen in radiosensitivity between S-phase cells and late G2/M phase cells is up to a magnitude of 3 in terms of a dose modifying factor [13].

859

860

56


Chemical factors. Malignant tumors often are characterized by radiographic and pathologic evidence of necrosis. Necrosis develops as tumors cells grow away from the vascular supply faster than new blood vessels are formed. A gradient of nutrients and oxygen develops between the supplying blood vessel and the areas of necrosis where the lack of oxygen and nutrients reach critically low levels. It has clearly been demonstrated that cells irradiated in this ‘‘hypoxic’’ state are resistant to radiation. At oxygen levels below about 2.5% a threefold dose of radiation may be required to achieve the same biological effect as under ‘‘normoxic’’ conditions. The phenomenon is a manifestation of the ‘‘oxygen effect’’ resulting from the radiosensitizing properties of elemental oxygen. Radiobiologic hypoxia is thought to be important only in tumors and not in normal tissues. Fractionated radiotherapy can effectively overcome hypoxia to some extent since irradiated areas of tumors are know to undergo ‘‘reoxygenation’’ in the intervals between individual fractions [2]. There is evidence that hypoxia may affect the response of brain metastases to radiosurgery [14].

Normal Tissue Radiobiology Model Predicting Normal Tissue Complications Treating physicians must be concerned not only with effects of treatment on tumor, but also with normal tissue effects. The normal tissues of particular interest in the treatment of benign brain tumors are spinal cord and brain stem, optic apparatus and other cranial nerves and brain parenchyma. Also of interest are effects on the vasculature within both normal and tumor tissue. The probability of normal tissue complication (NTCP) following radiotherapy is, like tumor control probability, a function of dose and dose per fraction (BED), the tissue at risk (radiosensitivity) and the volume irradiated.

NTCP has been shown to be well represented by the model NTCP ¼ 1 expR where R is the variable related to dose and volume k R ¼ d=d0 with d0 determining the slope of the NTCP versus dose curve and k being a constant accounting for volume effects [15,16]. This is represented graphically as a sigmoid-shaped curve similar to that obtained for tumor cure (> Figure 56-8). Curves for a wide variety of normal tissue endpoints have been generated. Although each has a similar shape the relative placement of these curves along the dose axis may be quite different. In clinical radiotherapy, the relative positions of the curves for tumor cure and normal tissue complication defines what is known as the therapeutic ratio. The therapeutic ratio may be calculated as Probabality of tumor cure : Probabality of complication An ideal therapeutic ratio would be described by curves that allow 100% tumor cure without appreciable probability of normal tissue complication. The opposite extreme would be exemplified by a tumor requiring high dose radiation for cure located within a critical normal structure with a low tolerance to radiation. In practice a regimen that maximizes the probability of an uncomplicated cure is optimal (> Figure 56-8b). For the situation where a/b for tumor is higher than that for critical normal tissue, dose fractionation will always serve to separate the TCP and NTCP curves and increase the therapeutic ratio. Single dose radiosurgery relies on localizing dose in a way that minimizes dose to normal tissue. The tolerance dose for specific tissues is a function of the selected toxicity endpoint, volume irradiated, total dose, dose per fraction used and the level of acceptable risk [17–20]. For example, the total dose to cause cerebral necrosis


in 5% of patients treated with a single dose of radiosurgery is vastly different than the dose associated with the same risk when conventional fractionation (1.8–2 Gy/day) is employed [10]. Tolerance doses may be expressed as D5/5, or the dose expected to produce complication in 5% of patients within 5 years of treatment [18]. This concept may be useful for effects such as necrosis or pituitary dysfunction, but is not useful for effects such as optic neuropathy or myelitis. Ideally, when dealing with benign tumor treatment, a dose regimen thought to be ‘‘safe’’ to the optic apparatus or spinal cord would seem the better choice. A dose and fractionation scheme effective in achieving tumor control and below the ‘‘threshold’’ for producing clinical neurologic dysfunction would be optimal [21,22]. The tolerance of normal structures in the CNS to single doses of radiosurgery and various fractionation regimens indicates that these structures are typical of late-responding tissues and demonstrate exquisite sparing through dose fractionation consistent with a low a/b. Estimates of a/b for spinal cord (myelopathy), brain parenchyma (cerebral necrosis) and optic nerve and chiasm (neuropathy) are 1–2 Gy [17,22,23].

Models to Compare BED of Different Fractionation Regimens It is important when investigating nonconventional fractionation regimens to have some basis for the choice of fraction size, total dose and interval between fractions. If a/b were well established for all tumors and normal tissues, the linear-quadratic model could provide such a basis. The formula for BED can be used to compare dose regimens of varying total doses and dose per fraction in a particular tissue. The equation may also be used to determine isoeffective total doses D associated with different doses per fraction d D1 =D2 ¼ ða=b þ d2 Þ=ða=b þ d1 Þ

56

Sheline et al. described a model for predicting the risk of brain necrosis as a function of total dose and number of fractions [23]. The model defined an isoeffect line for total dose as a function of fraction number. They defined the neuret, similar to BED, as Neuret ¼ D N0:41 XT0:03 ; where D is the total dose in cGy, N the number of fractions and T the overall time in days. This relationship demonstrated the strong dependence on N, a surrogate for fraction size, and the very weak dependence on overall time, T. These data may also be well fit to the linear –quadratic model using an a/b of 2.0 without a time factor. Although this model was not based on single dose data, it predicts a tolerance dose for brain necrosis of approximately 10 Gy in a single fraction. The literature would indicate that the formula derived by Sheline et al. could approximate isoeffect curves for other CNS effects such as optic neuropathy and spinal cord injury [21,24,25]. Common features of these models are an exponent of N similar to that found by Sheline et al. and an exponent of time (T) that is nearly 0. This emphasizes the importance of the number of fractions, or fraction size, in determining the tolerance dose of normal tissues in the CNS. For normal tissue, and probably most benign brain tumors, the overall time of treatment is relatively unimportant within the range normally encountered in a single radiation course, up to about 8 weeks at which time repopulation may begin in the normal CNS tissue [24]. Attempts to model isoeffective dose regimens for the risk of optic neuropathy following fractionated radiotherapy have lead to a similar model published by Goldsmith et al. [21] who defined the optic ret as Optic ret ¼ D N0:53 : They defined a threshold for optic neuropathy, defining a ‘‘safe’’ regimen as one resulting in no more than 890 optic ret. This corresponds to

861

862

56


5,400 cGy in 30 fractions, 3,750 cGy in 15 fractions and 3,000 cGy in 10 fractions, all commonly used fractionation schedules. Although not based on single fraction data, the optic ret model predicts that a single radiosurgery dose of 8.9 Gy or less would be safe to the optic apparatus. This agrees very well with single fraction tolerance doses proposed in the literature, which range from 8 to 10 Gy [26,27]. The total dose predicted by the optic ret model to be safe may be calculated as DðGyÞ ¼

8:9Gy N0:53

emphasizing the well established importance of fraction size in determining optic nerve tolerance to radiotherapy [28,29]. Leber et al. also demonstrated a relationship between radiosurgery dose to optic nerve and the risk of developing optic neuropathy, data that conform to the expected sigmoid-shaped curve with a steep slope relative to dose (> Figure 56-9). The tolerance of cranial nerves III-VI appears to be substantially higher than for the optic nerves. Leber et al. studied 210 nerves among 50 patients who received single doses of up to 30 Gy with no patient developing neuropathy [26].

. Figure 56-9 NTCP curve constructed for optic neuropathy following radiosurgery. The data points are from [26]

Volume Effects in CNS Normal Tissue Tolerance In animal models there is a clear effect of volume on spinal cord tolerance to radiation [17,30]. This effect is closely related to length of cord irradiated rather than volume. Experiments done on large animal models, including primates, have shown that the effect of increasing the irradiated volume of spinal cord or brainstem is to lower the threshold and increase the slope of the sigmoidshaped dose response curve [17]. The volume effect is much more important in the high dose region than in the low dose region. For example, increasing the volume receiving dose associated with a very low risk of myelopathy or necrosis will not affect the risk very much. However, in the high dose region, volume reduction may substantially lower the risk of toxicity. The dose-response work of Sheline et al. was based largely on patients treated with whole brain radiotherapy [23]. Recent work has clearly shown volume of brain irradiated to be a factor in development of post radiation toxicity. A study undertaken by the Radiation Therapy and Oncology Group (RTOG) examined the maximum tolerated dose (MTD) of single fraction radiosurgery as a function of irradiated volume [31]. Volumes ranging up to 34 cc (40 mm diameter) were included and all tumors were recurrent following previous radiotherapy. Single dose MTDs were established for volumes 4.2–14 cc and those 14.1–34 cc as 18 Gy and 15 Gy to the target margin, respectively. The MTD for tumors smaller than 4.2 cc was not reached at 24 Gy. 15, 18 and 24 Gy are well in excess of the tolerance dose predicted by the neuret model. The corresponding BEDs are 127.5, 180 and 312 Gy2, respectively. Equivalent doses given in 2 Gy fractions would be 64, 90 and 156 Gy, respectively. Volume effects for optic neuropathy are not well established. It has been proposed that small volumes of optic nerve or chiasm may tolerate


higher doses of radiation [32,33], but clear dosevolume guidelines do not exist.

Long-term Recovery of Radiation Damage in the CNS Re-irradiation of critical CNS structures presents an all too common dilemma in radiation oncology. Most commonly of concern is the tolerance of the optic pathway or spinal cord to re-irradiation. Long-term repair of radiation damage in the spinal cord has been clearly demonstrated in animal models. Hornsey and colleagues found longterm repair of rat spinal cord re-irradiated 100 days following an initial dose of radiation [34]. The degree of residual damage present at re-irradiation was dependent on the magnitude of the first dose. Wong and Hao reported similar results, also in rat spinal cord, finding up to 50% recovery following sufficient time, about one year for the maximum effect [35]. Ang and colleagues have reported on similar recovery of occult radiationinduced spinal cord injury in rhesus monkeys [17]. Significant recovery occurred at 1, 2 and 3 years following fractionated radiation to the cervical and thoracic spine. Based on a 5% incidence of myelopathy, recovery was quantified as 76, 85 and 101% at the 1, 2 and 3 year intervals, respectively. Histologic analysis revealed a mixture of white matter necrosis and vascular injury in the symptomatic animals, whereas histologically normal spinal cords were found in asymptomatic animals. Nieder et al. reported on clinical experience with spinal cord reirradiation [36]. They found a very low risk of reirradiation in a group of patients with a cumulative BED of 135 Gy2, an interval of at least 6 months between courses and neither course exceeding a dose equivalent to 98 Gy2. There are no good animal models for radiation optic neuropathy. Clinical data support long-term repair of radiation damage in the visual pathway, however. Schoenthaler et al. [37]

56

reported on 15 patients who received a second course of radiotherapy for recurrent pituitary tumors 2–17 years following initial treatment with radiotherapy. With follow-up of 1–30 years after the second course of treatment, no patient experienced optic neuropathy. Total doses ranged from 5,865 cGy to 10,400 cGy (median 7935 cGy). Flickinger et al. [38] reported on ten patients retreated for suprasellar tumors 1–17 years following initial treatment. Total doses ranged from 7,600 cGy to 9,865 cGy (median 8,500 cGy). One patient developed optic neuropathy 1.5 years following the second radiation course. Using these data it was estimated that 40% of the initial radiation damage remained as residual and recommended that this value be used cautiously in retreating tumors near the optic nerves and chiasm. Overall there is compelling evidence that longterm repair of radiation damage occurs in the CNS, in particular the spinal cord and optic apparatus. This phenomenon is likely at least partly due to repopulation of normal cells from surviving stem cell populations or migration of cells from unirradiated tissue [24]. However, extrapolation from the preclinical and retrospective clinical data available to clinical practice should be taken with caution. The potential benefit of reirradiation, treatment alternatives, and the risk of serious permanent sequelae need to be considered.

Radiobiological Considerations in Treatment Planning for Radiosurgery The best approach to avoiding radiation-induced CNS toxicity is always to minimize dose and volume of irradiated normal tissue. Careful treatment planning based on modern CT and/or MR imaging, effective patient positioning and immobilization and accurate treatment delivery all contribute to the ability to minimize the treated volume and to accurately measure doses to

863

864

56


normal structures. Dose-volume histogram analysis is an essential element in the optimization of treatment planning. Inverse treatment planning and intensity modulated radiotherapy have contributed to reduction in the volume of normal tissue receiving high dose as well as reduction in the dose per fraction and total dose. However, in many situations, when normal structures, especially optic chiasm, brain stem or spinal cord, lie adjacent to the tumor to be irradiated, dose fractionation is the better and perhaps only method to allow safe and efficacious treatment. The high BED associated with radiosurgery at once represents the key to the efficacy of

. Table 56-1 Classification of radiosurgery targets [10] Category

Target

Embedded/ surrounded

Example

I II III IV

Late Late Early Early

Embedded Surrounded Embedded Surrounded

AVM Meningioma Glioma Metastasis

. Figure 56-10 Physical dose (blue) and biologically effective dose (red) as a function of distance from the 50% isodose line for a gamma knife radiosurgery plan. Negative distance is toward target center and positive values are outward from target surface. a/b of 2 Gy is assumed

radiosurgery and the potential for complications associated with its use. Larson et al. described four categories of radiosurgery targets depending upon whether the target is early (high a/b) or late (low a/b) responding and whether it is embedded in or simply surrounded by normal tissue (assumed to always be late responding tissue, > Table 56-1) [10]. The steep dose gradients associated with radiosurgery treatment plans represent an even steeper gradient in BED (> Figure 56-10). This emphasizes the importance of accurate delineation of target and normal structures, precise delivery of treatment as planned and knowledge of the potential error involved in treatment delivery, as even a small error on the order of a millimeter or less can deliver a subtherapeutic or toxic dose.

Conclusion Stereotactic radiosurgery is a powerful tool in the treatment of intracranial tumors. The biological effectiveness of a single dose of radiation provides excellent outcomes in terms of tumor control, when tumor size and location are appropriate. When normal structures such as optic nerve or brain stem are near the target structure, the ability to deliver a therapeutic dose to the target may be hampered by risk to normal tissue. When large volumes are considered for radiosurgery, account must be taken of the large volume of brain parenchyma at risk for necrosis. Understanding of basic radiobiological principles and the accepted tolerance doses for normal tissues are essential for safe and efficacious use of stereotactic radiosurgery.

References 1. Munzenrider JE, Crowell C. Charged Particles. In: Mauch PM, Loeffler JS, editors. Radiation oncology: biology and technology. Philadelphia: W.B. Saunders; 1994. p. 34-55.


2. Hall EJ, Giaccia AJ. Radiobiology for the radiologist. 6th ed. Philadelphia: Lippincott Williams and Wilkins; 2006. 3. Fertil B, Malaise EP. Intrinsic radiosensitivity of human cell lines correlated with radioresponsiveness of human tumors: analysis of 101 published survival curves. Int J Radiat Oncol Biol Phys 1985;11(9):1699-707. 4. Taghian A, Suit H, Pardo F, et al. In vitro intrinsic radiation sensitivity of glioblastoma multiforme. Int J Radiat Oncol Biol Phys 1992;23:55-62. 5. Weichselbaum RR, Nove J, Little JB. X-ray sensitivity of human tumor cells in vitro. Int J Radiat Oncol Biol Phys 1980;6:437-40. 6. Elkind MM, Sutton H. X-ray damage and recovery in mammalian cells in culture. Nature 1959;184:1293-5. 7. Withers H. Biologic basis for altered fractionation schemes. Cancer 1985;55(9):2086-95. 8. McBride WH, Withers HR. Biological basis of radiation therapy. In: Perez CA, Brady LW, Halperin EC, SchmidtUllrich RK, editors. Principles and practice of radiation oncology. 4th ed. Philadelphia: Lippincott, Williams and Wilkins. 2004. p. 96-136. 9. Travis EL, Tucker SL. Isoeffect models and fractionated radiation therapy. Int J Radiat Oncol Biol Phys 1987;13:283-7. 10. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25(3):557-61. 11. Shrieve DC. Basic principles of radiobiology applied to radiotherapy of benign intracranial tumors. Neurosurg Clin N Am 2006;17:67-78. 12. Qi XS, Schultz CJ, Li XA. Possible fractionated regimens for image-guided intensity-modulated radiation therapy of large arteriovenous malformations. Phys Med Biol 2007;52(18):5667-82. 13. Sinclair WK. Cyclic X-ray responses in mammalian cells in vitro. Radiat Res 1968;33:620-43. 14. Goodman KA, Sneed PK, McDermott MW, et al. Relationship between pattern of enhancement and local control of brain metastases after radiosurgery. Int J Radiat Oncol Biol Phys 2001;50(1):139-46. 15. Alber N, Nusslin F. A representation of an NTCP function for local complication mechanisms. Phys Med Biol 2001;46:439-47. 16. Lyman JT, Wolbarst AB. Optimization of radiation therapy. III. A method of assessing complication probabilities from dose-volume histograms. Int J Radiat Oncol Biol Phys 1987;13:103-9. 17. Ang KK. Radiation injury to the central nervous system: clinical features and prevention. In: Meyer JL, editor. Radiation injury: advances in management and prevention, vol 32. Basil: Karger; 1999. p. 145-54. 18. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21(1):109-22. 19. Marks JE, Baglan RJ, Prassad SC, Blank WF. Cerebral radionecrosis: incidence and risk in relation to dose, time,

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

56

fractionation and volume. Int J Radiat Oncol Biol Phys 1981;7:243-52. Marks LB, Spencer DP. The influence of volume on the tolerance of the brain to radiosurgery. J Neurosurg 1991; 75:177-80. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neuropathy after irradiation of meningioma. Radiology 1992;185:71-6. Shrieve DC, Hazard L, Boucher K, Jensen RL. Dose fractionation in stereotactic radiotherapy for parasellar meningiomas: radiobiological considerations of efficacy and optic nerve tolerance. J Neurosurg 2004; 101:390-5. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain injury. Int J Radiat Oncol Biol Phys 1980;6(9):1215-28. van der Kogel AJ. Central nervous system radiation injury in small animals. In: Gutin PH, Leibel SA, Sheline GE, editors. Radiation injury in the nervous system. New York: Raven Press; 1991. p. 91-111. Wara WM, Phillips TL, Sheline GE, Schwade JG. Radiation tolerance of the spinal cord. Cancer 1975; 35:1558-62. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathways and cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neurosurg 1998;88: 43-50. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993;27(2):215-21. Harris JR, Levene MB. Visual complications following irradiation for pituitary adenomas and craniopharyngiomas. Ther Radiol 1976;120:167-71. Parsons JT, Bova FJ, Fitzgerald CR, Mendenhall WM, Million RR. Radiation optic neuropathy after megavoltage external-beam irradiation: analysis of time-dose factors. Int J Radiat Oncol Biol Phys 1994; 30(4):755-63. Hopewell JW, Morris JH, Dixon-Brown A. The influence of field size on the late tolerance of the rat spinal cord to single doses of X-rays. Br J Radiol 1987;60:1099-108. Shaw E, Scott C, Souhami L, et al. Single dose radiosurgical treatment of recurrent previously irradiated primary brain tumors and brain metastases: final report of RTOG protocol 90–05. Int J Radiat Oncol Biol Phys 2000;47:291-8. Habrand JL, Austin-Seymour M, Birnbaum S, et al. Neurovisual outcome following proton radiation therapy. Int J Radiat Oncol Biol Phys 1989;16:1601-6. Martel MK, Sandler HM, Cornblath WT, et al. Dosevolume complication analysis for visual pathway structures of patients with advanced paranasal sinus tumors. Int J Radiat Oncol Biol Phys 1997;38:273-84. Hornsey S, Myers R, Warren P. Residual injury in the spinal cord after treatment with X-rays or neutrons. Br J Radiol 1982;55:516-9.

865

866

56


35. Wong CS, Hao Y. Long-term recovery kinetics of radiation damage in rat spinal cord. Int J Radiat Oncol Biol Phys 1997;37:171-9. 36. Nieder C, Grosu AL, Andratschke NH, Molls M. Proposal of human spinal cord reirradiation dose based on collection of data from 40 patients. Int J Radiat Oncol Biol Phys 2005;61:851-5.

37. Schoenthaler R, Albright NW, Wara WM, Phillips TL, Wilson CB, Larson DA. Reirradiation of pituitary adenoma. Int J Radiat Oncol Biol Phys 1992; 24(2):307-14. 38. Flickinger JC, Deutsch M, Lunsford LD. Repeat megavoltage irradiation of pituitary and suprasellar tumors. Int J Radiat Oncol Biol Phys 1989;17(1):171-5.

64 Radiosensitizers in Neurooncology D. Khuntia . A. Chakravarti . H. I. Robins . K. Palanichamy . M. P. Mehta

Introduction Both brain metastasis and malignant gliomas represent a significant problem in neurooncology with very poor outcomes with current therapies. Brain metastasis, which affects around 200,000 patients a year portends a poor prognosis with median survivals ranging between two and seven months [1,2]. Malignant gliomas are comprised of World Health Organization (WHO) Grade III and IV gliomas. Median survival times for patients with Grade IV tumors, also referred to as glioblastoma (GBM), is noted to be especially poor, remaining just over one year with current therapeutic regimens [3]. The treatment regimen for malignant glioma patients has traditionally involved maximal surgical debulking, followed by radiation +/ chemotherapy. Cooperative group trials over the past several decades have demonstrated that adjuvant radiation significantly prolongs survival compared to surgery alone in malignant glioma patients [4]. The role of cytotoxic chemotherapy with radiation for malignant gliomas and brain metastasis has been less welldefined until more recently. More contemporary cytotoxic chemotherapeutic agents such as temozolomide (TMZ) have demonstrated activity in malignant gliomas, both when used as single agents as well as in combination with radiotherapy. Further, targeted therapies have been developed which inhibit specific molecular pathways required for tumor-specific survival, proliferation, migration, and angiogenesis. The ensuing discussion will focus on a description of ‘‘classic’’ radiosensitizers as well as novel radiosensitizers currently being evaluated. Further, we will describe #


the possible biological mechanisms of treatment resistance in malignant gliomas and the roles of chemotherapeutic and biotherapeutic modifiers of radiation response in patients with malignant gliomas.

Traditional Radiation Sensitizers Radiosensitizers concurrently enhance the effect of ionizing radiation on tumor cells while sparing the effects of radiotherapy on normal, healthy cells. Traditional radiation sensitizers are generally classified into the following three categories: hypoxic cell sensitizers, hypoxic cytotoxins, and nonhypoxic cell sensitizers.

Hypoxic Sensitizers and Hypoxic Cytotoxins Hypoxic cell sensitizers are able to increase the radiosensitivity of tumor cells deficient of oxygen without adversely affecting normal cells. Their mechanism of action centers around the ability to induce the formation and stabilization of toxic DNA radicals, mimicking the effects of oxygen [5]. Tumor cells are hypoxic in relation to the surrounding normal tissue secondary to obstruction of blood flow, defective or inadequate angiogenesis, or because of outstripping of capillary blood supply. These include the drugs nitroimadazole, misonidazole, etanidazole, nimorazole, and efaproxaril. A small randomized trial by Urtasun and colleagues initially reported a survival advantage with the use of metronidazole; however,

988

64

Radiosensitizers in neurooncology

. Table 64-1 Study

Therapy

N

Median survival (wks)

Urtasan [6]

RT (3000 rad/9F/3 wk) RT and MNZ RT (5800 rads/30F/6 wk) RT (3900 rads/9F/3 wk) and MISO RT (3900 cGy/9F/3 wk) and MNZ RT (5656 cGy/5.5 wks) RT (4352 cGy/12 F/4 wks) RT and MISO RT (4950 cGy/15F) RT (4950 cGy) and MISO RT (6000 cGy/7 wks) and MISO RT (6650 cGy/31F) RT (6650 cGy/31F) and MISO RT (4500 cGy/20F) and placebo RT (4500 cGy/20F) and MISO

15 16 17 15 17 20 18 17 81 82 54 27 18 195 188

15 26 25a 24a 25a 36 31 39 46.1 44.5 33+ 43 60 36 33

Urtasan [7] (2nd study)

Bleehen [8] (Cambridge)

EORTC [9] and placebo RTOG 78–01 [10,11] Vienna Study Group [11] Medical Research Council [12] a

Number of patients still alive at time data was reported Source: Adapted from [13]

subsequent studies failed to show an advantage (> Table 64-1) [6,13]. Hyperbaric oxygen (HBO) also has been used as a hypoxic cell sensitizer. Initial clinical trials included the use of hyperbaric oxygen to potentially reduce the hypoxia present in radioresistant GBM cells [14]. There is some improvement in local control for certain tumors, such as cervix and head and neck cancers, but the results are not accepted as conclusive. A non-randomized pilot study evaluated survival in 38 glioma patients treated with HBO + RT [15]. Overall survival was not significantly different versus historical controls (median survival of 46 weeks). Given this disappointing finding along with the cumbersome and expensive nature of the treatment, further development of HBO as a sensitizer was largely abandoned. Efaproxiral (also known as RSR-13), is a synthetic allosteric modifier of hemoglobin that functions by noncovalently binding to the hemoglobin tetramer and decreasing the hemoglobinoxygen binding capacity, allowing more oxygen to be available to the tissues [16]. This hypoxic

sensitizers is novel in that the radiation-enhancing effect does not rely on the direct diffusion of the drug into the tumor cells. Initial results in both GBM and brain metastasis were promising [17–19]. A phase II study of RSR-13 with radiation therapy for GBM reported a median survival of 12.3 months with 1-year and 18-month survivals of 54 and 24%, respectively [19]. In a phase III study looking at WBRT with or without RSR13, patients with metastatic breast cancer were found to have improved survival [17]. As a result, a large phase III study including only patients with metastatic breast cancer to the brain was conducted randomizing patients to WBRT with or without RSR-13. Results have shown that the addition of RSR-13 improves median survival (4.47 vs. 9 months, p = 0.001), quality of life (p = 0.019), and quality adjusted survival (p = 0.001) [20]. As a result of this trial, a confirmatory open-label, phase III study, Enhancing Whole-brain Radiation Therapy in Patients With Breast Cancer and Hypoxic Brain Metastases (ENRICH) has recently been completed. This follow up study was negative and has called into the question the value of this drug.


Hypoxic cytotoxins act as radiosensitizers by selectively killing hypoxic cells, which are generally more resistant to radiation. Hypoxic cytotoxins fall into three classes: quinone antibiotics such as mitomycin C, nitroaromatic compounds, and benzotriazine di-N-oxides such as tirapazamine. As with many of the other radiation sensitizers, these bioreductive agents have not been widely used because they are inconvenient and have minimal, if any, clinical benefit [15]. Tirapazamine, arguably the most attractive agents of the hypoxic cytotoxins, is a bioreductive agent with enhanced toxicity of hypoxic tumor cells. The RTOG conducted a phase II trial of two dose levels of tirapazamine (RTOG 94-17) in patient with GBM’s and compared their survivals with RPA matched patients within the RTOG database [21]. Unfortunately, there was no improvement in survival with the use of tirapazamine.

Non-hypoxic Sensitizers Promising non-hypoxic sensitizers include the halogenated pyrimidines such as 5-iododeoxyuridine (IUdR) and 6-bromodeoxyuridine (BUdR). These agents function as DNA base analogues, becoming incorporated in newly synthesized DNA, and rendering tumors more sensitive to damage from ionizing radiation. Halogenated pyrimidines are preferentially incorporated into tumor cells, presumably based on the higher proliferation index of tumors cells compared to normal tissue. Bromodeoxyuridine (BUdR) is one of the earliest halogenated pyrimidines to be studied. Because of concerns about rapid catabolism of halopyrimidines after intravenous infusion, early clinical trials focused on inter-arterial infusion of BUDR. In a nonrandomized Japanese study of intra-arterial BUdR, survival outcomes were encouraging, but at the expense of significant toxicity from carotid artery catheterization [22,23]. Subsequently, it was determined that prolonged

64

intravenous infusion can achieve radiosensitization equivalent to intra-arterial administration [23], with adequate steady-state plasma concentration and acceptable toxicities [24,25]. Other studies of have evaluated the use of continuous infusions of BUdR or iododeoxyuridine (IUdR) concurrently with radiation. Initial studies included a combined series of four phase I studies at the National Cancer Institute reported a median survival of 13 months for GBM patients treated with infusional BUdR or IUDR [26]. Other phase I and II studies evaluating continuous infusion IUdR with hyperfractionated RT reported median survivals of 11–15 months [25,27,28]. In another series, the NCOG (Northern California Oncology Group) treated newly diagnosed GBM patients with continuous infusion BUdR during RT and observed a median survival of 56 weeks. This was followed by adjuvant chemotherapy with PCV (procarbazine, CCNU, and vincristine). A subgroup of patients receiving higher cumulative doses of BUdR, showed improvement in progression-free survival [29,30]. The results of the NCOG study led to a subsequent single-institution trial investigating significantly higher doses of BUdR with hyperfractionated RT. However, median survival was not improved at 50 weeks, and significant toxicities with the escalated dose of BUdR were observed [31]. The RTOG also conducted a phase III study looking at radiation plus procarbazine, lomustine, and vincristine with or without BUdR in the management of anaplastic astrocytomas [32]. In this 286 patient randomized trial, BUdR was shown to have no survival advantage in this patient population, with a 4-year survival rate of 51% in both arms.

Contemporary Radiosensitizers One of the more promising radiation sensitizers under active clinical investigation is motexafin gadolinium (MGd). MGd is in a class of

989

990

64


drugs referred to as redox modulators. It is a metallotexaphyrin that catalyzes the oxidation of intracellular-reducing metabolites and generates reactive oxygen species that selectively concentrate in tumor cells to promote apoptosis. Because it contains the paramagnetic compound gadolinium, this drug allows excellent visualization of the tumor on MRI [33,34]. Original studies with the drug were performed in patients with brain metastasis and have found the drug safe and well tolerated [35–37]. Further, the drug has the ability to pick up additional metastasis as MGd accumulates over time after all of the dose is delivered (11.4% increase after ten doses). A subsequent phase III study (SMART trial) compared WBRT with or without MGd. The primary endpoints of the study were survival and time to neurologic progression. Secondary endpoints included time to loss of functional independence, radiologic response rate, time to radiologic progression, tome to progression of brain specific quality of life (using the Functional Assessment of Cancer Therapy-Brain [FACT-Br]). Unfortunately, the study did not show a benefit for the primary endpoints of the study, however there was an improvement for time to neurologic progression and neurocognitive function for patients with nonsmall cell lung cancer [38]. This was a pioneering study in that it was able to do very thorough neurocognitive testing on a large scale. It was found that over 90% of patients had impairment of at least one neurocognitive domain at presentation. This was highly correlated with baseline volume of disease [39]. As a result of the SMART trial, a follow-up international phase III study was conducted randomizing 554 patients with brain metastasis from NSCLC to WBRT with or without MGd [40]. The primary endpoint of this study was time to neurologic progression (TNP). TNP improved with MGd from 10 to 15.4 months, however this was not statistically significant (p = 0.12). Fewer patients in the MGd arm required salvage

brain surgery or radiosurgery, as well. It has been shown that the reason there was not a statistically significant benefit to neurologic progression was secondary to the fact that patients treated outside of North America, did not start WBRT soon after diagnosis (2.2 weeks vs. 6.5 weeks for Europe and Australia). When the entire cohort was reanalyzed with patients receiving WBRT within 3 weeks of radiotherapy, TNP was significantly prolonged with the use of MGd (p = 0.006). MGd has also been studied in malignant gliomas. Initial studies have shown the drug to be safe and tolerable with the maximum tolerated dose of 22 doses of 5 mg/kg [41]. Subsequently, a phase II study evaluated the safety and efficacy of MGd with 60 Gy radiation in 25 patients [42]. With 6-month follow-up, the Kaplan-Meier estimate of survival was 80% and median survival has not yet been reached. With these promising results, other studies are underway within the RTOG evaluating MGd with temozolomide in patients with GBM. Currently, the drug is being investigated in head and neck cancer, pediatric pontine gliomas, pancreatic carcinoma, and lung cancer [33].

Chemotherapeutic Agents as Radiosensitizers The use of systemic chemotherapy in conjunction with ionizing irradiation is now evolving as a therapeutic modality for both brain metastases and primary brain neoplasms. Common mechanisms for radiosensitization include: inhibition of DNA repair, (e.g., by pyrimidine or purine analogs); perturbation of cell cycling, (e.g., by paclitaxel) to optimize the fraction of G2/M-phase cells; specific effects on hypoxic cells, (e.g., mitomycin). There are multiple complicating factors which require investigative explication for such a multi-modality approach to be successful. Most obvious is the therapeutic index, because there is the potential for normal central nervous system (CNS) injury,


and/or radiation induced necrosis secondary to drug induced radiosensitization. Systemic chemotherapeutic toxicities present another area for potential morbidity. Additionally, the blood-brain barrier (BBB) may represent a significant obstacle to drug delivery. Two recent reports concerning the BBB highlight the significance of drug exclusion from the central nervous system by glycoprotein efflux pumps [43,44] beyond the traditional consideration of lipid solubility. It is of interest to note that such efflux pump drug extrusion may be less efficient in brain metastases than primary gliomas [43]. This point is dramatically illustrated by Gerstner and Fine in the case of paclitaxel [43]. It is notable that the BBB may fluctuate during therapy. Thus, a drug which may initially penetrate a neoplastic focus that has disrupted the BBB, may become less permeable as successful anti-neoplastic therapy allows reconstitution of the BBB [45]. This is illustrated in a report of the use of chemotherapy in newly diagnosed small cell lung cancer (SCLC) in which the relative response rate for CNS and systemic disease was 27 and 73%, respectively [46]. It is predictable to have similar results among other chemotherapy responsive tumors, (e.g., germ cell tumors, lymphoma, medulloblastoma/primitive neuro-ectodermal tumor) with chemotherapy that does not readily cross the BBB, i.e., an initial response, but diminished drug penetration after BBB reconstitution. This concept is particularly relevant to any clinical scenario involving radiotherapy, as a radiation therapeutic effect is highly likely for most neoplasms. Taking the aforementioned considerations collectively, the ideal chemotherapeutic radiosensitizer should have good penetration across the BBB, a demonstrated preclinical radiosensitizer effect, inherent tumor activity, and at least a presumption of therapeutic index. Other caveats relate to the fact that BBB penetration is particularly relevant in any clinical scenario addressing CNS prophylaxis. In spite of the difficulty of accessing long term neurological toxicity,

64

it remains a critical clinical endpoint for future studies. The use of various radiosensitizers, and in particular chemo-therapeutic agents, require long term follow up in any patient population expected to have a survival benefit measured in years. The risk of a potential radiation induced dementia [47] aggravated by a drug effect is a significant concern. A classic example of a deleterious drug/radiation interaction is the experience with high dose methotrexate [48,49]. An excellent example of the type of prospective neurocognitive studies that are possible has been reported by Meyers et al. [50]. Emerging laboratory data suggest that one possible mechanism of relative resistance of brain metastases to chemotherapy might be the florid astrocytic response induced by tumor cells in the brain, and laboratory data suggest that tumor cell-astrocyte cell-to-cell communication dramatically increases chemoresistance.

Specific Agents Temozolomide (TMZ) TMZ is an oral alkylating agent. It is highly bioavailable, crosses the BBB, and achieves significant levels in the cerebrospinal fluid [51]. It dissociates to form the active alkylating agent methyltriazenoimidazole-arboxamide (MTIC) at physiologic pH, which methylates the O-6 position of guanylic acid in DNA. TMZ is taken orally and is absorbed rapidly and completely after oral administration. Compared to other alkylating agents, its doselimiting toxicity, i.e., myelosuppression, is modest [52]. The first speculation that TMZ might have a radiosensitization effect related to the results of a positive phase II study [53] in newly diagnosed glioblastoma multiforme (GBM) patients performed by Stupp and colleagues. This was followed by a confirmatory phase III study [3] (performed by the European Organization for Research and Treatment of Cancer/National Cancer Institute of

991

992

64


Canada) showing a survival advantage for administering daily TMZ (75 mg/m2) during the course of radiotherapy, followed by maintenance TMZ [on a 5/28 day schedule (150–200 mg/m2)] starting one month post radiation. This approach has become standard of care for this patient population. Interestingly, a contemporary phase II study in which TMZ was only delivered with radiation (and not post-radiotherapy) at a somewhat lower dose of 50 mg/m2 per day of radiation (vs. 75 mg/m2 used in the Stupp regimen) had comparable results [54]. Parenthetically, preradiation TMZ tested in the context of another phase II study did not appear efficacious comparing results across studies [55]. The results of these clinical trials taken collectively (as well as results in the setting of brain metastases discussed below) resulted in hypothesis that TMZ was in part acting as a radiation sensitizer. The concept of TMZ radio-sensitization at the time of the aforementioned clinical trials was only weakly supported by a preclinical study which essentially observed an additive effect between these medications [56]. A convincing laboratory demonstration of radiosensitization, however, has recently reported by Chakravarti et al. [57]. Interestingly, these investigators found TMZ improves radiation response most effectively in MGMTnegative GBM cell lines by increasing radiationinduced double-stranded DNA damage. These results are also consistent with clinical results showing patients with a silent MGMT gene have a better clinical outcome [58]. The antitumor activity of TMZ has been attributed primarily to the methylation of DNA, which is highly dependent upon the formation of a reactive methydiazonium cation [59]. Nearly 70% of total DNA methylation by TMZ occurs at the N7-guanine, 9 and 5% of adducts are formed at the N3-adenine and O6-guanine, respectively. The cytotoxicity of TMZ is influenced by three DNA repair activities in particular. The first is O6-alkylguanine-DNA alkyltransferase (AGT). There is accumulating evidence that

the cytotoxicity of TMZ is highly dependent on the formation of O6-methylguanine, despite the fact that this lesion accounts for only a small percentage of the total DNA adducts formed. Adducts produced at the O6-position of guanine have been found to be especially mutagenic and cytotoxic. Methyl adducts at the O6-guanine in DNA are repaired by the cytoprotective DNA repair protein, MGMT, which transfers the methyl group to an internal cysteine acceptor residue. This reaction results in an irreversible inactivation of MGMT, requiring increased de novo protein synthesis to restore repair activity. Depletion of MGMT via pretreatment with substrate analogs such as O6-benzylguanine (O6-BG) has been investigated. It has been demonstrated in preclinical models that O6-BG can increase the cytotoxicity of TMZ by several fold. It has been found that continuous administration of O6-BG is more effective than intermittent dosing, indicating that regeneration of MGMT activity following MGMT inhibitor may be of clinical significance. The second mechanism of resistance to TMZ involves DNA mismatch repair pathways. One mechanism involves binding by a heterodimer complex consisting of hMSH2 and GTBP/p160 proteins and subsequent DNA incision following recruitment of an additional heterodimer consisting of hPMS2 and hMLH1 proteins. A section of DNA is then removed between the incision and the mismatch, and replaced by resynthesis and ligation. When this pathway is targeted to the strand directly opposite O6-MG, its unsuccessful attempt to find a complementary base results in continued excision/insertion which produces persistent damage to the DNA. The resulting interruptions in the daughter strands prevent replication in the subsequent S-phase and may account for two cell divisions being required before the emergence of TMZ toxicity. Since the cytotoxicity of TMZ is dependent upon a functional DNA mismatch repair pathway, resistance may be conferred by a mutation in any of the genes encoding for a


protein involved in mismatch recognition/incision (e.g., germline mutations in hereditary non-polyposis colorectal cancer). Such abnormalities result in a TMZ ‘‘tolerant’’ phenotype which is unaffected by MGMT activity. The third mechanism of TMZ resistance involves base excision repair and poly(ADP-ribose) polymerase. Methyl adducts produced at N7-guanine and N3-adenine by TMZ may also hinder DNA replication, as enzymatic or spontaneous depurination will ultimately result in DNA strand breakage. Preclinical data also suggests that TMZ has at least additive activity with radiation in human glioblastoma cells [60]. It was determined that TMZ and radiation had at least an additive effect. In U373MG GBM cells, it was determined that the addition of 10 mM of TMZ to 1–2 Gy of radiation increased cell kill by 2.5–3.0-fold. However, in a cell line with 100-fold greater MGMT activity, there was actually an antagonistic effect observed. This antagonistic effect was mitigated by co-incubation with O6-BG, revealing a strategy to enhance the additive effect of TMZ and radiation. The combination of TMZ/radiation has also been studied in the context of brain metastases. In a small randomized phase II study Antonadou and associates treated patients with TMZ (75 mg/m2/day) concurrent with 40 Gy of fractionated conventional radiotherapy (2 Gy, 5day/week) for 4 weeks versus radiotherapy alone [61]. Results demonstrated a statistically significant increase in response with TMZ plus whole brain radiotherapy (WBRT) (96%) versus WBRT alone (67%) with a trend toward improved survival. A confirmatory phase III study has also been performed [62] in a series of 108 randomized patients (82% with NSCLC) the difference in response rates for brain metastasis was statistically significant favoring WBRT plus TMZ (response rate, 53%) versus WBRT alone (response rate, 33%), with a trend toward increased survival (8.3 vs. 6.3 months). It is noteworthy that several studies (reviewed by Chang et al. [63]) have

64

demonstrated modest response to TMZ for brain metastases following post-radiotherapy relapse. In part, based on data regarding TMZ summarized above, the Radiation Therapy Oncology Group (RTOG) is now conducting a 3 arm phase III clinical trial in non-small cell lung cancer (NSCLC) patients with one to three brain metastases. This clinical trial compares WBRT and stereotactic-radiosurgery alone versus with radiosurgery/WBRT with TMZ or erlotinib. Details regarding this can be obtained online at the web-site: clinicaltrials.gov.

Camptothecins Topotecan is a topoisomerase 1 inhibitor with demonstrated activity in both NSCLC [64] and SCLC [65]. It has known radiosensitization effects [66,67] and high brain capillary permeability [68,69]. Topotecan has a response rate of 33–63%, as mono-therapy in patients with brain metastases from SCLC [70]. Several phase I/II clinical trials of topotecan and radiation therapy for brain metastases have been performed [71–73] with response rates ranging from 46 to 72%. Toxicities were variable depending on topotecan dosing with grade 3/4 hematological toxicity being predominant. To date a definitive randomized study has not been reported with this agent as an adjunct to radiation in the setting of CNS metastases. It is of interest to note that topotecan has been reported to have some activity in recurrent gliomas [74,75]. It was tested as a radiosensitizer in newly diagnosed GBM by the RTOG. The median survival was 9.3 months, not significantly different from historical controls evaluated with recursive partitioning analysis using the RTOG database [76]. Another camptothecan, irinotecan, may prove more effective for high grade glioma [77]. Recent data regarding the use of irinotecan with bevacizumab has been promising in recurrent

993

994

64


disease [78]. (It should be noted that irinotecan is administered intravenously (IV); topotecan has the advantage of also having an oral formulation, which may be advantageous for daily dosing as a radiosensitizer).

Taxanes Preclinical studies have shown activity for paclitaxel as a radiosensitizing agent in malignant cell lines [79,80]. The RTOG performed a phase II study evaluating the efficacy and feasibility of conventional radiotherapy and concurrent weekly paclitaxel in newly-diagnosed GBM [81]. The results, (i.e., a median survival of 9.7 months) did not represent an improvement when compared to the historical RTOG database. However, the concurrent use of anticonvulsant therapy may have contributed to increased paclitaxel metabolism compromising the results of this study [82]. In a later study in recurrent GBM patients in which there was adjustment for the use of anticonvulsant therapy, paclitaxel was not found to have significant activity [83]. Paclitaxel has also been studied as a radiosensitizing agent in the setting of fractionated stereotactic radiotherapy for recurrent GBM [84]. The results of this study suggested tumor volume was a significant prognosticator. Overall median survival was 7 months; 1- and 2-year actuarial survival rates were 17 and 3.4%, respectively [84].

in the setting of CNS disease. They are, however, excellent candidates for future study.

Platinum Agents Preclinical studies have demonstrated that platinum agents can inhibit repair of radiation induced damage, and exert direct cytotoxic effects on a variety of neoplasms including glioma cell lines [89,90]. These observations were applied to a study of fractionated stereotactic radiotherapy with cis-platinum in patients with recurrent high grade glioma with the possible suggestion of benefit [91]. A phase III intergroup trial of standard radiotherapy with and without continuous infusion cis-platinum in newly diagnosed GBM patients, however, found no significant difference in survival between groups [92]. A randomized phase III study of carboplatin and WBRT versus WBRT was initiated in patients with NSCLC brain metastases [93]. This study closed prematurely do to poor patient accrual (n = 44). The observed median survival in the WBRT alone arm was 4.4 months, and 3.7 months in the combined arm (p = 0.64). There was no difference in response rate between arms.

Molecularly Targeted Agents as Sensitizers

Pyrimidines

Signal Transduction Pathways Involved in Treatment Resistance

It is of passing interest to note the fluropyrimidines including capecitabine, and 5-flurouracil (its parent drug) both have good penetration into the CNS [63]. They are often used for radiosensitization in a variety of tumor types for which there is a firm preclinical rationale [85–88]. We are not aware of definitive controlled clinical studies using either these drugs

It is known that certain genetic events underlie disregulation of key signaling pathways in gliomagenesis. In secondary GBM’s, which arise from lower-grade gliomas, initial events involve p53 loss or mutation combined with Rb and PTEN mutations as tumors progress in grade. Primary GBM’s do not have clear track-records of arising from lower-grade tumors, and have


certain molecular features that distinguish them from secondary GBM. For instance, amplification of the epidermal growth factor receptor (EGFR) and overexpression of the EGFR protein are relatively common events in primary GBM’s, occurring in upwards of 50% of cases. Often, along with enhanced signaling through EGFR, there is concomitant loss of the INK4 gene, which codes for both p16 and p14ARF in primary GBM’s. Many primary GBM’s express the constitutively active EGFRvIII mutant that lacks the extracellular binding domain. The prognostic value of EGFR in GBM has been controversial. Wild-type EGFR expression, as determined by immunohistochemical analysis does not appear to be of prognostic value in GBM’s. There are reports that the constitutively active EGFRvIII mutant has prognostic value in selected series; however, has yet to be rigorously confirmed in a prospective manner. Loss of heterozygosity of chromosome 10 is a common event in GBM’s, occurring in upwards of 90% of cases. PTEN (phosphatase and tensin homology gene), which is a 30 phosphoinositol phosphatase is located on Chromosome 10q23.3 and is commonly lost in both primary and secondary GBM’s. Loss of PTEN results in constitutive activation of downstream mediators of the phosphatidyl 3-inositol kinase (PI3K) pathway, including AKT, which is a potent pro-survival molecule. > Figure 64-1 illustrates this pathway, as well as upstream regulators (e.g., EGFR). PI3K is a lipid kinase that promotes diverse biological functions including cellular proliferation, survival and motility [94]. The PI3K signaling pathway is frequently disregulated in glioblastoma, [95,96] often in combination with the ERK pathway, and mouse genetic studies suggest a causal role of this pattern [97]. Upwards of 40% of GBM’s contain alterations of the PTEN tumor suppressor gene, a negative regulator of PI3K signaling, which results in constitutive activation of the PI3K pathway [96]. Upstream of PI3K, the epidermal growth factor receptor (EGFR)

64

is commonly over-expressed, which may lead to disregulated PI3K and RAS/ERK signaling [98–104]. Other receptor tyrosine kinases such as PDGFR and c-MET are also commonly overexpressed in glioblastomas, and may deregulate these same pathways [98–105]. The PI3K and RAS/ERK pathways connect richly to other signaling cascades, thereby integrating signals associated with other cell surface events, stress activation pathways and extracellular matrix proteins. RAC1 is one such protein that links PI3K and RAS signaling with integrin-linked signaling, potentially playing a key role in promoting glioblastoma growth and survival [106]. Therefore, the PI3K and ERK pathways, provide important therapeutic targets. > Figure 64-1 illustrates various points within this pathway that can be targeted. The connectivity of the PI3K signaling pathway in GBM has recently been demonstrated by Mischel et al. It was demonstrated that PTEN loss was tightly linked to AKT activation. Further, it was determined that in a subset of glioblastoma patients treated by radiation alone, activation of PI3K pathway members was associated with adverse clinical outcome, providing direct clinical evidence of the role of PI3K signaling in radiation resistance in GBM’s.

Targeting EGFR Pathway Signaling in GBM’s – Clinical Data Phase I/II studies on the safety and efficacy of anti-EGFR agents in the setting of GBM suggest modest activity in the recurrent setting [107,108]. Interestingly, EGFR status, either measured by levels of IHC-detected EGFRwt or EGFRvIII or EGFR gene amplification, was not associated with outcome as measured by overall survival. In a separate Phase II study of another EGFR tyrosine kinase inhibitor, erlotinib (OSI-774) for recurrent high-grade glioma, [109] there were 6/25partial responses. Therefore, the emerging clinical data suggests that

995

996

64


. Figure 64-1 Angiogenesis, signaling pathway and inhibitors in clinical trails. Binding of growth factors to its receptors leads to an activation of receptor tyrosine kinase activity and binding of adaptor and Ras activating proteins and other nucleotide exchange factor. Mature Ras-GDP will be converted to Ras-GTP, thereby activating Ras and active form (GTP) can be converted to inactive form (GTP) by guanosine triphosphate activating protein (GAP). Ras-GTP stimulates the downstream effector molecules like Raf, Rac, MEK, PI3K, etc., which promotes cell proliferation and survival. The Inhibitor used in the phaseI-III trials as well as in the clinical setting to inhibit the key signaling molecule which mediates treatment resistance is also shown. Vascular endothelial growth factor (VEGF) functions by activating two receptor tyrosine kinases, Flt-1 (VEGFR-1) and KDR (VEGFR-2), both of which are selectively expressed on the primary vascular endothelium. KDR is responsible for VPF/VEGF-stimulated endothelial cell (EC) proliferation and migration. KDR mediates cell survival through downstream targets such as PLG and CSK leading to proliferation, migration, vasculature and survival

there may be a subset of patients with malignant gliomas who are responsive to anti-EGFR therapies. As response to anti-EGFR therapies when used as single agents in the recurrent setting appears to be independent of EGFR status, the role of downstream signaling pathways must be more carefully scrutinized. In one report, activation of phosphatidylinositol 3-kinase members (PI3K, AKT, p70s6k) was associated with significantly worse survival times in glioma patients as a whole [110]. More specifically, in GBM patients treated with radiation as the primary adjuvant

therapy without chemotherapy, activation of these critical PI3K pathway members was associated with adverse survival, indicating a direct role for this pathway in treatment resistance. As more evidence that activation of downstream pathways may mediate resistance to anti-EGFR agents, in a study of two primary GBM cell lines with equivalent expression of EGFR, the two cell lines were found to have very different sensitivities to the anti-EGFR agent AG1478 [111]. Further investigation revealed that upregulation of the insulin-like growth factor receptor 1 (IGFR1) contributed to resistance to AG1478. Upon dual


inhibition of IGFR1 and EGFR, it was possible to more significantly downregulate activity of PI3K/ AKT and reduce cellular survival as well as sensitivity to radiotherapy. In a correlative study on patients with recurrent GBM’s treated by antiEGFR agents, it was determined that patients most responsive to anti-EGFR therapies are those with expression of EGFRvIII and PTEN. Tumors that were PTEN-deficient were found to be resistant to anti-EGFR agents as a rule. It can be conceptualized that EGFRvIII drives signaling through PI3K in tumors with intact PTEN and would therefore be responsive to anti-EGFR therapies. In contrast, tumors with PTEN deficiency have constitutive signaling through PI3K/AKT. Since neither EGFR nor EGFRvIII is driving PI3K pathway activation in PTEN-deficient tumors, it stands to reason that anti-EGFR agents would have limited efficacy in suppressing pro-survival signaling through PI3K/AKT. The RTOG has investigated the safety and efficacy of Gefitinib, an EGFR tyrosine kinase inhibitor, in combination with radiation for newly diagnosed GBM patients (RTOG 0211) [112]. However, the preliminary results from the phase II portion of the study revealed similar survivals with the use of Gefitinib when compared to historical controls [113].

mTor Pathway Inhibition – CCI-779 Preclinical data suggests that the mammalian target of rapamycin (mTor) pathway, which is downstream of EGFR and PI3K/AKT, represents an attractive therapeutic target. CCI-779 is a small molecule inhibitor (Rapamycin analog) of mTor. In a recently reported NCCTG Phase II trial of CCI-779 in recurrent GBM, 41 patients with recurrent GBM were treated by CCI-779 at a dose of 250 mg i.v. qweek [114]. Assessment of tumor Akt and p70s6k kinase phosphorylation pretreatment demonstrated activation in the majority of

64

patients (14/17 and 11/17 patients, respectively). Post-treatment, an analysis of peripheral blood mononuclear cell phosphorylation of p70s6k showed post-treatment inhibition in 7/10 patients. There were signs of some activity, with a significant decrease in T2 abnormality in five patients and a significant decrease in T1 gadolinium enhancement in three patients, which fell short of partial response. Investigation of mTor inhibitors in combination with radiation is presently ongoing.

Angiogenesis Pathways There is ample evidence that angiogenesis plays an important role in the pathogenesis and treatment resistance in gliomas. One of the key underlying factors of angiogenesis may be the hypoxic environment that results when tumor growth outstrips blood and, hence, nutrient supply. A key transcription factor involved in angiogenesis in gliomas is hypoxia-induced factor 1 (HIF-1). HIF-1 is a heterodimer protein consisting of HIF-1a and a constitutively expressed aryl hydrocarbon receptor nuclear translocator (ARNT, also known as HIF1b). In pathologic specimens taken from GBM patients, HIF-1 expression has been commonly detected in the leading edge of invading tumor cells, as well as in the necrotic core of tumor [115,116]. In hypoxic conditions, HIF-1 binds to hypoxia response elements (HRE’s), thereby inducing expression of hypoxia-responsive gene that are involved in angiogenesis, invasion, and survival. Under normoxic conditions, HIF-1a is rapidly degraded by the proteosome. It is known that HIF-1a interacts with the Von Hippel Landau (VHL) protein, which helps to target HIF-1a for proteosomal degradation. It is also important to realize that hypoxia-independent factors may serve to increase HIF-1a expression, including loss of PTEN [117]. The vascular endothelial growth factor (VEGF) has also been found to play an important

997

998

64


role in angiogenesis in malignant gliomas. VEGF is a highly specific endothelial mitogen. In addition to VEGF (VEGF-A), there are other important family members of this pathway including VEGF-B, VEGF-C, VEGF-D, VEGF-E, and placenta growth factor (PGF). There appears to be an accumulation of VEGF that increases with tumor grade and tumor size in gliomas [118]. Extracellular VEGF is known to bind to its tyrosine kinase receptors, VEGFR1 and VEGFR2, which are strongly expressed on endothelial cells present in the vasculature surrounding high-grade gliomas, but absent in normal brain vasculature. Hypoxia serves to increase VEGF expression levels via the HIF family. It has been found that VEGF-A and VEGF-B mRNA are commonly overexpressed in GBMs [118]. One report suggests that expression of VEGF family members is tightly linked to EGFR expression levels [119]. Further, it has been found that PI3K pathway activation serves to increase VEGF mRNA expression levels independent of hypoxia. Given the potential importance of angiogenic pathways in mediating radiation resistance, these molecules present themselves as being attractive targets in GBM’s.

PTK 787 (Vatalanib): Clinical Data Given the promise of anti-angiogenic strategies in human tumors, studies have been conducted in the setting of recurrent GBM. PTK 787/ZK 222584 is a drug that has been found to inhibit all known VEGF receptors and therefore inhibits signaling by VEGF’s (A-D). In a study of 55 patients with recurrent GBM treated by PTK787, [120] there were 2 (4%) partial responses, 31 (56%) stable disease, and 14 (25%) disease progression. Median duration of stable disease was 12.1 weeks. Dynamic contrast-enhancedenhanced (DCE) and dynamic susceptibility change (DSC) MRI revealed decreases in vascular permeability and cerebral blood volume. In a separate study, PTK-787 was combined with

either TMZ or CCNU [121]. Among 51 patients evaluable for response, 4 had a partial response and 27 patients had stable disease. The median time to progression was 15.7 weeks for the PTK787+TMZ arm and 10.4 weeks for the PTK-787+ CCNU arm. Phase I/II studies are currently planned investigating PTK-787 + TMZ in combination with radiotherapy for newly-diagnosed GBM patients. In an ongoing trial, vatalanib is evaluated in combination with radiotherapy and TMZ chemotherapy [122]. Preclinical observations of potential increase in hypoxia and radiation resistance raise concern as to the optimal timing of anti-angiogenic therapy [123]. The EORTC randomized phase II trial (EORTC 26041-22041) therefore evaluates safety and efficacy of the addition of vatalanib (PTK787) either concurrent with TMZ/RT, or by adding the VEGFR inhibitor only after completion of concomitant chemoradiotherapy [122].

Bevacizumab and AZD2111 Increased anti-tumor activity with the addition of bevacizumab to chemotherapy has been demonstrated for several types of malignancies including colorectal cancer, renal cell carcinoma, breast cancer, and non-small cell lung cancer. A recent study in recurrent malignant glioma reports a relatively high radiographic response rate of 60% with the combination of bevacizumab and irinotecan [78,124]. However, it remains unclear whether this increased response rate translates into a true prolongation of survival. Therefore, a definitive phase III trial is about to be initiated by the RTOG.A clinical trial with correlative imaging and biological endpoints suggests that VEGF receptor inhibition by AZD2171 leads to normalization of tumor vasculature and restoration of the blood-brain barrier, thus reducing contrast enhancement and edema [125]. These findings are consistent with prior observations in rectal cancer, that anti-VEGF therapy leads to


normalization of the vasculature, decreased tumor interstitial pressure, better oxygenation and drug delivery [126]. A trial with the combined VEGFR and EGFR tyrosine kinase inhibitor vandetinib (ZD6474) is about to start accrual.

Cilengitide Integrins are heterodimer transmembrane receptors for the extracellular matrix, regulating cell adhesion and migration. In vessels, integrins interact with the basal membrane, thereby maintaining vascular quiescence; during angiogenesis they are essential for endothelial cell migration, proliferation and survival [127,128]. In preclinical models inhibition of integrin function efficiently suppresses angiogenesis and inhibits tumor progression, the integrins aVb3 and

64

aVb5 were identified as specific for tumor angiogenesis. Cilengitide, a synthetic RGD-motif peptide binds to the aVb3 and aVb5 integrin receptors. In a phase I study in recurrent glioma single agent activity of cilengitide was observed [129]. A phase II trial of cilengitide added to standard TMZ-based chemoradiation has recently been completed. Therapy was associated with little or no additional toxicity, initial results suggest efficacy in a subgroup of patients [128].

Summary Malignant gliomas and brain metastasis remain among the most treatment-refractory tumors. Traditional upfront treatment regimens have incorporated nitrosurea-based chemotherapy. This strategy has evolved to include temozolomide-based approaches. Promising Phase I/II

. Figure 64-2 Future treatment strategy. Over one-quarter of the patients enrolled on the TMZ+RT arm survived beyond 2-years, there appears to be a finite percentage of patients who derive long-term benefit from this treatment regimen. To this end, targeted therapies have emerged as an attractive option. Accumulating evidence suggests that certain molecular pathways are selectively upregulated in tumor vs. normal cells. Some of these pathways have been shown to be instrumental in proliferation, migration, invasion, angiogenesis, and/or survival in preclinical models. These would appear to represent ideal therapeutic targets, as their antagonism may lead to an improvement in the therapeutic ratio of radiation. Emerging data from clinical studies on ‘‘first generation’’ targeted therapies appear to demonstrate benefit for select patients. Further molecular/genetic profiling must be undertaken to identify exactly which patients benefit

999

1000

64


data with TMZ in the recurrent setting prompted a Phase III EORTC study of TMZ in combination with RT for patients with newly-diagnosed GBM. The landmark EORTC 26981-22981/NCIC CE3 study demonstrated a significant improvement in not only median survival, but also in terms of 2-year survival. Given that over one-quarter of the patients enrolled on the TMZ+RT arm survived beyond 2 years, there appears to be a finite percentage of patients who derive long-term benefit from this treatment regimen. Given that the EORTC-based regimen represents an incremental improvement in the standard of care, rather than a truly curative solution for most patients, further efforts must be expended to identify novel therapeutic approaches. To this end, targeted therapies have emerged as an attractive option. > Figure 64-2 illustrates some of the promising targeted therapy strategies in gliomas. Accumulating evidence suggests that certain molecular pathways are selectively upregulated in tumor vs. normal cells. Some of these pathways have been shown to be instrumental in proliferation, migration, invasion, angiogenesis, and/or survival in preclinical models. These would appear to represent ideal therapeutic targets, as their antagonism may lead to an improvement in the therapeutic ratio of radiation. Emerging data from clinical studies on ‘‘first generation’’ targeted therapies appear to demonstrate benefit for select patients. Further molecular/genetic profiling must be undertaken to identify exactly which patients benefit.

References 1. Khuntia D, Brown P, Li J, Mehta MP. Whole-brain radiotherapy in the management of brain metastasis. J Clin Oncol 2006;24(8):1295-304. 2. Gaspar L, Scott C, Rotman M, et al. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997;37(4):745-51. 3. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352(10):987-96.

4. Walker MD. The contemporary role of chemotherapy in the treatment of malignant brain tumor. Clin Neurosurg 1978;25:388-96. 5. Rowinsky EK. Novel radiation sensitizers targeting tissue hypoxia. Oncology (Williston Park) 1999;13 10 Suppl 5:61-70. 6. Urtasun R, Band P, Chapman JD, Feldstein ML, Mielke B, Fryer C. Radiation and high-dose metronidazole in supratentorial glioblastomas. N Engl J Med 1976; 294(25):1364-7. 7. Chang CH, Housepian EM. Tumors of the central nervous system: modern radiotherapy in multidisciplinary management. New York: Mason Publishing; 1982. 8. Bleehen NM, Wiltshire CR, Plowman PN, et al. A randomized study of misonidazole and radiotherapy for grade 3 and 4 cerebral astrocytoma. Br J Cancer 1981; 43(4):436-42. 9. Misonidazole in radiotherapy of supratentorial malignant brain gliomas in adult patients: a randomized double-blind study. Eur J Cancer Clin Oncol 1983; 19(1):39-42. 10. Carabell SC, Bruno LA, Weinstein AS, et al. Misonidazole and radiotherapy to treat malignant glioma: a phase II trial of the radiation therapy oncology group. Int J Radiat Oncol Biol Phys 1981;7(1):71-7. 11. Stadler B, Karcher KH, Kogelnik HD, Szepesi T. Misonidazole and irradiation in the treatment of high-grade astrocytomas: further report of the Vienna Study Group. Int J Radiat Oncol Biol Phys 1984;10(9): 1713-7. 12. A study of the effect of misonidazole in conjunction with radiotherapy for the treatment of grades 3 and 4 astrocytomas. A report from the MRC Working Party on misonidazole in gliomas. Br J Radiol 1983;56(669): 673-82. 13. Chang JE, Khuntia D, Robins HI, Mehta MP. Radiotherapy and radiosensitizers in the treatment of glioblastoma multiforme. Clin Adv Hematol Oncol 2007;5(11): 894-902, 7-15. 14. Nelson DF, Urtasun RC, Saunders WM, Gutin PH, Sheline GE. Recent and current investigations of radiation therapy of malignant gliomas. Semin Oncol 1986;13(1):46-55. 15. Chang CH. Hyperbaric oxygen and radiation therapy in the management of glioblastoma. Natl Cancer Inst Monogr 1977;46:163-9. 16. Suh JH. A phase 3, randomized, open-label, comparative study of standard whole brain radiation therapy (WBRT) with supplemental oxygen (O2), with or without RSR13, in patients with brain metastases. Neuro Oncol;2003:345-6. 17. Suh J, Stea B, Nabid A. Standard whole brain radiation therapy (WBRT) with supplemental oxygen (O2), with or without RSR-13 (efaproxiral) in patients with brain metastasis: results of the randomzied REACH (RT-009) study. Proc Am Soc Clin Oncol;22 Suppl 14:115s.


18. Shaw E, Scott C, Suh J, et al. RSR13 plus cranial radiation therapy in patients with brain metastases: comparison with the Radiation Therapy Oncology Group Recursive Partitioning Analysis Brain Metastases Database. J Clin Oncol 2003;21(12):2364-71. 19. Kleinberg L, Grossman SA, Carson K, et al. Survival of patients with newly diagnosed glioblastoma multiforme treated with RSR13 and radiotherapy: results of a phase II new approaches to brain tumor therapy CNS consortium safety and efficacy study. J Clin Oncol 2002;20(14): 3149-55. 20. Scott C, Suh J, Stea B, Nabid A, Hackman J. Improved survival, quality of life, and quality-adjusted survival in breast cancer patients treated with efaproxiral (Efaproxyn) plus whole-brain radiation therapy for brain metastases. Am J Clin Oncol 2007;30(6):580-7. 21. Del Rowe J, Scott C, Werner-Wasik M, et al. Single-arm, open-label phase II study of intravenously administered tirapazamine and radiation therapy for glioblastoma multiforme. J Clin Oncol 2000;18(6):1254-9. 22. Sano K, Hoshino T, Nagai M. Radiosensitization of brain tumor cells with a thymidine analogue (bromouridine). J Neurosurg 1968;28(6):530-8. 23. Goffinet DR, Brown JM. Comparison of intravenous and intra-arterial pyrimidine infusion as a means of radiosensitizing tumors in vivo. Radiology 1977;124(3): 819-22. 24. Kinsella TJ, Mitchell JB, Russo A, et al. Continuous intravenous infusions of bromodeoxyuridine as a clinical radiosensitizer. J Clin Oncol 1984;2(10):1144-50. 25. Kinsella TJ, Collins J, Rowland J, et al. Pharmacology and phase I/II study of continuous intravenous infusions of iododeoxyuridine and hyperfractionated radiotherapy in patients with glioblastoma multiforme. J Clin Oncol 1988;6(5):871-9. 26. Jackson D, Kinsella T, Rowland J, et al. Halogenated pyrimidines as radiosensitizers in the treatment of glioblastoma multiforme. Am J Clin Oncol 1987;10(5): 437-43. 27. Freese A, O’Rourke D, Judy K, O’Connor MJ. The application of 5-bromodeoxyuridine in the management of CNS tumors. J Neurooncol 1994;20(1):81-95. 28. Sullivan FJ, Herscher LL, Cook JA, et al. National Cancer Institute (phase II) study of high-grade glioma treated with accelerated hyperfractionated radiation and iododeoxyuridine: results in anaplastic astrocytoma. Int J Radiat Oncol Biol Phys 1994;30(3): 583-90. 29. Phillips TL, Levin VA, Ahn DK, et al. Evaluation of bromodeoxyuridine in glioblastoma multiforme: a Northern California Cancer Center Phase II study. Int J Radiat Oncol Biol Phys 1991;21(3):709-14. 30. Phillips TL, Bodell WJ, Uhl V, Ross GY, Rasmussen J, Mitchell JB. Correlation of exposure time, concentration and incorporation of IdUrd in V-79 cells with radiation response. Int J Radiat Oncol Biol Phys 1989;16(5): 1251-5.

64

31. Groves MD, Maor MH, Meyers C, et al. A phase II trial of high-dose bromodeoxyuridine with accelerated fractionation radiotherapy followed by procarbazine, lomustine, and vincristine for glioblastoma multiforme. Int J Radiat Oncol Biol Phys 1999;45(1):127-35. 32. Prados MD, Seiferheld W, Sandler HM, et al. Phase III randomized study of radiotherapy plus procarbazine, lomustine, and vincristine with or without BUdR for treatment of anaplastic astrocytoma: final report of RTOG 9404. Int J Radiat Oncol Biol Phys 2004;58(4):1147-52. 33. Forouzannia A, Richards GM, Khuntia D, Mehta MP. Motexafin gadolinium: a novel radiosensitizer for brain tumors. Expert Rev Anticancer Ther 2007;7(6):785-94. 34. Khuntia D, Mehta M. Motexafin gadolinium: a clinical review of a novel radioenhancer for brain tumors. Expert Rev Anticancer Ther 2004;4(6):981-9. 35. Carde P, Timmerman R, Mehta MP, et al. Multicenter phase Ib/II trial of the radiation enhancer motexafin gadolinium in patients with brain metastases. J Clin Oncol 2001;19(7):2074-83. 36. Viala J, Vanel D, Meingan P, Lartigau E, Carde P, Renschler M. Phases IB and II multidose trial of gadolinium texaphyrin, a radiation sensitizer detectable at MR imaging: preliminary results in brain metastases. Radiology 1999;212(3):755-9. 37. Rosenthal DI, Nurenberg P, Becerra CR, et al. A phase I single-dose trial of gadolinium texaphyrin (Gd-Tex), a tumor selective radiation sensitizer detectable by magnetic resonance imaging. Clin Cancer Res 1999;5(4):739-45. 38. Mehta MP, Rodrigus P, Terhaard CH, et al. Survival and neurologic outcomes in a randomized trial of motexafin gadolinium and whole-brain radiation therapy in brain metastases. J Clin Oncol 2003;21(13):2529-36. 39. Meyers CA, Smith JA, Bezjak A, et al. Neurocognitive function and progression in patients with brain metastases treated with whole-brain radiation and motexafin gadolinium: results of a randomized phase III trial. J Clin Oncol 2004;22(1):157-65. 40. Mehta M, Gervais R, Chabot P, et al. Motexafin gadolinium (MGd) combined with prompt whole brain radiation therapy (RT) prolongs time to neurologic progression in non-small cell lung cancer (NSCLC) patients with brain metastases: results of a phase III trial. Proc Am Soc Clin Oncol 2006;24(18s):7014. 41. Ford JM, Seiferheld W, Alger JR, et al. Results of the phase I dose-escalating study of motexafin gadolinium with standard radiotherapy in patients with glioblastoma multiforme. Int J Radiat Oncol Biol Phys 2007;69(3): 831-8. 42. Regine WF, Schmitt FA, Scott C, et al. Feasibility of neurocognitive outcome evaluations in patients with brain metastases in a multi-institutional cooperative group setting: results of radiation therapy oncology group (RTOG) trial BR-0018. Int J Radiat Oncol Biol Phys 2004;58(5):1346-52. 43. Gerstner ER, Fine RL. Increased permeability of the blood-brain barrier to chemotherapy in metastatic brain

1001

1002

64 44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

54.

55.

56.

57.


tumors: establishing a treatment paradigm. J Clin Oncol 2007;25(16):2306-12. Muldoon LL, Soussain C, Jahnke K, et al. Chemotherapy delivery issues in central nervous system malignancy: a reality check. J Clin Oncol 2007;25(16): 2295-305. Ott RJ, Brada M, Flower MA, Babich JW, Cherry SR, Deehan BJ. Measurements of blood-brain barrier permeability in patients undergoing radiotherapy and chemotherapy for primary cerebral lymphoma. Eur J Cancer 1991;27(11):1356-61. Seute T, Leffers P, Wilmink JT, ten Velde GP, Twijnstra A. Response of asymptomatic brain metastases from smallcell lung cancer to systemic first-line chemotherapy. J Clin Oncol 2006;24(13):2079-83. DeAngelis LM, Mandell LR, Thaler HT, et al. The role of postoperative radiotherapy after resection of single brain metastases. Neurosurgery 1989;24(6):798-805. Bleyer WA. Neurologic sequelae of methotrexate and ionizing radiation: a new classification. Cancer Treat Rep 1981;65 Suppl 1:89-98. Evans AE. Central nervous system workshop. Cancer Clin Trials 1981;4 Suppl:31-5. Meyers CA, Smith JA, Bezjak A, et al. Neurocognitive function and progression in patients with brain metastases treated with whole-brain radiation and motexafin gadolinium: results of a randomized phase III trial. J Clin Oncol 2004;22(1):157-65. Ostermann S, Csajka C, Buclin T, et al. Plasma and cerebrospinal fluid population pharmacokinetics of temozolomide in malignant glioma patients. Clin Cancer Res 2004;10(11):3728-36. Brada M, Judson I, Beale P, et al. Phase I dose-escalation and pharmacokinetic study of temozolomide (SCH 52365) for refractory or relapsing malignancies. Br J Cancer 1999;81(6):1022-30. Stupp R, Dietrich PY, Ostermann Kraljevic S, et al. Promising survival for patients with newly diagnosed glioblastoma multiforme treated with concomitant radiation plus temozolomide followed by adjuvant temozolomide. J Clin Oncol 2002;20(5):1375-82. Combs SE, Gutwein S, Schulz-Ertner D, et al. Temozolomide combined with irradiation as postoperative treatment of primary glioblastoma multiforme. Phase I/II study. Strahlenther Onkol 2005;181(6):372-7. Gilbert MR, Friedman HS, Kuttesch JF, et al. A phase II study of temozolomide in patients with newly diagnosed supratentorial malignant glioma before radiation therapy. Neuro Oncol 2002;4(4):261-7. Wedge SR, Porteous JK, Glaser MG, Marcus K, Newlands ES. In vitro evaluation of temozolomide combined with X-irradiation. Anticancer Drugs 1997;8 (1):92-7. Chakravarti A, Erkkinen MG, Nestler U, et al. Temozolomide-mediated radiation enhancement in glioblastoma:

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

a report on underlying mechanisms. Clin Cancer Res 2006;12(15):4738-46. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 2005;352(10):997-1003. Newlands ES, Stevens MFG, Wedge SR, Wheelhouse RT, Brock C. Temozolomide: a review of its discovery, chemical properties, pre-clinical development, and clinical trials. Cancer Treat Rev 1997;23:35-61. Wedge SR, Porteous JK, Glaser MG, Marcus K, Newlands ES. In vitro evaluation of temozolomide combined with X-irradiation. Anticancer Drugs 1997;8(1): 92-7. Antonadou D, Paraskevaidis M, Sarris G, et al. Phase II randomized trial of temozolomide and concurrent radiotherapy in patients with brain metastases. J Clin Oncol 2002;20(17):3644-50. Antonadou D, Coliarakis N, Paraskevaidis M, et al. Whole Brain Radiotherapy Alone or in Combination with Temozolomide for Brain Metastases. A Phase III Study. Int J Radiat Oncol Biol Phys 2002;54 Suppl 1:93-4. Chang JE, Robins HI, Mehta MP. Therapeutic advances in the treatment of brain metastases. Clin Adv Hematol Oncol 2007;5(1):54-64. Stewart DJ. Update on the role of topotecan in the treatment of non-small cell lung cancer. Oncologist 2004;9 Suppl 6:43-52. von Pawel J, Schiller JH, Shepherd FA, et al. Topotecan versus cyclophosphamide, doxorubicin, and vincristine for the treatment of recurrent small-cell lung cancer. J Clin Oncol 1999;17(2):658-67. Kim JH, Kim SH, Kolozsvary A, Khil MS. Potentiation of radiation response in human carcinoma cells in vitro and murine fibrosarcoma in vivo by topotecan, an inhibitor of DNA topoisomerase I. Int J Radiat Oncol Biol Phys 1992;22(3):515-8. Lamond JP, Mehta MP, Boothman DA. The potential of topoisomerase I inhibitors in the treatment of CNS malignancies: report of a synergistic effect between topotecan and radiation. J Neurooncol 1996; 30(1):1-6. Baker SD, Heideman RL, Crom WR, Kuttesch JF, Gajjar A, Stewart CF. Cerebrospinal fluid pharmacokinetics and penetration of continuous infusion topotecan in children with central nervous system tumors. Cancer Chemother Pharmacol 1996;37(3):195-202. Sung C, Blaney SM, Cole DE, Balis FM, Dedrick RL. A pharmacokinetic model of topotecan clearance from plasma and cerebrospinal fluid. Cancer Res 1994;54 (19): 5118-22. Wong ET, Berkenblit A. The role of topotecan in the treatment of brain metastases. Oncologist 2004;9(1): 68-79. Gruschow K, Klautke G, Fietkau R. Phase I/II clinical trial of concurrent radiochemotherapy in combination


72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

with topotecan for the treatment of brain metastases. Eur J Cancer 2002;38(3):367-74. Hedde JP, Neuhaus T, Schuller H, et al. A phase I/II trial of topotecan and radiation therapy for brain metastases in patients with solid tumors. Int J Radiat Oncol Biol Phys 2007;68(3):839-44. Kocher M, Eich HT, Semrau R, Guner SA, Muller RP. Phase I/II trial of simultaneous whole-brain irradiation and dose-escalating topotecan for brain metastases. Strahlenther Onkol 2005;181(1):20-5. Kadota RP, Stewart CF, Horn M, et al. Topotecan for the treatment of recurrent or progressive central nervous system tumors – a pediatric oncology group phase II study. J Neurooncol 1999;43(1):43-7. Macdonald D, Cairncross G, Stewart D, et al. Phase II study of topotecan in patients with recurrent malignant glioma. National Clinical Institute of Canada Clinical Trials Group. Ann Oncol 1996;7(2):205-7. Fisher B, Won M, Macdonald D, Johnson DW, Roa W. Phase II study of topotecan plus cranial radiation for glioblastoma multiforme: results of Radiation Therapy Oncology Group 9513. Int J Radiat Oncol Biol Phys 2002;53(4):980-6. Castellino RC, Elion GB, Keir ST, et al. Scheduledependent activity of irinotecan plus BCNU against malignant glioma xenografts. Cancer Chemother Pharmacol 2000;45(4):345-9. Vredenburgh JJ, Desjardins A, Herndon JE II, et al. Bevacizumab plus irinotecan in recurrent glioblastoma multiforme. J Clin Oncol 2007;25(30):4722-9. Tishler RB, Schiff PB, Geard CR, Hall EJ. Taxol: a novel radiation sensitizer. Int J Radiat Oncol Biol Phys 1992; 22(3):613-7. Tishler RB, Geard CR, Hall EJ, Schiff PB. Taxol sensitizes human astrocytoma cells to radiation. Cancer Res 1992;52(12):3495-7. Langer CJ, Ruffer J, Rhodes H, et al. Phase II Radiation therapy oncology group trial of weekly paclitaxel and conventional external beam radiation therapy for supratentorial glioblastoma multiforme. Int J Radiat Oncol Biol Phys 2001;51(1):113-9. Chang SM, Kuhn JG, Rizzo J, et al. Phase I study of paclitaxel in patients with recurrent malignant glioma: a North American Brain Tumor Consortium report. J Clin Oncol 1998;16(6):2188-94. Chang SM, Kuhn JG, Robins HI, et al. A Phase II study of paclitaxel in patients with recurrent malignant glioma using different doses depending upon the concomitant use of anticonvulsants: a North American Brain Tumor Consortium report. Cancer 2001;91(2): 417-22. Lederman G, Wronski M, Arbit E, et al. Treatment of recurrent glioblastoma multiforme using fractionated stereotactic radiosurgery and concurrent paclitaxel. Am J Clin Oncol 2000;23(2):155-9.

64

85. Crane CH, Sargent DJ. Substitution of oral fluoropyrimidines for infusional fluorouracil with radiotherapy: how much data do we need? J Clin Oncol 2004;22(15): 2978-81. 86. Rich T. Capecitabine and radiation therapy for advanced gastrointestinal malignancies. Oncology (Williston Park);2002;16(12 Suppl 14):27-30. 87. Rich TA, Shepard RC, Mosley ST. Four decades of continuing innovation with fluorouracil: current and future approaches to fluorouracil chemoradiation therapy. J Clin Oncol 2004;22(11):2214-32. 88. Shewach DS, Lawrence TS. Antimetabolite radiosensitizers. J Clin Oncol 2007;25(26):4043-50. 89. Coughlin CT, Richmond RC. Biologic and clinical developments of cisplatin combined with radiation: concepts, utility, projections for new trials, and the emergence of carboplatin. Semin Oncol 1989;16 (4 Suppl 6):31-43. 90. Peterson K, Harsh Gt, Fisher PG, Adler J, Le Q. Daily low-dose carboplatin as a radiation sensitizer for newly diagnosed malignant glioma. J Neurooncol 2001;53(1): 27-32. 91. Glass J, Silverman CL, Axelrod R, Corn BW, Andrews DW. Fractionated stereotactic radiotherapy with cisplatinum radiosensitization in the treatment of recurrent, progressive, or persistent malignant astrocytoma. Am J Clin Oncol 1997;20(3):226-9. 92. Grossman SA, O’Neill A, Grunnet M, et al. Phase III study comparing three cycles of infusional carmustine and cisplatin followed by radiation therapy with radiation therapy and concurrent carmustine in patients with newly diagnosed supratentorial glioblastoma multiforme: Eastern Cooperative Oncology Group Trial 2394. J Clin Oncol 2003;21(8):1485-91. 93. Guerrieri M, Wong K, Ryan G, Millward M, Quong G, Ball DL. A randomised phase III study of palliative radiation with concomitant carboplatin for brain metastases from non-small cell carcinoma of the lung. Lung cancer 2004;46(1):107-11. 94. Vivanco I, Sawyers CL. The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer 2002;2(7):489-501. 95. Choe G, Horvath S, Cloughesy TF, et al. Analysis of the phosphatidylinositol 30 -kinase signaling pathway in glioblastoma patients in vivo. Cancer Res 2003;63(11): 2742-6. 96. Ermoian RP, Furniss CS, Lamborn KR, et al. Dysregulation of PTEN and protein kinase B is associated with glioma histology and patient survival. Clin Cancer Res 2002;8(5):1100-6. 97. Holland EC, Celestino JC, Dai C, Schaefer L, Sawaya RE, Fuller GN. Combined activation of Ras and Akt in neural progenitors induce glioblastoma formation in mice. Nat Genet 2000;25(1):55-7. 98. von Deimling A, von Ammon K, Schoenfeld D, Wiestler OD, Seizinger BR, Louis DN. Subsets of

1003

1004

64 99.

100.

101.

102.

103. 104.

105.

106.

107.

108.

109.

110.

111.

112.


glioblastoma multiforme defined by molecular genetic analysis. Brain Pathol 1993;3(1):19-26. Watanabe K, Tachibana O, Sata K, Yonekawa Y, Kleihues P, Ohgaki H. Overexpression of the EGF receptor and p53 mutations are mutually exclusive in the evolution of primary and secondary glioblastomas. Brain Pathol 1996; 6(3):217-23; discussion 23-4. Kleihues P, Ohgaki H. Primary and secondary glioblastomas: from concept to clinical diagnosis. Neuro Oncol 1999;1(1):44-51. Frederick L, Eley G, Wang XY, James CD. Analysis of genomic rearrangements associated with EGRFvIII expression suggests involvement of Alu repeat elements. Neuro Oncol 2000;2(3):159-63. Kleihues P, Louis DN, Scheithauer BW, et al. The WHO classification of tumors of the nervous system. J Neuropathol Exp Neurol 2002;61(3):215-25; discussion 26–9. Mischel PS, Cloughesy TF. Targeted molecular therapy of GBM. Brain Pathol 2003;13(1):52-61. Rao RD, Uhm JH, Krishnan S, James CD. Genetic and signaling pathway alterations in glioblastoma: relevance to novel targeted therapies. Front Biosci 2003;8: e270-e280. Abounader R, Lal B, Luddy C, et al. In vivo targeting of SF/HGF and c-met expression via U1snRNA/ribozymes inhibits glioma growth and angiogenesis and promotes apoptosis. FASEB J 2002;16(1):108-10. Senger DL, Tudan C, Guiot MC, et al. Suppression of Rac activity induces apoptosis of human glioma cells but not normal human astrocytes. Cancer Res 2002;62(7): 2131-40. Prados M, Yung WK, Wen PY, et al. Phase I study of ZD1839 plus temozolomide in patients with malignant glioma. J Clin Oncol 2004;22(14s):1504. Rich JN, Reardon DA, Peery T, et al. Phase II trial of gefitinib in recurrent glioblastoma. J Clin Oncol 2004;22(1):133-42. Prados M, Chang S, Burton EC, et al. Phase I study of OSI-774 alone or with temozolamide in patients with malignant glioma. Proc Am Soc Clin Oncol 2003;22:99. Chakravarti A, Zhai G, Suzuki Y, et al. The prognostic significance of phosphatidylinositol 3-kinase pathway activation in human gliomas. J Clin Oncol 2004; 22(10):1926-33. Chakravarti A, Loeffler JS, Dyson NJ. Insulin-like growth factor receptor I mediates resistance to antiepidermal growth factor receptor therapy in primary human glioblastoma cells through continued activation of physphoinositide 3-kinase signaling. Cancer Res 2002; 62(1):200-7. Chakravarti A, Seiferheld W, Robins IA, et al. An update of phase I data from RTOG 0211: A phase I/II clinical study of gefitinib + radiation for newly-diagnosed GBM patients. J Clin Oncol 2004;22(14S): 1571.

113. Chakravarthy A, Berkey B, Robins HI, et al. An update of phase II results from RTOG 0211: A phase I/II study of gefitinib with radiotheapy in newly diagnosed glioblastoma. J Clin Oncol 2006;24 Suppl 18:1527. 114. Galanis E, Buckner JC, Maurer M, et al. NCCTG phase II trial of CCI-779 in recurrent glioblastoma multiforme (GBM). J Clin Oncol 2004;22(14S):1503. 115. Zagzag D, Capo V. Angiogenesis in the central nervous system: a role for vascular endothelial growth factor/ vascular permeability factor and tenascin-C. Common molecular effectors in cerebral neoplastic and nonneoplastic ‘‘angiogenic diseases’’. Histol Histopathol 2002; 17(1):301-21. 116. Zagzag D, Amirnovin R, Greco MA, et al. Vascular apoptosis and involution in gliomas precede neovascularization: a novel concept for glioma growth and angiogenesis. Lab Invest 2000;80(6):837-49. 117. Zundel W, Schindler C, Haas-Kogan D, et al. Loss of PTEN facilitates HIF-1-mediated gene expression. Genes Dev 2000;14(4):391-6. 118. Gollmer J, Ladoux A, Gioanni J, et al. Expression of vascular endothelial growth factor-b in human astrocytoma. Neuro Oncol 2000;2(2):80-6. 119. Mischel PS, Shai R, Shi T, Horvath S, Lu KV, Choe G, Seligson D, Kremen TJ, Palotie A, Liau LM, Cloughesy TF, Nelson SF. Identification of molecular subtypes of glioblastoma by gene expression profiling. Oncogene 2003;22(15):2361-73. 120. Conrad C, Friedman HS, Reardon DA, et al. A phase I/II trial of single-agent PTK 787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in patients with recurrent GBM. J Clin Oncol 2004;22(14S):1512. 121. Reardon DA, Friedman HS, Yung WK, et al. A phase I/II trial of PTK-787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in combination with either temozolomide or lomustine for patients with recurrent GBM. J Clin Oncol 2004;22(14S):1513. 122. Brandes AA, Stupp R, Hau P, et al. EORTC study 26041–22041: phase I/II study on concomitant and adjuvant temozolomide (TMZ) and radiotherapy (RT) with or without PTK787/ZK222584 (PTK/ZK) in newly diagnosed glioblastoma – results of a phase I trial. J Clin Oncol 2007; ASCO Annual Meeting Proceedings 2007; 25(18S):2026. 123. Wachsberger PR, Burd R, Marero N, et al. Effect of the tumor vascular-damaging agent, ZD6126, on the radioresponse of U87 glioblastoma. Clin Cancer Res 2005; 11(2 Pt 1):835-42. 124. Vredenburgh JJ, Desjardins A, Herndon JE II, et al. Phase II trial of bevacizumab and irinotecan in recurrent malignant glioma. Clin Cancer Res 2007;13(4):1253-9. 125. Batchelor TT, Sorensen AG, di Tomaso E, et al. AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell 2007;11(1):83-95.


126. Willett CG, Boucher Y, di Tomaso E, et al. Direct evidence that the VEGF-specific antibody bevacizumab has antivascular effects in human rectal cancer. Nat Med 2004;10(2):145-7. 127. Alghisi GC, Ruegg C. Vascular integrins in tumor angiogenesis: mediators and therapeutic targets. Endothelium 2006;13(2):113-35.

64

128. Stupp R, Ruegg C. Integrin inhibitors reaching the clinic. J Clin Oncol 2007;25(13):1637-8. 129. Nabors LB, Mikkelsen T, Rosenfeld SS, et al. Phase I and correlative biology study of cilengitide in patients with recurrent malignant glioma. J Clin Oncol 2007;25 (13):1651-7.

1005

73 Radiosurgery for Functional Neurosurgery D. Kondziolka

Introduction The history of stereotactic radiosurgery goes back almost as far as the history of functional neurosurgery. Leksell initially conceived the idea of closed-skull, single-session irradiation of a precisely defined intracranial target in 1951 and applied this concept immediately to functional neurosurgery [1]. At a time when functional destruction of normal brain required thermal energy or chemical injection, Leksell crossfired photon or proton radiation beams to achieve a similar goal. The initial radiosurgical concept was to create a small, precisely defined focal lesion, which was defined by image guidance. The procedure would not completely avoid brain penetration since contrast encephalography provided the information for identification of the targets. Whereas the ganglionic portion of the trigeminal nerve could be indirectly located using plain radiographs or cisternograms, deep brain targets required air or positive contrast ventriculography; direct visualization of the target for functional radiosurgery required the later development of computed imaging technology. Thus, radiosurgical techniques were used to create image-guided, physiologic inactivity or focally destructive brain lesions without neurophysiologic guidance. This was controversial, and the lack of neurophysiologic guidance remained the greatest argument against the use of radiosurgery for selected disorders. Nevertheless, the current use of radiosurgery as a ‘‘lesion generator’’ is based on

#


extensive animal studies that defined the dose, volume, and temporal response of the irradiated tissue. The utility of radiosurgery has now been compared to microsurgical, percutaneous injection, and electrode-based techniques used for functional neurologic disorders. Current anatomic targets include the trigeminal nerve and other ganglia (facial pain syndromes – discussed in a separate chapter), the thalamus (for tremor or pain), the cingulate gyrus or anterior internal capsule (for pain or psychiatric illness), the hypothalamus (for cancer pain), and the hippocampus or other brain targets (for epilepsy) [2,3]. Leksell first coupled an orthovoltage X-ray tube to his early generation stereotactic frame, a concept used for trigeminal neuralgia but not for intraparenchymal brain targets [4]. Thus, he began work with physicist Borje Larsson to crossfire proton beams [5], and subsequently used a modified linear accelerator. His decision to build and then use the first Gamma Knife in 1967 reflected his frustration with particle beam technology which required travel of the patient to a special cyclotron center. As originally designed, the first Gamma Knife collimator helmets created a discoid volume of focal irradiation that could ‘‘section’’ white matter tracts or brain tissue in a manner similar to a leucotome or other instrument. Later models of the Gamma Knife have provided more flexibility in the creation of lesions or effects of different volumes (more suitable for tumors or vascular malformations), together with precise robotic delivery and efficiency.

1192

73

Radiosurgery for functional neurosurgery

The Beginning Years Before 1978, all uses of radiosurgery remained limited due to the lack of high-resolution, neuroimaging techniques to identify brain lesions or functional brain regions. Angiographic targeting of arteriovenous malformations proved successful, but proved to be limited by the two-dimensional estimates of complex three-dimensional target volumes. Functional radiosurgery was performed for a limited number of patients with intractable pain related to malignancy [6,7], movement disorders [2], psychiatric dysfunction [8,9], and trigeminal neuralgia [4]. On the sidelines, percutaneous retrogasserian glycerol rhizotomy was developed during an observation made during the refinement of the Gamma Knife technique for trigeminal neuralgia. Haka¨nsson and Leksell attempted to localize the trigeminal nerve within its cistern using glycerol mixed with tantalum powder (as a radiopaque marker) placed before radiosurgery. However, after injection of the glycerol, trigeminal neuralgia pain was relieved. For intractable pain related to malignancy, radiosurgery was used both for hypophysectomy as well as for medial thalamotomy. Although the procedure was non-invasive, the latency interval for lesion generation and pain relief was one obvious drawback. Steiner and colleagues presented results from an autopsy study after radiosurgery for cancer pain in 1980 [7]. The ablative dose for tissue volumes had been estimated during animal experiments through the 1960s using protons and photons [5,10,11]. Initial patients who had radiosurgery for tissue ablation received maximum doses of 100–250 Gy. At small volumes, doses in excess of 150 Gy provided consistent tissue necrosis in animal models. Since these first patients were treated for pain from a terminal malignancy, they did not live long enough to sustain a potential complication from such high doses. The clinical use of such doses proved to be the foundation for later use in tremor management.

Dose-Selection for Parenchymal Functional Radiosurgery Early animal experiments showed consistent lesion creation at doses at or above 150 Gy [10,11]. Clinical data showed that pain relief occurred usually within 3 weeks after radiosurgery [7]. In rat experiments at 200 Gy using a single 4 mm isocenter, we found a consistent relationship for lesion generation that substantiated observations from that human study [12]. Doses of 200 Gy were delivered to the rat frontal brain and then the brain was studied at one, 7, 14, 21, 60, and 90 days after irradiation. At 1 and 7 days, we noted that the brain continued to appear normal. By 14 days, the parenchyma appeared slightly edematous within the target volume. However, by 21 days, a complete circumscribed volume of necrosis was identified within the radiation volume (4 mm diameter). This remained consistent thereafter. Thus, the clinical observation of pain relief at 21 days noted by Steiner et al. was correlated with laboratory findings at the 200 Gy dose. The ablative radiosurgery lesion appears as a punched-out, circumscribed volume of complete parenchymal necrosis with cavitation. Within a 1–3 mm rim that characterizes the steep falloff in radiation dose, normalization of the tissue appearance is found. In this zone, blood vessels appear thickened and hyalinized, and often protein extravasation can be identified. The brain is edematous in this region, either from an increase in extracellular fluid, or from the intracellular swelling of gliosis. Acute or chronic inflammatory cells are present. MR imaging demonstrates all of these features after radiosurgical thalamotomy – a sharply defined, contrast-enhanced rim that defines the low signal lesion (on short TR images) surrounded by a zone of high-signal (on long TR images) brain tissue [13]. Friehs and colleagues collected imaging data from four centers who created functional radiosurgery lesions (n = 56). They found that maximum doses in excess of 160 Gy were more


likely to produce lesions larger than expected and recommended single 4 mm isocenter lesions at doses below 160 Gy [14]. Studies at the University of Pittsburgh found that in both large and small animal models, doses at or above 100 Gy caused necrosis, but the delay to necrosis was longer [12,15]. To identify the effect of increasing volume, we used an 8 mm collimator in a baboon model and found that half of the animals developed an 8 mm diameter necrotic lesion at doses as low as 50 Gy [16]. Dose, volume and time are the three key factors that determine the nature of the functional ablative lesion. Once created, this lesion remains stable over years [10]. The limitation of radiosurgery technology as a lesion generator stems from the inability to reliably control the effects of dose and volume. When a larger brain target may be desirable, the sharp fall-off in dose outside the target becomes less steep with increasing volume. The risk of an adverse radiation effect outside the target volume becomes problematic [17]. At small volumes (i.e., single 4 mm collimator), the radiosurgerycreated lesion appears more consistent.

Imaging in Functional Surgery Since physiological information is excluded from the targeting component of a functional radiosurgery procedure, high-quality stereotactic neuroimaging must be performed. The imaging must be accurate since small volumes are irradiated. In addition, the imaging must be of sufficient resolution to identify the target structure but also show important regional tissues. Magnetic resonance imaging is the preferred imaging tool for functional radiosurgery [18–20]. Accurate stereotactic MRI-based localization should be confirmed at each institution [21]. The use of fast inversion recovery or other long relaxation time MR sequences helps to separate gray and white matter structures. However,

73

to this day the targeting of physiologically abnormal brain regions such as groups of kinesthetic thalamic tremor cells or epileptic foci using imaging alone remains indirect. We believe that with improvements in subcortical imaging using higher field strength magnets, gamma knife radiosurgery will play an expanded role in movement disorders.

Radiosurgical Thalamotomy Ventrolateral thalamic surgery for the management of tremor related to Parkinson’s Disease remains a proven and time-honored concept within functional neurosurgery. Traditionally, this has involved imaging definition of the thalamic target, placement of an electrode into the thalamus, physiologic recording and stimulation at the target site, and creation of a lesion or providing electrical stimulation. Radiosurgical thalamotomy by definition avoids placement of the electrode and evaluation of the physiologic response. In radiosurgery, imaging definition alone is used to determine lesion placement. Through the use of contrast ventriculography, computed tomographic imaging, and more recently stereotactic MRI, thalamotomy using the Gamma Knife has been performed at centers across the world [18,22–24]. As discussed above, the issues of lesion volume and dose-selection remain important. Although radiosurgery can abolish tremor, many surgeons currently believe although adequate results might be obtained, better results may be possible with deep brain stimulation (DBS). The challenges inherent in choosing the best possible ablative target using imaging alone are significant. Radiosurgical thalamotomy, if performed, should be performed by surgeons experienced in radiofrequency thalamotomy or DBS. Due to the absence of electrophysiological information, the inability to stop the lesion during surgery, and the latency to the clinical

1193

1194

73


response, most surgeons use radiosurgery primarily for patients with advanced age or medical disorders where electrode placement would be associated with higher-risk. Ohye began to perform radiosurgical thalamotomy contralateral to a prior radiofrequency lesion, or to enlarge a previously mapped lesion [22]. Duma et al. reported a 5-year experience with 38 thalamotomies using the Gamma Knife and 28 month mean follow-up [18]. Complete tremor abolition was noted in 24%, excellent relief in 26%, good improvement in 29%, and little to no benefit in 21%. The median time to improvement was 2 months, consistent with data from previous animal experiments. They used a dose range of 110–165 Gy with better results at higher doses. Such higher doses may exert effects on a larger surrounding tissue volume of kinesthetic tremor cells (outside the sharply defined necrotic volume) that translates into tremor reduction and overcomes any limitations in target selection. Young et al. reported that 88% of 27 patients who had radiosurgical thalamotomy for tremor (120–160 Gy) became tremor free or ‘‘nearly’’ tremor free [20]. Hirato et al. also found tremor suppression after Gamma Knife thalamotomy in a small patient series [22]. Friehs et al. reported an experience of radiosurgical thalamotomy (n = 3) and caudatotomy (n = 10) with clinical improvement in most patients and no morbidity [25]. The mean age of 77 years in the Pittsburgh radiosurgical series is older than the mean age of 60 years in their DBS series [26]. Gamma Knife radiosurgery proved to be effective in improving medically-refractive essential tremor in a predominantly elderly patient series (> Figure 73-1) [27]. Eighteen patients (69%) improved both action tremor and writing scores, and an additional six (23%) improved their action tremor scores. Thirteen patients (50%) had either no or only slight intermittent tremor in the affected extremity and 90% had some degree of clinically significant tremor improvement. Overall, the

. Figure 73-1 CT scan 4 months after a left gamma knife thalamotomy (140 Gy) in a 90 year-old woman with essential tremor. The radiosurgical lesion is shown at the v.i.m target

mean Fahn-Tolosa-Marin tremor score improved from 3.8 to 1.7 (p < 0.000015). After radiosurgery, MRI usually showed a 4–5 mm round, well-circumscribed lesion with peripheral contrast enhancement surrounding a low signal region (> Figure 73-2). A localized area of high signal (seen on long relaxation time studies) demonstrated the peripheral neuronal effect, manifested as an increase in intra- or extracellular water. Two patients with complications had different MRI findings. Although the enhancing lesion created in one patient was unexpectedly large, complete resolution was seen on subsequent imaging. This indicated that the response was related to temporary blood brain barrier changes, and not to permanent radiation necrosis. As noted earlier, the target volume is crucial. Early results with larger target volumes using an 8 mm collimator were reported by Lindquist et al. [2]. Delayed cerebral edema and regions of radiation necrosis at high doses testified to the


73

. Figure 73-2 (a) MRI scan (flair) 6 months after gamma knife radiosurgery in a 56-year-old man with essential tremor who had refused deep brain stimulation. Significant tremor abolition was noted without side effects. (b) the MRI scan with contrast 3 years later is shown depicting a stable lesion

volume effects of radiosurgery [13]. Similar problems have been noted using combinations of 4 mm isocenters to construct a cylindrical rather than spherical target volume [17,20]. Nevertheless, the ability to create a small-volume lesion using radiosurgery without placement of a craniostomy or the invasive placement of an electrode remain attractive considerations. To that end, several surgeons have evaluated the use of radiosurgery for medial thalamotomy and for pallidotomy, procedures where the usefulness of physiologic recording or stimulation initially was less clear.

Radiosurgical Pallidotomy There was a resurgence in the use of radiofrequency-based stereotactic pallidotomy for patients with advanced Parkinson’s disease (PD) beginning in 1992. Some investigators then performed Gamma Knife pallidotomy using imageguidance alone as an alternative to electrode techniques. Rand et al. reported their preliminary results after radiosurgical pallidotomy and

noted relief of contralateral rigidity in 4 of 8 patients [28]. No patient in their series sustained a complication. Friedman et al. reported on four patients after Gamma Knife pallidotomy (180 Gy) with improvement in only one patient [29]. They noted heterogeneity in lesion volumes on MRI, a finding also documented by others. In contrast to the thalamus, where small radiosurgery lesions appeared consistent, pallidotomy lesions may be more variable due to effects on perforating arteries that supply that region of the basal ganglia. The lesion volumes and contrastenhancement patterns seem less consistent [20]. At our center only one radiosurgical pallidotomy has been performed. At present, this technique is performed rarely and DBS remains a much more valuable concept for most patients with an array of PD symptoms [30].

Radiosurgery for Pain The use of radiosurgery as an ablative tool to treat pain has a long history. Unfortunately too few patients have been managed to draw any

1195

1196

73


strong conclusions. Since the case report by Leksell in 1968 and the larger series by Steiner et al. in 1980, little has been written [6,7]. In Leksell’s two patients with carcinoma, the centrum medianum target received doses of 250 and 200 Gy. The second patient had bilateral radiosurgery spaced by 2 months and became pain-free. In Steiner’s series, doses as high as 250 Gy were believed unnecessary because of the sharp dosegradient. Young et al. performed medial thalamotomy for the treatment of chronic non-cancer pain in patients who had failed comprehensive medical, surgical, and behavioral therapies [31]. In 1996 they described that two-thirds of their 41 patient series had at least a 50% reduction in pain intensity estimates with improvements in physical and social functioning [3]. As might be expected, patients with deafferentation pain responded poorly, but more encouraging results were identified in patients with nociceptive syndromes. Again they cautioned on the use of larger volumes above that obtained with a single 4 mm isocenter, and on the use of doses above 160 Gy. Hayashi et al. performed pituitary glandstalk ablation by Gamma Knife radiosurgery, targeting the border between the pituitary stalk and gland with a maximum dose of 160 Gy using the 8-mm collimator to control cancer pain. They enrolled nine patients that had bone metastases and pain controlled well by morphine, KPS >40, and no previous radiation therapy [32]. All patients had failed the previous pain treatments except morphine. All patients became pain-free within a few days after radiosurgery, and which maintained as long as they lived. No recurrence of pain occurred. In addition, there was no panhypopituitarism and diabetes insipidus in the patients. This strategy of pituitary gland-stalk ablation for pain control also showed a good initial response (87.5%) of 8 patients with thalamic pain syndrome, however, the majority of patients (71.4%) experienced pain recurrence during the 6 month follow-up [33].

Radiosurgery for Psychiatric Disorders There is renewed interest in radiosurgical lesioning of the anterior internal capsule (anterior capsulotomy) in patients with medically refractory obsessive compulsive disorder (OCD). Radiosurgery for obsessive-compulsive and anxiety neurosis has been performed for over 45 years [8]. The first radiosurgical capsulotomy was performed by Leksell in 1953 using 300 kV X-rays [34]. Initially, pneumoencephalography was used for target definition in the placement of bilateral anterior internal capsule lesions. Five of the initial 7 patients had long-term benefit after 7 years of follow-up [2]. Since 1988, an additional 10 patients were treated at the Karolinska Institute using stereotactic MRI guidance. The initial use of an 8 mm collimator resulted in excessive edema, so these authors recommended the use of only 4 mm isocenters [2]. The results seem to be as efficacious as when conventional radiofrequency lesioning is performed [35]. Kihlstrom et al. described the stable imaging appearance of radiosurgical lesions 15–18 years after capsulotomy [36]. Oval shaped radiosurgical lesions in the anterior internal capsule or cingulate gyrus may impact on affective disorders or anxiety neuroses. Recently, Ruck et al. reported on long-term follow-up in 25 patients, 16 with an electrode and 9 with gamma knife surgery [37]. Response rates did not differ between methods and they concluded that capsulotomy was effective in reducing OCD symptoms. A series of patients from Brown University and the University of Pittsburgh have been presented at national meetings. The radiosurgical capsulotomy is performed only after comprehensive psychiatric evaluation and management leading to a diagnosis of severe OCD, and after failure of non-surgical approaches. In Pittsburgh, we performed Gamma Knife surgery on three patients with severe, medically intractable OCD (> Figure 73-3). According to our protocol,


. Figure 73-3 Bilateral anterior capsulotomies are shown on coronal contrast-enhanced MRI, 1 year after gamma knife radiosurgery (140 Gy) in a patient with obsessivecompulsive disorder

all patients were evaluated by at least two psychiatrists who recommended the capsulotomy procedure. The patient had to request the procedure, and have severe OCD according to the Yale Brown Obsessive Compulsive Scale (YBOCS). Patient ages were 37, 55, and 40, and pre-radiosurgery YBOCS scores were 32/40/39/ 40, and 39/40. Bilateral lesions were created with two 4 mm isocenters to create an oval volume in the ventral capsule at the putaminal midpoint. A maximum dose of 140–150 Gy was used. There was no morbidity after the procedure and all returned immediately to baseline function. All three patients had described functional improvements, and reduction in OCD behavior. One patient with compulsive skin picking and an open wound had later healing of their chronic wound and a reduction in the YBOCS score from 39 to 8/40 [38]. We believe this technique should be evaluated further in patients with severe and disabling behavioral disorders.

73

Radiosurgery for Epilepsy There is current interest in the use of radiosurgery for patients with focal epilepsy. The observation that brain irradiation (via radiation therapy or radiosurgery) could lead to cessation of seizures has spurred several groups to work in this field despite the lack of a consistent approach to defining the target volume. In 1985, BarciaSalorio et al. reported on six patients with epilepsy who had low dose radiosurgery. The epileptic focus was localized by means of conventional scalp electroencephalogram (EEG), subarachnoid electrodes, and depth electrodes. Radiosurgery (a 10 mm collimator to deliver an estimated dose of 10 Gy) was performed using a cobalt unit coupled to a stereotactic localizer. They hypothesized that this low radiation dose provided a specific effect on epileptic neurons, without inducing tissue necrosis. In 1994, they provided a long-term analysis in a series of 11 patients using a dose range of 10–20 Gy. Five patients had complete cessation of seizures, and an additional five were improved. Seizures began to decrease gradually after 3–12 months following radiosurgery [39]. Following this work, Lindquist and colleagues at the Karolinska Institute began to perform epilepsy radiosurgery using advanced localization techniques that included magnetoencephalography (MEG) to define interictal activity [2,40]. In some patients, the epileptic dipole activity identified on MEG before radiosurgery later resolved along with seizure cessation. At the same time, radiosurgery was evaluated in animal models of epilepsy. We used the kainic acid model of hippocampal epilepsy in the rat, and were able to stop seizures and improve animal behavior [41,42]. Rats were randomized to control or radiosurgery arms (20, 40, 60, or 100 Gy) and then evaluated with serial EEG, behavioral studies, functional MRI, and histology. More recently, radiosurgery has been of value in patients with gelastic or generalized seizures related to hypothalamic hamartomas

1197

1198

73


. Figure 73-4 Gamma knife radiosurgery plan for a patient with gelastic seizures and complex epilepsy related to a hypothalamic hamartoma

(> Figure 73-4) [43]. A larger indication may rest with the use of epilepsy to create an amygdalohippocampal lesion for patients with mesial temporal sclerosis as proposed by Regis et al. [44,45]. In 1993, Re´gis and associates in Marseille performed selective amygdalohippocampal radiosurgery for mesial temporal lobe epilepsy. Gamma Knife radiosurgery was used to create a conformal volume of radiation for the amygdala and hippocampus. This approximate 7 ml volume represented the largest functional target irradiated to that time. They delivered a margin dose of 25 Gy to the 50% isodose line, a dose that later caused target necrosis. The first patient became seizure free immediately and the second

after a latency of almost 1 year. Serial MR scans showed target contrast-enhancement that corresponded to the 50% isodose line [44]. Patients managed at their center have been part of a multidisciplinary prospective evaluation and treatment protocol. A recently published longer-term evaluation with 8 year mean follow-up (margin dose of 24 Gy), found that 9 of 16 patients were seizure free [46]. The first prospective multicenter clinical trial in the United States was recently completed (> Figure 73-5). A number of questions remain to be addressed regarding the role of radiosurgery for mesial temporal sclerosis-related epilepsy. The optimal target may include both


73

. Figure 73-5 Coronal MRI images at amygdalohippocampal radiosurgery (a) in a patient with mesial temporal sclerosis and at 15 months (b), showing the left sided lesion. Significant reduction in his seizures began at 10 months, with no adverse effects

amygdala and hippocampus, but the total target volume remains debated. Target volume helps to determine dose selection, including the dose received by regional structures such as the brainstem or optic tract. Finally, investigators need to determine whether the balance between seizure response and morbidity is acceptable, particularly in comparison to surgical resection. A randomized trial comparing radiosurgery to resection is planned. ‘‘Radiosurgery has been evaluated in animal models and in clinical use for epilepsy. The opportunity to use radiosurgery to disrupt an epileptic focus, or to change abnormal neurophysiologic patterns is of interest. We used the kainic acid model of hippocampal epilepsy in the rat, and were able to stop seizures and improve animal behavior (41,42). In clinical use, radiosurgery has been of value in patients with gelastic or generalized seizures related to hypothalamic hamartomas (47). A larger indication may rest with the use of epilepsy to create an amygdalohippocampal lesion for patients with mesial temporal sclerosis. First tested by Regis and colleagues from Marseille, a first prospective

clinical trial in the United States was recently completed.’’ Current issues that remain important for epilepsy radiosurgery include dose-selection (necrotizing vs. non-necrotizing), localization methods for non-lesional epilepsy, the target volume necessary for irradiation, and the expected short and long-term outcomes. It is not known what kind of tissue effect is required to stop the generation or propagation of seizures. Some groups have used low doses (i.e., 10–20 Gy) where few if any histologic changes would be expected. Others have used doses as high as 100 Gy that cause target necrosis and regional brain edema [48]. A typical amygdalohippocampal radiosurgery maximum dose to a volume less than 7.5 ml is 40–50 Gy. If focal hippocampal (or any other brain tissue) irradiation can eliminate seizures without the need for complete tissue destruction, then radiosurgery may become an important therapy for patients with intractable epilepsy. At the same time we await improvements in tools for localization of the seizure focus.

1199

1200

73


Functional Imaging and Radiosurgery Improvements in functional imaging eventually will impact on radiosurgery. Functional magnetic resonance imaging to localize cortical function prior to radiosurgery have been evaluated in pilot studies. Localization of hand motor function, leg motor function, and speech areas surrounding arteriovenous malformations and brain tumors before radiosurgery has assisted dose planning [49]. This information can be used to restrict the radiosurgery dose away from functional areas. Advancements in functional imaging may improve the localization of epileptic foci, and perhaps even regions of excitation in the basal ganglia. Magnetoencephalography is an exciting tool to identify functional activation. An ability to identify hyperactivity in deep brain structures would be valuable. With further improvements in neuroimaging and non-invasive physiologic studies, the future will see a significant linkage between functional brain disorders and stereotactic radiosurgery.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951;102:316-19. 2. Lindquist C, Kihlstrom L, Hellstrand de. Functional neurosurgery-a future for the gamma knife? Stereotact Funct Neurosurg 1991;57:72-81. 3. Young RF, Vermeulen S, Posewitz A, Grimm P, Blasko J, Jacques D, Rand R, Copcutt B. Functional neurosurgery with the Leksell gamma knife. Radiosurgery 1996;1:218-25. 4. Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand 1971;137:311-14. 5. Larsson B, Leksell L, Rexed B, Surander P, Mair W, Andersson B. The high-energy proton beam as a neurosurgical tool. Nature 1958;182:1222-3. 6. Leksell L. Cerebral radiosurgery I. Gammathalamotomy in two cases of intractable pain. Acta Chir Scand 1968;134:585-95. 7. Steiner L, Forster D, Leksell L, Meyerson B, Boethius J. Gammathalamotomy in intractable pain. Acta Neurochir 1980;52:173-84.

8. Leksell L, Backlund EO. Stereotactic gammacapsulotomy. In: Hitchcock ER, Ballantine HT, Meyerson BA, editors. Modern concepts in psychiatric surgery. Elsevier; Amsterdam: 1979. p. 213-16. 9. Rylander G. Stereotactic radiosurgery in anxiety and obsessive-compulsive states: Psychiatric aspects. In Hitchcock ER, Ballantine HT, Meyerson BA, editors. Modern concepts in psychiatric surgery. Amsterdam: Elsevier; 1979. p. 235-40. 10. Andersson B, Larsson B, Leksell L, Mair W, Rexed B, Sourander P, Wennerstrand J. Histopathology of late local radiolesions in the goat brain. Acta Radiologica 1970;9:385-94. 11. Rexed B, Mair W, Sourander P, Larsson B, Leksell L. Effect of high energy protons on the brain of the rabbit. Acta Radiologica 1960;53:289-99. 12. Kondziolka D, Lunsford LD, Claassen D, Maitz A, Flickinger J. Radiobiology of radiosurgery. Part I: the normal rat brain model. Neurosurgery 1992;31:271-9. 13. Leksell L, Herner T, Leksell D, Persson B, Lindquist C. Visualization of stereotactic radiolesions by nuclear magnetic resonance. J Neurol Neurosurg Psych 1985;48:19-20. 14. Friehs G, Noren G, Ohye C, Duma C, Mark R, Plombon J, Young RF. Lesion size following Gamma Knife treatment for functional disorders. Stereotact Funct Neurosurg 1996;66 Suppl 1:320-8. 15. Kondziolka D, Linskey ME, Lunsford LD. Animal models in radiosurgery. In: Alexander E, Loeffler JS, Lunsford LD, editors. Stereotactic radiosurgery. New York: McGraw-Hill; 1993. p. 51-64. 16. Lunsford LD, Altschuler EM, Flickinger JC, Wu A, Martinez AJ. In vivo biological effects of stereotactic radiosurgery: a primate model. Neurosurgery 1990; 27:373-82. 17. Kihlstrom L, Guo WY, Lindquist C, Mindus P. Radiobiology of radiosurgery for refractory anxiety disorders. Neurosurgery 1995;36:294-302. 18. Duma C, Jacques D, Kopyov O, Mark R, Copcutt B, Farokhi HK. Gamma knife radiosurgery for thalamotomy in Parkinsonian tremor: a five-year experience. Neurosurg Focus 1997;2(3). 19. Kondziolka D, Lunsford LD, Flickinger JC, Young R, Vermeulen S, Duma C, Jacques D, Rand RW, Regis J, Peragut JC, Manera L, Epstein M, Lindquist C. Stereotactic radiosurgery for trigeminal neuralgia: a multiinstitution study using the gamma unit. J Neurosurg 1996;84:940-5. 20. Young RF, Shumway-Cook A, Vermeulen S, Grimm P, Blasko J, Posewitz A. Gamma knife radiosurgery as a lesioning technique in movement disorder surgery. Neurosurg Focus 1997;2(3):e11. 21. Kondziolka D, Dempsey PK, Lunsford LD, Kestle J, Dolan E, Kanal E, Tasker RR. A comparison between magnetic resonance imaging and computed tomography


22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

for stereotactic coordinate determination. Neurosurgery 1992;30:402-7. Hirato M, Ohye C, Shibazaki T, Nakamura M, Inoue H, Andou Y. Gamma knife thalamotomy for the treatment of functional disorders. Stereotact Funct Neurosurg 1995;64 Suppl 1:164-71. Otsuki T, Jokura H, Takahashi K, Ishikawa S, Yoshimoto T, Kimura M, Yoshida R, Miyazawa T. Stereotactic gamma-thalamotomy with a computerized brain atlas: Technical case report. Neurosurgery 1994;35:764-8. Pan L, Dai J, Wang BJ, Xu W, Zhou L, Chen XR. Stereotactic gamma thalamotomy for the treatment of Parkinsonism. Stereotact Funct Neurosurg 1996;66 Suppl 1:329-32. Friehs G, Ojakangas CL, Pachatz P, Schrottner O, Ott E, Pendl G. Thalamotomy and caudatomy with the Gamma Knife as a treatment for Parkinsonism with a comment on lesion sizes. Stereotact Funct Neurosurg 1995;64 Suppl 1:209-21. Kondziolka D, Ong J, Lee JYK, Moore R, Flickinger J, Lunsford LD. Gamma knife thalamotomy for essential tremor. J Neurosurg 2008;108:111-17. Kondziolka D, Ong J, Lee JYK, Moore R, Flickinger J, Lunsford LD. Gamma knife thalamotomy for essential tremor. J Neurosurg 2008;108:111-17. Rand RW, Jacques DB, Melbye RW, Copcutt B, Fisher M, Levenick M. Gamma knife thalamotomy and pallidotomy in patients with movement disorders: preliminary results. Stereotact Funct Neurosurg 1993;61 Suppl:65-92. Friedman J, Epstein M, Sanes J, Lieberman P, Cullen K, Lindquist C, Daamen M. Gamma Knife pallidotomy in advanced Parkinson’s disease. Ann Neurol 1996; 39:535-8. Kwon Y, Whang CJ. Stereotactic Gamma knife radiosurgery for the treatment of dystonia.2. Stereotact Funct Neurosurg 1995;64 Suppl 1:222-7. Young RF Jacques DB Rand RW, Copcutt B. Medial thalamotomy with the Leksell gamma knife for treatment of chronic pain. Acta Neurochir (Suppl) 1994;62:105-10. Hayashi M, Taira T, Chernov M, Fukuoka S, Liscak R, Yu CP, Ho RT, Regis J, Katayama Y, Kawakami Y, Hori T. Gamma knife surgery for cancer pain-pituitary glandstalk ablation: a multicenter prospective protocol since 2002. J Neurosurg 2002;97:433-7. Hayashi M, Taira T, Chernov M, Izawa M, Liscak R, Yu CP, Ho RT, Katayama Y, Kouyama N, Kawakami Y, Hori T, Takakura K. Role of pituitary radiosurgery for the management of intractable pain and potential future applications. Stereotact Funct Neurosurg 2003;81:75-83. Leksell L, Herner T, Liden K. Stereotaxic radiosurgery of the brain. Report of a case. Kungl Fysiogr Sallsk Lund Forhandl 1955;25:1-10.

73

35. Alexander E, Lindquist C. Special indications: radiosurgery for functional neurosurgery and epilepsy. In: Alexander E, Loeffler JS, Lunsford LD, editors. Stereotactic radiosurgery. New York: McGraw-Hill; 1993. p. 221-5. 36. Kihlstrom L, Hindmarsh T, Lax I, Lippitz B, Mindus P, Lindquist C. Radiosurgical lesions in the normal human brain 17 years after Gamma knife capsulotomy. Neurosurgery 1997;41:396-402. 37. Ruck C, Karlsson A, Steele JD, Edman G, Meyerson B, Ericson K, Nyman H, Asberg M, Svanborg P. Capsulotomy for obsessive-compulsive disorder. Long-term followup of 25 patients. Arch Gen Psych 2008;65:914-22. 38. Kondziolka D, Hudak R. Management of obsessivecompulsive disorder-related skin picking with gamma knife radiosurgical anterior capsulotomies: a case report. J Clin Psychiatry 2008;69:1337-40. 39. Barcia-Salorio JL, Barcia JA, Hernandez G, Lopez Gomez L. Radiosurgery of epilepsy. Long-term results. Acta Neurochir (Suppl) 1994;62:111-13. 40. Hellstrand DE, Abraham-Fuchs K, Jernberg B, Kihlstrom L, Knutsson E, Lindquist C, Schneider S, Wirth A. MEG localization of interictal epileptic focal activity and concomitant stereotactic radiosurgery. A non-invasive approach for patient with focal epilepsy. Physiol Meas 1993;14:131-6. 41. Maesawa S, Kondziolka D, Dixon E, Balzer J, Fellows W, Lunsford LD. Subnecrotic stereotactic radiosurgery controlling epilepsy produced by kainic acid injection in rats. J Neurosurg 2000;93:1033-40. 42. Mori Y, Kondziolka D, Balzer J, Fellows W, Flickinger JC, Lunsford LD, Thulborn KR. Effects of stereotactic radiosurgery on an animal model of hippocampal epilepsy. Neurosurgery 2000;46:157-68. 43. Mathieu D, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC. Gamma knife radiosurgery for epilepsy caused by hypothalamic hamartomas. Stereotact Funct Neurosurg 2006;84:82-7. 44. Re´gis J, Peragut JC, Rey M, Samson Y, Levrier O, Porcheron D, Regis H, Sedan R. First selective amygdalohippocampal radiosurgery for mesial temporal lobe epilepsy. Stereotact Funct Neurosurg 1995;64 Suppl 1: 193-201. 45. Regis J, Rey M, Bartolomei F, Vladyka V, Liscak R, Schrottner O, Pendl G. Gamma knife surgery in mesial temporal lobe epilepsy: a prospective multicenter study. Epilepsia 2004;45:504-15. 46. Bartolomei F, Hayashi M, Tamura M, Rey M, Fischer C, Chauvel P, Regis J. Long-term efficacy of gamma knife radiosurgery in mesial temporal lobe epilepsy. Neurology 2008;70:1658-63. 47. Mathieu D, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC. Gamma knife radiosurgery for epilepsy

1201

1202

73


casued by hypothalamic hamartomas. Stereotact Funct Neurosug 2006;84:82-87. 48. Whang CJ, Kim CJ. Short-term follow-up of stereotactic Gamma Knife radiosurgery in epilepsy. Stereotact Funct Neurosurg 1995;64 Suppl 1:202-8.

49. Witt TC, Kondziolka D, Baumann S, Noll D, Small S, Lunsford LD. Pre-operative cortical localization with functional MRI for use in stereotactic radiosurgery. Stereotact Funct Neurosurg 1996;66:24-9.

70 Radiosurgery for Metastases M. Maarouf . C. Bu¨hrle . M. Kocher . V. Sturm

Introduction Brain metastases are the most common intracranial neoplasms in adults. Between 20 and 40% of patients with systemic cancer develop brain metastasis over the course of their disease [1–4]. Certainly, the recognized incidence of brain metastasis has risen over the past 30 years, partly as a result of the advent of MRI, and partly as tumor patients have longer survival owing to more effective systemic therapies. Metastatic brain tumors derived from lung cancer are the most common type, making up 40–60% of the total, followed by those derived from breast cancer (15–20%) and melanoma (10–20%). Colorectal and renal cell carcinomas account for 5–10% each and unknown primary tumor for 15% of patients [5,6]. The prognosis of patients with brain metastases is generally poor. Without any treatment the median survival is 4 weeks [7]. The addition of steroids to reduce the edema induced by brain metastases improves this to 2–3 months [8]. External-beam radiotherapy applied to the whole brain further extends median survival to 4–6 months [8,9]. The choice of the therapeutic strategy for patients with intracranial metastases must be made not only with regard to expected survival times but also in terms of the quality of the remaining life span. The quality of life of these patients is nearly always significantly impaired by progressing focal neurological and neuropsychiatric symptoms, symptoms arising from raised intracranial pressure, symptoms caused by extracranial systemic disease, and/or psychological problems. Prolonged hospitalization #


associated with exhaustive treatment (surgery plus radiation therapy) can further impair the quality of life. The prognosis for patients with brain metastases is associated strongly with their recursive partitioning analysis (RPA) class. Using data from three Radiation Therapy Oncology Group (RTOG) brain metastases trials involving more than 1,200 patients treated by WBRT, RPA was performed to define prognostic factors [8]. The Karnofsky performance status (KPS), age, primary tumor status (controlled vs. uncontrolled), and extracranial metastases were the most relevant prognostic factors. On the basis of these factors, 3 prognostic classes were defined: RPA Class 1 (KPS 70%; age Table 70-1). In the first prospective randomized trial, published 1990 by Patchell et al. [10], 48 patients with known systemic cancer were treated with either biopsy of the suspected brain metastasis plus WBRT or complete surgical resection of the metastasis plus WBRT. The radiation doses were the same in both groups and consisted of a total dose of 36 Gy given as 12 daily fractions of 300 cGy each. There was a statistically significant increase in survival in the surgical group (9.2 vs.

3.5 months). In addition, the time to recurrence of brain metastases, freedom from death due to neurologic causes and duration of functional independence were significantly longer in the surgical resection group. The one month mortality was 4% in each group, indicating that there was no extra mortality from surgery. While the results obtained by Vecht et al. [12] are similar to those of the group of Patchell and confirm a significant benefit of surgery on overall survival and on quality of life, the results of Mintz et al. [11] were divergent with respect to benefit of surgery on overall survival (> Table 70-1). In the study of Mintz the benefit of surgery may be lost in patients with poor prognostic factors such as advanced extracranial disease or lower performance status. Of the patients in the study by Mintz et al. [11], 45% had extracranial metastases; in the trials by Patchell et al. [10] and by Vecht et al. [12] this number was only 37.5 and 31.7%, respectively. In the trial by Mintz et al. [11], 21% of patients had a KPS of 3.5 cm in diameter) and unresectable metastasis as primary treatment modality or salvage therapy for

. Table 70-1 Randomized trials of surgery plus WBRT as compared with WBRT alone Variables

Patchell et al. [10]

Vecht et al. [12]

Mintz et al. [11]

Treatment Patients (n) Eligibility criteria Steroids Median survival (months) Local recurence (%) Median functionally independent survival (months) CNS death (%)

WBRT Surg. + WBRT 23 25 KPS 70, age 18 All 3.5 9.2; p < 0.1 52 20; p < 0.02 1.8 8.8; p < 0.005

WBRT Surg. + WBRT 31 32 WHO PS 2, age 18 Most 6 10; p < 0.04 NR 3.5 7.5; p < 0.06

WBRT Surg. + WBRT 43 41 KPS 50, age < 80 All 6.3 5.6;p < 0.24 NR NR

50

33

63

29

35

46

WBRT = Whole-brain radiation therapy; Surg. = surgery; KPS = Karnofsky performance status; WHO PS = World Health Organisation performance status; NR = not reported

Radiosurgery for metastases

recurrent tumor after failed surgery or fractionated radiotherapy or radiosurgery. Our results (only available in abstract form) and the report of Ostertag and Kreth [13] demonstrate a high rate of local tumor remission with low radiation toxicity. Chemotherapy with newer drugs, such as oral temozolomide, is used occasionally as salvage therapy in those patients with recurrent non-germ cell and non-small cell metastases who, for a variety of reasons, cannot be treated surgically or radiosurgically. In most patients, however, chemotherapy yields little improvement in survival or quality of life and is usually a last resort after other therapies have failed [14].

Radiosurgery Compared to brain gliomas the intracranial metastases are usually more or less spherical, relatively small, and typically have discrete borders that are clearly visualized with cCT and MRI. On the basis of these facts, stereotactic radiosurgery is an appropriate treatment option for patients harboring single or few (up to 3–4) brain metastases. The goal of radiosurgery for the treatment of intracranial metastases is the complete destruction of tumor or at least a local control and improvement or complete remission of the tumor-related symptoms. Since the first published report on the efficacy of radiosurgery for brain metastases by Sturm et al. 1987 [15], ten thousands of patients have been treated worldwide either by Gamma Knife- or LINAC radiosurgery. Currently, radiosurgery is a widely accepted treatment modality for brain metastases. When used as initial therapy, radiosurgery is applied either as a monotherapy or as a boost with whole-brain radiotherapy. Radiosurgery has also been used as salvage treatment for progressive intracranial disease after whole-brain radiotherapy.

70

Numerous retrospective studies [16–21] have shown improved local control rates after radiosurgery (73–94%) when compared to whole brain radiotherapy (WBRT) alone (55– 60%) [9], and most of these studies also reported an increase in median survival from 3–6 months (WBRT) to 7–12 months (radiosurgery) for selected cases [14,17–24]. Two representative cases with complete or partial remission are shown in > Figures 70-1– > 70-2. For single metastases with a maximum diameter of approximately 3 cm, the optimum therapeutic dose is 20–24 Gy (radiosurgery alone) [9,23,25]. The incidence of side effects (symptomatic adjacent edema and/or radiation necrosis) is generally low, ranging between 2 and 5% in most series. A possible reason for the low incidence of radiation-induced side effects, in particular radiation necrosis, is that brain metastases are usually small and spherical, thus being more easily treatable with a conformal field than, e. g., in complex configurated arteriovenous malformation or skull base tumor. Voges et al. [26] have shown that the volume of perilesional brain tissue receiving 10 Gy and more is a major risk factor for radiation necrosis. The incidence of symptomatic radiation necrosis increases exponentially, if the volume of perilesional tissue receiving 10 Gy and more exceeds 10 cc. The better the dose conformation to the target volume, the smaller is the exposure of the surrounding normal brain tissue to radiation and to the risk of radiation necrosis.

Radiosurgery Versus Surgery and Whole Brain Radiotherapy For solitary metastasis, there are one prospective randomized study [27] and three retrospective trials comparing radiosurgery with surgery plus WBRT [28–30]. In the randomized study by Muacevic et al. [27] 64 patients harboring a single, resectable metastasis 3 cm in diameter with a Karnofsky performance score (KPS) 70, and

1141

1142

70


. Figure 70-1 MR images of a 61-year-old female with a single lung cancer (NSCC) metastasis in the insular area left-sided treated with SRS (target surface dose, 20 Gy). The images (left image before, right image 3 years after SRS) illustrate a complete tumor remission

. Figure 70-2 MR images of a 55-year-old patient with 2 cerebellar metastases of mamma carcinoma, left image before SRS (target surface dose, 17 Gy) and right image 21 months thereafter with a significant size reduction of both tumors (partial remission)

stable systemic disease were randomly assigned to microsurgery plus WBRTor Gamma Knife radiosurgery alone. Both treatment groups (31 SRS alone and 33 surgery and WBRT) did not differ in terms of age, KPS, tumor size, tumor location, site of the primary tumor, RPA classification, and

time to metastasis. Length of survival did not differ between the treatment groups: median survival was 9.5 months after surgery plus WBRT and 10.3 months after radiosurgery alone. The 1-year local tumor control rate was statistically not significant: 82% in the surgery group and


96.8% in the radiosurgery group. Patients of the radiosurgery group more often experienced distant recurrences (1-year distant recurrence rate: 25.8% vs. 3%); after adjustment for the effects of salvage radiosurgery this difference was lost. The retrospective study by Muacevic et al. [28] reviewed 108 patients with a single metastasis no larger than 3.5 cm in diameter and stable systemic disease who received SRS alone or surgery plus WBRT. Patients in the SRS group had significantly smaller tumors than did the patients in the surgery plus WBRT group (mean size: 2.07 vs. 2.7 cm). The SRS group also contained a higher proportion of patients with melanoma. Although median survival was 15.7 months in the surgery plus WBRT group and 8.1 months in the SRS group, the survival difference was not statistically significant. No significant differences in local control or complications were observed between the groups, but a higher incidence of distant recurrences was reported in the SRS group. According to these results the authors concluded that radiosurgery provided local tumor control rates as good as resection and WBRT in selected patients. The review by Scho¨ggl et al. [29] retrospectively matched 133 patients who received WBRT and either Gamma Knife SRS or surgery for the treatment of a single brain metastasis less than 3 cm in diameter. Median survival and 1-year overall survival did not differ significantly between the groups; however, the authors reported that SRS was superior in terms of local control and treatment-related morbidity. To be included in the review by O’Neill et al. [30], patients had to be candidates for both SRS and surgical resection. Tumor size had to be no larger than 3.5 cm in diameter, and patients with deep-seated tumors or ventricular obstruction were excluded. These inclusion criteria were met by 23 patients who had received SRS and 74 patients who had received surgery, most of whom had additionally been treated with WBRT. In the SRS group significantly fewer patients had

70

a good performance score. No significant differences in survival or cause of death were detected between the groups, and the authors concluded that neither SRS nor surgical resection was superior in that study.

Radiosurgery as Boost with Whole Brain Radiotherapy (WBRT) Versus WBRT Alone There have been three randomized trials [16,31,32] examining the use of whole-brain radiotherapy plus radiosurgery boost as compared with whole-brain radiotherapy alone in selected patients with brain metastases. Kondziolka et al. [32] reported the first trial on the subject. Patients with two to four brain metastases all 25 mm were randomized to whole-brain radiotherapy alone (30 Gy in 12 fractions) or whole-brain radiotherapy plus radiosurgery. The study was stopped at 60% accrual. Only 27 patients were randomized and the results were reported for 14 patients treated with whole-brain radiotherapy alone and 13 patients treated with whole-brain radiotherapy and additional radiosurgery boost. The two groups were well balanced with respect to age, sex, tumor type, number of tumors, and extent of extracranial disease. The largest randomized study to date has been reported by Andrews et al. [31], which randomized 164 patients to whole-brain radiotherapy and radiosurgery boost versus 167 patients randomized to whole brain radiotherapy alone. Patients with one to three newly diagnosed brain metastases were included. The brain metastases could have a maximum diameter of 4 cm for the largest lesion and the additional lesions could not exceed 3 cm (based on contrast-enhanced magnetic resonance imaging). The arms of the trial were well balanced for baseline characteristics known to affect survival such as age, Karnofsky performance status

1143

1144

70


(KPS) and status of extracranial metastases. The study was powered to establish possible differences in outcome and stratified for number of brain metastases (one vs. two to three) and extent of extracranial disease (none vs. present). The Chougule et al. trial [16] was published in abstract form. Patients with one to three brain metastases with tumor volume 30 cc and minimum life expectancy of 3 months were included. The randomization arms were Gamma Knife radiosurgery alone (30 Gy to the tumor margin), whole brain radiotherapy (30 Gy in 10 fractions) plus gamma knife boost (20 Gy to the tumor margin), and whole-brain radiotherapy alone (30 Gy in 10 fractions). Overall median survival was 7, 5, and 9 months for the arms, respectively (not significant). All three randomized trials [16,31,32] revealed an improvement in local brain tumor control in patients treated with whole-brain radiotherapy and radiosurgery boost as compared with whole-brain radiotherapy alone. But none of the three trials reports a statistically significant improvement in overall survival in the radiosurgery boost arms as compared with whole-brain radiotherapy alone. Median survival for the whole-brain radiotherapy alone arm ranged from 5.7 to 9 months. Median survival for the whole-brain radiotherapy and radiosurgery boost arms ranged from 5 to 11 months. However, the RTOG study, Andrews et al. [31], demonstrated that radiosurgery is beneficial in patients with single metastasis. The rates of local control and overall survival with radiosurgery and WBRT are comparable to the rates achieved in patients treated with conventional surgery plus WBRT. In contrast, other reports have questioned the usefulness of WBRT for patients having radiosurgery [14,33]. It has been argued that if patients develop additional brain tumors following radiosurgery, the procedure can be repeated to minimize the risk for delayed cognitive decline frequently noted after WBRT [34].

Radiosurgery or Whole Brain Radiotherapy The randomized trial by Chougule et al. [16] investigated this subject and was only reported in abstract form. The authors randomized patients to gamma knife radiosurgery alone versus whole-brain radiotherapy and gamma knife radiosurgery versus whole-brain radiotherapy alone (> Table 70-2). The local tumor control rate was higher in the two radiosurgery arms (87% for gamma knife radiosurgery alone and 91% for gamma knife radiosurgery and wholebrain radiotherapy) than in the whole-brain radiotherapy arm (62%). Median survival was 7, 5, and 9 months for gamma knife radiosurgery alone versus whole-brain radiotherapy and Gamma Knife radiosurgery versus whole-brain radiotherapy, respectively. Survival was not different among the three arms. Hasegawa et al. suggested that brain metastases are well controlled with SRS alone and that WBRT, in addition to SRS, may be omitted to reduce the risk of radiation-related toxicity [40]. Those authors suggested that SRS alone was a reasonable alternative to WBRT plus SRS. Sneed et al. suggested in two retrospective studies that the omission of WBRT in the initial management of patients who undergo radiosurgery does not appear to compromise survival [14,38]. However, the studies by Sneed et al. and Hasegawa et al. did not compare SRS alone with WBRT alone. Rades et al. [35] compared SRS alone with WBRT alone. They investigated whether stereotactic radiosurgery alone improved outcomes for patients in recursive partitioning analysis (RPA) Classes 1 and 2 who had 1–3 brain metastases compared with whole-brain radiotherapy. The authors found that SRS alone is associated with improved entire brain control and local control compared with 30–40 Gy of WBRT alone for patients in RPA Classes 1 and 2 who have 1–3 brain metastases, whereas overall survival and distant


control do not differ significantly. They concluded that SRS alone appears to be more effective than WBRT alone. Furthermore, it is less time consuming: SRS takes only 1 day, which means that the patients have to spend only very little time of their limited life span receiving treatment. For our institution [41], the survival times in patients with inoperable brain metastases treated by radiosurgery alone (117 patients, median age 60 years) were compared with those of a historical collective (138 patients, median age 58 years), treated by WBRT alone. Only patients with one to three metastases were included and treated with LINAC radiosurgery. Details of the radiosurgical procedure have been described before [41–43]. The most frequent primary tumor type in the SRS group was non-small-cell lung cancer (30%), followed by malignant melanoma (27%), renal cell carcinoma (13%), breast carcinoma (12%), and other types (18%). Patients with singular metastases in this group had either deep-seated tumors not suitable for resection or were referred for radiosurgery rather than surgery because of the physician’s or patient’s preference. In the WBRT group, non-small-cell lung cancer (28%) was also dominant, but breast cancer (19%) which is thought to have a slightly better prognosis [44] was more frequent (melanoma 6%, renal cell carcinoma 5%, others 42%). Patients with singular metastases in this group were either unresectable because of localization or size of the lesion. All patients in both groups were classified into the three RPA prognostic classes. In RPA class I patients (Karnofsky performance score 70, primary tumor controlled, no other metastases, age Table 72-1) [24–39]. Most studies reported a greater than 90% control of tumor size (range

89.8–100%). A weighted average adenoma control rate for all published nonsecretory adenoma series detailing such findings and encompassing a total of 696 patients was 94.8%. Tumor growth cessation, not amount of volume reduction, is still considered successful treatment. Some series have even demonstrated improvement in visual function following radiosurgery related to shrinkage of the tumor [28,29,32,35,40,78,79]. Most pituitary adenomas are slow growing lesions. As such, it is important to perform longterm neuro-imaging follow-up of radiosurgical patients. Such follow-up will help to differentiate between the natural history of an adenoma and true radiosurgical induced volume control. It has been our experience that the longer the follow-up the more likely a patient’s pituitary

. Table 72-1 Radiosurgery for patients with nonfunctioning pituitary adenomas

Authors (year)

Radiosurgery unit

No. of patients

Mean or median follow-up (months)

Martinez et al. [24] Lim et al. [25] Mitsumori et al. [26] Witt et al. [27] Yoon et al. [28] Hayahsi et al. [29] Inoue et al. [30] Mokry et al. [31] Izawa et al. [32] Shin et al. [33] Feigl et al. [34] Sheehan et al. [35] Wowra and Stummer [36] Petrovich et al. [37] Muramatsu et al. [38] Pollock and Carpenter [39] Losa et al. [40] Muacevic et al. [41] Picozzi et al. [42] Kajiwara et al. [43] Iwai et al. [44] Mingione et al. [45]

GK GK LINAC GK LINAC GK GK GK GK GK GK GK GK GK LINAC GK GK GK GK CK GK GK

14 22 7 24 8 18 18 31 23 3 61 42 30 56 8 33 54 51 51 14 28 100

36 26 47 32 49 16 >24 21 28 19 55 31 58 41 30 43 41 21.7 40.6 35.3 36.4 44.9

Maximal dose (Gy)

Margin dose (Gy)

Growth control (%)

28 48 19 38 21 NR 43 28 NR NR NR 32 29 30 26.9 36 33 NR NR NR NR 41.5

16 25 15 19 17 20 20 14 22 16 15 16 16 15 15 16 17 16.5 16.5 12.6 12.3 18.5

100 92 100 94 96 92 94 98 94 100 94 98 93 100 100 97 96 95 89.8 93 93 92

Note: NR, not reported; GK, Gamma Knife radiosurgery; LINAC, linear accelerator radiosurgery unit; CK, Cyberknife

Radiosurgery for pituitary tumors

72

. Table 72-2 Radiosurgery for patients with Cushing’s disease Mean or median followup (months)

Endocrine remission rate (%)

Endocrinological criteria for remission

Maximal dose (Gy)

Margin dose (Gy)

NR

150

NR

86

4

18

58.25

25

50

GK

3

36

40

24

100

GK LINAC

4 5

26 47

48 19

25 15

25 40

GK

6

20

NR

28

67

GK LINAC

25 1

32 49

38 21

19 17

28 NR

UFC < 90 mg/24 h; normal ACTH and cortisol Normal 24 h UFC NR

GK

10

16

NR

24

10

NR

GK

3

>24

43

20

100

NR

GK

8

27

55

29

62

GK GK

50 5

NR 56

NR 35

NR 17

58 33

UFC < 100 mg/24 h Normal 24 h UFC NR

GK

12

28

NR

22

17

NR

GK

43

44

47

20

63

Normal 24 h UFC

GK GK

7 18

88 204

NR 60–240

32 NR

50 83

UFC < 90 mg/24 h Normal 24 h UFC

GK GK

4 20

55 64

NR 49

15 29

NR 35

GK

9

42

40

20

78

NR ACTH < 50 pg/mL; cortisol < 10 mg/dL UFC < 90 mg/24 h

LINAC

5

38

16–20

14.8– 19.2

80

Authors (year)

Radiosurgery unit

No. of patients

Levy et al. [46]

Proton/helium beam

64

Ganz et al. [47]

GK

Martinez et al. [24]

Lim et al. [25] Mitsumori et al. [26] MorangeRamos et al. [48] Witt et al. [27] Yoon et al. [28] Hayashi et al. [29] Inoue et al. [30] SH Kim et al. [49] Laws et al. [50] Mokry et al. [31] Izawa et al. [32] Sheehan et al. [4] Shin et al. [33] Hoybye et al. [51] Feigl et al. [34] Kobayashi et al. [52] Pollock et al. [18] Wong et al. [53]

Normal basal cortisol and dexamethasone test AM UC < 650 nmol/24 h and PM UC < 250 nmol/ 24 h ACTH < 10 microg/L; UFC < 650 nmol/ 24 h NR NR

Normal urinary cortisol and dexamethasone suppression test

1175

1176

72


. Table 72-2 (continued) Mean or median followup (months)


Maximal dose (Gy)

Margin dose (Gy)

41

30

15

50

5 35

42.5 42

54.1 33.7

28.5 14.7

56 49

CK

2

35.3

NR

17.5

50

GK

40

54.7

NR

29.5

42.5

GK

90

45

49

23

54

Authors (year)

Radiosurgery unit

Petrovich et al. [37]

GK

Choi et al. [54] Devin et al. [55] Kajiwara et al. [43] Castinetti et al. [56]

GK LINAC

Jagannathan et al. [57]

No. of patients 4

Endocrinological criteria for remission Normal serum cortisol, ACTH, and 24-h UF cortisol UFC < 90 mg/24 h Normal cortisol Normal ACTH and cortisol Normal UFC and dexamethasone suppression test Normal 24 h UFC

Note: NR, not reported; GK, Gamma Knife radiosurgery, LINAC, linear accelerator radiosurgery unit; CK, Cyberknife

adenoma falls into one of two categories—tumor enlargement or tumor reduction. No growth may simply be a limitation of inadequate follow-up following radiosurgery.

Radiosurgery for Cushing’s Disease Cushing’s disease, perhaps the most famous of pituitary disorders, was described by Harvey Cushing in 1912 as a polyglandular disorder [80]. It was not until 1933 that Cushing first performed neurosurgery to treat a patient with a basophilic pituitary adenoma presumed to be secreting excess ACTH [80]. Over the years, neurosurgeons and endocrinologists have debated the criteria for defining ‘‘cure’’ for Cushings disease [81]. Many advocate the use of a 24-h urine free cortisol (UFC) determination as the ‘‘gold standard.’’ Others measure ACTH or basal serum or salivary cortisol and factor these into the evaluation of endocrinological success or failure in Cushing’s disease. In fact in a recent consensus statement by leading endocrinologists, there was no consensus

regarding the definition of endocrinological cure, and the remission rates vary according to the criteria used and the time assessed [82]. Most centers define an endocrinological remission as a 24-h UFC in the normal range coupled with the resolution of clinical stigmata, or a series of normal post-operative serum cortisol levels obtained throughout the day (range 5.4–10.8 mg/dL or 150–300 nmol/L) [82]. Twenty-seven series have reported the results for 486 patients with Cushing’s disease treated surgically with postoperative radiosurgery (> Table 72-2) [2,5,18,24–26,28–34, 37,46,48,51,52,54,83,84]. Mean radiosurgical margin doses for these series range from 15 to 32 Gy. Fifteen series utilized the urine cortisol collection as part of the criteria for endocrinological evaluation. Unfortunately, another eight of these studies do not report the methodology employed to establish endocrinological remission or failure. The others utilize a combination of the aforementioned endocrinological tests. Endocrinological remission rates vary from 10 to 100%. In those series with at least ten patients and a median follow-up of 2 years, endocrinological


72

. Table 72-3 Radiosurgery for Patients with Acromegaly

Maximal dose (Gy)

Margin dose (Gy)



Authors (year)

Radiosurgery unit

Ganz et al. [47] Martinez et al. [24] Landolt et al. [58]

GK

4

18

54.25

19.5

25

GH < 5 mIU/L

GK

7

36

39

25

71

Normal IGF-1

GK

16

NR

50

25

81

GK LINAC

20 1

26 47

48 19

25 15

38 0

GH < 10 mIU/L and IGF-1 < 50 mIU/L GH < 2 ng/mL NR

GK

15

20

NR

28

20

GH < 5 ng/mL; normal IGF-1

GK LINAC

20 2

32 49

38 21

19 17

20 50

Normal IGF-1 GH < 5 ng/mL

GK

22

16

NR

24

41

NR

GK

12

>24

43

20

58

NR

GK

2

12

36

22

0

NR

GK

11

27

55

29

46

GH < 5 ng/mL

GK

56

NR

NR

NR

25

Mokry et al. [31]

GK

16

46

33

16

31

Izawa et al. [32] Shin et al. [33]

GK

29

28

NR

22

41

Normal IGF-1 for gender and age GH < 7 ng/mL; IGF-1 < 380 IU/ mL NR

GK

6

43

NR

34

67

Zhang et al. [60] Fukuoka et al. [61] Ikeda et al. [62] Feigl et al. [34] Pollock et al. [18]

GK

68

34

NR

31

96

GK

9

42

41

20

50

GK

17

48

NR

25

82

GK GK

9 26

55 42

NR 40

15 20

NR 42

GK

30

46

40

20

37

Lim et al. [25] Mitsumori et al. [26] MorangeRamos et al. [48] Witt et al. [27] Yoon et al. [28] Hayashi et al. [29] Inoue et al. [30] MS Kim et al. [59] SH Kim et al. [49] Laws et al. [50]

Attanasio et al. [63]

No. of patients

Mean or median followup (months)

GH < 10 mIU/L and IGF-1 < 450 ng/mL GH < 12 ng/mL GH < 5 ng/mL and normal IGF-1 Normal IGF-1 for age NR GH < 2 ng/mL and normal IGF-1 for age GH < 2.5 microg/ L

1177

1178

72


. Table 72-3 (continued)

Maximal dose (Gy)

Margin dose (Gy)



Authors (year)

Radiosurgery unit

Petrovich et al. [37] Muramatsu et al. [38] Choi et al. [54] Gutt et al. [64] Kajiwara et al. [43] Castinetti et al. [65]

GK

6

41

30

15

100

LINAC

4

30

24.5

15

50

Normal IGF-1 and GH NR

Koybayashi et al. [66] Jezkova et al. [67] Pollock et al. [68] Roberts et al. [69]

No. of patients

Mean or median followup (months)

GK GK CK

12 44 2

42.5 22.8 35.3

54.1 36 NR

28.5 18 17.5

50 48 0

GH < 5 mIU/L Normal IGF-1 NR

GK

82

49.5

NR

25

17

GK

67

63.3

35.3

18.9

GH < 2 ng/mL and normal IGF-1 for age GH < 1 ng/mL

GK

96

43.2

70

35

50

GK

46

63

43.5

20

60

CK

9

25.4

NR

21

44.4

4.8

GH < 2.5 mg/L and normal IGF-1 GH < 2 ng/mL and normal IGF-1 for age Normal IGF-1

Note: NR, not reported; GK, Gamma Knife radiosurgery; LINAC, linear accelerator radiosurgery unit; CK, Cyberknife

remission rates range from 17 to 83%. The wide range of endocrinological remission rates is likely a function of the testing utilized and the practice of some centers to perform testing while patients remain on steroid synthesis inhibiting agents (e.g. ketoconazole). Rare cases of delayed recurrence following initial radiosurgical remission have been observed at the University of Virginia. This only stands to reason as such delayed recurrences can also occur after microsurgery. Delayed recurrence underscores the need for long-term periodic endocrinologic testing.

Radiosurgery for Acromegaly Just as the endocrinological criteria for Cushing’s disease remain the subject of debate, the criteria for curing acromegaly have also been inconsistent. The most widely accepted guidelines for

remission in acromegaly consist of a GH level less than 1 ng/mL in response to a glucose challenge and a normal serum IGF-1 when matched for age and gender [85–88]. Thirty-two studies detail the results of radiosurgical treatment for 766 patients with acromegaly (> Table 72-3) [2,18,24–34, 37,38,47–49,54,58– 63]. The mean radiosurgery margin doses in these series range from 15 to 35 Gy. Eight studies did not report the criteria utilized to define an endocrinological remission. Of the remaining twenty-four studies, a myriad of criteria are employed to define remission. Remission rates following radiosurgery vary from 0 to 100%. In those series with at least ten patients and a median follow-up of 2 years, endocrinological remission rates still vary with a wide range (4.8–96%). Certainly, some of the variation in endocrinological remission rates with acromegaly may be attributed to the inconsistent criteria defining remission. Another confounding


72

. Table 72-4 Radiosurgery for patients with prolactinomas Mean or median follow-up (months)

Maximal dose (Gy)

Margin dose (Gy)


20

12

150

NR

60

3

18

40

13.3

0

Authors (year)

Radiosurgery unit

Levy et al. [46] Ganz et al. [47]

Proton/ helium beam GK

Martinez et al. [24] Lim et al. [25] Mitsumori et al. [26] Witt et al. [27] Yoon et al. [28] Hayashi et al. [29] Inoue et al. [30] MS Kim et al. [59] SH Kim et al. [49] Laws et al. [50] Mokry et al. [31] MorangeRamos et al. [70] Izawa et al. [32] Landolt et al. [16]

GK

5

36

43

33

0

GK

19

26

48

25

56

4

47

19

15

0

NR

GK

12

32

38

19

0

NR

LINAC

11

49

21

17

84

24

43

20

50

NR

GK

20

12

36

22

19

NR

GK

18

27

55

29

17

74-6). This is similar to the evolution of indications for radiosurgery for intracranial lesions that occurred during the last

1211

1212

74

Whole body and spinal radiosurgery

. Figure 74-5 Case example of a 42-year-old man with a painful colon metastasis of the L2 vertebral body. He had received prior conventional fractionated irradiation to the lesion with temporary improvement of symptoms. The treatment plan was designed to treat the tumor with a prescribed dose of 18 Gy in a single fraction that was calculated to the 80% isodose line; the maximum tumor dose was 22.5 Gy (CyberKnife Radiosurgical System, Accuray, Inc., Sunnyvale, CA). Axial and sagittal images of the treatment plan (a and b). The tumor volume was 58.8 cm3 and the cauda equina received a maximum dose of 12 Gy (bold red line = contoured tumor; light orange line = 80% isodose line; yellow line = 60% isodose line; dark purple line = 40% isodose line). The dose-volume histogram (DVH) is shown (c). Notice the conformality of the isodose line around the cauda equina


74

. Figure 74-6 Case example of a 57-year-old woman with progression on MR imaging of a T4 breast metastasis after prior conventional fractionated irradiation treatment. The prescribed dose to the planned tumor volume (green line) was 13.5 Gy using 12 co-planar beams (Synergy S, Elekta Inc., Atlanta, GA). Axial and sagittal images of the treatment plan (a and b). The gross tumor volume was 19.3 cm3 and the spinal cord received a maximum dose of 9 Gy (red line). Digitally reconstructed radiograph of one coplanar beam demonstrating the position of the leaves of the multileaf collimator. Cone beam images obtained during patient setup and positioning for treatment provide extremely high spatial resolution of both bony structures as well as soft tissue, making possible the setup of sites with submillimeter targeting errors (d and e)

1213

1214

74


decade. > Table 74-1 summarizes the candidate lesions for spine radiosurgery. > Table 74-2 summarizes some of the primary indications for spine radiosurgery. Spine radiosurgery can deliver radiation to anywhere along the spine, both extradural as well as intradural. Candidate lesions may be those that would require difficult surgical approaches for adequate resection. Candidate patients may have significant medical comorbidities precluding open surgical intervention or a relatively short life expectancy that would deem them inappropriate for open surgical intervention. Another indication would be to halt tumor progression that might lead to spinal instability or neural compromise. Radiosurgery as initial therapy may decrease the need for reirradiation by improving tumor control compared to conventional techniques.

. Table 74-1 Candidate lesions for spine radiourgery Minimal spinal cord compromise Previously irradiated lesions Radioresistant lesions that would benefit for a radiosurgical boost Residual tumor after surgery Recurrent tumor after prior surgical resection Lesions requiring difficult or morbid surgical approaches Relatively short life-expectancy as an exclusion criteria for open surgical intervention Significant medical co-morbidities precluding open surgical intervention No overt spinal instability

. Table 74-2 Indications for spine radiosurgery Pain Primary treatment modality Prevention of tumor progression Radiation boost for radioresistant tumors Progressive neurologic deficit Treatment of residual tumor after surgery Postsurgical tumor progression

Pain The most frequent indication for the treatment of spinal tumors is pain, and pain was the primary indication for spinal radiosurgery in 70% of our cases. Radiation is well known to be effective as a treatment for pain associated with spinal malignancies. Conventional external beam irradiation may provide less than optimal pain relief since the total dose is limited by the tolerance of adjacent tissues (e.g., spinal cord). Spinal radiosurgery was found to be highly effective at decreasing pain in this difficult patient population, with an overall long-term improvement of pain in 374 of the 435 cases (86%), depending upon primary histopathology. Durable pain improvement was demonstrated in 96% of women with breast cancer, 96% of cases with melanoma, 94% of cases with renal cell carcinoma, and 93% of lung cancer cases [82–85]. Pain usually decreases within weeks after treatment, and occasionally within days. Spinal radiosurgery is also effective at alleviating radicular pain caused by tumor compression of adjacent nerve roots. In some cases, post-treatment imaging revealed pathologic fractures, likely the cause of pain and the reason for radiosurgical failure. Such fractures require either open or closed internal fixation to alleviate the pain due to spinal instability. The most extensive published work to date on pain control and quality of life improvement after spinal radiosurgery has been from Georgetown University Hospital [23,37]. Using visual analog scales (VAS) for pain assessment and the 12-item Short Form Health Survey (SF-12), radiosurgery was demonstrated to statistically improve pain control and maintain quality of life with follow-up to 24 months. Early adverse events in their experience were infrequent and minor. The Memorial Sloan Kettering group demonstrated a 90% excellent palliation of symptoms with a median follow-up of 12 months [65]. Other series report similar pain improvement results [17,19,21,22,25,26,86].


Radiographic Tumor Progression Spine radiosurgery is frequently used to treat radiographic tumor progression after conventional irradiation treatment or after prior surgery. The majority of these lesions have received irradiation with significant spinal cord doses, precluding further conventional irradiation delivery. Currently, spine radiosurgery is often being used as a ‘‘salvage’’ technique for those cases in which further conventional irradiation or open surgery are not appropriate. The ideal lesion should be well circumscribed such that the lesion can be easily outlined (contoured) for treatment planning. Overall long-term radiographic tumor control for progressive spinal disease in a series of 500 cases was 88% [46]. Radiographic tumor control differed based upon primary pathology: breast (100%), lung (100%), renal cell (87%) and melanoma (75%). Yamada et al. [65] reported a 90% long term radiographic control rate. Similar radiographic control rates have been reported by others [17,20–22,37,64]. As greater experience is gained, the technique will likely evolve into an initial upfront treatment for spinal metastases in certain cases (e.g., oligometastases). This is similar to the evolution that occurred for the treatment of intracranial metastases using radiosurgery that occurred over the past decade.

Primary Treatment Modality or Boost Radioresistant lesions (e.g., renal cell, sarcoma) that have completed external beam irradiation may undergo radiosurgery as a boost treatment. Another option may be to use radiosurgery as the sole radiation treatment. This clinical scenario is often encountered in a patient undergoing treatment to a symptomatic spine lesion with other significant but asymptomatic spine metastases. The additional asymptomatic lesions may be

74

treated with radiosurgery to avoid further irradiation to the neural elements as well as to avoid further bone marrow suppression and permit subsequent systemic therapy. The benefits for this approach include a single treatment which is radiobiologically larger than can be delivered with standard radiotherapy, with minimal radiation dose to adjacent normal tissue. When used as a primary treatment modality, long-term radiographic tumor control was demonstrated in 90% of cases (in all breast, lung and renal cell carcinoma metastases, and 75% of melanoma metastases) [46,82,83]. Degen et al. [37] reported a 100% tumor control rate in lesions that had not previously undergone irradiation. Radioresistant tumors (e.g., renal cell carcinoma, melanoma, sarcoma) may be treated with spinal radiosurgery after conventional irradiation as a boost treatment with excellent long-term radiographic control. Ryu et al. [26,65] found this to be a highly effective treatment paradigm.

Progressive Neurological Deficit Spine radiosurgery may be used to treat patients with progressive neurological deficits when open surgical intervention is felt to be contraindicated. In most of these particular cases, conventional irradiation has already been delivered to the symptomatic spinal lesion. In our experience, 36 of 42 patients (86%) with a progressive neurological deficit prior to treatment experienced at least some clinical improvement [46]. In most of these cases, open surgical decompression was precluded because of medical co-morbidities. Yamada reported 90% and 92% palliation of symptoms in patients treated for weakness and paresthesias, respectively [65]. Degen reported neurological deficits improved in 16 patients, were unchanged in 24 patients, and worsened in 11 patients in their series [37]. In a series from the Henry Ford Hospital, 12 of 16 cases (75%) with neurologic deficit from spinal

1215

1216

74


cord compression were clinically improved or stable improved after spine radiosurgery (unpublished data).

After Open Surgery If a tumor is only partially resected during an open surgery, radiosurgery may be used to treat the residual tumor at a later date. The spinal tumors can be removed away from neural structures allowing for immediate decompression, the spine can be instrumented if necessary, and the residual tumor can be safely treated at a later date with radiosurgery, thus further decreasing surgical morbidity. Anterior corpectomy procedures in certain cases can be successfully avoided by posterior decompression and instrumentation alone followed by radiosurgery to the remaining anterior lesion. With the ability to effectively perform spinal radiosurgery, the current surgical approach to these lesions might change. Given the steep falloff gradient of the target dose with negligible skin dose, such treatments can be given early in the postoperative period as opposed to the usual significant delay before standard external beam irradiation is permitted. Open surgery for spinal metastases will likely evolve in a similar manner in which malignant intracranial lesions are debulked in such a way as to avoid neurological deficits and minimize surgical morbidity. Rock et al. [87] specifically evaluated the combination of open surgical procedure followed with adjuvant radiosurgery. They found this to be a successful treatment paradigm that was associated with a significant chance of stabilizing or improving neurological function. The technique was well tolerated and associated with little to no morbidity. Our institution has also found radiosurgery combined with open surgery to be a safe and highly successful treatment of the residual tumor bed [46].

Intradural Benign Tumors There is less clinical experience with radiosurgery for benign tumors of the spine than for metastatic lesions. Benign spinal tumors represent a group of intradural extramedullary neoplasms that include meningiomas, schwannomas, and neurofibromas. Microsurgical resection is the primary established treatment for these benign spinal tumors and most spinal meningiomas, schwannomas, and neurofibromas are non-infiltrative. Recurrence is unlikely with complete extirpation [36,88]. However, spinal radiosurgery may be a reasonable alternative treatment in certain clinical scenarios for patients who are less than ideal candidates for open resection because of age, medical comorbitidies, and recurrent tumor [89]. Multiple benign spinal tumors, as are common in familial neurocutaneous disorders, may be another pattern of spinal pathology better suited for the less invasive radiosurgical option. Finally, tumors that have recurred after open surgical resection may make safe surgical resection challenging or not possible. Radiosurgery has been used to successfully treat benign intradural extramedullary lesions with excellent long-term radiographic response, similar to the experience for intracranial radiosurgery [89]. In our institution’s experience with great than one hundred benign intradural extramedullary spinal tumors, long-term pain improvement was demonstrated in 73% of cases and long-term radiographic tumor control was demonstrated in all cases. While surgical extirpation remains the primary treatment option for most benign spinal tumors, radiosurgery has been demonstrated to have both short-term as well as long-term clinical benefits for the treatment of such lesions. Its role in patients with neurofibromatosis will also be further defined with greater clinical experience. Most importantly, spine radiosurgery has been found to be extremely safe at the doses currently used, even for intramedullary lesions. Although a risk for


secondary malignant transformation with radiosurgery of benign tumors is theoretically possible, such a case has never been reported.

Meningiomas Spinal meningiomas are arachnoid cap cellderived tumors of the 5th–7th decade which have a female predominance and occur mainly in the thoracic region. Gross total surgical resection optimizes outcome [90]. Using the CyberKnife, Dodd et al. treated 16 spinal meningiomas (mean dose 20 Gy, mean tumor volume 2.4 cm3, mean follow-up 27 months) and demonstrated radiologic stabilization in 67% and radiologic tumor decrease in 33% of the 15 who had radiographic follow-up. Most patients experienced an improvement in pain and strength with radiosurgery [89]. From our institution’s published series, 13 spinal meningiomas were treated using a singlefraction technique (mean dose 21 Gy, mean tumor volume 4.9 cm3). Eleven of 13 patients had radiosurgery as an adjunctive treatment for residual or recurrent tumor following open surgical resection. Radiographic tumor control was demonstrated in all cases with a median follow-up of 17 months [91].

Schwannomas Schwannoma is the most common spinal tumor and has no proclivity for spinal region or gender [92]. The Stanford series comprised 30 tumors (mean dose 19 Gy, mean tumor volume 5.7 cm3, mean follow-up 26 months) and all but one patient had radiographic tumor control after radiosurgery; one-third of patients reported improvement in pain, weakness, or sensation, but 18% had a clinical decline after treatment [89]. Forty percent of patients in this series had spinal schwannomas in the setting of NF2. From our institution, 35 schwannomas have been treated

74

with spine radiosurgery (mean dose 22 Gy, mean tumor volume 11.0 cm3). Fourteen of seventeen patients (82%) described significant pain improvement for whom pain was the primary indication. Radiographic tumor control was demonstrated in six of seven patients (86%) for which radiosurgery was utilized as the primary treatment. A total of 3 patients went on to undergo upon surgical resection for new or persistent neurological deficits [91].

Neurofibromas Neurofibromas of the spine are often multiple, predominate in the cervical region, and are commonly associated with NF1. At Stanford, 9 neurofibromas in 7 patients with NF1 (mean dose 11 Gy, mean tumor volume 4.31 cm3, mean follow-up 19.9 months) were treated with radiosurgery and tumor stabilization on imaging was documented in six of seven (86%) patients; however, no patients described an improvement in symptoms after radiosurgery and half of the patients complained of a worsening in pain, weakness, or numbness at their last follow-up [89]. Our institution reported 25 neurofibroma cases (mean dose 21 Gy, mean tumor volume 12.6 cm3) [91]. No patient had evidence of radiographic tumor progression on follow-up. Twenty-one of these patients had NF1 and 9 had NF2. Radiosurgery ameliorated discomfort in 8 of 13 (62%) patients treated for pain; however all patients who saw no improvement in pain had NF1. These findings echo outcomes from the Stanford report that found pain control in NF1-associated spinal neurofibromas to be recalcitrant to radiosurgery [89]. Poorer microsurgical results for neurofibromas have also been observed in patients with NF1 [93]. The multiplicity of neurofibromas in NF1 may be partially to blame as this factor makes identifying the symptomatic neurofibroma in need of treatment more difficult. Moreover, the infiltrating of

1217

1218

74


neurofibromas, in contrast to the other benign extramedullary intradural spinal tumors, may engender more irreversible neural damage and increase the susceptibility of the native nerve root to injury from both microsurgical and radiosurgical treatments. Finally, future genomic investigations may reveal that intrinsic genetic differences in NF1-associated neurofibromas predispose to a weaker radiobiologic response.

Arteriovenous Malformations Stereotactic radiosurgery has emerged over the last three decades as a successful alternative to microsurgical resection and embolization for cerebral arteriovenous malformations. Since this technique was first described by Steiner in 1972 [94], over 5,000 patients have successfully been treated with this modality. Radiosurgery causes gradual hyperplasia of the endothelial tissue within the arteries of the nidus of the AVM which leads to vessel occlusion [95]. Numerous reports show that cerebral AVMs have an 80–85% obliteration rate for those vascular malformations smaller than 2.5 cm [45,96–101]. With the success with radiosurgery in the management of brain AVMs, the treatment of spinal cord AVMs with radiosurgery would be a logical next step. Spinal arteriovenous malformations have been successfully treated with radiosurgery [102]. Given the various subtypes of spinal cord AVMs, those with a relatively compact nidus would represent the optimal targets [103]. Type I and IV spinal cord AVMs are dural and perimedullary arteriovenous fistulas, and are often optimally treated with endovascular embolization and or microsurgical resection. Type III spinal cord AVMs, also called juvenile type spinal AVMs, are characterized by large and diffuse intramedullary nidus which can also extend into the extramedullary space. They are less well defined spinal cord AVMs, and also do not represent optimal

radiosurgical targets. Type II spinal cord AVMs, also called glomus AVMs, represent a compact vascular nidus and are often suitable radiosurgery targets. These type II spinal cord AVMs are also difficult if not impossible to treat with embolization and microsurgery alone and often were not treated prior to the development of spinal radiosurgery. The largest experience with spinal AVM radiosurgery comes from the Stanford series [103]. Twenty-three patients with spinal cord AVMs have been treated at Stanford using the Cyberknife technology. Twenty-two of these patients had type II or glomus AVMs while one patient had a type III juvenile spinal cord AVM. Spinal AVM locations included the cervical spine (12 patients), thoracic spine (eight patients), and conus region (three patients). Thirteen of these twenty-three patients (57%) presented with hemorrhage, which was multiple in six of these patients. In the remaining 10 patients, progressive neurologic dysfunction was the presenting symptom. Ten of the 23 patients (43%) underwent prior embolization to their AVM. A median of three treatment sessions was performed. Mean target volume was 2.8 cm3 with a range of 0.26–15 cm3. The mean marginal dose used was 20.3 Gy with a range of 16–21 Gy. A median of two sessions was utilized in this cohort with a range of one to four treatments. The mean maximal internidal dose was 25.8 Gy and ranged from 22.5 to 30 Gy. Clinical follow-up ranged from four to 83 months with a mean of 35 months. Radiographic follow-up averaged 25 months. Although postoperative MRI imaging demonstrated noticeable reduction in the volume of all AVMs over the course of follow-up, only eight of the 23 patients underwent formal spine angiography. Of these eight patients, three patients had complete angiographic obliteration. To date, no patients suffered a rebleed after spine radiosurgery and the clinical outcome was improved or unchanged in 96% of patients. A single patient


who was neurologically severely impaired prior to radiosurgery deteriorated further at 9 months following radiosurgery. MR imaging demonstrated a significant decrease in flow void as well as high signal in the adjacent spinal cord on T2 imaging consistent with radiation induced edema. One other patient experienced significant onset of radiation edema (for a conus AVM) but this AVM obliterated rapidly over 9 months and the patient was one of the three patients who was documented to have complete obliteration of his AVM. The future of radiosurgery for the treatment of spinal AVM has still not been determined.

Potential Advantage of Spine Radiosurgery Spine radiosurgery has several advantages over other treatment alternatives. For patients with intradural tumors, there are obvious advantages to this treatment over the potential complications associated with open surgical techniques. Spine radiosurgery avoids the need to irradiate large segments of the spinal cord as well as delivery of a minimal radiation dose to adjacent normal tissue. Early stereotactic radiosurgery treatment of spinal lesions may obviate the need for extensive spinal surgeries for decompression and fixation in these often debilitated patients. A much larger radiobiologic dose can be delivered compared to external beam irradiation. It may also avoid the need to irradiate large segments of the vertebral column, known to have an important deleterious effect on bone marrow reserve in these cancer patients who frequently require bone marrow suppressive systemic therapy. Avoiding open surgery as well as preserving bone marrow function facilitates continuous chemotherapy in patients with cancer. Furthermore, improved local control, such as has been the case with intracranial radiosurgery. Could translate into more effective palliation and potentially longer survival. With greater

74

clinical experience, upfront radiosurgery perhaps will become more commonly used in certain cases such as patients with single symptomatic spine metastases of a radioresistant histology. Radiosurgery allows for the treatment of lesions previously irradiated using conventional external beam techniques. An advantage to the patient of using single-fraction radiosurgery is that the treatment can be completed in a single day rather than over a course of several weeks, which may be beneficial for patients with a limited life expectancy from cancer. In addition, cancer patients may have difficulty being transported to a radiation treatment facility for prolonged, daily fractionated therapy. A large single fraction of irradiation may be more radiobiologically advantageous to certain tumors such as renal cell carcinoma compared to prolonged fractionated radiotherapy. The radiobiology of large fraction radiotherapy to certain tumor types is currently under investigation. Clinical response such as pain or improvement of a neurological deficit might also be more rapid with a radiosurgery technique. This rapid clinical response is becoming welldocumented in the peer-reviewed literature [19,37]. Finally, the procedure is minimally invasive compared to open surgical techniques and can be performed in an outpatient setting.

Summary Spine radiosurgery is a feasible, safe and clinically effective for the treatment of a variety of spinal tumors using a wide variety of different technologies. Spine radiosurgery represents a logical extension of the current state-of-the-art radiation therapy. It has the potential to substantially (would only use the word significantly if we are speaking of statistics) improve local control of cancer of the spine, which could translate into better palliation. The major potential benefits of radiosurgical ablation of spinal lesions are relatively short treatment time in an outpatient

1219

1220

74


setting combined with potentially better local control of the tumor with minimal risk of side effects, however, further studies confirming these presmed benefits are needed. Spine radiosurgery offers a new and important alternative therapeutic modality for the treatment of spinal tumors in patients who are medically inoperable or who are poor surgical candidates, have had prior irradiation to their tumor, have lesions not amenable to open surgical techniques or as an adjunct to surgery.

References 1. Horn E, Henn J, Lemole J, GM. Thoracoscopic placement of dural-rod instrumentation in thoracic spinal trauma. Neurosurgery 2004;54:1150-4. 2. Yamada Y, Lovelock D, Bilsky M. A review of imageguided intensity-modulated radiotherapy for spinal tumors. Neurosurgery 2007;61:226-35. 3. Clain A. Secondary malignant disease of bone. Br J Cancer 1965;19:15-29. 4. Black P. Spinal metastasis: current status and recommended guidelines for management. Neurosurgery 1979;5:726-46. 5. Gokaslan Z, York J, Walsh G. Transthoracic vertebrectomy for metastatic spinal tumors. J Neurosurgery 1998;89:599-609. 6. Gerszten PC, Welch WC. Current surgical management of metastatic spinal disease. Oncology 2000;14:1013-36. 7. Faul CM, Flickinger JC. The use of radiation in the management of spinal metastases. J Neurooncol 1995;23:149-61. 8. Kim YH, Fayos JV. Radiation tolerance of the cervical spinal cord. Radiology 1981;139:473-8. 9. Markoe AM, Schwade JG. The role of radiation therapy in the management of spine and spinal cord tumors. In: Rea GL, editor. Spine tumors. American Association of Neurological Surgeons; 1994. p. 23-35. 10. Shapiro W, Posner JB. Medical vs surgical treatment of metastatic spinal cord tumors. In: Thompson R, Green J, editors. Controversies in Neurology. New York: Raven Press; 1983. p. 57-65. 11. Sundaresan N, Digiacinto GV, Hughes JEO, et al. Treatment of neoplastic spinal cord compression: results of a prospective study. Neurosurgery 1991;29:645-50. 12. Sundaresan N, Krol G, Digiacinto CV, et al. Metastatic tumors of the spine. In: Sundaresan B, Schmidek H, Schiller A, et al., editors. Tumors of the spine. Philadelphia, PA: W.B. Saunders; 1990:279-304.

13. Lu C, Stomper PC, Drislane FW, et al. Suspected spinal cord compression in breast cancer patients: a multidisciplinary risk assessment. Breast Cancer Res Treat 1998;51:121-31. 14. Vitaz T, Oishi M, Welch W, et al. Rotational and transpositional flaps for the treatment of spinal wound dehiscence and infections in patient populations with degenerative and oncological disease. J Neurosurg Spine 2004;100:46-51. 15. Patchell RA, Tibbs PA, Regine WF, et al. Direct decompressive surgical resection in the treatment of spinal cord compression caused by metastatic cancer: a randomised trial. The Lancet 2005;21:1-6. 16. Loblaw DA, Laperriere NJ. Emergency treatment of malignant extradural spinal cord compression: an evidence-based guideline. J Clin Oncol 1998;16: 1613-24. 17. Ryu S, Chang S, Kim D, et al. Image-guided hypofractionated stereotactic radiosurgery to spinal lesions. Neurosurgery 2001;49:838-46. 18. Amendola B, Wolf A, Coy S, et al. Gamma knife radiosurgery in the treatment of patients with single and multiple brain metastases from carcinoma of the breast. Cancer J 2000;6:88-92. 19. Benzil DL, Saboori M, Mogilner AY, et al. Safety and efficacy of stereotactic radiosurgery for tumors of the spine. J Neurosurgery 2004;101:413-8. 20. Bilsky MH, Yamada Y, Yenice KM, et al. Intensitymodulated stereotactic radiotherapy of paraspinal tumors: a preliminary report. Neurosurgery 2004;54:823-30. 21. Chang EL, Shiu AS, Lii M-F, et al. Phase I clinical evaluation of near-simultaneous computed tomographic image-guided stereotactic body radiotherapy for spinal metastases. Int J Rad Onc Biol Phys 2004;59:1288-94. 22. Desalles AA, Pedroso A, Medin P, et al. Spinal lesions treated with Novalis shaped beam intensity modulated radiosurgery and stereotactic radiotherapy. J Neurosurgery 2004;101:435-40. 23. Gagnon GJ, Henderson FC, Gehan EA, et al. Cyberknife radiosurgery for breast cancer spine metastases: a matched-pair analysis. Cancer 2007;110:1796-802. 24. Jin J-Y, Chen Q, Jin R, et al. Technical and clinical experience with spine radiosurgery: a new technology for management of localized spine metastases. Technol Cancer Res Treat 2007;6:127-33. 25. Milker-Zabel S, Zabel A, Thilmann C, et al. Clinical results of retreatment of vertebral bone metastases by stereotactic conformal radiotherapy and intensitymodulated radiotherapy. Int J Rad Onc Biol Phys 2003;55:162-7. 26. Ryu S, Yin FF, Rock J, et al. Image-guided and intensitymodulated radiosurgery for patients with spinal metastasis. Cancer 2003;97:2013-8. 27. Shiu AS, Chang EL, Ye J-S. Near simultaneous computed tomography image-guided stereotactic spinal


28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

radiotherapy: an emerging paradigm for achieving true stereotaxy. Int J Rad Onc Biol Phys 2003;57:605-13. Yin FF, Ryu S, Ajlouni M, et al. Image-guided procedures for intensity-modulated spinal radiosurgery. J Neurosurgery 2004;101:419-24. Auchter R, Lamond J, Alexander E. A multinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastasis. Int J Rad Onc Biol Phys 1996;35:27-35. Chang S, Adler J, Hancock S. The clinical use of radiosurgery. Oncology 1998;12:1181-91. Flickinger J, Kondziolka D, Lunsford L. A multiinstitutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Rad Onc Biol Phys 1994;28:797-802. Kondziolka D, Patel A, Lunsford L. Stereotactic radiosurgery plus whole brain radiotherapy vs radiotherapy alone for patients with multiple brain metastases. Int J Rad Onc Biol Phys 1999;45:427-34. Loeffler J, Alexander EI. Radiosurgery for the treatment of intracranial metastasis. New York: McGraw-Hill; 1993. Loeffler JS, Kooy HM, Wen PY, et al. The treatment of recurrent brain metastases with stereotactic radiosurgery. J Clin Oncol 1990;8:576-82. Sperduto P, Scott C, Andrews D. Stereotactic radiosurgery with whole brain radiation therapy improves survival in patients with brain metastases: report of radiation therapy oncology group phase III study 95–08. Int J Rad Onc Biol Phys 2002;54:3a. Conti P, Pansini G, Mouchaty H, et al. Spinal neurinomas: retrospective analysis and long-term outcome of 179 consecutively operated cases and review of the literature. Surg Neurol 2004;61:34-44. Degen JW, Gagnon GJ, Voyadzis J-M, et al. CyberKnife stereotactic radiosurgical treatment of spinal tumors for pain control and quality of life. J Neurosurg Spine 2005;2:540-9. Hamilton AJ. Linear accelerator (LINAC) – based stereotactic spinal radiosurgery. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. p. 857-69. Hamilton A, Lulu B. A prototype design for linear accelerator based radiosurgery. Acta Neurochir (Wien) 1995;63:40-3. Blomgren H, Lax I, Naslund I, et al. Stereotactic high dose fraction radiation therapy of extracranial tumors using an accelerator: clinical experience of the first thirty-one patients. Acta Oncol 1995;34:861-70. Lax I, Blomgren H, Naslund I, et al. Stereotactic radiotherapy of malignancies in the abdomen: methodological aspects. Acta Oncol 1994;33:677-83. Hamilton A, Lulu B, Fosmire H, et al. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery 1995;36:311-9.

74

43. Baisden JM, Benedict SH, Sheng K, et al. Helical tomotherapy in the treatment of central nervous system metastasis. Neurosurg Focus 2007;22:1-6. 44. Chang S, Adler J. Current status and optimal use of radiosurgery. Oncology 2001;15:209-21. 45. Colombo F, Pozza F, Chierego G. Linear accelerator radiosurgery of cerebral arteriovenous malformations: an update. Neurosurgery 1994;34:14-21. 46. Gerszten PC, Burton SA, Ozhasoglu C, et al. Single fraction radiosurgery for spinal metastases: clinical experience in 500 cases from a single institution. Spine 2007;32:193-9. 47. Gerszten PC, Welch WC. CyberKnife radiosurgery for metastatic spine tumors. Neurosurg Clin N Am 2004;15:491-501. 48. Hitchcock E, Kitchen G, Dalton E, et al. Stereotactic linac radiosurgery. Br J Neurosurg 1989;3:305-12. 49. Medin P, Solberg T, DeSalles A. Investigations of a minimally invasive method for treatment of spinal malignancies with LINAC stereotactic radiation therapy: accuracy and animal studies. Int J Rad Onc Biol Phys 2002;52:1111-22. 50. Pirzkall A, Lohr F, Rhein B, et al. Conformal radiotherapy of challenging paraspinal tumors using a multiple arc segment technique. Int J Rad Onc Biol Phys 2000;48:1197-204. 51. Ryu S, Rock J, Rosenblum M, et al. Patterns of failure after single-dose radiosurgery for spinal metastasis. J Neurosurgery 2004;101:402-5. 52. Pieters RS, Niemierko A, Fullerton BC, et al. Cauda equina tolerance to high dose fractionated irradiation. Int J Rad Onc Biol Phys 2006;64:251-7. 53. Ryu S, Jin J-Y, Jin R, et al. Partial volume tolerance of the spinal cord and complications of single dose radiosurgery. Cancer 2007;109:628-36. 54. Hopewell J, Morris A, Dixon-Brown A. The influence of field size on the late tolerance of the rat spinal cord to single doses of X rays. Br J Radiol 1987;60:1099-108. 55. Emami B, Lyman A, Brown JT, et al. Tolerance of normal tissue to therapeutic irradiation. J Radiat Oncol Biol Phys 1991;21:109-22. 56. Gerszten PC, Bilsky MH. Spine radiosurgery. Contemp Neurosurg 2006;28:1-8. 57. Tong D, Hendrickson F. The palliation of symptomatic osseous metastases; final results of the study by the radiation therapy oncology group. Cancer 1982;50:893-9. 58. Wara WM, Phillips TL, Sheline GE, et al. Radiation tolerance of the spinal cord. Cancer 1975;35:1558-62. 59. Hatlevoll R, Host H, Kaalhus O. Myelopathy following radiotherapy of bronchial carcinoma with large single fractions: a retrospective study. Int J Radiat Oncol Biol Phys 1983;9:41-4. 60. Abbatucci JS, Delozier T, Quint R, et al. Radiation myelopathy of the cervical spinal cord: time, dose and volume factors. Int J Radiat Oncol Biol Phys 1978;4:239-48.

1221

1222

74


61. McCuniff AJ, Liang MJ. Radiation tolerance of the cervical spinal cord. Int J Radiat Oncol Biol Phys 1989;16:675-8. 62. Phillips TL, Buschke F. Radiation tolerance of the thoracic spinal cord. AJR 1969;105:659-64. 63. Klish MD, Watson GA, Shrieve DC. Radiation and intensity-modulated radiotherapy for metastatic spine tumors. Neurosurg Clin N Am 2004;15:481-90. 64. Rock JP, Ryu S, Yin FF. Novalis radiosurgery for metastatic spine tumors. Neurosurgery Clinics of North America 2004;15:503-9. 65. Yamada Y, Lovelock M, Yenice KM, et al. Multifractionated image-guided and stereotactic intensity modulated radiotherapy of paraspinal tumors: a preliminary report. Int J Rad Onc Biol Phys 2005;62:53-61. 66. Ho AK, Fu D, Cotrutz C, et al. A study of the accuracy of cyberknife spinal radiosurgery using skeletal structure tracking. Neurosurgery 2007;60:147-56. 67. Muacevic A, Staehler M, Drexler C, et al. Technical description, phantom accuracy and clinical feasibility for fiducial-free frameless real-time image-guided spinal radiosurgery. J Neurosurg Spine 2006;5:303-12. 68. Adler J, Murphy M, Chang S, et al. Image-guided robotic radiosurgery. Neurosurgery 1999;44:1-8. 69. Adler JR, Chang SD, Murphy MJ, et al. The CyberKnife: a frameless robotic system for radiosurgery. Stereotactic and Functional Neurosurgery 1997;69:124-8. 70. Murphy MJ, Cox RS. Frameless radiosurgery using realtime image correlation for beam targeting. Med Phys 1996;23:1052-3. 71. Welsh J, Mehta M, Mackie T, et al. Helical tomotherapy as a means of delivering scalpsparing whole brain radiation therapy. Technol Cancer Res Treat 2005;4:661-2. 72. Extracranial Stereotactic Radiotherapy and Radiosurgery. Slotman BJ, Solberg TD, Verellen D, editors. New York: Taylor and Francis Group; 2006. 73. Stereotactic Body Radiation Therapy. Kavanagh BD, Timmerman RD, editors. Philadelphia, PA: Lippincott Williams and Wilkins; 2005. 74. Timmerman RD, Papiez L, McGarry R. Extracranial stereotactic radioablation: results of a phase I study in medically inoperable stage I non-small cell lung cancer. Chest 2003;124:1946-55. 75. Timmerman RD, McGarry R, Yiannoutsos C. Excessive toxicity when treating central tumors in a phase II study of stereotactic body radiation therapy for medically inoperable early-stage lung cancer. J Clin Oncol 2006; 24:4833-9. 76. Nagata Y, Takayama K, Matsuo Y. Clinical outcomes of a phase I/II study of 48Gy stereotactic body radiotherapy in 4 fractions for primary lung cancer using a stereotactic body frame. Int J Radiat Oncol Biol Phys 2005;63:1427-31. 77. Onishi H, Araki T, Shirato H. Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanes multiinstitutional study. Cancer 2004;101:1623-31.

78. Uematsu M, Shioda A, Tahara K. Computed tomographyguided frameless stereotactic radiotherapy for stage I nonsmall cell lung cancer: 5-year experience. Int J Radiat Oncol Biol Phys 2001;51:666-70. 79. Timmerman RD. Stereotactic body radiation therapy. Curr Probl Cancer 2005;29:120-57. 80. Timmerman R, Kavanagh B, Chos CL. Stereotactic body radiation therapy in multiple organ sites. J Clin Oncol 2007;25:947-52. 81. Shefter TE, Cardenes HR, Kavanagh BD. Stereotactic body radiotherapy for liver tumors. In: Kavanagh BD, Timmerman RD, editors. Stereotactic Body Radiation Therapy. Philadelphia, PA: Lippincott Williams and Wilkins; 2005. 82. Gerszten PC, Burton S, Ozhasoglu C, et al. Stereotactic radiosurgery for spine metastases from renal cell carcinoma. J Neurosurg Spine 2005;3:288-95. 83. Gerszten PC, Burton S, Welch WC, et al. Single fraction radiosurgery for the treatment of breast metastases. Cancer 2005;14:2244-54. 84. Gerszten PC, Burton SA, Belani C, et al. Radiosurgery for the treatment of spinal lung metastases. Cancer 2006;107 (11):2653-61. 85. Gerszten PC, Burton SA, Quinn AE, et al. Radiosurgery for the treatment of spinal melanoma metastases. Stereotact Funct Neurosurg 2006;83:213-21. 86. Ryken TC, Meeks SL, Pennington EC, et al. Initial clinical experience with frameless stereotactic radiosurgery: analysis of accuracy and feasibility. Int J Rad Onc Biol Phys 2001;51:1152-8. 87. Rock J, Ryu S, Shukairy MS, et al. Postoperative radiosurgery for malignant spinal tumors. Neurosurgery 2006;58:891-8. 88. Cohen-Gadol A, Zikel O, Koch C, et al. Spinal meningiomas in patients younger than 50 years of age: a 21-year experience. J Neurosurg Spine 2003;98:258-63. 89. Dodd RL, Ryu MR, Kammerdsupaphon P, et al. CyberKnife radiosurgery for benign intradural extramedullary spinal tumors. Neurosurgery 2006;58:674-85. 90. Peker S, Cerci A, Ozgen S, et al. Spinal meningiomas: evaluation of 41 patients. J Neurosurg Sci 2005;49:7-11. 91. Gerszten PC, Burton SA, Ozhasoglu C, et al. Radiosurgery for benign intradural spinal tumors. J Neurosurgery 2007;106:A742. 92. Seppala M, Haltia M, Sankila R, et al. Long term outcome after removal of spinal schwannoma: a clinicalopathological study of 187 cases. J Neurosurgery 1995;83:621-6. 93. Rampling R, Symonds S. Radiation myelopathy. Curr Opin Neurol 1998;11:627-32. 94. Steiner L, Lesksell L, Greitz T, et al. Stereotaxic radiosurgery for cerebral arteriovenous malformations. Report of a case. Acta Chir Scand 1972;138:459-64. 95. Chang S, Shuster D, Steinberg G, et al. Stereotactic radiosurgery of arteriovenous malformations: pathologic changes in resected tissue. Clin Neuropathol 1997;16:111-6.


96. Betti O, Munari C, Rosler R. Stereotactic radiosurgery with the linear accelerator: treatment of arteriovenous malformations. Neurosurgery 1989;24:311-21. 97. Coffey R, Lunsford L, Bissonette D, et al. Stereotactic gamma radiosurgery for intracranial vascular malformations and tumors: report of the initial North American experience in 331 patients. Stereotact Funct Neurosurg 1990;54:535-40. 98. Colombo F, Benedetti A, Pozza F, et al. Linear accelerator radiosurgery of cerebral arteriovenous malformations. Neurosurgery 1989;24:833-40. 99. Friedman W, Bova F. Linear accelerator radiosurgery for arteriovenous malformations. J Neurosurgery 1992;77: 832-41.

74

100. Steinberg G, Fabrikant J, Marks M, et al. Stereotactic heavy-charged-particle Bragg-peak radiation for intracranial arteriovenous malformations (see comments). N Engl J Med 1990;323:96-101. 101. Steiner L. Radiosurgery in cerebral arteriovenous. In: Flamm ES, Fein J, editors. Cerebrovascular surgery. New York: Springer; 1985. p. 1161-215. 102. Sinclair J, Chang SD, Gibbs IC, et al. Multisession CyberKnife radiosurgery for intramedullary spinal cord arteriovenous malformations. Neurosurgery 2006;58: 6 1081-9. 103. Chang S, Hancock S, Gibbs I, et al. Spinal cord arteriovenous malformation radiosurgery. In: Gerszten PC, Ryu SI, editors. Spine radiosurgery. New York: Thieme; (in press).

1223

80 Anesthesia for Functional Neurosurgery P. H. Manninen . N. Apichatibutra

Stereotactic functional neurosurgery is used for the treatment of many movement and functional disorders. Historically, the awake craniotomy was used for epilepsy surgery, but now it is used increasingly for patients with brain tumors. The role of the anesthesiologist is critical in assuring success for these patients. This chapter will cover the anesthetic considerations and management of patients undergoing surgery for functional disorders, stereotactic biopsy, epilepsy, and brain tumors. The anesthesia management includes the awake patient with monitored anesthesia care or conscious sedation and the use of general anesthesia.

Stereotactic Functional Neurosurgery Functional neurosurgery with the use of deep brain stimulation (DBS) is used in the treatment of patients with movement disorders and other chronic illnesses [1,2]. The initial success was with patients with Parkinson’s disease [3,4]. Now the applications and indications have expanded to many other disorders such as dystonia, tremors, movement disorders, depression, obsessive-compulsive disorder, epilepsy, chronic pain, and other newer areas that are under investigation [1,5,6]. Deep brain stimulation is a minimally invasive procedure that enables structures in the brain such as the subthalamic nucleus to be stimulated electrically by an implanted pacemaker. There are many steps to this procedure including the placement of a rigid frame to the #


patient’s head for stereotaxis and radiological imaging to localize brain structures with references to external coordinates. The intraoperative course includes making a burr hole for insertion of the DBS for microelectrode recordings (MER) and clinical testing of an awake patient. The DBS is then secured to the skull and at the same setting or at a future date it is connected to a lead that is tunneled to the chest or abdomen where a pulse generator pacemaker is implanted. In most centers the anesthesiologist plays an important role in the care and monitoring of the patients and in providing sedation or anesthesia during the intraoperative course [4,7–10]. However, the use of sedation and/or general anesthesia remains controversial. The preferred technique of anesthetic management of the patient in many centers is local anesthesia with monitored anesthesia care, that is, no sedation [4,10]. The reasons for having a completely awake patient are to preserve the MER of a single unit, to use stimulation testing for localization, and to observe for improvement in symptoms, adverse effects, and neurological examination. There are concerns that even mild sedative drugs may interfere with neurophysiology and MER. However, performing these procedures without sedation in certain patients may be problematic. Some patients are reluctant to be awake for these long procedures. The patient may have such severe continuous movements, kyphosis, pain, depression, psychotic symptoms, or be uncooperative that they are unable to lie still and/or to cooperate for neurophysiologic testing. The patient’s symptoms may also be worse due to the ‘‘weaning off ’’

1332

80

Anesthesia for functional neurosurgery

from their usual medications. Some of these patients will require sedation or even general anesthesia. There are many challenges and demands for the anesthesiologist whether the procedure is performed with or without sedation. These include keeping the patient comfortable, responsive, and cooperative for a long period of time and if providing sedation, ensuring that it does not interfere with electrophysiological brain mapping and clinical testing. Many of these patients also present with complex medical problems, may be elderly, and on multiple medications. The role of the anesthesiologist is to also monitor the patient’s vital signs, especially cardiovascular and respiratory, ensuring that they remain stable and within normal range. Continuous vigilance is required to rapidly diagnose and treat all complications some of which may be life-threatening.

Preoperative Assessment Prior to the surgical procedure, the anesthesiology team should review all patients scheduled for functional neurosurgery. This can be done in a preoperative anesthesia consult clinic if patients are admitted to the hospital on the day of surgery. All routine assessments including laboratory studies and considerations for anesthesia are reviewed. The patient should be in stable condition with respect to all other ongoing medical conditions. Optimal preparation includes the continuation of routine medications. The anesthesiologist should emphasize to the patient to follow instructions given by the neurologist/neurosurgeon regarding discontinuation of medications for their functional disorder. All patients must follow standard ‘‘nothing by mouth’’ orders. Psychological preparation of the patient is important for the long and potentially difficult awake procedures. Patient should be advised as to what will happen during the course of the day

and the role of the anesthesiologist in providing sedation, if possible, and maximum comfort and support. Specific anesthetic considerations will apply to patients with different disorders. Patients with neurodegenerative disorders such as Parkinson’s disease are often elderly and may have respiratory, cardiovascular, and autonomic system compromise [11–13]. Potential drug interactions and adverse effects from anti-Parkinson’s medications may also occur during anesthesia. Ergotderived dopamine agonists are associated with a higher incidence of cardiac valvular disorders [14]. Drugs that are selective monoamine oxidase inhibitors may have serious interactions with meperidine and sympathomimetic agents [15]. The effect of sedative and anesthetic agents must also be considered in patients with Parkinson’s disease such as propofol induced dyskinesia and the suppression of tremors with remifentanil [16–18]. Patients with dystonia or torticollis may present difficulties in airway management. Pyschiatric patients will also have their own set of challenges and patients with chronic pain will need special consideration in management of their pain medications perioperatively.

Procedure In the operating room or radiology suite the surgeons will apply a rigid head frame to the patient’s skull. There may or may not be an anesthesiologist present and sedation may or may not be used. Local anesthesia is used as a subcutaneous infiltration at the pin sites. Supraorbital and greater occipital nerve blocks are an alternative as they have been shown to be less painful than subcutaneous infiltration but did not result in any difference at the time of pin placement or during surgery [19]. In the radiology suite during imaging no sedative medication is usually required except in cases of extreme anxiety or claustrophobia or if general anesthesia is required for the entire


procedure. The patient is then transferred to the operating room fully awake or if under sedation or general anesthesia with monitoring and continuation of anesthesia. Proper positioning of patients on the operating table is an important step to ensure maximum comfort and cooperativeness. After the head frame is fixed to the operating room table one needs to ensure that the patient’s neck is comfortable, they are able to open their mouth so that breathing, speech, and swallowing are not restricted. The legs should be flexed and supported under the knees to maintain stability when the head and back are elevated to a sitting position. Slipping downward on the table while head is fixed has resulted in complete upper airway obstruction even in nonsedated patients [10]. All pressure points should be padded. Specific treatment modalities, such as physiotherapy, may be used in specific patients [20]. If a patient with Parkinson’s disease has excessive tremors and can not lie still, a small dose of levodopa may help to decrease the intensity of the tremors. Local anesthesia is infiltrated at the site of the burr hole for insertion of electrodes for the MER.

Intraoperative Anesthetic Considerations Standard monitors used for all procedures include an electrocardiogram, noninvasive blood pressure, oxygen saturation, end-tidal CO2, and respiratory rate. The need for additional monitors such an intra-arterial catheter for blood pressure, temperature, and the insertion of a urinary catheter are dependent on the practice of each institution, condition of the patient, and depth of sedation or anesthesia. Administration of supplemental oxygen is mandatory even for awake patients. This may be delivered via nasal prongs or a mask with an outlet for end-tidal CO2 and respiratory rate monitoring. A secure intravenous catheter is used for medications and fluid administration. Omission of the urinary

80

catheter will be more comfortable for the patient, but fluid administration needs to be restricted. The patient may initially feel very cold in the operating room and may even require a warming blanket. However, with prolonged surgery and with increased levels of anxiety some patients become too hot. Patients who are not sedated will require frequent reassurance, additional comforts such as moving arms and legs, and even ice chips to wet their mouth. Awareness of an awake patient by all operating room personnel and the administration of psychological support by all members of the team will help the patient in this stressful environment. Communication is essential between the neurosurgeon and the anesthesiologist throughout the procedure. The anesthesiologist needs to know when it is appropriate to give the patient sedation or analgesia. The timing of the induction of general anesthesia, if it is to be used, must be clarified. Intubation of the patient’s trachea will be more difficult after the stereotactic frame is pinned to the patient’s skull. If general anesthesia is to be started in the radiology suite, the anesthesiologist must have adequate equipment and support to care for the patient in this potentially ‘‘remote’’ site.

Local Anesthesia The local anesthetic agents usually used for infiltration of the pin sites for the frame and the incision sites are bupivacaine, ropivacaine, and lidocaine with and without epinephrine [21,22]. Compared with lidocaine, bupivacaine has slower onset of action but longer duration. Ropivacaine has less cardiac and central nervous system toxicity compared with bupivacaine. Complications of local anesthesia include toxic blood levels resulting in seizures, respiratory, and cardiac arrest. The maximum doses are based on the patient’s weight and for lidocaine this is 5 mg/kg and up to 7 mg/kg with epinephrine, for bupivacaine 2 mg/kg and

1333

1334

80


3mg/kg with epinephrine, and for ropivacaine 3.6 mg/kg with epinephrine. If the procedure has been long, additional infiltration of local anesthesia may be required for closure.

other agents. However, they may cause profound respiratory depression. The short acting potent opioid, remifentanil has been reported to decrease tremors in patients with Parkinson’s disease [18].

Conscious Sedation General Anesthesia Sedation during functional neurosurgery is controversial in terms of whether to use it at all, when to use it, how much, and which agents. Sedation may be given throughout the whole procedure or just at the beginning and/or at the end. Most frequently all sedation is stopped during MER and stimulation testing. Commonly agents used are midazolam and propofol with or without opioids such as fentanyl or remifentanil [7,10,23–25]. Benzodiazepines are good anxiolytic agents with amnesic properties but they may alter the threshold for stimulation or prevent precise clinical assessment of the patient. Midazolam has a short duration of action and may be used in low doses prior to insertion of head pins for the frame. Propofol also has a short duration of action and can be given as a bolus or infusion. Propofol has been reported to reduce tremors, as well as to produce myoclonus movements [16,17,26]. Some patients may respond to sedative agents with paradoxical agitation or disinhibition. The management of these patients is to wake up the patient or change the sedative agent. A newer agent, dexmedetomidine, an alpha-2 agonist, has a more favorable pharmacological profile for sedation [27]. Dexmedetomidine provides adequate sedation while maintaining good ventilation and airway patency due to minimal respiratory depression and stable blood pressure [28]. Dexmedetomidine was shown not to impair the intensity of movement disorder in patients with Parkinson’s disease, not to interfere with MER, to provide hemodynamic stability, and decreased use of anti-hypertensive medications. Opioids are often added for analgesia and to improve sedation in conjunction with

The ability to use general anesthesia in specific patients may broaden the scope of patients that are suitable for functional neurosurgery. It would be difficult to treat some patients with severe continuous movements, severe pain and anxiety, or those who are uncooperative and children without the use of general anesthesia [29–31]. There are a few reports of the successful use of general anesthesia. Malteˇte et al. retrospectively reviewed the effects of general anesthesia on the postoperative outcome of patients with Parkinson’s disease following DBS insertion [29]. Fifteen patients with severe anxiety, respiratory problems and ‘‘off medication’’ dystonia requiring general anesthesia were matched with 15 patients treated under local anesthesia. General anesthesia included the administration of propofol sedation with no airway manipulation. The improvement in motor disability was greater in patients who received local anesthesia compared with those who received general anesthesia. Yamada et al. compared 15 patients who required general anesthesia because of severe psychosis and anxiety to 10 patients with local anesthesia [30]. General anesthesia was started after placement of the frame and imaging. Patients were intubated with fiberoptic guidance and had a balanced anesthetic with propofol, fentanyl, nitrous oxide, and sevoflurane. General anesthesia did not adversely effect postoperative improvements in motor and daily activity scores, except for ‘‘off medication’’ bradykinesia. Hertel et al. also looked at the feasibility of a DBS surgery with general anesthesia [31]. General anesthesia was used in nine patients for reasons of fear,


scoliosis, vertebral pain and heavy coughing. The entire procedure was with total intravenous anesthesia with remifentanil and propofol and endotracheal intubation was with rocuronium. Five minutes before the start of MER and test stimulation, remifentanil was stopped and propofol was lowered as far as possible. The authors concluded that MER were possible with their technique and clinical improvement was achieved in all patients. The adverse effects of general anesthesia in patients with functional disorders must be recognized. With Parkinson’s disease these are well documented and include respiratory and cardiovascular complications and neurological exacerbations including rigidity, confusion, and prolonged recovery time [11–13]. The ability to test the patient during the procedure is lost as well as the ability to identify potential complications and collateral effects of stimulation such as dysarthria and motor responses.

Complications The role of the anesthesiologist also includes the rapid recognition and treatment of complications during and immediately after the procedure. Respiratory complications, though not common, are of great concern [10]. Oversedation may result in decreased respiratory rate, desaturation, and loss of airway patency. There are difficulties in airway management when the patient is in a fixed stereotactic frame, which is rigidly attached to the operating room table. Also the frame may cover some or all of the patient’s mouth and nose. During the procedure if the patient becomes restless or attempts to move, acute airway obstruction may occur as the body shifts but the head remains fixed to the bed. Should this happen one needs to immediately release the frame from the operating room table to release the patient’s airway. This will usually resolve the problem. Supplemental oxygen

80

during the entire procedure will help decrease the incidences of desaturation, especially in sedated patients. Securing the airway is not often required but the anesthesiologist should always be ready. Patients may develop a sudden loss of consciousness from an intracranial bleed or from repetitive seizures. Appropriate airway equipment such as oral and nasal airways, laryngeal mask airways, endotracheal tubes, laryngoscope, and fiberoptic bronchoscopy should be readily available. Ideally, if possible, one should attempt to secure the airway without the removal of the patient’s head frame, so the surgery can be continued if needed. Other respiratory complications relate to the diseases of the patients. Patients with Parkinson’s disease may have restrictive pulmonary dysfunction from poor respiratory muscles function. This may lead to reduced forced vital capacity, reduced baseline arterial oxygen saturation, upper airway obstruction, and obstructive sleep apnea [11,32,33]. Respiratory insufficiency due to the absence of anti-Parkinson’s medications in the postoperative period may also occur. Patients with functional disorders are also at risk as they may have increased sensitivity to sedatives. Cardiovascular parameters are carefully monitored especially blood pressure ensuring that it remains within normal limits. This is frequently difficult in patients who are not sedated and may become anxious, agitated, and are uncomfortable. Hypertension is a common complication [8]. Vasoactive agents may be required to control the blood pressure during the insertion of the electrodes, as higher incidences of intracranial hemorrhage have been associated with higher arterial pressures [34,35]. The optimal level of blood pressure is controversial; one may use a systolic pressure below 160 mmHg systolic or 20% of patient’s preoperative blood pressure. Sedation will help to control blood pressure increases but, if this is not an option, then other agents need to be used. Commonly used agents include labetalol, hydrazaline, nitroglycerine,

1335

1336

80


sodium nitroprusside, or esmolol. Orthostatic hypotension may occur whenever a patient is placed in a sitting position and this may be aggravated by anti-Parkinson’s drugs [36]. Hypotension is more common after induction of general anesthesia as most anesthetic agents also have peripheral vasodilatation effects, adding to the possible effects of orthostatic hypotension, preoperative hypovolemia, and autonomic dysfunction. Other less common complications include venous air embolism [37–39]. During creation of the burr hole in awake patients, sudden vigorous coughing may be a sign of venous air embolism. Other signs are unexplained hypoxia and hypotension. Earlier detection may be possible with precordial Doppler monitoring. Tension pneumocephalus has also been reported during DBS insertion [40]. Neurological complications may occur during or after the procedure [34,35,41]. Focal deficits such as extremity weakness or confusion may not require any acute treatment by the anesthesiologist. Seizures may occur especially during stimulation testing. Most seizures are focal and do not require treatment. In cases of tonic clonic seizures small doses of midazolam and/or propofol can be given and the procedures can be resumed after control of the seizure. A sudden loss of consciousness due to an intracranial bleed or from a major neurological injury will require rapid treatment. Securing of the airway if needed can be done by any technique the anesthesiologist is comfortable with and may require releasing the head and/or removing the head from the frame. It may be necessary to take the patient for a computerized tomography scan to rule out a hematoma and even to plan for a craniotomy. All this may require the transfer of an unstable or anesthetized patient to other areas of the hospital. The internalization of the wires and pacemaker may be done on the same day as the DBS procedure. Otherwise the patient will return to the operating state at a later date during the same hospitalization. The internalization procedure is

performed under general anesthesia. At this time the influence of anesthetic agents on the DBS is not of concern. The anesthesiologist may use any technique of general anesthesia that is appropriate for each individual patient.

Vagal Nerve Stimulation Vagal nerve stimulation is used for the treatment of medically refractory epilepsy and some neuropsychiatric disorders. The procedure involves the surgical placement of an electrode wrapped around the left vagus nerve and then tunneled and connected to a generator pacemaker inserted into the chest wall. The whole procedure is performed under general anesthesia with endotracheal intubation [42]. The concerns of the anesthetic management relate to the medical condition of the patient. Possible complications during the procedure include cardiovascular events such as bradycardia or atrio-ventricular block, which will respond to appropriate treatment such as atropine, but may result in termination of the procedure. Postoperative complications include lower facial paralysis, laryngeal dysfunction including vocal cord paralysis, and other possible respiratory problems such as decreased respiratory effort, risk of aspiration, and worsening of obstructive sleep apnea.

Patients with DBS for Other Surgery With increased use of DBS for a variety of disorders, the anesthesiologist will encounter patients with an implanted DBS coming for other surgeries [42–45]. This also includes patients with vagus nerve stimulators. The electrical pulse generator pacemaker is usually located in the anterior chest wall of the patient. There is confusion and controversy on the safety of electrocautery use and magnetic resonance imaging (MRI).


Electrocautery may damage the leads and can temporarily suppress the neurostimulator output and/or reprogram the neurostimulator. During surgery a bipolar mode of electrocautery should be used if possible. When unipolar cautery is needed the ground plate should be as far away as possible from the neurostimulator and leads. External defibrillation may also damage the stimulator. The lowest possible energy for the current should be used and the placement of the defibrillator pads should be as far away as possible from the pacemaker and leads. The effects of the MRI include heating and torque of the DBS and leads, interference with the pacemaker program and image distortion. The stimulator should be turned off for the MRI and all safety recommendations for placement of the radiofrequency coils and parameter selection adhered to. Whenever the DBS is turned off, it important to resume its activity as soon as possible to prevent complications from lack of stimulation, especially if preoperative medications have been withheld. The correct use of a magnet to activate or deactivate the stimulators is complex and differs from cardiac pacemakers. One needs to consult the manufactures information for each generator.

Sterotactic Biopsy Surgery Stereotactic biopsy surgery uses computerized tomography or MRI for obtaining coordinated references to an extracranial system to guide the biopsy needle for accurate localization and sampling of intracranial lesions such as tumor or abscess. Traditionally, the procedure involved the placement of a head frame to the cranium, but fewer frame-based procedures are now performed due to the development of frameless neurological navigation imaging systems. These frameless stereotactic systems use scalp markers as fiducial points to relate the surgical instruments to a computer generated image. Most of these procedures are performed with the patients

80

awake or with minimal sedation to allow for neurological examination of the patient [46,47]. The role of the anesthesiologist is to provide sedation, if needed, and to monitor the patient’s vital signs and neurological system. In some circumstances the anesthesiologist may also need to travel with the patient to or from the radiology suite to the operating room. Acute intracranial complications can lead to a sudden change in level of consciousness. Respiratory complications may be potentially difficult to manage as the patient’s head frame is fixed to the operating room table and the frame may cover the nose and mouth of the patient, limiting access to the airway.

Epilepsy Surgery Epilepsy, a common chronic disorder, is usually treated with a variety of antiepileptic drugs. Medical treatment is deemed refractory if unacceptable side effects associated with the medications preclude adequate seizure control. Some of these patients are candidates for surgical resection of the epileptogenic focus. Surgery for partial seizure disorders involves the resection of a specific epileptogenic focus or a form of temporal lobectomy. Generalized seizures are treated by interruption of the seizure circuits by a corpus callosotomy or a hemispherectomy. A multidisciplinary evaluation including invasive and noninvasive investigations is performed to identify the origin of seizure activity and to evaluate the feasibility of performing surgery safely with minimal risk of neurological or cognitive injury [48,49].

Preoperative Localization of Epileptogenic Focus Many advances in neurological imaging techniques have reduced the need for invasive evaluation. Intracarotid sodium amytal injection (Wada test) is used to test for the lateralization of

1337

1338

80


language and memory. The drug is injected into the carotid artery via a femoral artery puncture. These tests are often done without the presence of an anesthesiologist. However, more recently the drug of choice has become etomidate, an intravenous anesthetic agent and this has necessitated the presence of an anesthesiologist to monitor and care for the patient. Invasive preoperative investigations for the localization of the epileptic focus in some patients may require surgery for the placement of epidural electrodes, or subdural grids and strip electrodes through burr holes or a craniotomy. These procedures are usually performed under general anesthesia. The anesthetic plan should take into consideration the concerns of a patient with epilepsy and the precautions that apply to any craniotomy. All anesthetic agents may be used as there are no electroencephalogram recordings at this time. Electrode plates or large grids are quite bulky and might require brain shrinkage with the use of mannitol and hyperventilation. These patients might develop postoperative problems with brain edema and require urgent removal of the grid due to the development of intracranial hypertension. After the insertion of intracranial electrodes a thiopental test may be conducted to evaluate some patients before surgery. An anesthesiologist is needed to be present for these tests as potential adverse effects, especially airway compromise, may occur [50].

Preoperative Assessment Epilepsy surgery may be with general anesthesia or with conscious sedation for an awake craniotomy [51,52]. The decision is usually made by the surgeon and is dependent on the location of the seizure focus, the need for intraoperative testing or localization of the seizure focus and eloquent brain function, and the ability of the patient to withstand an awake procedure. The preoperative anesthetic preparation of the patient is performed in the preoperative anesthesia consult clinic and

includes the routine assessment of the patient’s medical conditions, appropriate laboratory tests and investigations. Medications that the patient is on prior to surgery, such as anti-hypertensive agents, should be continued. The administration of anticonvulsant agents prior to surgery is done in consultation with the neurologist and surgeon. Premedication for the purpose of sedation is rarely required as these patients are usually well informed and agents that may influence the electroencephalogram, such as benzodiazepines, should be avoided. Appropriate preparation of the patient for the chosen technique of anesthesia is carried out. The patient should be informed of what to expect during the procedure including a rehearsal of stimulation testing. Specific considerations in patients with epilepsy include the associated medical problems such as psychiatric disorders, rare syndromes such as neurofibromatosis, multiple endocrine adenomatosis, and history of trauma. There are many adverse effects of antiepileptic drugs, which are dose-dependent and usually associated with long-term therapy [48]. Drugs may have neurological side effects such as sedation, confusion, learning impairment, ataxia, and gastrointestinal problems such as nausea and vomiting. Most anticonvulsants are metabolized by the liver. Thus, long-term usage will cause liver enzyme induction, which increases the rate of metabolism of other drugs, particularly anesthetic agents. Long-term therapy with phenytoin causes gingival hyperplasia with poor dentition and, potentially, difficult airway management. Carbamazepine can depress the hemopoietic system and, in rare cases, causes cardiac toxicity. Valproic acid might occasionally result in thrombocytopenia and platelet dysfunction.

Awake Craniotomy The reasons for having an awake patient for all or part of the procedure are for better


electrocorticographic localization of the seizure focus without the influence of general anesthetic agents, direct electrical stimulation of the cerebral cortex to delineate eloquent areas of brain function in order to preserve them during surgical resection, and for continuous clinical neurological monitoring of the patient [53–57]. The success of an awake craniotomy depends on the proper selection and preparation of the patient. Psychological preparation of the patient by the neurologist and surgeon and should be continued by the anesthesiologist. The challenge is to have the patient comfortable enough to remain immobile through a long procedure, but sufficiently alert and cooperative to comply with testing. The analgesic and sedative drugs employed must have minimal interference with electroencephalographic and stimulation testing.

Intraoperative Anesthetic Considerations The operating room is an unfamiliar and frightening place for most patients. The environment should be made quiet and comfortable with an appropriate room temperature. The operating room table should be as soft as possible with padding for all pressure points and adequate supports. The position of the patient for surgery will be dependent on the procedure. Ideally for an awake patient, the lateral position is more comfortable, allows for better visualization of the patient’s face and treatment of complications such as airway obstruction, nausea and vomiting or seizures. However, the supine position is possible as long as one maintains a good view of the patient’s face and has the ability to manipulate the airway. This can be accomplished by appropriate draping. Careful positioning of the head and neck is required when the headframe is secured to the operating room table to ensure patient comfort.

80

Standard monitors include an electrocardiogram, blood pressure, pulse oximeter, and end tidal capnography. Other monitors such as invasive arterial monitoring are added depending on the practice within each institution and patient needs. To decrease patient discomfort, urinary catheters can be avoided for shorter procedures and if intravenous fluid administration is kept to a minimum. Supplemental oxygen is given to all patients by mask or nasal prongs with an outlet for the monitoring of end tidal CO2 and respiratory rate. Nasal prongs are often tolerated better by the patient and also allow the patient to speak freely for speech testing.

Anesthetic Agents and Techniques Local anesthesia is used for insertion of head pins and for the surgical incision. Scalp nerve blocks are performed using long-acting local anesthetic agents, such as bupivacaine or ropivacaine with the addition of epinephrine [21,22]. Lidocaine, which has a faster onset, may be added and also used to infiltrate areas that are still painful during the procedure, such as dura. There are two commonly used anesthetic techniques for sedation and analgesia; ‘‘conscious sedation’’ and ‘‘asleep awake asleep.’’ The conscious sedation technique implies a level of sedation where the patient has a minimally depressed level of consciousness but is able to maintain their own airway and respond to verbal stimulation [55]. Generally, there is no manipulation of the airway other than the administration of supplemental oxygen via nasal prongs or cannula, or a facemask. Traditionally ‘‘neurolept anesthesia’’ was employed using fentanyl and droperidol [53,54]. Opioids used for analgesia have included fentanyl, sufentanil, and alfentanil [58,59]. Now more commonly an infusion of the short acting potent opioid, remifentanil, is used as this drug is easy to titrate and does not accumulate [60–63]. Though opioids have been shown to produce

1339

1340

80


epileptiform activity in patients with epilepsy, and may also produce myoclonic movements that may clinically resemble seizure-like activity, this does not appear to be a problem during awake craniotomies as the doses used are generally small [48,63]. Propofol is the usual drug of choice for sedation [60–62,64–66]. It has a rapid onset of action and fast offset so that the level of sedation can be quickly changed and can be given without airway protection. However, propofol may suppress epileptogenic activity and has to be discontinued at least 15–30 min prior to electrocorticographic recordings [65,66]. Propofol may also cause tonic–clonic movements mimicking seizures, but will also effectively stop seizures [26,65]. All these agents can be administered as continuous infusions, intermittent boluses or in any combination. Target controlled sedation and the use of patient controlled anesthesia have also been described [60,66]. Dexmedetomidine has also been used as an adjunct for sedation and analgesia. Its benefits include minimal risk of respiratory depression, hemodynamic stability, and anesthetic sparing effects [27,67–71]. The addition of dexmedetomidine may improve the safety of patients during awake craniotomy and still allow for electrocorticography, cortical mapping, and neurological testing though it is unclear whether it has any anticonvulsant effect [70]. Anti-emetic agents (dimenhydrinate, prochlorperazine, metoclopramide, odansteron, granisetron) may also be needed and do not affect electrocorticography. Sedation is stopped just prior to stimulation testing and may be resumed after all testing is complete. During resection of the lesion heavy sedation should be avoided if ongoing neurological testing is required. In addition to the use of sedation and analgesic medications, continuous communication and reassurance with the patient is critical. Many nonpharmacological measures can help the patient tolerate these procedures. This includes warning the patient in advance of any painful or disturbing stimuli, especially the loud noise

caused by drilling of the bone, which can be frightening though not painful. Often small comforts such as allowing the patient to move intermittently, wetting their lips, ensuring that they are not too hot or cold and just holding the patient’s hand will go a long way to help the patient through the procedure. Another approach to the awake craniotomy is the use of the ‘‘asleep awake asleep’’ technique [67–69,72]. After arrival in the operating room, the patient is administered general anesthesia and the airway is secured with a laryngeal mask or endotracheal tube. The advantage of the laryngeal mask airway is its easier placement, less coughing and possibly less laryngospasm with removal. Both inhalation and intravenous anesthetic agents may be used, with or without controlled ventilation. When the craniotomy is complete, the patient is awakened and the airway removed to allow for the electrocorticographic recordings and stimulation testing. After the testing is completed, general anesthesia is resumed and the airway may or may not be reinserted. The reinsertion of the airway is the main disadvantage of the asleep awake asleep technique. Advantages of this technique include increased patient comfort and tolerance during surgery especially for longer procedures and a secured airway with the ability to use hyperventilation.

Complications Intraoperative complications include respiratory and cardiovascular changes, seizures, the restless patient, nausea and vomiting and the requirement to induce general anesthesia during conscious sedation [73,74]. Other less frequent complications include air embolism [75]. Oxygen desaturation and/or airway obstruction may result from oversedation, seizures, mechanical obstruction, or loss of consciousness from an intracranial event. The anesthesiologist needs to have a preplanned approach to deal with these


problems. If the problem is just simply oversedation then the treatment is to stop the administration of anesthesia agents and if needed, support the airway with a chin-lift or with a mask and assisted ventilation. However, if the loss of the airway is due to continuous seizures and/or their treatment, or loss of consciousness due to an intracranial event, one may need to secure a definite airway for the remainder of the procedure. The choice of the airway and the technique of insertion will depend on the skill of the anesthesiologist, the anatomy of the patient, and the requirements for the remainder of the procedure. Insertion of a laryngeal mask or endotracheal tube may require the induction of anesthesia, and the use of a direct laryngoscopy or fiberoptic bronchoscopy. During intubation the surgeon can assist by protecting the sterile areas and changing the position of the patient’s head if needed. During the procedure seizures can occur at any time, especially if the patient has been off their usual anticonvulsant medications. Short seizures may not require any treatment. Seizures that are convulsive or generalized need to be treated by protecting the patient from injury, ensuring patent airway, adequate oxygenation, and circulatory stability. Prior to electrocorticographic recordings, seizures can be treated with a small dose of thiopental or propofol, afterward benzodiazepines or longer acting anti-convulsants may be used. Other intraoperative problems include excessive pain and discomfort some of which can be predicted and the patient warned about such as the scalp block and the noise of the drill. Many factors can influence the incidence of nausea and vomiting, such as anxiety, medications, and surgical stimulation, especially the stripping of dura and manipulation of the temporal lobe or meningeal vessels. Intraoperative nausea and vomiting can be treated with any of the common anti-emetic agents. Some patients may become extremely restless, agitated, or uncooperative during the procedure. If the patient

80

becomes disinhibited from the sedative agents, especially propofol, the treatment is to ‘‘lighten’’ the patient’ s level of sedation so one can communicate with them, change the anesthetic agents being used, deepen the level of sedation, or in some situations one may have to convert to general anesthesia. The incidence of conversion to general anesthesia is low [54,73,74]. Other less common complications include local anesthetic toxicity, a tight brain, and cardiovascular changes.

General Anesthesia The reason for choosing general anesthesia is the preference of the surgeon and/or the inability of the patient to tolerate an awake craniotomy. However, with the advances in preoperative neurological imaging, functional testing, and the use of frameless stereotactic surgery for localization of the epileptic focus, the need for an awake patient has greatly decreased. The challenge for general anesthesia if intraoperative localization of the epileptic focus is needed is to provide good conditions for electrocorticography and for motor testing, ensuring that the influence of the anesthetic agents be kept at a minimum, but also avoiding long periods of potential awareness on the part of the patient. Specific preoperative preparation is to inform patients of the possibility that awareness and recall may occur at the time of electrocorticographic recording, but reassuring them that this will be brief and painless. All anesthetics will affect electrocorticography but the use of the shorter acting anesthetic agents, either inhalation agents and/or intravenous agents will allow for faster changes in the depth of anesthesia. During the time of recording all or most of the anesthetic agents are stopped, as much as possible. Another approach would be to do the initial craniotomy and electrocorticography and cortical mapping with the patient awake and then induce general anesthesia.

1341

1342

80


The difficulty of this technique is the induction of anesthesia and securing of the airway with an exposed brain and the patient’s head in a frame fixed to the operating room table. If no intraoperative testing or recording is to be performed, the anesthetic techniques and agents are at the discretion of the anesthesiologist and the management of the patient is as for any patient undergoing a craniotomy. Some specific concerns with general anesthesia in patients with epilepsy include the effects of long-term anticonvulsant therapy, which may lead to increased dosage requirements of opioids and neuromuscular blocking agents [48]. If the patient has had a recent craniotomy or burr holes for electrode placement, intracranial air might still be present and nitrous oxide should be avoided to prevent complications from an expanding pneumocephalus. Complications that may occur during epilepsy surgery are similar to those for any craniotomy, however, severe bradycardia is a common occurrence with resection of the amygdalo-hippocampus [76].

Intraoperative Recordings and Activation Electrocorticography is performed during surgery after opening of the dura by the placement of electrodes directly on the cortex over the area predetermined to be epileptogenic as well as on adjacent cortex. Additional recordings can be performed with microelectrodes placed into the cortex or depth electrodes into the amygdala and hippocampal gyrus. Stimulation of epileptogenic focus is possible pharmacologically if insufficient information to define the seizure focus adequately is obtained during routine electrocorticography. Traditionally methohexital or thiopental were administrated to gradually increase beta activity in normal functioning neural tissue, but not in the seizure focus. Other agents used include propofol, etomidate, or small dose of opioids in patients who are awake [48,49,63,77,78]. If the patient is

under general anesthesia, other agents can be used such as alfentanil, remifentanil or inhalation agents such as enflurane, sevoflurane with or without hypocarbia [79–82].

Awake Craniotomy for Tumor Surgery Awake craniotomy for tumor surgery is an accepted procedure that allows for mapping of brain function by cortical stimulation to delineate areas of eloquent function, such as speech, sensory, and motor in order to preserve them during resection of the tumor [83–86]. Another possible advantage is the avoidance of general anesthesia and earlier discharge from hospital [87]. The anesthetic management of patients for an awake craniotomy for tumor surgery is in most respects similar to that for epilepsy surgery except the concerns of the anesthetic effects on electroencephalography are not present. The goals of the anesthetic management are to have a comfortable patient who is able to stay immobile on an operating room table for the duration of the procedure and yet be alert and cooperative to comply with cortical mapping. These goals can be accomplished by adequate preparation of the patient, a comfortable environment, appropriate administration of analgesic, and sedative medications, ongoing communication and support of the patient, and rapid treatment of complications.

Preoperative Assessment The success of an awake craniotomy for tumor surgery depends on the proper selection and preparation of the patient. Not all patients are suitable for awake craniotomy such as the extremely anxious patient. Language barriers may make the procedure more difficult, but should not rule out an awake craniotomy if a suitable interpreter is available. The patient should be


informed of what to expect during the procedure including a rehearsal of stimulation testing. Otherwise the preoperative assessment is similar to any patient with a brain tumor and the concerns of intracranial pathology with respect to intracranial pressure and neurological deficits. Medications that the patient is on prior to surgery, especially steroids (dexamethasone) and anti-convulsants, should be continued. Premedication is not usually necessary. Frequently patients with brain tumors may have only had a short period of time to accept their diagnosis in contrast to patients with epilepsy who have a chronic disorder. The concept of having brain surgery and being awake in the operating room can be terrifying. Prior to bringing the patient in, the operating room should be completely ready so that all members of the team, anesthesiologist, nursing staff and surgeon, can devote their attention to the patient. For the anesthesiologist, preparation of the operating room also includes the preparation of all anesthetic drugs, monitors, and equipment that are required for an awake procedure as well as for general anesthesia should this be required and for the treatment of any complication. The position of the patient for surgery will be dependent on where the lesion is. Supine, lateral, and semi-prone have all been used successfully. It is important to maintain a good view of the patient’s face and have the ability to manipulate the airway. Similar to epilepsy surgery only standard monitors, electrocardiogram, noninvasive blood pressure, pulse oximeter, and end tidal capnography are essential. Other monitors such as invasive blood pressure, urinary catheter are added as indicated by the procedure, the patient and routine practice in each institution.

Anesthetic Agents and Techniques Overall the anesthetic management of the patient is similar to that described above for epilepsy

80

surgery. Many different anesthetic agents and techniques have been successfully used for tumor surgery, as the influence of drugs on the electroencephalogram is not of concern. For a conscious sedation technique the most common combinations used include propofol and/or midazolam and fentanyl or infusion of remifentanil [88,89]. Dexmedetomidine has also been used for tumor surgery [67–69]. Generally there is no manipulation of the airway other than the administration of supplemental oxygen via nasal prongs or cannula, or a facemask. Once the patient is positioned on the operating table, sedation and analgesia can be started. If the surgical approach includes the use of navigational equipment the head will be placed in a head frame with pins using infiltration with local anesthetic. The incision area is infiltrated with local anesthetic agents or scalp nerve blocks may be used. When the dura is approached the amount of sedation should be decreased so that the patient will be alert during testing. In addition to the use of sedation and analgesic medications, continuous communication, reassurance, and the use of nonpharmacological measures is important. In some patients where the resection margins of the tumor are very close to critical eloquent areas, it is useful to continue testing, for speech or motor function, during resection and thus heavy sedation should not be used. The asleep awake asleep technique is a commonly used for tumor surgery [90–92]. The patient is administered general anesthesia with the use of a laryngeal mask or endotracheal tube for airway management during the craniotomy, then awakened and the airway is removed for cortical mapping and testing. After the testing is completed, general anesthesia is resumed with or without reinsertion of the airway for the resection of the tumor and closure. Advantages of this technique for tumor surgery are the increased patient comfort and tolerance for long tumor resections and the ability to use hyperventilation to treat increased brain mass.

1343

1344

80


Complications Most intraoperative complications are similar to those occurring during epilepsy surgery. Respiratory changes such as desaturation or decrease in respiratory rate may occur from oversedation, an intracranial event or seizures. Rapid treatment is required to prevent airway obstruction. The techniques (describe in section above) used will be dependent on the condition of the patient and the expertise of the anesthesiologist. Some patients with brain tumors initially present with a seizure and thus it is possible for the patient to have an intraoperative seizure. However, most seizures, including patients with no prior history, will occur when electrical stimulation is performed during cortical mapping. Treatment includes protection of the patient, and if needed, a small dose of propofol or midazolam. Repeated doses may be required if seizures continue and longer acting anticonvulsants may also be required such as phenytoin. The incidence of nausea and vomiting during awake craniotomy for tumor surgery is very low compared with epilepsy surgery [73,74]. Whether this is a result of the anesthetic drugs used or the type of surgery is unclear. Most tumor surgery does not involve resection or pulling of the deep brain structures, which are known to cause vomiting. Venous air embolism has also been reported [93]. Other common complications include excessive pain, restlessness, agitation, or uncooperativeness. Rarely one may have to convert to general anesthesia.

possible in the older mature child. In children coexisting conditions with multiple organ system involvement or significant psychological, and behavior problems needs to be reviewed preoperatively. As well the parents may be very closely involved in the child’s management and will require consideration as well. Cerebral hemispherectomy and corpus callosotomy are used in children for the treatment of generalized seizures. These procedures are usually done under general anesthesia as they involve a large craniotomy. The major concerns of these lengthy procedures are the possibility of extensive blood loss and air emboli as the surgical site is close to major vessels and sinuses. The process of Gamma Knife radiosurgery for children with a brain tumor or AVM is a long procedure, and most children are under general anesthesia for most of the procedures. The whole process usually involves the application of the head frame, angiography, scanning in the MRI and CT prior to the radiosurgery and traveling between these sites. The anesthetic agents used may be inhalation or intravenous depending on the supply of gases and scavenging within each of the locations. Standard monitoring is required at all sites and during transfers. The usual anesthetic practice differs in the setting of Gamma Knife room as the anesthesiologist must remain outside of the room and can only observe the patient from monitors displaying real-time pictures of the patient and vital signs.

Postoperative Care Pediatric Patients The management of pediatric patients differs in that very few would be able to tolerate an awake procedure. Most functional neurosurgical procedures are performed under general anesthesia including the use of Gamma Knife. Awake craniotomy, especially using the asleep awake asleep technique of anesthesia, and DBS procedures are

In the initial postoperative period all patients should be closely observed in the postanesthetic care unit and then transferred to an observational unit or ward accustomed to caring for neurosurgical patients with functional disorders and epilepsy. Hemodynamic and respiratory parameters and neurological signs should be repeatedly checked and recorded according to the


routine practice of the institution. The patient should be nursed in 30 degree head up position to improve ventilation, reduce face, neck, and airway edema and to facilitate cerebral venous and cerebrospinal fluid drainage. Discharge from the postanesthetic care unit to a suitable location should be only when the patient is stable and discharge criteria have been met. Potential for rapid neurological deterioration requires regular and frequent monitoring of the neurological status. Any change in the neurological status of the patient should be immediately assessed. A CT Scan may be needed to rule out an intracranial hemorrhage. Aggressive rapid management of oxygenation, airway, and any ventilation problem is important considering the detrimental effects of hypoxia and hypercarbia on the brain. Cardiovascular monitoring will help to avoid prolonged episodes of hypotension and hypertension both of which may lead to adverse neurological outcomes. Other routine physiological parameters should be considered and monitored for such as temperature, blood glucose, intravascular volume, osmolality, cerebral perfusion pressure and intracranial pressure. Complications may also occur such as seizures, which require immediate attention to the airway, protection of the patient, and pharmacological treatment with benzodiazepines or propofol initially, and then barbiturates and/or phenytoin. Appropriate postoperative analgesia suitable for each patient is important. Some patients may not require any analgesia; others will have a headache or incision pain with varying intensity. A scalp block that is still present may reduce postoperative pain. Initially intravenous fentanyl is a good analgesic. Codeine still remains the drug of choice in many neurosurgical centers [94,95]. Traditionally, potent opioids such as morphine are avoided due to concerns with hypercarbia, miosis, sedation, and nausea. However, modest intravenous doses are generally believed to be safe. Postoperative nausea and vomiting are common complications in neurosurgical patients but

80

the incidence is markedly reduced in patients undergoing awake procedures. Any standard anti-emetic agents are suitable.

Summary The role of the anesthesiologist will continue to be important in the treatment of patients with functional disorders, epilepsy, and brain tumors as more patients will be treated by these different neurosurgical techniques in stereotactic surgery, and awake craniotomy. With the development of new surgical, imaging and monitoring techniques, the anesthetic management of these patients will also continue to evolve and will require continuous change and development. The appropriate anesthetic management of these patients is critical to the success of the operations and will remain a challenge for the anesthesiologist.

References 1. Pereira EAC, Green AL, Nandi D, et al. Deep brain stimulation: indications and evidence. Expert Rev Devices 2007;4:591-603. 2. Halpern C, Hurtig H, Jaggi J, et al. Deep brain stimulation in neurologic disorders. Parkinsonism Relat Disord 2007;13(1):1-16. 3. Lozano AM, Mahant N. Deep brain stimulation surgery for Parkinson’s disease: mechanisms and consequences. Parkinsonism Relat Disord 2004;10:S49-57. 4. Machado A, Rezai AR, Kopell BH, et al. Deep brain stimulation for Parkinson’s disease: surgical technique and perioperative management. Mov Disord 2006;21 Suppl 14:S247-58. 5. Abosch A, Lozano A. Stereotactic neurosurgery for movement disorders. Can J Neurol Sci 2003;30:S72-82. 6. Marangell LB, Martinez M, Jurdi RA, et al. Neurostimulation therapies in depression: a review of new modalities. Acta Psychiatr Scand 2007;116:174-81. 7. Ollinet C, Bedague D, Carcey J, et al. Functional surgery for movement disorders: implications for anaesthesia. Ann Fr Anesth Reánim 2004;23:428-32. 8. Santos P, Valero R, Arguis MJ, et al. Preoperative adverse events during stereotactic microelectrode-guided deep brain surgery in Parkinson’s disease. Rev Esp Anestesiol Reanim 2004;51:523-30.

1345

1346

80


9. Fa`bregas N, Craen RA. Endscopic and stereotactic neurosurgery. Curr Opin Anesthesiol 2004;17:377-82. 10. Venkatraghavan L, Manninen P, Mak P, et al. Anesthesia for functional neurosurgery. Review of complications. J Neurosurg Anesthesiol 2006;18:64-7. 11. Mason LJ, Cojocaru TT, Cole DJ. Surgical intervention and anesthetic management of the patient with Parkinson’s disease. Int Anesthesiol Clin 1996;34:133-50. 12. Nicholson G, Pereira AC, Hall GM. Parkinson’s disease and anaesthesia. Br J Anaesth 2002;89(6): 904-16. 13. Burton DA, Nicholson G, Hall GM. Anaesthesia in elderly patients with neurodegenerative disorders: special considerations. Drugs Aging 2004;21:229-42. 14. Zanettini R, Antonini A, Gatto G, et al. Valvular heart disease and the use of dopamine agonist for Parkinson’s disease. N Engl J Med 2007;356(1): 39-46. 15. Gillman PK. Monoamine oxidase inhibitors, opioid analgesics and serotonin toxicity. Br J Anaesth 2005;95(4):434-41. 16. Anderson BJ, Marks PV, Futter ME. Propofol: contrasting effects in movement disorders. Br J Neurosurg 1994;8:387-8. 17. Krauss KJ, Akeyson EW, Giam P, et al. Propofol-induced dyskinesias in Parkinson’s disease. Anesth Analg 1996;83:420-2. 18. Bo¨hmdorfer W, Schwarzinger P, Binder, S, et al. Temporary suppression of tremor by remifentanil in a patient with Parkinson’s disease during cataract extraction under local anesthesia. Anaesthesist 2003;52(9):795-7. 19. Watson R, Leslie K. Nerve blocks versus subcutaneous infiltration for stereotactic frame placement. Anesth Analg 2001;92:424-7. 20. Chevrier E, Fraix V, Krack P, et al. Is there a role for physiotherapy during deep brain stimulation surgery in patients with Parkinson’s disease? Eur J Neurol 2006;13:496-8. 21. Heavner JE. Local Anesthetics. Curr Opin Anaesthesiol 2007;20:336-42. 22. Costello TG, Cormack JR, Hoy C, et al. Plasma ropivacaine levels following scalp block for awake craniotomy. J Neurosurg Anesthesiol 2004;16:147-50. 23. Fukuda M, Kameyama S, Noguchi R, et al. Intraoperative monitoring for functional neurosurgery during intravenous anesthesia with propofol. No Shinkei Geka 1997:25:231-7. 24. Murata J, Sawamura Y, Kitagawa M, et al. Minimally invasive stereotactic functional surgery using an intravenous anesthetic propofol and applying image Fusion and AtlasPlan. No To Skinkei 2001;53:457-62. 25. Gray H, Wilson S, Sidebottom P. Parkinson’s disease and anaesthesia. Br J Anaesth 2003;90:524-5. 26. Reynolds LM, Koh JL. Prolonged spontaneous movement following emergence from propofol/nitrous oxide anesthesia. Anesth Analg 1993;76:192-3. 27. Hsu Y-W, Cortinez LI, Robertson KM, et al. Dexmedetomidine pharmacodynamics: I. Crossover comparison

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

of the respiratory effects of dexmedetomidine and remifentanil in healthy volunteers. Anesthesiology 2004;101:1066-76. Rozet I, Muangman S, Vavilala SM, et al. Clinical experience with dexmedetomidine for implantation of deep brain stimulators in Parkinson’s disease. Anesth Analg 2006;103:1224-8. Malteˆte D, Navarro S, Welter ML. Subthalamic stimulation in Parkinson disease: with or without anesthesia? Arch Neurol 2004;61:390-2. Yamada K, Goto S, Kuratsu J, et al. Stereotactic surgery for subthalamic nucleus stimulation under general anesthesia: A retrospective evaluation of Japanese patients with Parkinson’s disease. Parkinsonism Relat Disord 2007;13:101-7. Hertel F, Zu¨chner M, Weimar I, et al. Implantation of electrodes for deep brain stimulation of the subthalamic nucleus in advanced Parkinson’s disease with the aid of intraoperative microrecording under general anesthesia. Neurosurgery 2006;59:1138-45. Easdown LJ, Tessler MJ, Minuk J, Upper airway involvement in Parkinson’s disease resulting in postoperative respiratory failure. Can J Anaesth 1995;42(4): 344-7. Fa`bregas N, Rapado J, Gambu´s PL, et al. Modeling of the sedative and airway obstruction effects of propofol in patients with Parkinson’s disease undergoing stereotactic surgery. Anesthesiology 2002;97:1378-86. Binder DK, Rau GF, Starr PA. Risk factors for hemorrhage during microelectrode-guided deep brain stimulator implantation for movement disorders. Neurosurgery 2005;56:722-32. Gorgulho A, De Salles AA, Frighetto L, et al. Incidence of hemorrhage associated with electrophysiological studies performed using macroelectrodes and microelectrodes in functional neurosurgery. J Neurosurg 2005;102:888-96. Bouhaddi M, Vuillier F, Fortrat JO, et al. Impaired cardiovascular autonomic control in newly and long-term-treated patients with Parkinson’s disease: involvement of L-dopa therapy. Auton Neurosci 2004;116:30- 8. Suarez S, Ornaque I, Fabregas N, et al. Venous air embolism during Parkinson surgery in patients with spontaneous ventilation. Anesth Analg 1999;88:793-4. Moitra V, Permut TA, Penn RM, et al. Venous air embolism in an awake patient undergoing placement of deep brain stimulators. J Neurosurg Anesthesiol 2004;16:321-2. Deogaonkar A, Avitsian R, Henderson MJ. Venous air embolism during deep brain stimulation surgery in an awake supine patient. Stereotact Funct Neurosurg 2005;83(1): 32-5. Jain V, Prabhakar H, Rath GP, et al. Tension pneumocephalus following deep brain stimulation surgery with bispectral index monitoring. Eur J Anaesthesiol 2007;24(2):203-4.


41. Higuchi Y, Iacono RP. Surgical complications in patients with Parkinson’s disease after posteroventral pallidotomy. Neurosurgery 2003;52:558-71. 42. Hatton KW, McLarney JT, Pittman T, et al. Vagal nerve stimulation: overview and implications for anesthesiologists. Anesth Analg 2006;103:1241-9. 43. Davies R.G. Deep brain stimulators and anaesthesia. Br J Anaesth 2005;95(3):424-7. 44. Nutt JG, Anderson VC, Peacock JH, et al. DBS and diathermy interaction induces severe CNS damage. Neurology 2001;56:1384-6. 45. Dagtekin O, Berlet T, Gerbershagen JH. Anesthesia and deep brain stimulation: postoperative akinetic state after replacement of impulse generators. Anesth Analg 2006;103:784. 46. Tan TK, Manninen PH. Anesthesia for stereotactic surgery. Semin Anesth 2000;19:292-9. 47. Bilgin H, Mogol EB, Bekar A, et al. A comparison of alfentanil, fentanyl and remifentanil on hemodynamic and respiratory parameters during stereotactic brain biopsy. J Neurosurg Anesthesiol 2006;18:179-84. 48. Kofke WA, Tempelhoff R, Dasheiff RM. Anesthetic implications of epilepsy, status epilepticus, and epilepsy surgery. J Neurosurg Anesthesiol 1997;9:349-72. 49. Sahjpaul RL. Awake craniotomy: controversies, indications and techniques in surgical treatment of temporal lobe epilepsy. Can J Neurol Sci 2000;27:S55-63. 50. Kofke WA, Dasheiff RM, Dong ML, et al. Anesthetic care during thiopental tests to evaluate epileptic patients for surgical therapy. J Neurosurg Anesthesiol 1993;5:164-70. 51. Herrick IA, Gelb AW. Anesthesia for temporal lobe epilepsy surgery. Can J Neurol Sci 2000;27:S64-7. 52. Arango MF, Steven DA, Herrick IA. Neurosurgery for the treatment of epilepsy. Clin Opin Anaesthesiol 2004;17:383-7. 53. Manninen PH, Contreras J. Anesthetic considerations for craniotomy in awake patients. Int Anesthesiol Clin 1986;24:157-74. 54. Archer DP, McKenna JMA, Morin L, et al. Conscioussedation analgesia during craniotomy for intractable epilepsy: a review of 354 consecutive cases. Can J Anaesth 1988;35:338-44. 55. Aglio LS, Gugino LD. Conscious sedation for intraoperative neurosurgical procedures. Tech Neurosurg 2001;7:52-60. 56. Frost EAM, Booj LHDJ. Anesthesia in the patient for awake craniotomy. Curr Opin Anaesthesiol 2007;20:331-5. 57. Hans P, Bonhomme V. Anesthetic management for neurosurgery in awake patients. Minerva Anestesiol 2007;73:507-12. 58. Welling EC, Donegan J. Neuroleptanalgesia using alfentanil for awake craniotomy. Anesth Analg 1989;68:57-60. 59. Gignac E, Manninen PH, Gelb AW. Comparison of fentanyl, sufentanil and alfentanil during awake craniotomy for epilepsy. Can J Anaesth 1993;40:421-4.

80

60. Hans P, Bonhomme V, Born JD, et al. Target-controlled infusion of propofol and remifentanil combined with bispectral index monitoring for awake craniotomy. Anaesthesia 2000;55:255-9. 61. Berkenstadt H, Perel A, Hadani M, et al. Monitored anesthesia care using remifentanil and propofol for awake craniotomy. J Neurosurg Anesthesiol 2001;13:246-9. 62. Keifer JC, Dentchev D, Little K, et al. A retrospective analysis of a remifentanil/propofol general anesthetic for craniotomy before awake functional brain mapping. Anesth Analg 2005;101:502-8. 63. Herrick IA, Craen RA, Blume WT, et al. Sedative doses of remifentanil have minimal effect on ECoG spike activity during awake epilepsy surgery. J Neurosurg Anesthesiol. 2002;14(1):55-8. 64. Silbergeld DL, Mueller WM, Colley PS, et al. Use of propofol (Diprivan) for awake craniotomies: Technical note. Surg Neurol. 1992;38:271-2. 65. Samra SK, Sneyd JR, Ross DA, et al. Effects of proprofol sedation on seizures and intracranially recorded epileptiform activity in patients with partial epilepsy. Anesthesiology. 1995;82:843-51. 66. Herrick IA, Craen RA, Gelb AW, et al. Propofol sedation during awake craniotomy for seizures: patient-controlled administration versus neurolept analgesia. Anesth Analg. 1997;84:1285-91. 67. Mack PF, Perrine K, Kobylarz E, et al. Dexmedetomidine and neurocognitive testing in awake craniotomy. J Neurosurg Anesthesiol. 2004;16:20-5. 68. Ard JL, Bekker AY, Doyle WK. Dexmedetomidine in awake craniotomy: a technical note. Surg Neurol. 2005;63:114-7. 69. Bekker, AY, Kaufman B, Samir H, et al. The use of dexmedetomidine infusion for awake craniotomy. Anesth Analg. 2001;92:1251-3. 70. Souter MJ, Rozet I, Ojemann JG, et al. Dexmedetomidine sedation during awake craniotomy for seizure resection: effects on electrocorticography. J Neurosurg Anesthesiol. 2007;19:38-44. 71. Oda Y, Toriyama S, Tanaka K, et al. The effect of dexmedetomidine on electrocorticography in patients with temporal lobe epilepsy under sevoflurane anesthesia. Anesth Analg. 2007;105:1272-7. 72. Huncke K, Van de Wiele B, Fried I, et al. The asleepawake–asleep anesthetic technique for intraoperative language mapping. Neurosurgery 1998;42;1312-7. 73. Nikas DC, Danks RA, Black PM. Tumor surgery under local anesthesia. Tech Neurosurg. 2001;7:70-84. 74. Skucas AP, Artru AA. Anesthetic complications of awake craniotomies for epilepsy surgery. Anesth Analg. 2006:102:882-7. 75. Scuplak SM, Smith M, Harkness WF. Air embolism during awake craniotomy. Anaesthesia 1995;50:338-40. 76. Sato K, Shamoto H, Yoshimoto T. Severe bradycardia during epilepsy surgery. J Neurosurg Anesthesiol 2001:13:329-32.

1347

1348

80


77. Smith M, Smith SJ, Scott CA, et al. Activation of the electrocorticogram by propofol during surgery for epilepsy. Br J Anaesth 1996;76:499-502. 78. Herrick IA, Craen RA, Gelb AW, et al. Propofol sedation during awake craniotomy for seizures: electrocorticographic and epileptogenic effects. Anesth Analg 1997;84:1280-4. 79. Manninen PH, Burke SJ, Wennberg R, et al. Intraoperative localization of an epileptogenic focus with alfentanil and fentanyl. Anesth Analg 1999;88:1101-6. 80. McGuire G, El-Beheiry H, Manninen PH, et al. Activation of electrocorticographic activity with remifentanil and alfentanil during neurosurgical excision of epileptogenic focus. Br J Anaesth 2003;91:651-5. 81. Watts AD, Herrick IA, McLachlan RS, et al. The effect of sevoflurane and isoflurane anesthesia on interictal spike activity among patients with refractory epilepsy. Anesth Analg 1999;89:1275-81. 82. Kurita N, Kawaguchi M, Hoshida T, et al. The effects of sevoflurane and hyperventilation on the electrocorticogram spike activity in patients with refractory epilepsy. Anesth Analg 2005;101:517-23. 83. Danks RA, Rogers M, Aglio LS, et al. Patient tolerance of craniotomy performed with the patient under local anesthesia and monitored conscious sedation. Neurosurgery 1998;42:28-36. 84. Taylor MD, Bernstein M. Awake craniotomy with brain mapping as the routine surgical approach to treating patients with supratentorial intraaxial tumors: a prospective trial of 200 cases. J Neurosurg 1999;90:35-41. 85. Danks RA, Aglio LS, Gugino LD, et al. Craniotomy under local anesthesia and monitored conscious sedation for the resection of tumors involving eloquent cortex. J Neurooncol 2001;49:131-9.

86. Berkenstadt H, Ram Z. Monitored anesthesia care in awake craniotomy for brain tumor surgery. IMAJ 2001;3:297-300. 87. Blanshard HJ, Chung F, Manninen PH, et al. Awake craniotomy for removal of intracranial tumor: considerations for early discharge. Anesth Analg 2001;92:89-94. 88. Johnson KB, Egan TD. Remifentanil and propofol combination for awake craniotomy: case report with pharmacokinetic simulations. J Neurosurg Anesthesiol 1998;10:25-9. 89. Manninen PH, Balki M, Lukitto K, et al. Patient satisfaction with awake craniotomy for tumor surgery: a comparison of remifentanil and fentanyl in conjunction with propofol. Anesth Analg 2006:102:237-42. 90. Sarang A, Dinsmore J. Anaesthesia for awake craniotomy evolution of a technique that facilitates neurological testing. Br. J Anaesth 2003;90:161-5. 91. Tongier WK, Joshi GP, Landers DF, et al. Use of laryngeal mask airway during awake craniotomy for tumor resection. J Clin Anesth 2000;12:592-4. 92. Fukaya C, Katayama Y, Yoshino A, et al. Intraoperative wake-up procedure with propofol and laryngeal mask for optimal excision of brain tumour in eloquent areas. J Clin Neurosci 2001;8:253-5. 93. Balki M, Manninen PH, McGuire, et al. Venous air embolism during awake craniotomy in a supine patient. Can J Anaesth 2003;50:835-8. 94. Leslie K, Williams DL. Postoperative pain, nausea and vomiting in neurosurgical patients. Curr Opin Anaesthesiol 2005;18:461-5. 95. Roberts GC. Post-craniotomy analgesia: current practices in British neurosurgical centres-a survey of postcraniotomy analgesic practices. Eur J Anaesthsiol 2005;22:328-32.

77 Evoked Potentials in Functional Neurosurgery J. L. Shils . J. E. Arle

Introduction Intraoperative evoked potentials have a long history in functional neurosurgery and have become a critical component in the success of many of these procedures. Functional neurosurgery seeks to modify or augment neural function in such a way as to bring about a clinical benefit for the patient. Prior to the availability of highquality microelectrode recording (MER) amplifiers, evoked potentials and their associated evoked responses were the sine qua non for reliably localizing relevant functional deep brain targets. Yet, even as MER and imaging capabilities advance, evoked potentials and response characterization remain an important aspect of localization and evaluation during stereotactic and functional procedures. An evoked response is defined as a measured response of the nervous system that is directly produced by either an external or internal stimulus. The stimulus may be mechanical, auditory, visual, thermal, or electrical and may either directly activate neural elements, as during electrical stimulation, or may indirectly activate neural elements as during mechanical movement of a joint or limb. Recording an evoked response is performed by placing a transducer at a point ‘‘downstream’’ from the stimulus that is either: (1) directly effected by the stimulus, or (1) located where the transducer can record far-field effects of the stimulation. The recording can be obtained either in a ‘‘free-running’’ mode or in a ‘‘triggered’’ mode. During functional neurosurgical procedures, both free-running and triggered responses are measured. The electrical recording #


of the evoked response is defined as the ‘‘evoked potential.’’ Evoked responses are used for localization during all modern surgical procedures for treating movement disorders and pain. Although the thalamus and its associated nuclei have been the primary targets for evoked potential and evoked response testing [1–28], all targets utilize evoked responses in one form or another. For example, Klostermann et al. [29] first used visual evoked potentials (VEPs) during pallidotomies to determine the location of the optic tract in relation to the base of the internal globus pallidus (GPi). Others have since also used VEPs to localize the optic tract during pallidal surgeries [30–34]. Median nerve somatosensory evoked potentials (SSEPs) have been studied for surgeries targeting the subthalamic nucleus (STN) [28,29,35–37]. And, additionally, free-running evoked responses are currently used for all surgical targets during both MER and non-MER methodologies in stereotactic targeting. Intraoperative language mapping uses a type of inverse evoked response whereby the stimulation is used directly on neural elements to disrupt briefly and help localize functionally relevant language regions in cortex during tumor removal or epilepsy surgery. Finally, motor cortex stimulation therapies use both peripherally activated evoked potentials and cortically activated evoked potentials, allowing the patient to be operated on under general anesthesia. We begin here with a succinct history of evoked response testing during stereotactic and functional procedures, followed by detailed methodologies used during evoked response and evoked potential testing in STN, GPi, thalamic, and

1256

77

Evoked potentials in functional neurosurgery

cortical procedures. Emphasis will be on technical aspects of equipment used for evoked potentials, and specific aspects of intraoperative technique.

History Since the late 1930s, when Meyers [38,39] was performing campotomies, physiology played a role in localizing critical structures1. Initially, intra-operative physiology consisted of direct electrical brain tissue stimulation evoked responses that were visually interpreted at the periphery. However, as technology (both in electronics and surgical methodology) improved, the quality and utility of intra-operative physiology increased. Dawson, in 1951, was the first to describe a reliable technique for detecting small evoked potentials in noise [42,43]. The first digital evoked potential device was based on the Computer of Average Transients in the 1960s which was specifically designed to compute recurring transient waveforms, improving the signal-to-noise ratio [44]. This device was primarily used to analyze evoked potential data and study weak NMR profiles in solid state physics. Another important early use of evoked responses includes work by Bucy [45] who in 1941 operated on two patients, one for severe abnormal movements in the left arm and leg and the other for ‘‘tremor’’-like movements of the right extremities following a traumatic fall and 14 week coma. The surgical procedure removed parts of the primary motor cortex to alleviate the abnormal movements. Bucy hoped to minimize trauma to the face and speech. Intraoperative stimulation to evoke responses in the limb muscles was performed, using 60 Hz, 2 ms, monophasic stimuli up to 20 V in the anesthetized patient. After mapping the cortex

1

It should be noted that many of Hans Berger’s early EEG studies were done on the exposed cortex from trauma [40] and pioneering work by Walker [41] in the 1940’s utilized electrocortocography for the treatment of epileptics.

with this technique, the cortical areas representing the upper limb where the abnormal movements were seen, were removed. The patient woke from surgery without tremor, but with right upper extremity paralysis. Over the next few weeks the tremor returned to the right leg and hand area while voluntary movements of the hand never returned. Sugita and Doi [46] investigated tremor augmentation and synchronization using cortical evoked responses (EEG evoked potential maximum) to localize the areas most important in tremor driving. By stimulating at the level of the anterior commissure-posterior commissure (AC-PC) line, in the ventral lateral nucleus of the thalamus, they could demonstrate a low amplitude evoked response in the sensorimotor cortex (contralateral frontoparietal region) with minimal tremor driving in the contra-lateral hand. At more dorsal locations, within this nucleus, they were able to demonstrate higher amplitude evoked responses with a higher degree of tremor driving. Of note [46] is the fact that when the stimulation was turned off, this stimulation frequency synchronized tremor activity continued for a few seconds, slowly returning to the patient’s resonant tremor frequency. Ablation at this area demonstrated the best control of tremor in patients with extrapyramidal disorders and continues to the present as one of the primary targets for DBS relief of tremor. In the 1960s Hassler et al. [47] discussed the absolute necessity of intraoperative evoked electrical stimulation testing prior to the making of any lesions in the brain. Also, of note is his discussion of the effect on the motor state (situation – e.g., resting, active, raised. . .) and consciousness of the patient with stimulation. Again, they noted that low frequency (4–8 Hz) stimulation will drive the tremor even during pallidal stimulation while high frequency stimulation (25–100 Hz) will reduce tremor [47]. More recent analysis of DBS treatment of tremor in the VIM region by Benabid and colleagues demonstrated a similar


high frequency efficiency in tremor suppression, specifically at frequencies greater than 100 Hz [48]. In 1963 Spiegel and Wycis presented their investigations of stimulation evoked responses as a result of stimulation in the thalamic part of Forel’s H field (prerubral area (H3)) [49]. Increases in tremor amplitude were noted with frequencies of 20 and 30 Hz, while inhibition of tremor was noted at stimulation frequencies greater than 100 Hz. They describe conjugate eye movements and various vegetative effects such as changes in heart rate and pupil dilation, but do not describe the stimulation parameters or location of these responses in any detail. There have also been interesting studies utilizing the peripheral H-reflex and the effect of thalamic stimulation on this measure [50,51]. The H-reflex consists of electrical stimuli delivered to afferent Ia sensory fibers which returns through the alpha motor fibers after traversing the monosynaptic connection in the ventral gray of the spinal cord. Laitinen and Ohno [51] found that with lesioning or continuous stimulation to the VL thalamus there was neither consistent facilitation nor inhibition of the H-reflex, but with single pulse thalamic stimulation, using a 2.5 ms delay relative to the H-reflex stimulation pulse, there was a distinct facilitation of the H-reflex. However, Stern et al. [50] found a suppression with continuous thalamic stimulation. It is interesting to note that neither group investigated this as a potential tool for functional localization during the procedure. Much early work on evoked potentials and neurophysiologic studies during functional surgery can be attributed to Albe-Fessard and colleagues [2–9] who developed many techniques for the practical application of evoked responses in the operating room and localization of evoked potential generators. These techniques include detailed microelectrode recordings and evaluation of kinesthetic and voluntary joint movements,

77

sensory responses, thalamic evoked potential generators and their use in localization in the thalamus. In one study, by Yamshiro et al. [2] median nerve SSEPs were recorded in the ventral caudal (VC) and ventral intermediate (VIM) nuclei of the thalamus. The most interesting finding was that they were able to demonstrate a difference in the P15 and N20 latencies with VC recordings of about 0.5 ms later than the VIM latencies. In another study, Albe-Fessard [3] found no latency difference between the thalamus and cortex when recording from an electrode in the VP area. There was an amplitude gradient in the P15 as the electrode moved through the base of the VC nucleus (increase) below the AC-PC line, out of the nucleus. There was no direct correlation to the P15 peak and the primary tremor area, unfortunately, as this would have been very helpful as a localization tool. Fukushima et al. [52] performed several SSEP studies in multiple thalamic nuclei utilizing not only wrist median nerve stimulation, but stimulation at other body locations including the shoulder, torso, ulnar notch, and the posterior tibial nerve at the ankle. They found the largest and most peaked response to peripheral stimulation in the VC nucleus, but could not determine a somatotopic arrangement due to the large overlap between SSEP responses from the different body locations across multiple thalamic nuclei and including the zona incerta and the subthalamic nucleus [52]. Fager [53] used 2.0 V and 25 Hz stimulation to determine a safe distance from the internal capsule when placing lesions in the thalamic or subthalamic areas by noting either delayed responses to stimulation which are considered safe, or non-delayed direct responses to stimulation which would indicate stimulation of the internal capsule. For these procedures Fager was not able to demonstrate benefit with the stimulation, but was able to use the direct motor evoked responses elicited from the internal capsule as an indicator of safe lesioning distance from the capsule.

1257

1258

77


Important Aspects of Equipment Used to Stimulate and Generate Evoked Responses Electrode Design Evoked potentials during functional neurosurgery are recorded using three primary types of electrodes: (1) microelectrodes, (2) permanent DBS electrodes, and (3) strip electrodes (> Figure 77-1). Each type of electrode has a distinct geometry and size, and so each recording property will differ. More importantly, the electrical properties and frequency response characteristics of each electrode type will require different amplifiers and conditioning electronics to (1) minimize artifact, (2) improve signal to noise ratios, and (3) optimize energy transfer. Clinical microelectrodes used today are manufactured either from tungsten, or platinum-iridium (Pt-Ir). The tips are etched to micron dimensions and then soldered to a semi-rigid stainless steel insulated wire. Though Pt-Ir electrodes have better signal transfer properties and lower impedances, in the frequency range of interest for evoked response and single unit recordings [54], tungsten is a mechanically stronger material and thus more suited to the harsh environment provided by ancillary operating room services. An evoked potential recording from the center of the ventral intermediate thalamic nuclei (trajectory shown in lower right figure inset) with stimulation of the median nerve is shown in > Figure 77-2. The recording electrode used was a Pt-Ir microelectrode (> Figure 77-1b). The reference electrode in the third trace is located at the tip of the guide canula (> Figure 77-1e) while the reference electrode in the fourth trace is at the 10–20 location Fpz. The top trace is the classical cortical SSEP recording (C3’-Fpz) and the second trace is the sub-cortical montage channel (CV-Fpz). For the recordings shown

in > Figure 77-2, as the electrode moves ventrally, the distance between the active and reference electrode increases and thus the active tip continuously moves away from the stationary reference electrode. This type of recording is called a referential recording. Differences between active electrode positions with referential recording will display as amplitude changes in each of the recordings. One potential drawback with referential recordings, however, is that large artifacts located at the referential electrode can both obscure biologic signals of interest at the active electrode and affect the ‘‘true’’ signal generator response at all positions of the active electrode recording. Utilizing a bipolar recording method keeps both the active and referential electrodes at the same distance by simultaneously moving both along the tract. The bipolar electrode recording technique allows the testing of responses by utilizing a common inter-electrode distance. Evoked potential generators can be localized more easily with this method because phase changes (the primary morphology of such a generator location) are more readily detected. Also, far-field potential effects are minimized with the bipolar montage. The above descriptions demonstrate the difference between referential (monopolar) and bipolar recording montages. In the referential case, changes in neural generator effects will manifest as changes in either amplitude or phase of responses, (> Figure 77-3a) while changes in generator location in the bipolar case will be seen as a phase change in (> Figure 77-3b). One advantage of referential recording is the ability to discriminate large spatial generators that may range over a distance within the same general range as the distance between the two recording surfaces. When changing from microelectrodes to the DBS or paddle type electrodes, (> Figure. 77‐1c, d), for example, one can reliably record both a referential and bipolar montage and thus benefit from the advantages of both methods.


77

. Figure 77-1 Examples of different electrodes used for evoked potential recordings during functional neurosurgical procedures. (a) A glass coated Pt-Ir electrode used for micro-electrode recording (Universal electrode, Dothan, AL). (b) A glass coated Pt-Ir electrode used for micro-electrode recordings (FHC, Bowdoinham, ME). (c) Example of a tungsten, nonglass coated, semi-microelectrode (Integra, Burlington, MA). (d) The deep brain stimulating electrode with 1.5 mm long electrodes separated by 1.5 mm (Medtronic model 3387, Minneapolis, MN). (e) The micro-electrode/guide tube system. (f) Paddle electrodes used for cortical or extra-dural surface recordings (Medtronic Resume electrodes, Minneapolis, MN)

Amplifier Design Once the electrode transduces the neural signal into an electrical signal it is passed through an amplifier/conditioning system. Most importantly, the input stage of the amplifier used for

microelectrodes needs to be configured differently than the input stage of an amplifier used to record from larger electrodes. This is related to the fact that the equivalent circuit of all electrodes have both resistive and capacitive elements and, therefore, electrodes themselves will

1259

1260

77


. Figure 77-2 Evoked potentials recorded from the scalp and a microelectrode placed in the VIM nucleus of the thalamus. The top trace is the standard cortical montage C3’-Fpz and the second trace is the sub-cortical montage channel CV-Fpz. The third trace shows the response in the thalamus at a distance of 5.0 mm from the base of the VIM nucleus. The distance from the tip of the recording electrode to the canula is 11.0 mm. The bottom trace shows the same active micro-recording electrode yet referenced to the scalp Fpz electrode. This figure is the combination of two recording windows from the Cadwell Cascade intra-operative evoked potential system

color the recorded bio-potential, primarily as a function of frequency (> Figure 77-4 [54,55] and > equation 77-1). > Figure 77-4 shows the equivalent electrical circuit of a metal microelectrode and the frequency response of this electrode type. The primary frequency, and thus impedance, effects are due to the electrode-electrolyte interface (Cma,, Rma, and Ema) and its capacitive nature, with minor contributions from the reference electrode. It should be noted that the distributed capacitance can be an issue with high frequency

noise, and in areas with a lot of wireless traffic this should be considered. Xc ¼

1 2pfC

ð1Þ

The size, shape and material of the electrode recording surface are all important factors that effect resistive and capacitive parameters and are the basis of these electrical differences. Therefore, it is critical that the design of the amplifier’s internal electronics do not adversely modify the signal characteristics; yet if the amplifier’s input stage could ‘‘correct’’ some of the distortion that would also be desirable. For example, microelectrodes act as low frequency filters [56], but as most noise is in the low frequency range, the amplifier electronics can potentially compensate for this if designed properly. Another important design feature of the amplifier relates the impedance of the electrode and its resistive/capacitive properties to the energy transfer from the electrode to the amplifier. The impedance of microelectrodes is typically between 500 kO and 1.5 MO. To maximize energy transfer, the front-end impedance of the amplifier should be at least one and preferably two orders of magnitude greater than the impedance of the electrode. To accomplish this, the amplifier front-end is designed with different configurations of filters and other components to match the characteristics of the electrode. Once the signal is conditioned by the amplifier front-end, the rest of the electronics can be of a more common design. This is particularly important now since most signals today are converted into the digital domain for analysis. Thus, current amplifier design has two primary goals [57]: (1)

(2)

Obtaining very high input impedance to minimize input loading and thus reduce loss of signal transfer from the electrode to the amplifier Assuring that the electrode/amplifier circuit is as highly resistive as possible to minimize frequency related distortions to the original bioelectrical signal


77

. Figure 77-3 Referential (a) and bipolar (b) evoked potential recordings in the VIM nucleus of the thalamus using a Medtronic model 3387 electrode. The top trace represents the most ventral electrodes. Note the large amplitude change and relative phase shift with electrode 0 just at the base of the thalamus in the referential montage. The bipolar montage shows an amplitude change, but no phase shift since no contact is fully out of the nucleus

Common Functional Neurosurgical Procedures Using Evoked Potentials Subthalamic Nucleus Procedures Functional neurosurgical procedures that involve targeting the subthalamic region and the subthalamic nucleus (STN) utilize evoked response testing in two ways: (1) recording in the STN region while performing kinesthetic and voluntary manipulation of the patient’s limbs, and

(2) stimulating in the STN region and looking for responses. > Figure 77-5 shows an example of single unit activity changes (firing rate increase) during flexion testing of the elbow joint in a patient with idiopathic Parkinson’s disease. The neural evoked response consists of increased firing rate each time the elbow is flexed which typically can be heard as well as seen. Location of this activity is often of significant value for locating the proper placement of the implanted stimulating electrode in the sensorimotor area of this nucleus.

1261

1262

77


. Figure 77-4 Equivalent circuit (modified from [54]) of the extracellular microelectrode recording circuit. Rma, Cma, and Ema represent the metal microelectrode-electrolyte interface parameters. Cd is the distributed capacitance that exists between the shaft of the metal microelectrode and the extracellular fluid resistance Rexc. Rmb, Emb and Cmb are the reference electrodes resistance, potential and capacitance respectively. Finally Cw represents the capacitance between the leads of the electrode. The Figure also shows the experimentally recorded frequency response of the metal micro-electrode [55]

. Figure 77-5 Evoked responses recorded with a Pt-Ir microelectrode showing the change in firing rate each time the patient’s wrist is flexed. The top trace is a recording of the accelerometer that is placed on the patient’s wrist. The shaded areas highlight the responses


To be considered appropriate for DBS electrode placement, our surgical protocol seeks a minimum of three distinct evoked responses in at least 4.5 mm of recorded STN under ideal circumstances. In addition to the neural recorded evoked responses to peripheral manipulation, electrical stimulation of the STN region also may be tried. Electrical stimulation in the STN and nearby regions is performed to evoke responses in peripheral muscles. One method uses stimulation threshold testing to help delineate the proximity of the electrode to the corticospinal tract by observing and recording time-locked triggered motor activity in the periphery. Another method looks for changes (both improvement and decrement) in the Parkinsonian symptoms via therapeutic stimulation. Symptomatic observations include changes in tremor, rigidity, speech, eye movements, and bradykineisia. Initial patient preparation and surgical set-up have been previously described [58]. Once MER or Semi(S)-MER has started, the STN is localized using both single/multi-unit physiology and recorded evoked responses as follows:

77

When a single unit is located, evoked response testing of the contralateral joints is begun. For the upper limb the wrist, elbow and shoulder are tested. For the lower limb, the ankle and knee are tested, independently, while the hip is usually tested with the knee. Similar to the wrist example of movement causing a firing rate increase in the related single unit, > Figure 77-6 shows a similar example related to ankle movement. The exact mechanism and pathway for joint activation of the STN is not known, though there are various theories about this action [59]. The most important aspect of these responses, for intra-operative localization purposes, is the fact that these evoked responses are both repeatable and indicative of the electrode being in the sensorimotor region of the STN. Our median number of cells per STN encountered tract where STN cells are recorded is five and the average number of trajectories per patient side is 1.7. In the STN, kinesthetic cells are mostly found in the dorsal and middle regions of the nucleus with few found in the ventral area, similar to findings of Rodriguez-Oroz et al. and Starr et al. [60,61].

. Figure 77-6 Evoked responses recorded with a Pt-Ir microelectrode showing the change in firing rate each time the patient’s ankle is flexed. The top trace is a recording of the accelerometer that is placed on the patient’s ankle. The shaded areas highlight the responses

1263

1264

77


As well, we have noted evoked responses mainly located between 9.0 and 14 mm lateral to the midline, corresponding to the sensorimotor area of the STN [60]. Consistent with data from Starr et al. [61], we have also found that the sensorimotor leg region is located more medial than the upper limb region and can be very useful in determining a subsequent movement of the electrode if necessary. During MER in the STN we have yet to find stimulation useful in evoking either beneficial or detrimental evoked responses, due primarily, we believe, to the very focal nature of micro stimulation. The exception to this observation has been testing to determine whether the electrode tip is located in the corticospinal tract. To test for corticospinal tract location, we use very low frequency stimulation (1–5 Hz) at 50–100 mA. Due to the small volume affected by micro-stimulation and the high current density at the tip, a microelectrode can elicit MEP responses in limb or face muscles. EMG electrodes placed on these regions will pick up these responses. We have not performed detailed current versus distance relationships since we commonly do not target the corticospinal tract, yet we find support for these findings from data mapping corticobulbar and corticospinal tracts in monkeys with micro stimulation [62,63]. In contrast, others [64] have found microstimulation useful in the determination of optimal DBS electrode placement. Even though we do not use micro stimulation in the STN, we find macro-stimulation evoked response testing to be critical for proper DBS placement. Stimulation parameters used for this testing are 60 ms and 185 Hz and voltages up to 4.0 V. Due to the design of the Screener Stimulation Device (Model 3682, Medtronic Inc., Minneapolis, MN) we can only test in bipolar mode. Newer devices, such as the ANS MTS trial system (Model 6510, Advanced Neuromodulation Systems, Plano, TX) allow for monopolar testing and constant current

testing. Our standard stimulation sequence includes testing each sequential electrode pair up to 4.0 V. We choose 4.0 V since this is greater than any value we have used in therapeutic stimulation with patients. Also, due to the voltage doubling circuit in the IPG we would rather avoid having the stimulator set above 3.7 V. In cases where we may obtain questionable evoked responses, specifically increased rigidity, or the potential need for tremor reduction at higher voltages, we will test a larger stimulation field using contact patterns 0,1,2,+3 and +0,1,2,3 and up to 4.0 V. Several have used median nerve SSEPs to localize the STN [28,29,35–37]. Pinter et al. [35] discussed the usefulness of SSEPs for localizing STN while Klostermann et al. concluded that SSEPs could not be helpful in localizing STN [29]. More recently Kitagawa et al. [37] did detailed analysis of SSEPs in the subthalamic region finding that electrodes placed posterior to the STN, in the ZI and medial lemniscus (ML), demonstrated a phase reversal at the ZI/ML boarder, while median nerve SSEPs, with the electrode in the STN proper, demonstrated no phase reversal at different locations in the STN [37]. Kitagawa et al. also discussed the fact that these are far-field potentials in this region but also there may be a potential contribution to the P16/N18 from the ZI in the STN region due to excitatory post-synaptic potentials in the ZI region. Dinner et al. [65] found peaks in the STN mirroring the far-field cortical P14(P16)/N18 in all referential DBS electrode recordings lacking specificity. A larger response is noted in the DBS contact 1, but no specific location of this electrode is given which in turn makes using this data for localization purposes undesirable. However, appreciating that the electrode is in the STN as opposed to another region lack of a phase reversal, or definitive amplitude peak localization would still be difficult. Since the primary goal of functional surgery in the STN is to modify extrapyramidal components of the system, it is critical to assure that the stimulation is focused


on the sensorimotor segment of the STN and associated sub-thalamic fiber tracts. We have found that SSEPs in this region are very ambiguous and offer no real benefit in localizing the proper functional region of the STN for final DBS localization. > Figure 77-7 shows an example of both bipolar and monopolar recordings from the DBS lead (ANS Libra Deep Brain Stimulation System model, Plano, TX). There are no specific defining features of the evoked potentials to help differentiate the STN from

77

other structures nearby, yet a potential N18 phase shift is noted. Note that the monopolar recordings show no amplitude gradient, as referenced to Fz, with contacts in the STN (0 and 1) and those above the STN (2 and 3). The bipolar recording shows almost a flat line near the 20–25 ms latency demonstrating the similar amplitudes in all channels. There is some phase reversal at 18 ms. in the electrodes around the base of the thalamus, similar to what Kitigawa had discussed [37].

. Figure 77-7 Referential (a) and bipolar (b) evoked potential recordings during placement of a DBS lead in the STN. The recordings are from the DBS lead (Medtronic model 3387, Minneapolis, MN). The top traces represent the most ventral electrodes by looking at the referential recording. It is difficult to see amplitude changes indicating there is no N20 generator nearby, as is known in the bipolar trace there is a small phase shift at a latency of 16 ms. in the most dorsal electrode contacts. This could be related to the electrodes at the dorsal edge of the ZI and the ventral edge of the thalamus

1265

1266

77


Internal Globus Pallidal Procedures Surgeries to modify the firing properties of the GPi include DBS and lesioning procedures for the treatment of Parkinson’s disease, Dystonia, Tourette’s, and multiple other movement related disorders [66–71]. With the internal capsule located medial and posterior to the GPi and the optic tract ventral to the GPi, recording of evoked responses is critical to assuring correct lesion or DBS placement. As with the STN procedures, joint movement evoked responses are helpful in localizing the functionally relevant area of the nucleus. As such, all cells recorded in the GPi are evaluated by testing the wrist, elbow, shoulder, ankle, knee, and hip motions of patients. There is lingering debate about the somatotopic nature of the GPi [72–74]. Guridi et al. observed a physiologically defined somatotopy of kinesthetic cells with the face and arm region located ventrolaterally and the leg dorsomedially [72]. Taha et al. found the leg in-between the arm in a rostral caudal axis (the arm representation is concave) [73]. Vitek et al. have found the leg to be medial and dorsal with respect to the arm while the face is more ventral [74]. For PD procedures, the evoked responses to joint manipulation are usually linked to only one joint while for dystonia procedures we have found multiple joints will at times generate an evoked response in a single cell. This data is consistent with some published findings [75], although others have found little multi-joint activity in a single-unit’s receptive field [76]. It should be noted that when testing for kinesthetic evoked responses one needs to be careful not to manipulate the patient in such a way as to shake the recording system generating spurious signals that can be incorrectly identified as inhibitory responses. Also, manipulation with vascular artifact needs to be carefully appreciated. The most common complication in early experience with posterior ventral pallidotomy

procedures was a visual field cut [77]. Mindful of this, locating the optic tract became paramount in reducing the morbidity of pallidotomy. Yokoyama et al. [30] and Tobimatsu et al. [31] used visual evoked potentials to localize the optic tract. Yokoyama et al. used LED goggles to present alternating flash stimuli [30] at various depths until the optic tract was located. The optic tract was localized when the evoked response amplitude increased in a manner similar to what was found in cat studies this group had previously performed [32]. We do use visual evoked responses, recorded at the end of the micro-recording tract. At the point the microelectrode passes the base of the GPi we turn the lights out in the operating room and in 0.5 mm increments stop the electrode and shine a flashlight on and off in a patients eyes. We then listen predominantly to the background noise of the electrode. When we are within 0.5 mm of the optic tract we hear high frequency changes in the neural noise. The closer the electrode tip gets to the optic tract the louder the change. For pallidotomies we would ideally like to be at least 1 mm from the optic tract when placing a lesion. For DBS procedures we try to keep the electrode 2–3 mm from the optic tract. Prior to removing the micro-electrode, stimulation is performed to activate the optic tract. The stimulation parameters are 300 Hz and 100 ms. If the electrode is in the optic tract, optic responses can be noted down to 10 mA. With the micro-electrode we find that we can be only 0.5 mm away from the optic tract even at a stimulation amplitude of 100 mA. Thus, using micro-electrode stimulation is not an optimal technique from a safety standpoint. During the recent pallidotomy era (roughly 1990–2000), after MER and prior to placing the lesion, we would stimulate with the lesioning probe using a 300Hz, 100 ms monophasic and/ or biphasic pulse as the stimulation current is slowly raised. If a visual response is evoked at a stimulation under 1 mA or 2.0 V, the probe is moved dorsally to avoid damage to the optic


tract. During DBS procedures, we stimulate through the lowest two contacts of the DBS lead using the Medtronic screener box. Stimulation is raised to 4 V at 210 ms and 130 Hz. If there are no visual evoked responses, defined as phosphenes or visual hallucinations, then the electrode is considered safe in position relative to the optic tract. It is important to note the difference in stimulation voltages between the two electrode types and stimulation devices. A second evoked response is used during pallidal surgeries to test for proximity to the internal capsule. Anatomically, the internal capsule is located posterior and medial to the GPi (> Figure 77-8). Capsular responses are more critical during DBS surgery due to the larger spread of electrical effects than with lesioning [58]. During pallidotomies it was important to localize the distance from the tip of the lesioning probe to that of the internal capsule for two reasons. First, to assure that the lesion would not impinge on the internal capsule and cause a

. Figure 77-8 Axial MRI image demonstrating the relationship of the Internal Capsule to the globus pallidus. The internal capsule is shaded in

77

permanent deficit. Second, contractions in the nasal labial fold are important to demonstrate the appropriate location for the lesioning tip [78]. Utilizing the CF-3 Lesion Generator (Integra, Inc., Burlington, MA) with a 3 mm long by 1.5 mm lesioning probe, the voltage is quickly raised to a maximum stimulation of 6 V while observing the nasal labial fold and contralateral thumb. If stimulation does not induce any changes up to 2.5 V, it is considered safe for the placement of a lesion. Moreover, if stimulation at 5–6 V induces a contraction of the contralateral nasal-labial fold and/or contraction of the contralateral thumb, then its position is considered optimal for lesioning. Such testing is performed prior to all lesions. Note that many patients will have a mild weakness in their nasal-labial fold for approximately 2–4 weeks following surgery. This would be an indicator of a properly placed lesion. This is due to a transient edematous zone around the lesion that will resolve over time. With DBS procedures, stimulation takes place once the DBS electrode is inserted, after MER is completed. At this point it is critical to make sure that therapeutic stimulation will not cause adverse events that will inhibit use of the therapy. Similar to the responses during the recording phase of the surgery, the primary evoked responses are motor and visual in nature. In dystonia patients there may also be some sensory related phenomena that are not as readily elicited in PD patients. Evoked response testing is done in similar fashion to STN testing utilizing a sequential bipolar pattern starting with contacts set at 0, 1+. The pulse width for PD patients is 60 ms, while for dystonia patients it is set at 330 ms. This difference is related to settings used for therapy. It is highly unlikely that PD therapy will require a 330 ms pulse, while it is possible that this may be needed for dystonia therapy. The testing frequency for both PD and dystonia is 130 Hz. First, optic tract evoked responses are tested by turning the OR lights off and asking

1267

1268

77


the patient to describe any visual phenomena they may have while raising the voltage. These responses may include ‘‘flashing lights,’’ ‘‘hazy lights,’’ or even visual hallucinations. Stimulation amplitudes are slowly raised to 5.0 V for dystonia patients and 4.0 V for PD patients. If optic responses are noted during cathodal stimulation of contact 0, but not for contact 1, then the electrode will be raised 1.0 mm before securing in place. Following these details in our procedures we have never had optic tract responses when stimulating above contact 0 in the operating room, nor any optic tract adverse events during therapeutic stimulation.

Thalamic Procedures Of all the deep brain structures involved in functional localization, the thalamus is most readily designed for both evoked response testing and evoked potential monitoring. This is because the thalamus contains multiple evoked potential generators and because microelectrode, semimicroelectrode and macro-electrodes can all record evoked responses (event related responses). Additionally, evoked responses are particularly helpful because the single-unit signatures within various thalamic nuclei are difficult to differentiate purely on firing pattern alone. The primary thalamic targets for both movement disorders and pain are in the ventral thalamus. Each of these areas has been neurophysiologically studied in detail [79]. Because the ventral oralis anterior (VOA) and the ventral oralis posterior (VOP) have similar microelectrode recording characteristics [79] it is critical to use evoked potential and evoked response testing when targeting this area. The ventral caudal nucleus (VC), which contains sensory receiving cells and lies just posterior to the ventral intermediate (VIM) nucleus, also has similar recording characteristics to the other three regions. Evoked response testing in the thalamus is separated into both recorded responses and stimulated responses. At our center we do all microelectrode recordings

moving dorsal to ventral while all stimulation testing is done in reverse during the removal of the electrode. There are historically four major naming conventions for the motor thalamic nuclei [80]: Jones [81], Hassler [82,83], Ilinsky and Kultal-Ilinsky [84], and Olszewski [85]. In this discussion we choose to use the nomenclature of Hassler [82,83]. For a helpful cross-reference between each of the nomenclatures see table 1 in reference [80] of this chapter. Beginning with the most posterior of the ventral nuclei, the VC nucleus is the primary receiving area for the spinothalamic and medial lemniscus tracts, both sensory. Microelectrode recorded evoked responses occur in response to light touch. > Figure 77-9 shows two examples of evoked responses in the VC thalamus to light brushing of the thumb. Receptive fields of the cells in this region vary in size with the region of the body. Areas of large receptive fields (4–5 cm) that will evoke a response in a single cell include the shoulder, torso and abdomen whereas areas of very focal (1 cm or less) receptive fields include the thumb, fingertips, mouth and jaw. Cellular responses to evoked skin stimuli in this region are arranged somatotopicaly. Responses follow a medial to lateral alignment (> Figure 77-10 [86]), with mouth responses most medial and leg responses most lateral. > Figure 77-11 shows the map of the responses from two patients who had very focal temporal head pain in which a DBS lead was being placed in VC for pain relief. The medial lateral angles for these two cases were 83 and 85 degrees. One can see the sagittal overlap of some of the more diffuse areas such as the forehead and temporal head area with those of the focal areas of the thumb and lip when stimulating with the DBS lead (1.1 mm diameter) as compared to the micro-electrode (10 mm). Despite identifying the precise area, it was impossible to alleviate any focal temporal head pain without also causing parasthesias in the hand and mouth areas of these patients. One patient was able to tolerate the extra sensations in his


77

. Figure 77-9 Two examples of microelectrode recordings in the VC nucleus of the thalamus showing evoked responses to light brushing of the thumb. In the bottom trace expanded time tracings are shown to demonstrate the increased firing rate

. Figure 77-10 Lateral representation of the ventral posterior nuclei of the thalamus. Taken from [86]

hand and mouth with a minor reduction in head pain, while the other patient could not tolerate the extra sensations and the therapy had to be halted. Work by Tasker and Kiss have found that intra-oral representation is between 10–11 mm from midline, face tactile cells are located between 11–14 mm lateral from midline, manual digits 13–16 mm from midline, and foot representation 15–17 mm

lateral to midline [79]. It is important to note that this overall representation is only found in the ventral one half to two-third of the nucleus [79]. Giorgi et al. found a similar arrangement with the addition of fingers being somewhat more caudal and posterior to the face and hand and the hip being just lateral to the shoulder with the thigh; foot and leg having minimal representation [87]. Recording of evoked responses is important during surgeries for the treatment of pain and for the treatment of tremor. Even though the final electrode for tremor is placed in the VIM nucleus, body representations in both VC and VIM are similar. Therefore, knowing the location in the VC can be an important guide for a subsequent electrode trajectory. Moreover, finding the VIM/ VC boarder is critical for tremor therapy since the best location for tremor treatment has been found just inside VIM at this border [88,89]. The VIM nucleus, along with the VOP nucleus, is considered the primary receiving nucleus from the dentate and anterior interposed nuclei of the cerebellum [90]. Similar to organization within the VC nucleus, recorded activity is primarily found in the ventral two-third of the

1269

1270

77


. Figure 77-11 Stimulation response locations from two patients who were being operated on for the treatment of focal temporal pain. Patient 1 (black tracts) was injured in a thirty foot fall while at work. Patient 2 (grey tracts) started to have pain after a previous surgery. In both cases post-operative stimulation caused sensations in the hand region at therapeutic levels for alleviation of pain in the temporal region

nucleus. A similar medial-to-lateral somatotopic arrangement is found in this nucleus with face being medial and leg lateral. Kinesthetic responses are the main evoked responses recorded in this region. > Figure 77-12 is an example of such an evoked response. Mechanical stimuli such as the squeezing of the muscle bellies, or the passive movement of joints will evoke response during MER and S-MER in this region. Observations by the authors find the receptive fields in this region to be larger than those of the VC, which is consistent with findings of others [79]. This observation is likely due to the fact that this area receives information from joints rather than from skin surface. In contrast to the purely excitatory nature of the VC evoked responses, VIM evoked responses can be either excitatory (demonstrating an increase in amplitude or firing frequency of a unit) or inhibitory. Inhibitory responses (> Figure 77-13) are found less frequently than excitatory ones, occurring in about 10–15% of the recordings. Lenz et al. has described both fast and slow adapting cells in the VIM to passive joint movements [23]. Thus,

when testing joints, one should test for a minimum of five manipulations. Table 4 in Lenz et al. [23] describes the distribution of evoked responses within different thalamic nuclei. VC stimulation evokes responses at low amplitude levels as compared to evoked responses from stimulation in VIM. Microelectrode stimulation of 5–10 mA can evoke sensations in the face and hand areas. Slightly higher stimulation (20–40 mA) is typically required to evoke responses in the proximal limbs or torso. AlbeFessard et al. [2,3] originally found responses in the VC during micro-stimulation studies from as low as 4 mA whereas at a distance of 2 mm, 100 mA would not generate an evoked response. Ohara et al. [14] performed detailed micro stimulation studies in the VC nucleus and found vibratory sensations ventral to the movement receptive fields in the upper extremity and face. Contrary to this study, however, we have found sensory evoked responses, defined as a tingling, numbness, vibration, or electrical shock, in all areas of the VC thalamus that are adjacent to the VIM, and have used this critical information


77

. Figure 77-12 A wrist kinesthetic excitatory evoked response. The top trace is the raster plot for the full test, the second trace is the raw data from the microelectrode recording while the third trace is the accelerometer on the hand. The middle raster plots are peri-event histograms representing the firing rates from 0.5 s before the motion of the wrist to 0.5 s after the wrist motion. At time 0 one can see the increase in firing frequency each time the wrist is flexed. The spike rate histogram at the bottom also shows this effect

in planning a next trajectory of the microelectrode or final electrode. If we do not find tremor cells in a MER trajectory, then VC responses are the next most important response. Using the fact that the somatotopy in the VC and VIM region is the same in a sagittal plane allows this mapping to be done and is a technique that has been used by others [91]. Macro electrode stimulations of 0.1–0.5 mV will, many times, evoke stimulation in the VC when using the DBS lead. When this occurs in an area where there are sensory related single units and that is also the area with the distal limb tremor, then the limb tremor will almost always disappear for a short period of time but return. The parasthesias in that area will most likely stay.

On the other hand if the macro electrode is placed in the center of a tremorgenic zone [89] (in the VIM area) a transient parasthesia may occur at low stimulation levels, but will disappear within a short time, while the distal limb tremor will continue to be reduced or eliminated. VIM stimulation evoked responses are more complex, however, than in the VC. During surgeries for tremor, the primary goal of surgery is to stop tremor, but other motor effects are not the primary intention. In the authors’ experience, tremor reduction or cessation is seen at levels that are much less than those seen for motor responses. Micro stimulation in the VIM at levels reaching 40–100 mA will cause parasthesias in the contralateral body and the majority of these response

1271

1272

77


. Figure 77-13 A wrist kinesthetic inhibitory evoked response. The top trace is the raster plot for the single unit recordings, the second trace is the raw data for the microelectrode recording while the third trace is the accelerometer on the hand. The middle raster plots are peri-event histograms representing the firing rates from 1 s before the motion of the wrist to 1 s after the wrist motion. At time 0 one can see the decrease in firing frequency each time the wrist is flexed. The spike rate histogram at the bottom also shows this effect

fields are larger than for VC responses, consistent with past studies[79]. These responses are most easily seen in the face and distal upper extremity. Investigators have also noted stimulation evoked movement sensations, even though no actual movement has occurred [92]. Macro-stimulation, however, evokes transient sensory responses and can also evoke sensations of dizziness or nausea [79]. In our experience these responses are more likely to occur at or close to the time of surgery. After surgery, dizziness and nausea tend to diminish over time and finally disappear. Most peripheral nerve evoked potential studies, and clinical testing related to functional neurosurgery, has occurred during surgeries in the thalamus. > Figure 77-14 shows the normal

path of the median nerve evoked potential with the presumed evoked response generators and their wave pattern. An important question remains although the VC thalamus is thought to be one of the SSEP generator sites, as to whether this knowledge can be used to localize a target during functional surgery. A basis for utilizing the SSEP as a localization tool is the ability to record the near-field components of the resulting waves. The VC nucleus would be the only thalamic nucleus where this could be performed because it is a primary component of the SSEP [93]. To use far field recordings as a consistent localizing tool, either multiple recordings would be needed to triangulate the source, or multiple studies differentiating location would need to be done.


. Figure 77-14 Shows the median nerve SSEP pathway from the spinal cord to the cortex and the representative median nerve SSEP traces. Note the N20 peak comes from the cortex, while the N18/P18 peak comes from the thalamus (Drawing used with permission from MJ Shils)

Albe-Fessard and colleagues performed multiple studies of median nerve SSEPs in the VC and VIM nuclei of the thalamus [2] and could demonstrate a difference in the P15 and N20 latencies with the VC recordings set about 0.5 ms later than the VIM latencies. > Figure 77-15 shows the results of microelectrode recorded evoked potentials in a trajectory that passes through the VIM and VC of an essential tremor patient. The microelectrode used was a Pt-Ir glass coated tip (impedance of 800 kO) with a diameter of 10 mm (FHC model MTBPBN(EN2), Bowdoin, ME). Recordings

77

were started at 7.4 mm above the base of the thalamus in the VIM. Stimulation was at the median nerve of the contralateral wrist utilizing 4.7 Hz stimulation, 300 s pulse width, 10 mA and 50 averages. The bandwidth of the signal was from 30 to 750 Hz. The signal was referenced to the canula that is 16.1 mm from the base of the thalamus (15 mm from the original stereotactic target). The amplitude of the N20 starts off low in the VIM and reaches a peak just at the entry to VC (2.2 mm). At the base of the thalamus the amplitude once again decreases. One primary difference in this recording is the large N20 response and small P18 response, possibly due to the reference including the whole length of the canula and not just its tip. Also, this recording comes primarily from VIM and the most anterior portion of the VC nucleus, not the more posterior part. > Figure 77-16 shows the same patient with the evoked potential being recorded from the Medtronic model 3387 DBS lead in a slightly different trajectory than the one used in > Figure 77-15. Stimulation parameters were the same as above. Note the large P18 in the contact 0 derivation, receiving more signal from the posterior part of the VC nucleus while in the other leads the P18 is indistinguishable from background though a lower amplitude N20 is seen. Post-operative testing on this patient demonstrated low intensity parasthesias when using contact 0 in either the monopolar (0.3 V) or bipolar (0.7 V) configuration with a 185 Hz/60 s pulse width signal. Tremor suppression was noted at contact 2 cathode and case positive with a stimulation voltage of 2.0 V. Stimulation in contact 1 was able to reduce the tremor, but caused non-transient parasthesias. This example portrays the utility of evoked potentials in thalamic functional surgery. The primary problem with using SSEPs in the awake patient is both speed and patient comfort. It can take 30–60 s to produce a reliable SSEP and many patients find the stimulation sensations to be disconcerting, whereas simple brush stroke

1273

1274

77


. Figure 77-15 Median nerve evoked potentials recorded in the VIM nucleus of the thalamus at different depths during microelectrode recordings. The active microelectrode tip is referenced to Fpz. No phase shift is noted in this recording due to the Fpz reference

and limb movement evoked responses can be done in about 5 s.

Cortical Localization and Stimulation Procedures Motor cortical surgeries for the treatment of pain and movement disorders likely began in the 1940s with Bucy’s direct extirpation of the primary motor cortex [94,95]. Bucy stimulated the cortex to identify appropriate regions for removal. Currently, we use two evoked responses: (1) somatosensory evoked phase-reversal potential mapping (> Figure 77-17a), and (2) direct epidural motor stimulation evoked response mapping (> Figure 77-17b). Intra-operative

phase reversal mapping makes use of the orientation of the pyramidal cells of the cortex and the location of the sensory cortex in relation to that of the motor cortex. Knowing that the cortical generator (N20/P25) of the median nerve somatosensory evoked potential is in the somatosensory cortex [93] allows us then to locate the motor strip just anterior across the central sulcus (> Figure 77-17a). One of the N20 dipoles is located on the posterior wall of the central sulcus [96]. Thus, recording on one side of the sulcus gives one polarity while recording from the other side shows the opposite polarity. The technique using both SSEP and motor mapping has been described in the literature [97]. Briefly, the electrode is placed parallel with and essentially overlying the M1 strip of


77

. Figure 77-16 Referential (a) and bipolar (b) median nerve evoked potentials recorded from the DBs electrode (Medtronic model 3387, Minneapolis, MN). The top trace in both montages are recorded from the most distal electrodes. There is a large amplitude change from contact 0 to contact 1, in the referential recording, indicating that contact 0 may be either out of the thalamus or just on the border of the thalamus. For the bipolar recording case there is a minor phase shift moving from contacts 0–1 to contacts 1–2. This indicates that contact 0 is probably just on the board of the thalamic nucleus

cortex and ultimately sutured directly to the dura to avoid post-operative electrode movement (> Figure 77-18). Somatosensory testing consists of placing the 4-contact Resume lead (Medtronic Inc., Minneapolis, MN) on the dura in several directions, mostly perpendicular to the suspected pre-central gyrus. Median and Ulnar nerve somatosensory evoked potentials (SSEPs) are then observed (Cascade EP machine, Cadwell Laboratories, Inc., Kennewick, WA) using a 20 mA, 100 ms monopolar square pulse at a rate of 4.32 Hz. SSEPs were recorded from the Resume lead in both a bipolar (contact 0–1, 1–2, 2–3) and monopolar (all referenced to the 10–20 location of Fz) recording montage. The central sulcus is the point where the N20 response phase reverses (> Figure 77-17a). This is performed across multiple locations to follow the course of the central sulcus. Motor mapping consists of placing an anodal 5 mm stimulation ball probe (Model E1564, Valleylab, Gosport, UK) over the M1 area referenced to a cathode placed at Fz. Stimulation

consists of trains of 5 stimuli each, at an interstimulus duration of 4 ms, a 500 ms pulse width. Stimulation amplitudes were slowly increased at each location starting at 5 mA and increasing to a maximum of 25 mA. Stimulation then stops when the first EMG responses are noted. EMG needles are placed in bipolar fashion (separated by 2 cm in all muscles except the orbicularis oris and orbicularis oculi) in the orbicularis oculi, orbicularis oris, trapezius, deltoid, biceps, triceps, flexor carpi ulnaris, abductor policis brevis, first dorsal interossious, quadraceps, anterior tibialis, and abductor hallicus muscles. Stimulation is performed with a Grass S-88 and two SIU-7 constant current stimulus isolators (Astromed-Grass, West Warwick, RI) and responses are recorded on a Cadwell Cascade (Cadwell Laboratories, Inc., Kwennewick, WA). > Figure 77-18 shows an example of the findings in the extensor and APB muscles with this technique. In this way two types of intraoperative physiology corroborate to locate the precentral gyrus.

1275

1276

77


. Figure 77-17 Example of phase reversal mapping during surgery for motor cortex stimulation. The top left traces show both the referential and bipolar phase reversals. The bottom left trace shows the electrode on the dural surface. The right image shows a representation of the location of each electrode over the cortical surface. Looking at the electrode position it can be seen that the P20 responses come from electrodes over the motor cortex, while the N20 responses come from the electrodes over the sensory cortex

The normal homunculus of the motor strip has face and upper limb on the lateral surface with leg on the superior edge and medial surface of the cortex. During our standard mapping cases with focal stimulation we do not generally activate the lower limb muscles. In one patient, a car accident had caused a brachial plexus avulsion and left her with no motor or sensory function below the elbow. The rationale for the surgery was to treat intense pain in the deltoid region. The accident occurred five years prior to the surgery and since then the motor cortex representation had shifted in relation to the new peripheral inputs. In this case, thigh and torso were adjacent to the shoulder and then the face following. The thigh and upper leg was now more lateral than in the normal motor cortex.

In pain patients we have found it beneficial to place the stimulating lead over the cortical motor region that is involved in the pain area. In our experience, we do not derive much benefit for lower limb pain and these findings are similar to some prior studies [98,99], but different from others findings [100,101]. In a study by Tsubokawa et al. [98] they noticed that to benefit patients with lower extremity pain they needed to use stimulation intensities that actually generated motor responses. In the patient described above, although we sought out the deltoid region specifically, we found face, then deltoid, abdominal muscles, and then leg, moving superiorly from a starting position lateral on the cortex. Tumor and epilepsy surgeries utilize both sensory and motor evoked potentials to map


77

. Figure 77-18 Example of responses obtained during motor mapping performed during surgery for the placement of a motor cortex stimulator. Each area is labeled as the location where a response is obtained

out eloquent areas of cortex prior to and during resections. In the case of tumor and epilepsy resections, the dura is open and the cortex is visible, which may make one think that locating the central sulcus should be an easy task visually, yet due to the structural abnormality normal cortical arrangement is often distorted. In order to determine true functional cortex, mapping remains a necessary tool, even with new complex imaging algorithms. Similar to motor cortical stimulation cases, median and ulnar nerve SSEPs are used to localize the central sulcus. Multiple positions are tested with either 4 or 6 contact strips of 64 channel grids. In a study with 230 patients, Romsto¨ck et al. [102] found that not only is the N20-P20 phase reversal important, but in large tumors located centrally and postcentrally, later waves (at 35 and 30 ms) are important as well. They found a reduction in amplitude in 18% of patients or complete loss of these waves, yet in about 55% of this group evaluation of potentials between 25 and 30 ms continued to make SSEP mapping a viable tool

with only 8% of patients not being able to be mapped intra-operatively. As with SSEP localization techniques, motor evoked potential mapping is needed to localize functional tissue in anatomy that may not conform to the textbooks. There are two primary methods to stimulate the cortex for mapping. The first is the one described above in the cortical stimulation section utilizes a very short train (5–9 pulses or 20–40 ms) of high frequency (250 Hz) energy [103]. The second method utilizes a long train (2–5 s) of low frequency (50–60 Hz) energy which was originally described by Penfield and Boldrey [104]. Reports of seizures, however, with the Penfield technique approximately 24% [104,105] while with the short train technique vary from 1.2 [106] to 1.6% [107] in patients with a history of epilepsy. As Szeleńyi et al. point out in their review of the literature there, is no greater incidence in patients with a history of seizures as compared to those without a history of seizures in either technique [107]. In one paper by Roux et al. [108] they found good central correlation between fMRI and

1277

1278

77


intraoperative motor mapping. One interesting conclusion from the paper was that fMRI localized a larger area of cortex than direct cortical stimulation mapping, thus potentially reducing the amount of tumor the surgeon could remove near the margins.

Conclusion Despite advances in imaging, functional imaging, and multi-channel micro-electrode recording, evoked response testing is still critically important in localizing functional neurosurgical targets. With the advent of neuromodulation therapies and the fact that all of these targets currently are for functional clinical treatment, evoked response testing not only helps in localization, but also in understanding how the therapy may work in a particular patient. Though pure evoked potential studies may have now taken a back-seat to micro-electrode recording, the use of evoked response testing during these procedures is still a vital component of their success. Evoked response testing consists of both peripheral manipulation of the sensory and motor systems to generate a response in the brain and the electrical, or even mechanical manipulation of the neural tissue to evoke responses in the periphery. Even as functional neurosurgery breaks new ground in neuromodulation therapies for treatment of depression, addiction, anorexia, obesity, obsessive compulsive disorder, etc, it will be the functional responses (i.e., evoked responses), be it beneficial or adverse, that may still remain an important measure of how well these therapy works [109,110].

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

References 14. 1. Lin YC. Lenz FA. Distribution and response evoked by micro stimulation of thalamus nuclei in patients with dystonia and tremor. Chin Med J 1994;107(4):265-70. 2. Yamashiro K, Tasker RR, Iwayama K, Mori K, AlbeFessard D, Dostrovsky JO, Chodakiewitz JW. Evoked

15.

potentials from the human thalamus: correlation with micro stimulation and single unit recording. Stereotact Funct Neurosurg 1989;52(2–4):127-35. Albe-Fessard D, Tasker R, Yamashiro K, Chodakiewitz J, Dostrovsky J. Comparison in man of short latency averaged evoked potentials recorded in thalamic and scalp hand zones of representation. Electroencephalogr Clin Neurophysiol 1986;65(6):405-15. Albe-Fessard D. Electrophysiological methods for the identification of thalamic nuclei. Zeitschrift fur Neurologie 1973;205(1):15-28. Albe-Fessard D, Arfel G, Guiot G, Derome P, Guilbaud G. Thalamic unit activity in man. Electroencephalogr Clin Neurophysiol 1967;Suppl 25:132. Albe-Fessard D, Arfel G, Guiot G, Derome P, Hertzog E, Vourc’h G, Brown H, Aleonard P, De la Herran J, Trigo JC. Electrophysiological studies of some deep cerebral structures in man. J Neurol Sci 1966;3(1):37-51. Albe-Fessard D, Arfel G, Guiot G, Derome P, Dela H, Korn H, Hertozog E, Vourch G, Aleonard P. Characteristic electric activities of some cerebral structures in man. Annales de Chirurgie 1963;17:1185-214. Santibanez G, Trouche E, Albe-Fessard D. Study of the evolution, in the course of time, of the amplitude of cortical and thalamic activities evoked by repetitive somatic stimuli. Journal de Physiologie 1963;55:335-6. Krauthamer G, Albe-Fessard D. Inhibition of induced cortical and sub cortical activities by stimulation of the capsule. Comptes Rendus des Seances de la Societe de Biologie et de Ses Filiales 1961;155:1443-50. Lenz FA, Gracely RH, Baker FH, Richardson RT, Dougherty PM. Reorganization of sensory modalities evoked by micro stimulation in region of the thalamic principal sensory nucleus in patients with pain due to nervous system injury. J Comp Neurol 1998;399 (1):125-38. Lin YC, Lenz FA. Effective response evoked by micro stimulation of thalamus nuclei in patients with tremor. Chin Med J 1993;106(5):372-4. Lenz FA, Dostrovsky JO, Kwan HC, Tasker RR, Yamashiro K, Murphy JT. Methods for micro stimulation and recording of single neurons and evoked potentials in the human central nervous system. J Neurosurg 1988;68 (4):630-4. Lee JI, Ohara S, Dougherty PM, Lenz FA. Pain and temperature encoding in the human thalamic somatic sensory nucleus (Ventral caudal): inhibition-related bursting evoked by somatic stimuli. J Neurophysiol 2005;94(3):1676-87. Ohara S, Weiss N, Lenz FA. Micro stimulation in the region of the human thalamic principal somatic sensory nucleus evokes sensations like those of mechanical stimulation and movement. J Neurophysiol 2004;91(2):736-45. Ohara S, Lenz FA. Medial lateral extent of thermal and pain sensations evoked by micro stimulation in somatic


16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26. 27. 28.

29.

sensory nuclei of human thalamus. J Neurophysiol 2003;90(4):2367-77. Garonzik IM, Hua SE, Ohara S, Lenz FA. Intraoperative microelectrode and semi-microelectrode recording during the physiological localization of the thalamic nucleus ventral intermediate. Mov Disord 2002;17 Suppl 3: S135-44. Hua SE, Garonzik IM, Lee JI, Lenz FA. Microelectrode studies of normal organization and plasticity of human somatosensory thalamus. J Clin Neurophysiol 2000;17 (6):559-74. Lee J, Dougherty PM, Antezana D, Lenz FA. Responses of neurons in the region of human thalamic principal somatic sensory nucleus to mechanical and thermal stimuli graded into the painful range. J Comp Neurol 1999;410 (4):541-55. Zirh TA, Reich SG, Perry V, Lenz FA. Thalamic single neuron and electromyographic activities in patients with dystonia. Adv Neurol 1998;78:27-32. Hua SE, Lenz FA, Zirh TA, Reich SG, Dougherty PM. Thalamic neuronal activity correlated with essential tremor. J Neurol Neurosurg Psychiatry 1998;64(2):273-6. Mandir AS, Rowland LH, Dougherty PM, Lenz FA. Microelectrode recording and stimulation techniques during stereotactic procedures in the thalamus and pallidum. Adv Neurol 1997;74:159-65. Lenz FA, Gracely RH, Romanoski AJ, Hope EJ, Rowland LH, Dougherty PM. Stimulation in the human somatosensory thalamus can reproduce both the affective and sensory dimensions of previously experienced pain. Nat Med 1995;1(9):910-3. Lenz FA, Kwan HC, Martin RL, Tasker RR, Dostrovsky JO, Lenz YE. Single unit analysis of the human ventral thalamic nuclear group. Tremor-related activity in functionally identified cells. Brain 1994;117 (Pt 3):531-43. Lenz FA, Seike M, Richardson RT, Lin YC, Baker FH, Khoja I, Jaeger CJ, Gracely RH. Thermal and pain sensations evoked by micro stimulation in the area of human ventrocaudal nucleus. J Neurophysiol 1993;70 (1):200-12. Tasker RR, Gorecki J, Lenz FA, Hirayama T, Dostrovsky JO. Thalamic microelectrode recording and micro stimulation in central and deafferentation pain. Appl Neurophysiol 1987;50(1–6):414-7. Bakay RAE, Vitek JL, DeLong MR. Thalamotomy for tremor. Neurosurg Atlas 1992;2(4):290-312. Lozano A. Thalamic deep brain stimulation for the control of tremor. Neurosurg Atlas 1998;7:125-135. Hanajima R, Dostrovsky JO, Lozano AM, Hutchison WD, Davis KD, Chen R, Ashby P. Somatosensory evoked potentials (SEPs) recorded from deep brain stimulation (DBS) electrodes in the thalamus and subthalamic nucleus (STN). Clin Neurophysiol 2004;115(2):424-34. Klostermann F, Vesper J, Curio G. Identification of target areas for deep brain stimulation in human basal ganglia

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40. 41.

42. 43.

44.

45.

77

substructures based on median nerve sensory evoked potential criteria. J Neurol Neurosurg Psychiatry 2003;74 (8):1031-5. Yokoyama T, Sugiyama K, Nishizawa S, Yokota N, Ohta S, Uemura K. Visual evoked potentials during posteroventral pallidotomy for Parkinson’s disease. Neurosurgery 1999;44(4):815-24. Tobimatsu S, Shima F, Ishido K, Kato M. Visual evoked potentials in the vicinity of the optic tract during stereotactic pallidotomy. Electroencephalogr Clin Neurophysiol 1997;104(3):274-9. Sugiyama K, Yokoyama T, Yamamoto T, Uemura K. Localization of the optic tract by using sub cortical visual evoked potentials (VEPs) in cats. Clin Neurophysiol 1999;110(1):92-6. Lozano A, Hutchison W, Kiss Z, Tasker R, Davis K, Dostrovsky J. Methods for microelectrode-guided posteroventral pallidotomy. J Neurosurg 1996;84(2):194-202. Yokoyama T, Sugiyama K, Nishizawa S, Yokota N, Ohta S, Yamamoto S, Uemura K. Visual evoked oscillatory responses of the human optic tract. J Clin Neurophysiol 1999;16(4):391-6. Pinter MM. Characteristics of SEP from subthalamic nucleus in human. Electroencephalogr Clin Neurophysiol Suppl 1999;50:364-72. Pesenti A, Priori A, Locatelli M, Egidi M, Rampini P, Tamma F, Caputo E, Chiesa V, Barbieri S. Subthalamic somatosensory evoked potentials in Parkinson’s disease. Mov Disord 2003;18(11):1341-5. Kitagawa M, Murata J-I, Uesugi H, Hanajima R, Ugawa Y, Saito H. Characteristics and distribution of somatosensory evoked potentials in the subthalamic region. J Neurosurg 2007;107:548-54. Meyers R. Surgical interruption of the pallidofugal fibers: its effect on the syndrome of paralysis agitans and technical considerations in its application. NY State J Med 1942;42:317-25. Hayne RA, Meyers R, Knott J. Characteristics of electrical activity of human corpus striatum and neighboring structures. J Neurophysiol 1949;12:185-95. Berger H. Uber das Elektrenzephalogramm des Menschen. Arch, Psychiat. Nevenkrankh 1929;87:527-70. Walker AE, Marshall C, Beresford EM. Electrocortocographic characteristics of the human cerebrum in posttraumatic epilepsy. Assoc Res Nerv Ment Dis 1947;26:502-15. Dawson GD. A summation technique for detecting small signals in large irregular background. J Physiol 1951:115:2-3 Dawson GD. A summation technique for the detection of small evoked potentials. Electroencephalogr Clin Neurophysiol 1954;6(1):65-84. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Meth 2004:9–21. Bucy PC. The surgical treatment of extrapyramidal diseases. J Neurol Neurosurg Psychiatry 1951;14:108-17.

1279

1280

77


46. Sugita K, Doi T. The effects of electrical stimulation on the motor and sensori systems during stereotaxix operations. Confin Neurol 1967;29:224-9. 47. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger JA. Physiological observations in stereotaxic operations in extrapyramidal motor disturbances. Brain 1960;83:337-54. 48. Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM, Hommel M, Perret JE, de Rougemont J. Long-term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991; 337:403-6. 49. Spiegel EA, Wycis HT, Szekely EG, Soloff L, Adams J, Gildenberg P, Zanes C. Stimulation of Forel’s field during stereotaxic operations in the human brain. Electroencephalogr Clin Neurophysiol 1963;16:537-48. 50. Stern J, Mendell J, Clark K. H reflex suppression by thalamic stimulation and drug administration. J Neurosurg 1968;29:393-6. 51. Laitinen LV, Ohno Y. Effects of thalamic stimulation and thalamotomy on the H reflex. Electroencephalogr Clin Neurophysiol 1970;28:586-91. 52. Fukusima T, Mayanagi Y, Bouchard G. Thalamic evoked potentials to somatosensory stimulation in man. Electroencephalogr Clin Neurophysiol 1976;40:481-90. 53. Fager CA. Evaluation of Thalamic and Subthalamic surgical lesions in the alleviation of Parkinson’s disease. J Neurosurg 1968;28:145-9. 54. Geddes LA. Electrodes and the measurement of bioelectric events. New York: Wiley; 1972. 55. Shils JL, Patterson T, Stecker MM. Electrical properties of metal microelectrodes. Am J Electrodiag Technol 2000;40:71-82. 56. Shils JL, Patterson T, Stecker MM. Electrical properties of metal microelectrodes. Am J Diagn Technol 2000;40:71-82. 57. Gedded LA and Baker LE, Principles of applied biomedical instrumentation. New York: Wiley; 1968. 58. Shils JL, Tagliati M, Alterman RL. Neurophysiological monitoring during surgery for movement disorders. In: Deletis V, Shils JL, editors. Neurophysiology in Neurosurgery: a modern intraoperative approach. New York: Academic Press; 2002. 59. Mink JW. The basal ganglia: focused selection and inhibition of competing motor programs. Prog Neurobiol 1996;50(4):381-425. 60. Rodriguez-Oroz MC, Rodriguez M, Guridi J, Mewes K, Chockkman V, Vitek J, DeLong MR, Obeso JA. The subthalamic nucleus in Parkinson’s disease: somatotopic organization and physiological characteristics. Brain 2001;124(9):1777-90. 61. Starr PA, Theodosopoulos PV, Turner R. Surgery of the subthalamic nucleus: use of movement-related neuronal activity for surgical navigation. Neurosurgery 2003;53 (5):1146-9; discussion 1149.

62. Follett KA, Mann MD. Effective stimulation distance for current from macro electrodes. Exp Neurol 1986;92 (1):75-91. 63. Alterman RL, Sterio D, Beric A, Kelly PJ. Microelectrode recording during posteroventral pallidotomy: impact on target selection and complications. Neurosurgery 1999;44 (2):315-21; discussion 321–3. 64. Pollak P, Krack P, Fraix V, Mendes A, Moro E, Chabardes S, Benabid AL. Intraoperative micro- and macro-stimulation of the subthalamic nucleus in Parkinson’s disease. Mov Disord 2002;17 Suppl 3:S155-61. 65. Dinner DS, Neme S, Nair D, Montgomery EB, Baker KB, Rezai A, Lu¨ders HO. EEG and evoked potential recordings from the subthalamic nucleus for deep brain stimulation of intractable epilepsy. Clin Neurophysiol 2002;113:1391-402. 66. Temel Y, Visser-Vandewalle V. Surgery in Tourette syndrome. Mov Disord 2004;19(1):3-14. 67. Visser-Vandewalle V, Ackermans L, van der Linden C, Temel Y, Tijssen MA, Schruers KR, Nederveen P, Kleijer M, Boon P, Weber W, Cath D. Deep brain stimulation in Gilles de la Tourette’s syndrome. Neurosurgery 2006;58(3):E590. 68. Alterman RL, Snyder BJ. Deep brain stimulation for torsion dystonia. Acta Neurochir Suppl 2007;97 (Pt 2):191-9. 69. Vidailhet M, Vercueil L, Houeto JL, Krystkowiak P, Benabid AL, Cornu P, Lagrange C, Tezenas du Montcel S, Dormont D, Grand S, Blond S, Detante O, Pillon B, Ardouin C, Agid Y, Destee A, Pollak P. French Stimulation du Pallidum Interne dans la Dystonie (SPIDY) study group. Bilateral deep-brain stimulation of the globus pallidus in primary generalized dystonia. New Engl J Med 2005;352(5):459-67. 70. Zhuang P, Li Y, Hallett M. Neuronal activity in the basal ganglia and thalamus in patients with dystonia. Clin Neurophysiol 2004;115(11):2542-57. 71. Coubes P, Cif L, El Fertit H, Hemm S, Vayssiere N, Serrat S, Picot MC, Tuffery S, Claustres M, Echenne B, Frerebeau P. Electrical stimulation of the globus pallidus internus in patients with primary generalized dystonia: long-term results. J Neurosurg 2004;101 (2):189-94. 72. Guridi J, Gorospe A, Ramos E, Linazasoro E, Rodriguez MC, Obeso JA. Stereotactic targeting of the globus pallidus internus in Parkinson’s disease: imagining versus electrophysiological mapping. Neurosurgery 1999;45:278-89. 73. Taha JM, Favre J, Baumann TK, Burchile KJ. Characteristics and somatotopic organization of kinesthetic cells in the globus pallidus of patients with Parkinson’s disease. J Neurosurg 1996;85:1005-12. 74. Vitek JL, Bakay RAE, Hashimoto T, Kaneoke Y, Mewes K, Yu Zhang J, Rye D, Starr P, Baron M, Turner R, Delong MR. Microelectrode guided pallidotomy:


75.

76.

77.

78.

79. 80.

81. 82.

83.

84.

85. 86.

87.

88.

Technical approach and application in medically intractable Parkinson’s disease. J Neurosurg 1998;88:1027-43. Vitek JL, Chockkan V, Zhang JY, Kaneoke Y, Evatt M, DeLong MR, Triche S, Mewes K, Hashimoto T, Bakay RA. Neuronal activity in the basal ganglia in patients with generalized dystonia and hemiballismus. Ann Neurol 1999;46(1):22-35. Hutchison WD, Lang AE, Dostrovsky JO, Lozano AM. Pallidal neuronal activity: implications for models of dystonia. Ann Neurol 2003;53(4):480-8. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76(1):53-61. Alterman RL, Kelly PJ. Pallidotomy technique and results: the New York University experience. Neurosurg Clin N Am 1998;9(2):337-43. Tasker RR, Kiss ZHT. The role of the thalamus in functional neurosurgery. Neurosurg Clin N Am 1995;6(1):73-194. Krack P, Dostrovsky J, Ilinsky I, Kultas-Ilinsky K, Lenz F, Lozano A, Vitek J. Surgery of the motor thalamus: Problems with the present nomenclatures. Mov Disord 2002;17(S3):S2-8. Jones EG. The thalamus. New York: Plenum; 1985. Hassler R. Anatomy of the Thalamus. In: Schaltenbrand G, Bailey P, editors. Introduction to stereotaxis with an atlas of the human brain. Theime; Stuttgart: p. 230-90. Hassler R. Architectonic organization of the thalamic nuclei. In: Schaltenbrand G, Walker EA, editors. Stereotaxy of the human brain: anatomical, physiological and clinical applications. New York: Theime; 1982. p. 140-80. Ilinsky IA, Kultas-Ilinsky K. Neuroanatomy organization and connections of the motor thalamus in primates. In: Kultas-Ilinsky K, Ilinsky IA, editors. Basal ganglia and thalamus in health and movement disorders. New York: Plenum; 2001. p. 77-91. Olszewski J. The thalamus of Macacca mulata. An atlas for use with stereotactic instruments. Basel: Karger; 1952. Albe-Fessard D. Evoked potential recording in functional neurosurgery: part I: use of thalamic evoked potentials to improve the stereotactic localization of electrodes in the human brain. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw-Hill; 1998. Giorgi C, Kelly PJ, Eaton DC, Guiot G, Dermoe G. A study of the tridimentional distribution of somatosensory evoked responses in human thalamus to aid the placement of stimulating electrodes for treatment of pain. Acta Neurochir Suppl 1980;30:279-87. Atkinson JD, Collins DL, Bertrand G, Peters TM, Pike GB, Sadikot AF. Optimal location of thalamotomy lesions for tremor associated with Parkinson’s disease: a probabilistic analysis based on postoperative magnetic resonance imaging and integrated digital atlas. J Neurosurg 2002;96 (5):854-66.

77

89. Lenz FA, Normand SL, Kwan HC, Andrews D, Rowland LH, Jones MW Seike M Lin YC, Tasker RR, Dostrovskey JO, Lenz YE. Statistical prediction of the optimal site for thalamotomy in Parkinsonian tremor. Mov Disord 1995;10(3):318-28. 90. Stranton GB. Topographical organization of ascending cerebellar projections from the dentate and interposed nuclei in Macaca mulatta: an anterograde degeneration study. J Comp Neurol 1980;190(4):699-731. 91. Kall BA, Goerss SJ, Kelly PJ. A new multimodality correlative imaging technique for VOP/VIM (VL) thalamotomy procedures. Stereotact Funct Neurosurg 1992;58:45-51. 92. Tasker RR, Organ LW, Hawrylyshyn PA. The thalamus and midbrain in man: a physiological atlas using electrical stimulation. Springfield: Charles C. Thomas; 1982. 93. Chiappa K. Evoked potentials in clinical medicine. 3rd ed. Lippincott Williams and Wilkins; 1997. 94. Gabriel EM, Nashold BS. Evolution of neuroablative surgery for involuntary movement disorders: an historical review. Neurosurgery 1998;42(3):575-91. 95. Vilesky JA, Gilman S. Using extirpations to understand the human motor cortex Horsley, Forester, Bucy. Arch Neurol 2003;60(3):446-51. 96. Cakmur R, Towle VL, Mullan JF, Suarez D, Spire JP. Intra-operative localization of sensorimotor cortex by cortical somatosensory evoked potentials: from analysis of waveforms to dipole source modeling. Acta Neurochir 1997;139:1117-25. 97. Arle JE, Apetauerova D, Zani J, Deletis V, Penny D, Hoyt D, Gould G, Shils JL. Motor Cortex Stimulation for Parkinson’s Disease: 12 month follow-up in 4 patients. J Neurosurg 2008;109(1):133-139. 98. Tsubokawa T, Katyama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation in patients with thalamic pain. J Neurosurg 1993;78:393-401. 99. Saitoh Y, Kato A, Ninomiya H, Baba T, Shibata M, Mashimo T, Yoshimine T. Primary motor cortex stimulation within the central sulcus for treating deafferentation pain. Acta Neurochir Suppl 2003;87:149-52. 100. Nguyen J-P, Lefaucher J-P, Decq P, Uchiyama T, Carpentier A, Fontaine D, Brugieres P, Pollin B, Feve A, Rostaing S, Cesaro P, Kervel Y. Chronic motor cortex stimulation in the treatment of central and neuropathic pain. Correlations between clinical, electrophysiological and anatomical data. Pain 1999;82:245-51. 101. Smith H, Joint C, Schlugman D, Nandi D, Stein JF, Aziz TZ. Motor cortex stimulation for neuropathic pain. Neurosurg Focus 2001;11(3):1-9. 102. Romsto¨ck J, Fahlbusch R, Ganslandt O, Nimsky C, Strauss C. Localization of the sensorimotor cortex during surgery for brain tumors: feasibility and waveform patterns of somatosensory evoked potentials. J Neurol Neurosurg Psychiatry 2002;72:221-9.

1281

1282

77


103. Taniguchi M, Cedzich C, Schramm J. Modification of cortical stimulation for motor evoked potentials under general anesthesia: technical description. Neurosurgery 1993;32:219-26. 104. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1969;60: 339-443. 105. Sartorius CJ, Wright G. Intraoperative brain mapping in a community setting: technical considerations. Surg Neurol 1997;47:380-8. 106. Sala F, Lanteri P. Brain surgery in motor areas: the invaluable assistance of intraoperative neurophysiological monitoring. J Neurosurg Sci 2003;47:79-88. 107. Szeleńyi A, Joksimovicˇ B, Seifert V. Intraoperative risk of seizures associated with transient direct cortical

stimulation in patients with symptomatic epilepsy. J Clin Neurophysiol 2007;24(1):39-43. 108. Roux FE, Boulanouar K, Ranjeva JP, Manelfe C, Tremoulet M, Sabatier J, Berry I. Cortical Intraoperative Stimulation in Brain Tumors as a Tool to Evaluate Spatial Data from Motor Functional MRI. Invest Radiol 1999;34(3):225-9. 109. Arle JE, Apetauerova D, Brophy S, Shils JL. Dyskinesia in Parkinson’s Disease treated by deep brain stimulation once electrode position was revised: case report. Neuromodulation 2007;10(3);238-43. 110. Yingling CD, Ojermann S, Dodson B, et al. Identification of motor pathways during tumor surgery facilitated by multichannel electromyographic recording. J Neurosurg 1999;91:922-7.

Functional Neurosurgery – Technical Aspects

76 Image Guided Functional Neurosurgery S. Khan . N. K. Patel . E. White . P. Plaha . S. Ashton . S. S. Gill

Introduction Since the early descriptions of human stereotactic surgery, image guidance has played an essential role in these procedures [1]. There has been a transition from skull radiographs, ventriculography through to CT scanning and MRI, providing assistance in the planning of surgery. Initially these imaging techniques gave an indirect location for the subcortical structure’s being targeted. The indirect methods are based on brain atlases and typically use the anterior commissure (AC) and posterior commissure (PC) as internal landmarks to co-register the atlas with the patient [2–4]. However there is substantial individual variation in AC-PC based coordinates of subcortical nuclei. Direct visualization of the target morphology using high resolution MRI can demonstrate significant variations in anatomy between individuals and also between different hemispheres in the same individual [5–7]. Normalization of atlas-based coordinates to the dimensions of the patient’s brain is based on the assumption that a linear transformation between two brains is possible, a concept that has never been proven. To compensate for the individual variations when using the indirect method, many centers have developed intraoperative clinical and electrophysiological monitoring procedures. The use of microelectrode recording and macroelectrode stimulation/recording of the target sites is regarded by many to be a pre-requisite for accurate delivery of therapy. Intraoperative stimulation is performed with the patient awake, in the off medication state, so that functional #


change can be observed. This can prolong operative time and requires additional staff to be present in theatre, i.e., neurologists. The passing of multiple microelectrodes is associated with greater expense, higher complication rates and potentially prolonged procedures, again exposing patients to many hours of awake surgery, which can be accompanied by significant stress for both the patient and operative staff [8–11]. These constraints highlight the need for more accurate, simpler and quicker targeting methods.

General Overview In this chapter, the authors discuss the issues surrounding the use of MRI for direct target visualization, determining the target position in stereotactic space, accurate targeting and confirmation of targeting. We aim to briefly cover the approaches used by our unit for direct targeting using high resolution MRI, employing an implantable guide tube (> Figure 76-1), with intraoperative MRI for confirmation of accurate placement.

Clinical Methods and Rationale Defining Functional Targets on MR-Images The direct visualization of small subcortical nuclei has in the past been limited by constraints in MR technology, the prolonged scan acquisition times required for high resolution MR

1240

76

Image guided functional neurosurgery

. Figure 76-1 Guide tube with a threaded cylindrical hub and a dome shaped proximal end; and adjacent stylette whose T-shaped proximal end fits within the guide tube’s hub and whose distal end projects beyond the guide tube

appropriate scan parameters in a standard 1.5 T MRI scanner.

MRI Scan Parameters In addition to the hardware related parameters MRI image quality is also dependent on scanning parameters that dictate image resolution (matrix size, slice thickness), SNR and lack of artifact.

images and motion artifact, especially in patients with movement disorders. However advances in MR technology in conjunction with manipulation of scan parameters has enabled the direct visualization of thalamic and subthalamic nuclei.

MRI Scanners MRI Image quality is affected by limitations in the field strength, field homogeneity and acquisition parameters. Advances in major aspects of MR imaging have improved these limitations enabling the direct visualization of brain targets for functional Neurosurgery. The first of these has been the increase in Magnet field strength from 1 to 1.5 and in limited circumstances 3 and 7 T platforms. Most modern scanners also automatically adjust the uniformity of the magnetic field, a process known as shimming, producing a homogonous field. In conjunction with these improvements, increases in magnetic field gradients and improved coil design have resulted in improved signal to noise ratio (SNR), spatial resolution and reduced acquisition times [12]. This has allowed for the direct visualization of functional targets such as the Globus pallidus interna and the subthalamic nucleus using

The SNR can be improved by increasing field of view and slice thickness, however this results in a loss of spatial resolution, which is counterproductive when attempting to visualize some of the small subcortical nuclei targeted in diseases such as PD. The commonest method of increasing SNR is by increasing the number of signal averages (NSA/NEX) performed during a scan. This results in the reduction of noise that is randomly generated and therefore cancelled out over successive averages. This strategy is limited by time constraints. Improvements in SNR are related to the square root of the increase in NSA, whereas increase in scan time is directly related to increase in NSA. Hence a fourfold increase in NSA, will result in a fourfold increase in scan time but only a doubling of SNR [13]. This can be particularly problematic in patients with movement disorders, where significant movement artifact prevents the use of long scan times. In our unit, we rely solely on high-resolution MRI images for planning of functional Neurosurgery and therefore we apply a modified Leksell frame and acquire pre-operative MRI images with the patients anesthetized and paralyzed, under strict stereotactic conditions. This prevents movement artifact, enables the use of long acquisition times with a high NSA and therefore SNR. Image quality is also degraded by ‘‘cross talk’’ from adjacent slices when acquiring data in a


continuous acquisition. This arises because the profile of a slice is never ideal. This results in signal from tissue adjacent to the selected slice affecting the image contrast/quality. Depending on the exact nature of the sequence, sequence timings (e.g., TE times) and relaxation constants of the tissue being imaged the extent of this cross-talk will vary. Placing a gap between slices reduces the ‘‘cross talk’’ from adjacent tissue [13]. In addition to optimizing SNR and therefore contrast between different tissues, MR sequences used for planning of functional Neurosurgery require a high spatial resolution in order to optimize the visualization of the small subcortical nuclei targeted in diseases such as PD. Image resolution is determined by voxol size, which is related to matrix size, field of view and slice thickness. MRI brain sequences are able to use a small field of view, however reductions in matrix size and slice thickness in order to optimize spatial resolution must be offset against loss of SNR due to these changes [12]. Acquiring images in the AC-PC plane enables correlation of the position of targets such as the STN with surrounding structures like the Red Nucleus and Substantia nigra. This also allows for the use of a brain atlas (e.g., Schaltenbrand and Warren, Stuttgart: Thieme), sliced in the AC-PC plane, as a reference tool during the planning of surgery.

Applying these principles has enabled the visualization of thalamic nuclei [14], the STN [6,15] and GPi [16] in a standard 1 and 1.5 T MR scanner. In our experience the subthalamic nucleus (STN) is best seen on high-resolution T2-weighted images (1.5 T TR 4000, TE 120, TSE 11, NSA 12, 2 mm slice, 0.4 mm gap, voxol size 0.45 0.45 mm) [17–19]. This sequence also provides good visualization of the Red nucleus and the mammillothalamic tract (> Figure 76-2).

76

. Figure 76-2 High-resolution axial T2-weighted MR images showing bilateral STN

We base target planning for the putamen and globus pallidus on a combination of high-resolution T2-weighted and proton density sequences (TR 4000, TE 15, TSE 7, NSA 8) [16].

Clinical Method 1.

2.

Our modified Leksell stereotactic frame (Elekta Instrument AB, Stockholm, Sweden), with non-conducting plastic posts, is positioned low on the head parallel to the orbito-meatal plane, which is used as an approximation of the AC-PC plane, and fixed to the skull using insulated pins. The frame is applied under general anesthesia, which is maintained during the imaging. A mid-sagittal plan scan is acquired and the AC and PC are visualized. Images are acquire parallel (axial) and perpendicular (coronal) to the AC-PC plane to allow direct comparison with an atlas. (> Figure 76-3a and b) If the frame is not parallel then the orientation can be readily adjusted by loosening the fixation of the anterior posts

1241

1242

76


. Figure 76-3 Midline Saggital MRI scans demonstrating the orientation of axial and coronal images in relation to the anterior and posterior commisure

3.

4.

to the frame and sliding them up or down as appropriate before re-tightening them. Deep brain targets are defined on long acquisition, high-resolution images acquired under strict stereotactic conditions, in both the axial and coronal planes. Different sequences are used to optimize visualization of the different targets. This use of appropriate sequences provides good visualization of the STN, Red nucleus and the mammillothalamic tract (> Figure 76-2) [17–19]. The location of other structures in the subthalamic region such as the zona incerta and prelemniscal radiation can be determined indirectly from these structures and with reference to a brain atlas used as a visual guide. Target visibility may be increased by adjusting the window setting of the T2-weighted images maximizing the gray/white matter contrast and by using magnified hard copy images. The boundaries of visible targets in the subthalamic region, such as the STN, may be enhanced by overlaying inverted images on

to standard T2 images, a method, which neutralizes the gray areas and allows for easier identification of the bright STN edges on a dark background. Often boundaries poorly seen in one imaging plane are better seen on another. With crosscorrelation of STN in axial, coronal and sagittal planes, the boundaries can be further identified, confirmed and a 3-D map of the target constructed.

Transposing the Target Position into Stereotactic Space The accuracy of MRI stereotaxis was problematic early in it development due to intrinsic distortion in the MR field, patient related artifact and image distortion from stereotactic frame systems. As detailed above newer MRI scanners have minimal intrinsic image distortion and can correct for global non-patient related distortion, resulting


in a homogonous field. Patient related distortion occurs due to the differences in magnetization of varying tissue (magnetic susceptibility artifact) and the local chemical environment of the protons in the structure being scanned (chemical shift). Both can generate artifact as well as image distortion. Magnetic susceptibility artifact can be minimized with the use of spin echo (SE) sequences, high receiver bandwidth (strong gradients) or short time to echo (TE). As the fat and fluid composition of the brain is largely homogenous there is negligible chemical shift, again this can be reduced with the use of strong gradients [20–25]. Distortion of fiducial systems can significantly affect accuracy of anatomical targeting. Fiducials are usually placed at the periphery of the image where there is maximal magnetic susceptibility artifact and fiducials using a fatty medium would also be subject to chemical shift artifact [26–28]. MRI fiducial systems should ideally be placed as close to the patients head as possible and consist of a non-fat containing medium. The Leksell frame has been demonstrated to generate minimal distortion [26]. Any residual distortion can be mapped by the use of Phantom studies. Our own unit currently uses two Phantoms, the first produced by Elekta Intstruments and the second has been designed by Renishaw plc (Wotton-under-edge, Gloucestershire, UK) to fit into the Leksell frame. These are self-contained cylinders of parallel rods, where the cylinders are a solid material and the rods are filled with copper sulfate solution. The first Phantom considers the Axial plane and the second can be rotated to consider both the Coronal and Sagittal planes. Other forms of Phantoms created for distortion correction have been investigated [21,29,30]. A comparison of the rod positions from each MR image and the actual geometric positions of the rods provide a spatial transformation. The transformation obtained is used to correct distorted images in

76

the same plane and location. Specialist software has been created by Renishaw plc to obtain the transformation and apply the distortion correction to the images. An alternate and commonly used method of stereotactic localization is to fuse MRI with CT. Stereotactic space is defined by the CT images and MR is used for target visualization. CTstereotactic accuracy is dependent on mechanical factors and image acquisition parameters. Mechanical movement of the CT Gantry and table must be checked as part of the CT imaging protocol [31]. Modern CT scanners allow the rapid acquisition of thin slices enabling a high degree of stereotactic accuracy [32]. However fusion of MR to CT can introduce an error of up to 1.3 mm [33]. Morphing of MR with CT is performed on a best match or least error basis and true fusion has not yet been demonstrated [34].

Clinical Methods 1.

2.

The 3-D coordinates of a selected target are determined by overlaying a transparency with a 1mm grid scaled to match the magnification of the magnified hard copy images (e.g., 1.6). The center of the grid is positioned in the center of the stereotactic space by aligning four reference points on the transparency with the fiducials visible on the image. The positions of the reference points on the transparency can be adjusted to accommodate for the geometric distortion in the particular MRI scanner after carrying out phantom studies. Alternatively, the specialist software can provide images where the distortions have been corrected, and stereotactic co-ordinates are produced taking this into account. The trajectory is planned. For DBS cases the electrode contacts are outlined on the images and typically the second contact of a quadripolar lead (contacts 1 or 5 of lead

1243

1244

76


3389 or 3387, Medtronic Inc.) is placed at the target site.

Surgical Procedure and Peri-Operative Confirmation of Accurate Targeting Having visualized and obtained accurate stereotactic co-ordinates for the surgical target a DBS electrode, or other therapeutic device, is delivered to the target. Minor displacement of DBS electrodes from the optimal target position can result in significant sensory, motor and emotional side effects and so confirmation of correct placement is essential. However errors in targeting can occur due to a number of reasons: 1. 2. 3.

Miscalculation of co-ordinates Human error in setting the stereoguide Intraoperative brain shift due to (a) Change in the patients head position relative to the frame (b) Loss of CSF and pneumocephaly [35] (c) Movement of the brain whilst advancing the therapeutic device/probe through the parenchyma (d) Deflection of the therapeutic device/ probe (e.g., dural edge)

It is therefore essential to minimize these factors. The stereotactic frame should be placed under sufficient tension or can be embedded into the outer table of the skull to prevent displacement. A second individual should independently check the target co-ordinates and stereoguide settings. Intraoperatively placing the patient at 45 head up places the burr holes uppermost and minimizes CSF loss. Whilst the dura is open the

burr hole can be constantly irrigated with saline in order to prevent air entry. Conventional methods of refining target position, accounting for intraoperative brain shift and confirming accuracy of placement include MER, intraoperative stimulation with assessment of clinical response and MRI/CT confirmation of electrode position. The shortcoming of these methods, increased hemorrhage risk with MER, prolonged operative time and therefore patient discomfort with both awake MER and intraoperative stimulation have been well described [8,9,11,36]. The use of postoperative MRI/CT for conformation of electrode placement relies on the assumption that the electrode is in the center of the signal void/artifact it produces on these images and image distortion produced by the electrode is minimal. In addition MRI scanning with the presence of an implanted electrode carries some risk of contact heating and tissue damage [37]. The technique described below avoids brainshift, minimize hemorrhage risk and enables intraoperative conformation of electrode placement without causing image distortion [38].

Clinical Methods 1.

2.

3.

Surgery is performed under general anesthesia in a semi-sitting position, such that the frontal burrholes are uppermost. The burrhole is made with a one-quarter inch drill, guided in a pre-planned trajectory by the stereoguide, and is of sufficient size to visualize and coagulate cortical vessels under continuous saline irrigation, minimizing any CSF loss (> Figure 76-4). A elongated stop (Renishaw PLC, UK) is fixed into the upper carriage and with the stereoguide set to the desired coordinates and trajectory, and rechecked, a probe is inserted through the stop to the level of the skull surface (> Figure 76-5).


. Figure 76-4 Burrhole is made with a one-quarter inch drill, guided in a pre-planned trajectory by the stereoguide

4.

5.

6.

The distance of the probe above the stop, which equates with the distance from the skull to the target, is measured and enables a carbothane guide tube (Renishaw PLC, UK) to be cut to an appropriate length. For the insertion of DBS electrodes the guide tube is generally shortened so that when inserted, its distal end will be 12 mm short of the target thus ensuring that when the DBS lead is implanted all the four contacts will be exposed. The guide tube is inserted into the split guide (Renishaw PLC, UK) that is fixed into the lower carriage of the stereoguide. The probe is then advanced to the target through the stop, and the appropriately sized guide tube, that is held in

76

. Figure 76-5 A probe is inserted through a elongated stop (Renishaw PLC, UK) fixed into the upper carriage to the level of the skull surface and enables the guide tube to be cut to an appropriate length

7.

8.

9.

the instrument carriers split guide block (> Figure 76-6). The split guide is then unclamped from the lower carriage and its halves removed allowing the guide tube to be advanced over the probe to the vicinity of the brain target. Cellulose gauze is laid over the dura around the guide tube and acrylic cement placed into the burrhole. The hub of the guide tube is seated in the acrylic cement, which once set secures the guide tube in place. The probe is now removed. The length from the top of the guide tube dome to the stereoguide datum is measured. The probe is replaced with a radio-opaque stylette, whose T shaped proximal end fits

1245

1246

76


. Figure 76-6 Probe inserted through stop and guide tube to the brain target

within the hub, cut to length such that its distal end projects beyond the distal end of the guide tube into the target (> Figure 76-7). 10. Our technique of guide tube and stylette implantation performed under GA typically takes 20–30 min per side. 11. The scalp wound is now closed and the patient is transferred to a MRI or CT scanner where the position of the stylette is defined in relationship to the desired target (> Figure 76-8) with images acquired under stereotactic conditions. The peri-operative images are laid onto the plan scans, formatted as inverted images, and any displacement of the stylette from the planned target is measured. 12. Prior to insertion of a DBS lead (e.g., DBS 3389 or 3387 lead Medtronic Inc., Minnea-

. Figure 76-7 Hub of the guide tube is fixed within the burrhole with acrylic cement. Probe is replaced with a radio-opaque stylette cut to the appropriate length such that its distal end projects beyond the guide tube into the target

polis), the length to be inserted is marked off by a sutured stop around the lead, defined by the length of the stylette that has been withdrawn from the guide tube. Once inserted the leads tungsten guide wire is removed and the lead is bent through a 90 arc conforming with that within the slotted hub of the guide tube. The lead is then secured to the skull with a miniplate and screws. The electrode connectors are counter sunk into channels made in the skull. This improves cosmesis by eliminating the prominent profile of the connectors but also reduces complications such as skin erosion and lead displacement


. Figure 76-8 Peri-operative inverted coronal T2 weighted image (inferior) verifying the position of the radio-opaque stylettes within the planned STN target (pre-operative high-resolution T2-weighted image (superior) from which the STN and surrounding structures can be visualized, and are outlined on the peri-operative inverted image, inclusive of the visible stylette). Perioperative images are obtained in the same slice configuration as the pre-operative planning images

76

. Figure 76-9 Radio-opaque stylette is replaced with a DBS lead with depth of insertion marked off by tying a suture around the lead, as defined by the length of stylette withdrawn. DBS connectors are placed into troughs drilled into the skull

errors that have arisen during target localization, coordinate calculation or during the operative procedure. In these circumstances if the error is large, for example 2 mm or more, then it may be corrected by the implantation of another guide tube and stylette, through another burrhole and trajectory, whilst the sub optimal guide tube and stylette remain in-situ and act as an internal reference and a brain anchor limiting brain shift. On repeat confirmation of target localization, the sub-optimal guide tube and stylette are then removed.

13.

that are associated with the connectors lying in the posterior-auricular position [37] (> Figure 76-9). Implantation of bilateral DBS leads and connection to an implanted generator adds approximately 45–60 min; with total operative time, inclusive of peri-operative MR-imaging and transfers being about 3–3½ h. The peri-operative image may demonstrate a displacement of the radio-opaque stylette from the chosen target. This may result from

If there is a small error in electrode position, for example below 2 mm, in future this will be compensated by the use of DBS electrodes with stearable electric fields (in development) enabling a shift in the center of field by 1.5 mm.

Clinical Indications The guide tube may act as a port for the implantation of electrodes not only for deep brain stimulation (DBS) but also for radiofrequency lesioning; catheters for drug or trophic substance

1247

1248

76


delivery; neural stem cell or encapsulated cell transplantation; and viral-vector delivery.

Deep Brain Stimulation This method is useful for localizing functional targets without the use of microelectrode recording and or macroelectrode stimulation/recording. Performing our pre-operative plan MRI scans under general anesthesia has enabled us to use long acquisition images, eliminate movement artifact and optimizing image sensitivity whilst maintaining high spatial resolution. This technique has enabled us to directly visualize our target structures, and by cross correlating axial and coronal images, produce a volumetric representation of the intended target. In conjunction with intraoperative MRI in order to confirm guide tube and indwelling stylette placement this technique has resulted in consistent clinical outcomes that have enabled us to discontinue awake surgery, a practice that can be stressful for both patients and surgeon. More recently, the authors have utilized this method for the implantation of deep brain stimulating electrodes into a novel target site, the pedunculopontine nucleus, that had no defined and validated microelectrode signature [39]. The implantable guide tube is also useful for gaining repeat access to functional targets without the need to repeat the whole stereotactic procedure. The device facilitates non-stereotactic replacement of hardware, for example DBS electrodes or catheters, following complications of migration, fracture, malfunction or infection of these therapeutic devices. The device facilitates repeated or staged radiofrequency lesioning at the same target. The radio-opaque stylette may be left in situ within the indwelling guide tube as a permanent marker of the lesion site. Should there be a decline in the functional benefit following a lesion it is simple for the surgeon to repeat the lesion without the extensive work-up and imaging that is usually required. For small

functional targets in eloquent areas, such as the subthalamic nucleus, it may be safer to make an undersized lesion in the first instance and repeat the lesion at a later date to optimize its size. Finally, the guide tube would facilitate replacement of a DBS electrode with a permanent radiofrequency lesion.

Clinical Outcomes The clinical outcome of our patients undergoing subthalamic nucleus stimulation is comparable to outcomes reported in literature. This method has also allowed us to retrospectively correlate the degree of motor improvement with the anatomical location of the optimal stimulation contact in patients with PD. We have subsequently identified the caudal zona incerta as a better target than the STN for deep brain stimulation in the treatment of Parkinson’s disease, into which we now implant deep brain stimulation electrodes [19]. The therapeutic effect of stimulating the pedunculopontine nucleus for the previously treatment resistant axial symptoms of PD has been recently reported [39]. Our results on subthalamic region stimulation for the treatment of essential tremor, and more recently the stimulation of the Zona incerta specifically for multiple forms of tremor have been published [40].

Accuracy of Surgical Technique and Post-Operative Complications Despite concerns regarding the use of the Leksell G frame in MRI stereotaxy, the accuracy of this technique has been repeatedly demonstrated. We have previously reported a mean error of 0.3 0.4 mm in the medial to lateral plane and 0.4 0.4 mm in the AP plane, prior to adopting the use of an indwelling guide tube [17]. Similar findings have been reported by Simon et al,


and more recently by our group whilst using implanted guide tubes [38,41]. This technique, in contrast to methods employing intraoperative clinical and electrophysiological monitoring procedures using microelectrodes and/or macroelectrode stimulation of the target sites, reduces the number of brain trajectories and the potential associated brain trauma and hemorrhage risk. Following implantation of 506 guide tubes in 250 patients, there were 10 procedure-related complications. One patient developed a non-hemorrhagic paresis with expressive dysphasia following which he continues to make a gradual recovery of function and at the 12-month follow-up he exhibited a milder deficit. One patient developed a small cortical hematoma with a postoperative paresis that has completely resolved after 6 months. Another patient developed dysphagia for 3-months as a consequence of mistargeting secondary to an error in frame relocation, with both initial guide tubes and stylettes implanted into the thalami bilaterally. There was one post-operative self-limiting grand mal seizure and two pulmonary emboli, one of which was fatal. There have been two infections requiring removal of implanted hardware.

MRI Guided Direct Intracranial Drug Delivery In order to circumvent the blood-brain barrier, a number of techniques have been developed to infuse therapeutic agents, including viral vectors, neurotrophic proteins, chemotherapeutics and stem cells, directly into the brain. With the exception of biodegradable carmustine wafers, these techniques have yet to reach routine clinical use, but have been employed in a number of clinical trials.

Injection of drug-producing cells: e.g., Fibroblasts, genetically-modified to produce nerve growth factor (NGF), have been injec-

76

ted into the cholinergic basal forebrain of eight patients with mild Alzheimer’s disease as part of a phase 1 clinical trial [42]. Limited clinical and PETevidence of therapeutic efficacy was demonstrated. Encapsulated cells: e.g., Encapsulated cells, genetically-modified to produce human ciliary neurotrophic factor (CNTF), have previously been implanted into the right lateral ventricle of patients with Huntington’s disease, as part of a phase 1 clinical trial [43]. No significant clinical benefit was observed. Biodegradable polymers: e.g., Carmustine (BCNU) wafers in patients with recurrent high-grade gliomas. Injection: Intraparenchymal: Numerous clinical trials have utilized intraparenchymal injection to administer therapeutic agents into the brain, most commonly to treat high-grade gliomas. This technique has largely been superseded by convection-enhanced delivery. Intraventricular: e.g., Glial cell linederived neurotrophic factor (GDNF) and nerve growth factor (NGF) have been administered intraventricularly in phase 1 clinical trials to patients with Parkinson’s disease [44] and Alzheimer’s disease [45], respectively. Neither trial demonstrated significant clinical benefit. Convection-enhanced delivery (CED): This represents an extremely promising approach to administering therapeutic agents directly into the brain, as in contrast to all other techniques of direct intracranial drug delivery, drug distribution is achieved by bulk flow down a pressure gradient, rather than by diffusion, down a concentration gradient. Consequently it is possible to deliver therapeutic agents, regardless of their molecular size or infused concentration, homogeneously over large, controlled volumes of brain. Unfortunately despite the potential of this approach

1249

1250

76


and its use in a number of large phase III clinical trials, including the recently completed PRECISE trial (Phase 3 Randomized Evaluation of Convection-Enhanced Delivery of IL-13PE38QQR Compared to Gliadel Wafer with Survival Endpoint in Glioblastoma Multiforme at First Recurrence) CED has yet to be successfully translated into routine clinical practice.

The Role of Imaging in Direct Intracranial Drug Delivery Imaging has a central role in the development of neurosurgical techniques of direct intracranial delivery for two fundamental reasons: 1.

2.

Drug distribution within disease-specific target structures is inextricably linked to drug efficacy. Drug penetration into surrounding structures may be linked to the occurrence of side-effects.

Imaging facilitates the accurate and controlled delivery of therapeutic agents into the brain in a number of important ways:

Conventional stereotactic techniques enable the accurate placement of catheters or needles into target structures prior to infusing therapeutic agents. Molecular imaging to visualize drug efficacy e.g., PET. Visualizing the distribution of therapeutic agents within the brain. This is a rapidly expanding area of research that offers enormous potential in enhancing the clinical utility of cell-based therapies and convectionenhanced delivery.

Imaging Cell-based Therapies Whereas the short half-life of isotopic tracers renders PET and single-photon emission computed tomography (SPECT) ineffectual for prolonged in vivo cellular imaging, it is possible to label cells including embryonic, mesenchymal and neural stem cells with MR contrast agents. To date, both gadolinium chelates and iron oxide particles have been used to label cells in preclinical studies, although due to a paucity of toxicity data, neither has been used in clinical trials. However the potential of this approach has been demonstrated in a number of animal models. For example, ultra-small particle iron oxide (USPIO) labeled embryonic stem cells have been visualized migrating, across the corpus callosum, in a rat stroke model, from their implantation site, into the region of the infarct in the contralateral hemisphere [46].

The Role of Imaging in Convection-Enhanced Delivery A number of imaging strategies are in development to facilitate the administration of therapeutic agents into the brain by CED.

MRI-based Predictive Modeling of Drug Distribution Using patient-specific data, derived from a range of MR-based imaging techniques, including diffusion tensor imaging (DTI), and computational fluid dynamics, software algorithms for predicting drug distribution, by CED, have been developed. These mathematical models facilitate surgical planning of CED, by allowing pre-operative visualization of the predicted drug distribution and extent of infusate reflux along the catheter-brain interface. FDA approved software (iPlan Flow – BrainLAB), incorporating these features and integrated with surgical


navigation systems, has been developed to facilitate the administration of therapies to patients with brain tumors, in clinical trials. The predictive value of these algorithms in clinical practice is at present relatively limited. For example, when the predicted volume of distribution of cintredekin besudotox in patients with high-grade glioma, was compared to the volume of distribution of 123I-human serum albumin (HSA), observed by SPECT imaging, there was a concordance of only 65.75% (95% CI: 52.0–79.5%) [47]. In addition to the potential limitations of the predictive model used, this disparity may have resulted from the different distribution properties of cintredekin besudotox and 123I-HAS, tissue heterogeneity in the vicinity of high-grade gliomas, and the low spatial resolution of SPECT imaging, compared to MRI.

Surrogate Markers of Drug Distribution An alternative to predicting drug distribution from pre-operative imaging is to coinfuse, with the drug, a contrast agent, designed to act as a surrogate marker of drug distribution. A number of strategies have been developed including the administration of gadolinium-DTPA with glucocerebrosidase to a patient with Gaucher’s disease [48], as well as liposomal gadolinium [49] and iron oxide nanoparticles coated with dextran (ferumoxtran-10) [50], to visualize the distribution of liposomally-encapsulated drugs and adeno-associated virus (AAV), respectively. The obvious limitation of this approach is that the distribution properties of the surrogate marker and therapeutic agent are likely to be significantly different. However this is likely to represent a useful and complementary approach to predictive models of drug distribution.

76

Imaging Drug Distribution There is limited evidence that certain MRI acquisitions, in isolation, may provide limited visualization of drug distribution by CED. For example, diffusion-weighted MRI has been shown to be predictive of the response to paclitaxel, administered by CED, to patients with high-grade gliomas.

Conclusions The use of an implantable guide tube and stylette in stereotactic functional neurosurgery allows the surgeon to verify target localization visually on perioperative high-resolution MR-images. For DBS implantation this method enables direct correlation of the anatomical location of the active contacts with patient outcome and with continuous audit leads to improved practice. The method avoids exposing the patients to long hours of awake surgery and limits the number of brain probings in comparison to techniques employing MER. The device acts as a port for the delivery of DBS and lesioning electrodes, catheters for drug delivery, viral-vector delivery and cell implantation; and for repeated non-stereotactic access to the functional brain targets.

Acknowledgments We wish to thank our nurse specialists, Mrs Lucy Mooney and Mrs Karen O’ Sullivan, for carrying out the programming and assessments of patients whose data has contributed to this book chapter. We also wish to thank Ms Becky Durham and Ms Ruth Blachford for the illustrations. Steven Gill, Sadaquate Khan and Edward White are consultants to Renishaw PLC. None of the other Authors have any conflict of interest to declare. The intended use of the Renishaw ‘neuro| guide” electrode introducer Kit as defined by the manufacturer is “to provides a conduit through

1251

1252

76


which Deep Brain Stimulation (DBS) electrodes can be delivered to allow stimulation of sub cortical targets within the brain.’ This device is CEmarked for this purpose only in Europe.

References 1. Spiegel EA, et al. Stereotaxic Apparatus for Operations on the Human Brain. Science 1947;106(2754):349-50. 2. Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computerized tomography-guided determinations of brain targets in functional stereotaxis. J Neurosurg 1990;73(4):565-71. 3. Hariz MI. Correlation between clinical outcome and size and site of the lesion in computed tomography guided thalamotomy and pallidotomy. Stereotact Funct Neurosurg 1990;54–55:172-85. 4. Holtzheimer PE III, Roberts DW, Darcey TM. Magnetic resonance imaging versus computed tomography for target localization in functional stereotactic neurosurgery. Neurosurgery 1999;45(2):290-7; discussion 297–8. 5. Brierley JB, Beck E. The significance in human stereotactic brain surgery of individual variation in the diencephalon and globus pallidus. J Neurol Neurosurg Psychiatry 1959;22:287-98. 6. Ashkan K, et al. Variability of the subthalamic nucleus: the case for direct MRI guided targeting. Br J Neurosurg 2007;21(2):197-200. 7. Patel NK, Khan S, Gill S. Comparison of atlas- and magnetic-resonance-imaging-based stereotactic targeting of the subthalamic nucleus in the surgical treatment of Parkinson’s Disease. Stereotact Funct Neurosurg 2008;86:153-61. 8. Chevrier E, et al. Is there a role for physiotherapy during deep brain stimulation surgery in patients with Parkinson’s disease? Eur J Neurol 2006;13(5):496-8. 9. Gorgulho A, et al. Incidence of hemorrhage associated with electrophysiological studies performed using macroelectrodes and microelectrodes in functional neurosurgery. J Neurosurg 2005;102(5):888-96. 10. Hariz M, Blomstedt P, Limousin P. The myth of microelectrode recording in ensuring a precise location of the DBS electrode within the sensorimotor part of the subthalamic nucleus. Mov Disord 2004;19(7):863-4. 11. Hariz MI, Fodstad H. Do microelectrode techniques increase accuracy or decrease risks in pallidotomy and deep brain stimulation? A critical review of the literature. Stereotact Funct Neurosurg 1999;72(2–4):157-69. 12. McRobbie DW, editor. MRI from picture to proton. Cambridge: Cambridge University Press; 2003. p. 372. 13. Westbrook C, editor. MRI in practice. Oxford: Blackwell Publishing; 2005. p. 422.

14. Deoni SC, et al. Visualization of thalamic nuclei on high resolution, multi-averaged T1 and T2 maps acquired at 1.5 T. Hum Brain Mapp 2005;25(3):353-9. 15. Rampini PM, et al. Multiple sequential imagefusion and direct MRI localisation of the subthalamic nucleus for deep brain stimulation. J Neurosurg Sci 2003;47(1):33-9. 16. Hirabayashi H, Tengvar M, Hariz MI. Stereotactic imaging of the pallidal target. Mov Disord 2002; 17 Suppl 3:S130-S134. 17. Patel NK, et al. MRI-directed subthalamic nucleus surgery for Parkinson’s disease. Stereotact Funct Neurosurg 2002;78(3–4):132-45. 18. Patel NK, et al. MRI directed bilateral stimulation of the subthalamic nucleus in patients with Parkinson’s disease. J Neurol Neurosurg Psychiatry 2003;74(12): 1631-7. 19. Plaha P, et al. Stimulation of the caudal zona incerta is superior to stimulation of the subthalamic nucleus in improving contralateral parkinsonism. Brain 2006;129 (Pt 7):1732-47. 20. Reinsberg SA, et al. A complete distortion correction for MR images. II. Rectification of static-field inhomogeneities by similarity-based profile mapping. Phys Med Biol 2005;50(11):2651-61. 21. Doran SJ, et al. A complete distortion correction for MR images. I. Gradient warp correction. Phys Med Biol 2005;50(7):1343-61. 22. Tanner SF, et al. Radiotherapy planning of the pelvis using distortion corrected MR images: the removal of system distortions. Phys Med Biol 2000;45(8):2117-32. 23. Woo JH, Kim YS, Kim SI. The correction of MR images distortion with phantom studies. Stud Health Technol Inform 1999;62:388-9. 24. Maurer CR Jr, et al. Effect of geometrical distortion correction in MR on image registration accuracy. J Comput Assist Tomogr 1996;20(4):666-79. 25. Young IR, et al. The benefits of increasing spatial resolution as a means of reducing artifacts due to field inhomogeneities. Magn Reson Imaging 1988;6(5):585-90. 26. Walton L, et al. A phantom study to assess the accuracy of stereotactic localization, using T1-weighted magnetic resonance imaging with the Leksell stereotactic system. Neurosurgery 1996;38(1):170-6; discussion 176–8. 27. Yu C, et al. A phantom study of the geometric accuracy of computed tomographic and magnetic resonance imaging stereotactic localization with the Leksell stereotactic system. Neurosurgery 2001;48(5):1092-8; discussion 1098–9. 28. Michiels J, et al. On the problem of geometric distortion in magnetic resonance images for stereotactic neurosurgery. Magn Reson Imaging 1994;12(5):749-65. 29. Menuel C, et al. Characterization and correction of distortions in stereotactic magnetic resonance imaging for


30.

31. 32.

33.

34.

35.

36.

bilateral subthalamic stimulation in Parkinson disease. J Neurosurg 2005;103(2):256-66. Wang D, Doddrell DM, Cowin G. A novel phantom and method for comprehensive 3-dimensional measurement and correction of geometric distortion in magnetic resonance imaging. Magn Reson Imaging 2004;22 (4):529-42. Kondziolka D, editor. Radiosurgery, vol 6, Pittsburgh: Karger; 2006. Landi A, et al. Accuracy of stereotactic localisation with magnetic resonance compared to CT scan: experimental findings. Acta Neurochir (Wien) 2001;143(6): 593-601. Duffner F, et al. Relevance of image fusion for target point determination in functional neurosurgery. Acta Neurochir (Wien) 2002;144(5):445-51. Rezai AR, et al. Deep brain stimulation for Parkinson’s disease: surgical issues. Mov Disord 2006;21 Suppl 14: S197-S218. Miyagi Y, Shima F, Sasaki T. Brain shift: an error factor during implantation of deep brain stimulation electrodes. J Neurosurg 2007;107(5):989-97. Hariz MI. Safety and risk of microelectrode recording in surgery for movement disorders. Stereotact Funct Neurosurg 2002;78(3–4):146-57.

76

37. Oh MY, et al. Long-term hardware-related complications of deep brain stimulation. Neurosurgery 2002;50 (6):1268-74; discussion 1274–6. 38. Patel NK, Plaha P, Gill SSMagnetic resonance imagingdirected method for functional neurosurgery using implantable guide tubes. Neurosurgery 2007;61 5 Suppl 2:358-65; discussion 365–6. 39. Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine nucleus for Parkinson’s disease. Neuroreport 2005;16(17):1883-7. 40. Plaha P, Khan S, Gill SS. Bilateral stimulation of the caudal zona incerta nucleus for tremor control. J Neurol Neurosurg Psychiatry 2008;79(5):504-13. 41. Simon SL, et al. Error analysis of MRI and leksell stereotactic frame target localization in deep brain stimulation surgery. Stereotact Funct Neurosurg 2005;83(1):1-5. 42. Tuszynski MH, et al. A phase 1 clinical trial of nerve growth factor gene therapy for Alzheimer disease. Nat Med 2005;11(5):551-5. 43. Bloch J, et al. Neuroprotective gene therapy for Huntington’s disease, using polymer-encapsulated cells engineered to secrete human ciliary neurotrophic factor: results of a phase I study. Hum Gene Ther 2004;15(10):968-75. 44. Nutt JG, et al. Randomized, double-blind trial of glial cell line-derived neurotrophic factor (GDNF) in PD. Neurology 2003;60(1):69-73.

1253

45. Eriksdotter Jonhagen M, et al. Intracerebroventricular infusion of nerve growth factor in three patients with Alzheimer’s disease. Dement Geriatr Cogn Disord 1998;9(5):246-57. 46. Hoehn M, et al. Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance imaging investigation of experimental stroke in rat. Proc Natl Acad Sci USA 2002;99(25): 16267-72. 47. Sampson JH, et al. Clinical utility of a patient-specific algorithm for simulating intracerebral drug infusions. Neuro Oncol 2007;9(3):343-53. 48. Lonser RR, et al. Image-guided, direct convective delivery of glucocerebrosidase for neuronopathic Gaucher disease. Neurology 2007;68(4):254-61. 49. Krauze MT, et al. Real-time visualization and characterization of liposomal delivery into the monkey brain by magnetic resonance imaging. Brain Res Brain Res Protoc 2005;16(1–3):20-6. 50. Szerlip NJ, et al. Real-time imaging of convectionenhanced delivery of viruses and virus-sized particles. J Neurosurg 2007;107(3):560-7.

79 Impedance Recording in Functional Neurosurgery L. Zrinzo . M. I. Hariz

Introduction

General Principles

The stereotactic technique, described by Horsley and Clarke in 1908 and introduced to clinical practice by Spiegel and Wycis in 1947, is essential to anatomical localization in neurosurgery. The introduction of CT and especially MRI guided stereotactic surgery during the last decades of the twentieth century greatly facilitated the use and accuracy of anatomical targeting. Nevertheless, correlation of supposed anatomical location with expected clinical, physiological or electrical properties remains an important corroborative tool. Meyer first measured the electrical properties of brain tissue to localize pathology during surgery in 1921 [1]. Two years later, Grant reviewed tissue impedance patterns in 12 neurosurgical procedures and supplemented these observations with data from fresh and formalin fixed brains as well as animal trials. He concluded that impedance measurements were useful in locating deep seated tumors [2]. These techniques were particularly useful in tumor localization in the pre-CT era but persisted after the introduction of cross sectional imaging [3–7]. Continuous measurement of electrical impedance of brain tissue is readily measured en route to the target during stereotactic procedures and provides safe and reliable information during functional neurosurgical procedures. This has resulted in the widespread availability of commercial lesion generators to measure tissue impedance.

Electrical impedance is a measure of opposition to the flow of an alternating current and is measured in Ohms. Impedance includes resistance, but also takes into account the effects of capacitance and inductance (reactance). Impedance is a more complex measurement than resistance because the effects of capacitance and inductance vary with the frequency of the current passing through the circuit; impedance therefore varies with frequency. Brain tissue impedance is dependent upon numerous anatomical, chemical and physiological factors with the amount of myelin, direction of fibers, number of glial cells, density of neurons, composition of the CSF and blood flow all affecting the absolute values [8]. It is for this reason that the variation or ‘‘pattern’’ of electrical impedance as the probe travels through its trajectory within the brain is more useful than absolute static values [2,4,9,10]. Impedance is highest in white matter, lowest in CSF and intermediate in gray matter [4,9,11,12]. The direction of impedance change in passing between tissues is reproducible and provides supplementary information about the properties of the trajectory and target [4,11,13,14]. Monopolar impedance between a small probe tip and a large indifferent electrode is an old and excellent method to locate different structures inside the brain [15,16]. The indifferent electrode is usually a plate electrode applied to the thigh. The smaller the tip of the active

#


1326

79

Impedance recording in functional neurosurgery

electrode in relation to the area of the indifferent electrode, the more accurately impedance changes reflect those in the vicinity of the probe tip [17]. However, safety factors place practical constraints on the size of the active electrode since the risk of hemorrhage increases as probe diameter falls below 0.5 mm [18]. Therefore, non-insulated tips of the order of 1–2 mm in length and diameter are recommended. The standard Elekta electrode (Elekta, Stockholm, Sweden) previously used by the authors measures 2.1 mm in diameter and the non-insulated tip measures 4 mm. Impedance values with that electrode are therefore lower than with the recommended 1.5 2 mm Elekta electrode, (custom made) all other parameters being equal. Nevertheless, this 2.1 mm thick electrode can still disclose differences between gray matter, white matter and CSF spaces, as shown in > Figure 79-1.

Laitinen initially described how a feeding current with a carrier frequency between 1 and 10 kHz resulted in the sharpest difference between gray and white matter impedance values [16]. Impedance measurements in animal and cadavers published shortly afterwards provided such detail that images could be constructed from the acquired data resulting in the term ‘‘impedography’’ [19–21]. A more recent study suggests that carrier frequencies of 8–10 kHz may even offer enough resolution to differentiate between different gray matter structures, a claim that deserves further study in clinical practice [11]. Despite the well documented importance of carrier frequency in the utility of impedance measurements, some commonly available radiofrequency lesion generators (Radionics Inc., Burlington, MA) use suboptimal feeding currents, compromising the utility of this technique in functional neurosurgical practice [22].

. Figure 79-1 Illustration of lesion generator (Elekta, Stockholm, Sweden) showing three different impedance values recorded with a 2.1 mm radiofrequency electrode: (a) gray matter impedance of 455 Ohms. (b) white matter impedance of 616 Ohm. (c) CSF impedance of 166 Ohms


It has been suggested that coaxial bipolar electrodes may provide a more accurate measurement of local tissue impedance [12,23]. However, others have noted that tissue or electrolyte bridges between the two poles may degrade the quality of the recordings [15]. Monopolar impedance recording is more popular for a number of reasons. Monopolar radiofrequency electrodes serve a dual purpose of impedance measurement and lesion production thus necessitating the passage of only one probe through the brain when performing lesional surgery. In addition, monopolar impedance recording provides information about structures that lie immediately ahead or in close proximity to the probe trajectory. For example, a reduction in monopolar impedance may indicate proximity of the trajectory to a ventricle that would not otherwise be acknowledged by a bipolar recording device [9].

Continual Impedance Monitoring in Functional Procedures Continual impedance monitoring allows the surgeon to assess the impedance pattern in ‘‘real time’’ on a panel meter and as an audio signal, as the stereotactic probe is introduced towards the target area. The acoustic signal allows the surgeon to judge when the probe is passing through gray matter, white matter or CSF boundaries. During functional procedures, the authors pass a custom designed electrode (Elekta, Stockholm, Sweden) with a non-insulated tip of 1.5 mm in diameter and 2 mm in length through a stereotactically defined trajectory and target under continual monopolar impedance monitoring using the Leksell1 Neuro Generator (Elekta, Stockholm, Sweden). With this arrangement, white matter impedance is typically in the range of 800–1,000 Ohms falling to 500–700 Ohms on crossing the border into gray matter with a value of around 300–400 Ohms in CSF. Due to the smaller size of

79

the non-insulated tip of the customized electrode, compared with the standard 2.1 4 mm Elekta electrode, the impedance values are higher and also more local-specific than with the larger electrode.

Pallidal Target When targeting the posteroventral pallidum from a frontal coronal burrhole, 2–2.5 cm from the midline, the probe travels roughly parallel to the sagittal plane traversing the white matter of the corona radiata/internal capsule into the gray matter of the putamen or pallidum depending on the angulation of the electrode trajectory. The fall in impedance is clearly noticed with an abrupt change over a distance of one mm between white matter tracts and the gray matter of the anterodorsal (rostral) putamen or pallidum. One can sometimes notice a slight rise in impedance as the electrode traverses the various medullary laminae or approaches the ansa lenticularis at the base of the pallidum before the probe reaches the cistern between the pallidum and the supra-amygdala where the impedance rapidly falls to 400 Ohms. Continual impedance monitoring would immediately alert the surgeon of a medial deviation in the planned trajectory as the pitch of the auditory signal rises to that typical of white matter with transgression into the internal capsule [22,24–27].

Thalamic Target During thalamic procedures, with a frontal approach, the probe penetrates the lateral part of the head of the caudate nucleus to give impedance values suggestive of gray matter only to rise again as the white matter of the internal capsule is traversed. Impedance then slowly falls as the thalamus is acquired. Once the probe approaches the area below the thalamus, impedance starts to

1327

1328

79


rise once more. The impedance profile is less sharp during ventrolateral thalamic surgery than during pallidal surgery due to the fact that the probe advances along the thalamocapsular border and reticular thalamic nucleus, during part of its trajectory before reaching the gray matter of the ventrolateral-ventral intermediate thalamus proper [16].

areas link the ventral striatum to the caudate, rendering a mixed impedance value that lies between that of gray and white matter.

Advantages of Impedance Monitoring Figure 79-1a shows the impedance values in a case of pallidal surgery where an electrode with a bare tip 2.1 mm thick and 4 mm long, ended up in capsular white matter. It was subsequently repositioned to end up in pallidal gray matter (> Figure 79-1b). The patient was awake during surgery, and intraoperative stimulation at first location (> Figure 79-1a) with higher impedance confirmed a capsular location: upon stimulating with 120 Hz the patient exhibited tonic cramp in the hand. Should the stereotactic probe traverse or come close to the ventricle, cistern or cystic cavity, a sharp decrease of impedance will be noted. > Figure 79-1c shows a CSF impedance value, with same electrode as in > Figure 79-1a,b, in a case where the electrode traverses a ventricular space. The expected transition from gray to white matter, as determined from preoperative trajectory images, can be corroborated with continual impedance measurements as the probe passes along the trajectory to the desired target. The surgeon is alerted to unexpected deviations or transgressions into CSF spaces without time being spent analyzing or interpreting the acquired data [25]. Reduced operative time minimizes the amount of CSF loss and resulting brain shift; this allows patients to better tolerate surgery under local anesthetic. Finally, the 1–2 mm diameter of the recording probes used tend to displace fibers and vessels away from the trajectory rather than transect them and this may contribute to lower rates of hemorrhage than in cases where microelectrode recording is used [28]. >

Subthalamic Nucleus as Target When targeting the subthalamic nucleus (STN) the impedance profile may vary en route to the target depending on the chosen trajectory. Usually, the trajectory incorporates variable amounts of internal capsule, and thalamus, before entering the dorsolateral zona incerta immediately before entering the STN. This gives a rise of impedance on exiting the thalamus into the immediate subthalamic area, followed by a decrease of impedance upon entering STN. Here, the impedance values are less distinctive than for pallidal surgery due to the fact that the trajectory immediately above the STN, in the dorsal zona incerta, consists of mixed white and gray matter.

Anterior Internal Capsule as Target During anterior capsulotomy for obsessive compulsive disorder, it is important that the electrode, and the subsequent radiofrequency lesions, remain within the confines of the anterior internal capsule between the head of the caudate nucleus and the putamen. Here, continual impedance monitoring should remain white matter impedance until the last 2 mm of the trajectory. At this most ventral point, it may fall slightly, since the most ventral target point is some 2 mm below the extended intercommissural point. At this level, scattered gray matter


Limitations of Impedance Monitoring As with most physiological and clinical methods, impedance can only provide corroboration of presumed anatomical localization. The existence in some patients of large perivascular (VirchowRobin) spaces in the putamen and posteroventral pallidum may confuse the interpretation of impedance monitoring. Also, a suboptimal current frequency of the radiofrequency machine being used will reduce the discriminative value of impedance in distinguishing between gray and white matter. In all cases, postoperative stereotactic imaging remains the gold standard for accurately documenting anatomical localization after a stereotactic intervention.

79

Conclusions Impedance measurements were first used in the 1920s as a navigational tool in the localization of deep-seated tumors. Functional neurosurgeons can employ continual impedance measurement during stereotactic targeting to reliably document transition from white to gray matter along the trajectory as well as transgression of CSF spaces. This technique provides a rapid, safe and useful corroboration of real-time electrode localization in the operating theatre. Impedance measurements of chronically implanted DBS hardware can assist in the identification of device-related problems as a cause of therapeutic failure.

References Measuring Impedance in Chronically Implanted Systems Currently available deep brain stimulation (DBS) pulse generators (from Medtronic, Inc.) provide impedance measurement as one of a number of tools that can help in identifying device-related problems as a cause of therapeutic failure of DBS. The physician programmer is used to measure impedance for each of the contacts of the quadripolar DBS electrode separately in monopolar mode. Typically, when postoperative edema has subsided, the measured impedance lies between 500 and 1,500 Ohms. For mono-channel devices (Itrel II, Soletra) standard settings of 1 V, 210 ms, and 30 Hz should be used for maximum accuracy in impedance measurement. The Itrel II neurostimulator cannot read impedance values above 2,000 Ohms whereas the dual-channel Kinetra neurostimulator reads values >4,000 Ohms. Values of less than 50 Ohms suggest a short circuit [29,30]. An impedance >2,000 Ohms suggests a broken cable, lead fracture or other connection problem although this has to be confirmed by radiological and other means [31].

1. Meyer AW. Methode zum Auffinden von Hirntumoren bei der Trepanation durch electrische Widerstandmessung. Zbl Chir 1921;18:1824-26. 2. Grant FC. Localization of brain tumors by determination of electrical resistance of the growth. JAMA 1923;81: 2169-71. 3. Bullard DE, Makachinas TT. Measurement of tissue impedence in conjunction with computed tomographyguided stereotaxic biopsies. J Neurol Neurosurg Psychiatry 1987;50:43-51. 4. Gorecki J, Dolan EJ, Tasker RR, Kucharczyk W. Correlation of CT and MR with impedance monitoring and histopathology in stereotactic biopsies. Can J Neurol Sci 1990;17:184-9. 5. Organ L, Tasker RR, Moody NF. Brain tumor localization using an electrical impedance technique. J Neurosurg 1968;28:35-44. 6. Bullard DE. Intraoperative impedance monitoring during CT-guided stereotactic biopsies. Stereotact Funct Neurosurg 1989;52:1-17. 7. Rajshekhar V. Continuous impedance monitoring during CT-guided stereotactic surgery: relative value in cystic and solid lesions. Br J Neurosurg 1992;6:439-44. 8. Birzis L, Aguilar JA, Tachibana S. Cerebral hemodynamics revealed by electrical impedance changes. Confin Neurol 1968;30:1-16. 9. Limonadi FM, Roberts DW, Darcey TM, Holtzheimer PE, III, Ip JT. Utilization of impedance measurements in pallidotomy using a monopolar electrode. Stereotact Funct Neurosurg 1999;72:3-21. 10. Spiegel EA. Methodological problems in stereoencephalotomy. Confin Neurol 1965;26:125-32.

1329

1330

79


11. Axer H, Stegelmeyer J, Graf von Keyserlingk D. Comparison of tissue impedance measurements with nerve fiber architecture in human telencephalon: value in identification of intact subcortical structures. J Neurosurg 1999;90:902-9. 12. Oran LW, Tasker RR, Moody NF. The impedance profile of the human brain as a localization technique in stereoencephalotomy. Confin Neurol 1967;29:192-6. 13. Tachibana S, Aguilar JA, Birzis L. Scanning the interior of living brain by impedography. J Appl Physiol 1970;28: 534-9. 14. Dierssen G, Marg E. The value of impedance measurements to aid in the localisation in stereotactic surgery. Confin Neurol 1965;26:407-10. 15. Laitinen LV, Johansson GG. Locating human cerebral structures by the impedance method. Confin Neurol 1967;29:197-201. 16. Laitinen L, Johansson GG, Sipponen P. Impedance and phase angle as a locating method in human stereotaxic surgery. J Neurosurg 1966;25:628-33. 17. Ragheb T, Riegle S, Geddes LA, Amin V. The impedance of a spherical monopolar electrode. Ann Biomed Eng 1992;20:617-27. 18. Robinson BW, Bryan JS, Rosvold HE. Locating brain structures. Extensions to the impedance method. Arch Neurol 1965;13:477-86. 19. Tachibana S. Cerebral impedography in cat, rabbit and cadaver. Electroencephalogr Clin Neurophysiol 1969;27:670. 20. Tachibana S. Impedography in tumor localization. Trans Am Neurol Assoc 1970;95:317-9. 21. Tachibana S. Impedance study of brain tissue changes after penetrating injury. Exp Neurol 1971;32:206-17. 22. Laitinen LV. Personal memories of the history of stereotactic neurosurgery. Neurosurgery 2004;55:1420-28; discussion 8-9.

23. Hobza V, Jakubec J, Nemeckova J, Nemecek S, Sercl M. Impedance monitoring in the stereotactic localization of intracranial structures. Sb Ved Pr Lek Fak Karlovy Univerzity Hradci Kralove 1995;38:33-46. 24. Laitinen LV. Pallidotomy for Parkinson’s disease. Neurosurg Clin N Am 1995;6:105-12. 25. Heilbrun MP, Koehler S, McDonald P, Faour F. Optimal target localization for ventroposterolateral pallidotomy: the role of imaging, impedance measurement, macrostimulation and microelectrode recording. Stereotact Funct Neurosurg 1997;69(1-4)(Pt 2):19-27. 26. Iacono RP, Carlson JD, Kuniyoshi SM, Li YJ, Mohamed AS, Maeda G. Electrophysiologic target localization in posteroventral pallidotomy. Acta Neurochir (Wien) 1997;139:433-41. 27. Siemionow V, Yue GH, Barnett GH, Sahgal V, Heilbrun MP. Measurement of tissue electrical impedance confirms stereotactically localized internal segment of the globus pallidus during surgery. J Neurosci Methods 2000;96:113-17. 28. Hariz MI. Safety and risk of microelectrode recording in surgery for movement disorders. Stereotact Funct Neurosurg 2002;78:146-57. 29. Volkmann J, Herzog J, Kopper F, Deuschl G. Introduction to the programming of deep brain stimulators. Mov Disord 2002;17 Suppl 3:S181-7. 30. Butson CR, Maks CB, McIntyre CC. Sources and effects of electrode impedance during deep brain stimulation. Clin Neurophysiol 2006;117(2):447-54. 31. Joint C, Nandi D, Parkin S, Gregory R, Aziz T. Hardware-related problems of deep brain stimulation. Mov Disord 2002;17 Suppl 3:S175-80.

81

Lesions Versus Implanted Stimulators in Functional Neurosurgery W. S. Anderson . R. E. Clatterbuck . K. Kobayashi . J.-H. Kim . F. A. Lenz

Introduction Surgery for movement disorders began with cortical resections by Victor Horsely for chorea in 1906 [1]. In the late 1930s, Meyers pioneered a series of transventricular procedures in the basal ganglia for Parkinson’s disease [2]. The next advance in surgery for movement disorders came with the application of stereotactic techniques to lesioning of structures in the basal ganglia [3]. Lesion targets were refined in the 1950s and 1960s leading to the practice of pallidotomy for the treatment of rigidity and akinesia [4] and thalamotomy for tremor [5]. However, surgery for the treatment of movement disorders was dramatically curtailed with the development of levodopa as pharmacologic therapy for Parkinson’s disease. With the long-term treatment of patients with levodopa, the complications of levodopa became evident leading to renewed interest in the surgical options for treatment of movement disorders [6]. Leksell’s posteroventral pallidotomy was re-established as an effective treatment option for rigidity and bradykinesia in patients with Parkinson’s disease [7]. Stimulation of lesioning targets had long been known to be capable of alleviating symptoms in the operating room. In the early part of this decade, the development of improved technologies for implantable stimulating electrodes made deep brain stimulation (DBS) a viable option to surgical lesioning. We review here several of the important studies demonstrating efficacy of lesioning in the #


treatment of movement disorders from the last century. The discussion will primarily be limited to posteroventral pallidotomy and ventral intermediate (Vim) thalamotomy as the most common lesioning operations with a brief discussion of radiosurgical treatments, and lesioning for psychiatric diseases. These procedures will be compared with the three most common DBS procedures: high frequency stimulation (hfs) of the globus pallidus internus (Gpi), Vim, and subthalamic nucleus (STN). The potential complications and benefits of lesioning and stimulation will be discussed and compared.

Posteroventral Pallidotomy and GPi Stimulation Posteroventral pallidotomy for Parkinson’s disease was re-introduced by Laitinen [7]. In his study, 38 patients with Parkinson’s disease and a primary complaint of akinesia underwent stereotactic posteroventral pallidotomy and mean follow up of 28 months. Motor function was assessed using writing, drawing, and gait tests. Near complete relief of rigidity and akinesia were reported in 92% of patients. Eighty-one percent of patients with tremor experienced dramatic improvement. Levodopa induced dyskinesias were also improved. Six of these patients, however, suffered a permanent central homonymous visual field defect. A series of well-designed studies of posteroventral pallidotomy [8–12] have followed the

1350

81

Lesions versus implanted stimulators in functional neurosurgery

lead of Laitinen et al. Several incorporated evaluation protocols in which observers reviewed patients’ exams by videotape and were blinded to the treatment patients had received [8,9,11]. These studies all employed Hoehn and Yahr patient staging [13] and assessment according to the Core Assessment Program for Intracerebral Transplantation (CAPIT) [14] which incorporates the Unified Parkinson’s Disease Rating Scale (UPDRS) [15]. No patient in these studies was better than a Hoehn and Yahr stage III in the on-state. Follow-up intervals from 3 months to 1 year demonstrated a range of improvement in the off-state in the UPDRS motor subscale score from 14 to 70%. These studies demonstrated marked contralateral improvements in rigidity, akinesia, tremor, gait, balance, levodopa related dyskinesias, and on-off fluctuations. These studies documented some ipsilateral improvements as well. The first of these studies reported an off-state improvement in the UPDRS motor score of 71% at 1 year post-operatively [8]. In the study of Lozano et al. [9] UPDRS total motor score in the off-state improved by 30% at 6 months, while the total akinesia score improved by 33%. This study also noted a 15% improvement in the gait score in the off-state and a 92% reduction in contralateral dyskinesias. In a study by Baron et al. [10], a 25% improvement in the UPDRS motor score in the off-state at 3 months was seen. Shannon et al. [12] noted a 15% improvement in the off-state mean motor UPDRS score at 6 months; and a contralateral off-state combined tremor, rigidity, and bradykinesia score improved by 26%. In this study the dyskinesia severity score was improved by 73% at 6 months. In the study of Ondo et al. [11], total off-state UPDRS motor scores improved 14%. Improvement in UPDRS total tremor subscore improved by 59%, gait scores by 22%, and body bradykinesia by 17%. Lang et al. have published a 2-year follow-up study on 11 patients after posteroventral pallidotomy [16]. The initial results in their group of

40 patients were similar to the above studies at 6 months with overall improvement in motor function of 28%. The effect of pallidotomy on contralateral bradykinesia, dyskinesia, and rigidity were maintained at 2 years, while ipsilateral effects were generally lost. Siegfried and others began work with chronic implantable central nervous system stimulation in the early 1980s. Much of the early clinical work with stimulation was done for intractable pain and for tremor. Siegfried and Lippitz, inspired by Laitinen’s work in the posteroventral pallidum, first reported bilateral GPi-hfs in three patients in 1994 [17]. All three patients with advanced Parkinson’s disease (Hoehn and Yahr stage IV or worse) experienced dramatic decreases in on-off phenomena and levodopa induced dyskinesias. Effects were reversed, though not immediately, when stimulation was discontinued. Several other groups have demonstrated marked improvement in patients treated with unilateral and bilateral GPi-hfs. Pahwa et al. reported the treatment of five patients with Parkinson’s disease with GPi-hfs, three with bilateral implants [18]. All patients had disabling symptoms with Hoehn and Yahr stage III disease or worse. At 3-month follow-up the amount of time in the on-state increased from 21 to 65%. UPDRS motor scores in the off-state without stimulation improved 24% at 3 months post-operatively compared to the preoperative off-state. UPDRS motor scores at 3 months post-operatively in the onstate with stimulation were improved 60% over the preoperative on-state. In the off-state at 3 months follow-up, turning on stimulation improved UPDRS scores 21%. In a similar series, Gross et al. reported seven patients with Hoehn and Yahr stage III-IV disease who underwent placement of unilateral GPi-hfs electrodes [19]. Mean improvement over off-state UPDRS motor scores (post-procedure) were 33% with levodopa, 35% with stimulation, and 63% with both. Off-state scores pre- and post-procedure were similar. In addition, four of five patients with


tremor experienced considerable improvement. No patients experienced ipsilateral effects. These therapeutic benefits were purchased at a cost. Among the fifteen patients undergoing pallidotomy reported by Baron et al. [10], two (13%) suffered subclinical frontal hemorrhages, one (7%) suffered transient dysarthria, one (7%) suffered persistent worsening of a baseline dysarthria, and one (7%) had a persistent superior quadrantanopsia. Several patients in this study also experienced transient confusion and several experienced transient facial weakness. Ondo et al. [11] reported that among 34 patients undergoing pallidotomy, five (15%) experienced transient side effects which included aphasia (3%) and altered mental status (12%). Shannon et al. [12] reported 26 patients who underwent pallidotomy. One patient (4%) had a fatal hemorrhage and three (12%) had nonfatal hemorrhages. In addition two patients (8%) had cognitive changes, one patient (4%) developed aphasia, three patients (12%) experienced persistent frontal lobe dysfunction, one patient (4%) developed a mild but persistent hemiparesis, and one patient (4%) experienced persistent increase in dysarthria. This study also reported some transient side effects including altered mental status, facial weakness, and dysarthria. Dogali et al. reported no significant complications related to pallidotomy in eighteen patients [8]. Kondziolka et al. reported a 5% rate of transient dysarthria lasting 1–3 weeks in a series of 120 pallidotomies, with no hemorrhages encountered [20]. Additionally, Iacono et al. reported in a series of 126 patients undergoing either unilateral (58) or bilateral (68) pallidotomies, a hemorrhage rate of 3.2% per lesion [21]. De Bie et al. provided a very thorough summary of pallidotomy complications recorded from 1992 to 2000 [22]. These authors found in 334 cases of published unilateral pallidotomies in prospective studies, that 13.8% suffered permanent adverse effects (which were most commonly changes in personality or behavior, dysarthria,

81

dysphagia, and visual field defects). A symptomatic infarction or hemorrhage occurred in 3.9%, and the mortality rate was 1.2%. There was also a group of recorded patients who had undergone bilateral pallidotomy (five cohort studies with 20 total patients). Fourteen suffered adverse effects which included effects on speech and cognition. Interestingly, the patients undergoing microelectrode recording were separately analyzed, and demonstrated an increased rate of adverse effects (14.4% higher) with a frequency of infarct that was also 4.9% higher [20]. In the five patients included in the GPi-hfs studies of Pahwa et al. [18], the only complications reported included a single asymptomatic intracranial hemorrhage, a transient speech difficulty and hemiparesis related to stimulation which resolved in the operating room, and a facial dystonia and paresthesia which required electrode repositioning. Adverse side effects of stimulation included one patient with a visual disturbance (transient) and one patient with stimulation related chorea of a foot. Gross et al. [19] reported seven patients undergoing pallidal stimulation procedures without any complications.

Ventral Thalamotomy and Vim Stimulation One of the first large post-levodopa series of stereotactic thalamotomies for Parkinson’s disease with medically intractable tremor was completed at the Mayo clinic [23]. In this study, 36 patients (mean Hoehn and Yahr stage 2.4) were treated with 37 thalamotomies and 31 (86%) experienced complete relief of their tremor. Another three (5%) were significantly improved. During the follow-up period, which ranged from 14 to 68 months, only two patients suffered from recurrent tremor (both within 3 months). Diederich et al. [24] blindly compared tremor on the operated and unoperated side in 17 patients with Parkinson’s disease at a mean of 10.9 years

1351

1352

81


following stereotactic thalamotomy/subthalamotomy using videotaped examinations. At follow-up the mean Hoehn and Yahr stage was 1.8 and UPDRS motor subscore was 17.8. Severity scores for upper extremity tremor were significantly better on the side contralateral to surgery than on the ipsilateral side. In all these patients, the surgical side was initially chosen to treat the side with the more severe tremor. The largest recent series [25] retrospectively reviewed the outcomes in 60 patients with parkinsonian tremor (42 patients), essential tremor (6), cerebellar tremor (6), and post-traumatic tremor (6). These patients all underwent unilateral stereotactic Vim thalamotomy with the exception of two Parkinson’s patients who underwent bilateral procedures, and one patient who underwent lesioning more anteriorly. With a mean follow-up of 53.4 months, Parkinson’s patients experienced moderate to marked improvement in 86% of the cases. Patients with essential tremor showed similar improvement in 83% of cases. Results were not as dramatic for those patients with cerebellar tremor (67%) or post-traumatic tremor (50%). Our own experience with stereotactic Vim thalamotomy for essential tremor provides confirmation of these reports with a blinded measure of pre- and postoperative status [26]. Patients were evaluated preoperatively and at 3 and 12 months postoperatively with a functional disability score and a blinded handwriting/drawing score. Significant improvements in both scores were found postoperatively. Within patients analysis demonstrated statistically significant improvement in 72% of patients. With the advent of chronic DBS in the central nervous system for movement disorders, stimulation of the Vim nucleus has been examined as a treatment for tremor. The largest study [27] included 117 patients (177 operated sides) with movement disorders, including 80 with Parkinson’s disease and 20 with essential tremor. Bilateral implantation was undertaken in 59

patients and 14 patients underwent implantation contralateral to a previous thalamotomy. The follow-up period was as long as 7 years for the earliest procedures. At last follow-up 88% of Parkinson’s patients had complete or near complete relief of tremor, and another 10% had slight to moderate improvement. Global scores including all four limbs were slightly lower. Rigidity and akinesia were not significantly affected. In this series the effect on essential tremor was less dramatic with only 61% of patients experiencing complete or near complete relief of tremor at last follow-up. The effect of Vim DBS was inconsistent for patients with other dyskinesias or tremors. Other studies have reported similar results. Koller et al. [28] studied unilateral Vim DBS in 24 patients with Parkinson’s disease and 29 patients with essential tremor in a multicenter trial using a blinded 3 month postoperative evaluation and an open label 1-year follow-up. Complete resolution of tremor was seen in 31% of essential tremor patients and 58.3% of Parkinson’s patients. Only 3.4% of essential tremor patients and only 4.2% of Parkinson’s disease patients had no change in their tremor. No ipsilateral effects were detected. One direct comparison of thalamotomy and thalamic DBS for tremor found essentially equal efficacy in abolishing tremor (63.7% vs. 62.5%), however, 27% of thalamotomies had to be repeated for tremor recurrence while none of the DBS procedures needed revision [29]. Transient complications of thalamotomy were seen in 58–70% of patients and included contralateral weakness, dysarthria, dysphasia, confusion, dystonia, and sensory disturbances [23,25]. One patient in these two series died at 7 days postoperatively from a pulmonary embolism. The number of patients with long term or permanent complications was much smaller, in the range of 14–23%. Among these complications weakness/dyspraxia and dysarthria figured most prominently. Only one patient in these series experienced permanent cognitive difficulties


after a unilateral procedure. Favre et al. reported one hematoma (out of 137 lesioning procedures) in the context of a thalamotomy in a case where the systolic blood pressure had remained elevated at 160 mm Hg [30]. In one series of Vim DBS [28], six of 53 procedures were aborted secondary to intraoperative complications including failure of stimulation to suppress tremor (2 patients), intracranial hemorrhage (1), and subdural hemorrhage (1). One patient in this series experienced a postoperative seizure. Transient side effects of stimulation included paresthesias, experienced by most patients, and gait disorders, that were much less common. The long-term complications seen in this series in the first year included two superficial wound infections, one extension wire erosion, and one failure of the implantable pulse generator. In the large series of Benabid et al. [27], 5.1% of patients experienced small hematomas as a result of electrode passage (half of these were transiently symptomatic), and 31.6% of patients experienced minor side effects that were reversible with discontinuation of stimulation. These included paresthesias (9%), foot dystonia (9%), dysequilibrium (9%), and contralateral dystonia (5%). Dysarthria was seen in 23 patients (19.6%). Interestingly, 14 of these patients were receiving bilateral stimulation and four had a contralateral thalamotomy. Three secondary scalp infections (3%) leading to hardware removal were observed.

Bilateral Subthalamic Nucleus Stimulation Two prospective studies from the 1990s have been published discussing subthalamic nucleus (STN) stimulation for the treatment of Parkinson’s disease symptoms [31]. The evolution of DBS has made STN procedures feasible. Kumar et al. carried out a double-blind evaluation of seven patients with end-stage Parkinson’s disease that underwent bilateral STN-hfs. Patients

81

experienced an improvement in mean UPDRS motor scores post-operatively of 58% in the off-state with stimulation. In the on-state with stimulation, patients’ mean UPDRS motor score improved by 41% compared to the preoperative on-state. Time in the off-state and dyskinesias were both decreased postoperatively. Two patients experienced transient hemichorea as a result of the procedure, and four patients experienced some degree of postoperative cognitive difficulty (one experienced a venous infarction and another a thalamic lesion). Limousin et al. [32] reported 24 patients with Parkinson’s disease (Hoehn and Yahr stage IV-V in the off-state) that underwent bilateral STN-hfs. At 1-year follow-up, mean UPDRS motor scores improved 60% with stimulation in the off-state and 10% in the on-state. As in the Kumar series, patients were significantly improved in the off-state with the stimulator off. Rigidity, akinesia, tremor, gait, and dyskinesias were all improved. Complications included a large intracerebral hematoma (4%) leading to paralysis and aphasia and a subcutaneous infection (4%) requiring hardware removal. Eight patients experienced transient cognitive difficulties, and one patient had permanent worsening of preoperative cognitive deficits. Five patients experienced an eyelid apraxia requiring treatment. In two larger series of DBS for chronic pain [33,34], among a total of 263 patients, ten intracerebral hemorrhages (4%) (including three deaths), 23 infections (9%), ten hardware erosions (4%), and seven foreign body reactions (3%) were seen.

Other Lesioning Procedures Stereotactic radiosurgery for functional procedures is not as widely used as stimulation or traditional thermocoagulation, but is practiced as a form of lesioning by some groups [35–37]. It is particularly useful for patients not tolerating

1353

1354

81


implanted hardware, and demonstrates efficacy rates similar to the previous thermocoagulation studies. For instance, Kondziolka et al. found that 69% of 31 patients treated with Gamma Knife thalamotomy for essential tremor demonstrated improvements in action tremor and writing [35]. Complications of radiosurgery for functional procedures have also been reported, and include dysphagia, hemiplegia, visual field deficits, dysarthria, and even pseudobulbar laughter in a set of patients in which poor targeting was to blame [38]. In their series of 31 patients undergoing Gamma Knife thalamotomy for essential tremor, Kondziolka et al. reported one patient who developed a permanent mild hemiparesis and dysarthria 7 months after treatment [35]. Additionally, radiosurgical treatment does not allow for the use of microelectrode mapping to further refine the target location, which might make for higher morbidity rates [39]. Lesioning procedures performed for the treatment of psychiatric disease or pain will also have similar morbidities found in lesioning for movement disorders, including hemorrhage and infection. Seizures as a complication are probably represented more in series of cingulotomies. Wilkinson et al. describe a series of 23 patients undergoing bilateral cingulotomy for chronic pain [40]. Two of these patients suffered intraoperative seizures, and five had late seizures. In the large series of Ballantine et al. involving 714 cingulotomies in 414 patients, there were no deaths reported and no infections [41]. Two of the patients suffered acute subdural hematomas resulting in hemiplegia, and one patient developed a chronic subdural collection. Five of the patients suffered chronic postoperative seizures controlled with phenytoin. Bilateral lesioning of the anterior cingulate can produce cognitive deficits as well. Ochsner et al. reported on deficits picked up in visual cognitive inventories after cingulotomy [38]. These patients had difficulties with sequencing cognitive operations needed to generate complex

or moving images or in rotating images. These deficits are most likely related to the executive system function which the cingulate gyrus participates in [42]. Similar findings were present in a group of chronic pain patients having undergone bilateral anterior cingulotomy [42]. This group of patients, when compared to untreated chronic pain patients, showed deficits in attention and executive function. Self-initiated behaviors as well as behavioral spontaneity were the most affected [42]. This same group of authors also described changes in emotional experience after bilateral anterior cingulotomy [43]. The greatest effects were seen on inventories that correlate most with levels of emotional tension and agitation. This agrees with the clinical effects seen in patients with anxiety disorders. In the recent series of cingulate stimulation for OCD, no conclusive cognitive disturbances were noted [44,45], although this will be an interesting comparison as these cohorts grow larger.

Conclusions In the last decade many new options have become available for the surgical treatment of movement disorders. The principal procedures currently in use include posteroventral pallidotomy, ventrolateral thalamotomy, and DBS in Vim, GPi and STN. With the development of technology allowing routine use of indwelling stimulating devices in the central nervous system, DBS has become a more viable option. Although the collective experience with DBS is considerably smaller (but ever growing) than that with lesioning procedures, present data suggests that stimulation is as effective as lesioning procedures. Posteroventral pallidotomy series demonstrate 14–70% improvement in off-state UPDRS motor scores following surgery. In patients treated with pallidal stimulation, turning on the stimulator in the off-state improved UPDRS motor scores by 21–35%. Similarly, most


series place significant improvement in tremor following ventrolateral thalamotomy and Vim stimulation at over 80%. One series that directly compared the two modalities found both effective at around 63% [29]. These studies all suggest that lesioning and stimulation are equally efficacious in the treatment of movement disorders. DBS carries the obvious advantage of being reversible and adjustable. These advantages make bilateral procedures more practical and safe. It remains to be seen if these theoretical advantages will be born out in larger series or randomized trials. Given these advantages however, there are situations in which lesioning is warranted. A lesioning procedure would be more appropriate for people who do not want an implantable device or did not tolerate it for a variety of wound healing or infection issues. Also, lesioning might be more appropriate in patients with no possibility of follow-up for generator programming or replacement, for instance patients living in a developing nation. In general though, stimulation should be used in all patients whenever possible. It would appear from the data presented here that clinically significant hemorrhage risks are similar for permanent lesioning paradigms and stimulator lead placements, about 1.7–4% [30]. The obvious disadvantages of permanent indwelling hardware, increased infection risk (3–9%) and mechanical failure, are born out to some degree with the present series. Other neurologic complications seen with stimulation such as paresthesias and dystonias generally resolved when stimulation was decreased. In comparison, complications such as aphasia, cognitive worsening, mild hemiparesis, and visual field deficits seen with lesioning were in some cases permanent. Such complications were seen in as high as 14–23% of patients. Another aspect to be considered is the increased costs associated with implanting and maintaining DBS. It will take a multicenter, prospective, randomized trial with outcome measures including a cost-benefit analysis to address all these issues.

81

References 1. Horsley V. The function of the so-called motor cortex. Br Med J 1909;2:125-32. 2. Meyers R. Surgical procedure for postencephalitic tremor, with notes on the physiology of the premotor fibers. Arch Neurol Psych 1940;44:455-9. 3. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human brain. Science 1947;106:349-50. 4. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotactic thermolesions in the pallidal region. A clinical evaluation of 81 cases. Acta Psych Scand 1960;35:358-77. 5. Hassler R, Riechert T. Indikationen und Lokalisationsmethode der gezielten Hirnoperationen. Nervenarzt 1954;25:441-7. 6. Marsden CD, Parkes JD. Success and problems of longterm levodopa therapy in Parkinson’s disease. Lancet 1977;1:345-9. 7. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61. 8. Dogali M, Fazzini E, Kolodny E, Eidelbert D, Sterio D, Devinsky O, Beric A. Stereotactic ventral pallidotomy for Parkinson’s disease. Neurol 1995;45:753-61. 9. Lozano AM, Lang AE, Galvez-Jimenez N, Miyasaki J, Duff J, Hutchison WD, Dostrovsky JO. GPi pallidotomy improves motor function in patients with Parkinson’s disease. Lancet 1995;346:1383-6. 10. Baron MS, Vitek JL, Bakay RAE, Green J, Kaneoke Y, Hashimoto R, Turner RS, Woodward JL, Cole SA, McDonald WM, DeLong MR. Treatment of advanced Parkinson’s disease with microelectrode-guided pallidotomy: 1 year pilot study results. Ann Neurol 1996;40:355-66. 11. Ondo WG, Jankovic J, Lai EC, Sankhla C, Khan M, Ben-Arie L, Schwartz K, Grossman RG, Krauss JK. Assessment of motor function after stereotactic pallidotomy. Neurology 1998;50:266-70. 12. Shannon KM, Penn MD, Kroin JS, Janko KA, York M, Cox SJ. Stereotactic pallidotomy for the treatment of Parkinson’s disease. Neurol 1998;50:434-8. 13. Hoehn MM, Yahr MD. Parkinsonism: onset, progression, and mortality. Neurol 1967;17:427-42. 14. Langston JW, Winder H, Goetz CG, Brooks D, Fahn S, Freeman T, Watts R. Core Assessment Program for Intracerebral Transplantations (CAPIT). Mov Disord 1992;7:2-13. 15. Fahn S, Elton RL, members of the UPDRS Development Committee. Unified Parkinson’s disease rating scale. In: Fahn S, Marsden CD, Calne DB, Goldstein M, editors. Recent developments in Parkinson’s disease, vol. 2. Florham Park: MacMillan Health Care Information; 1987. p. 153-64.

1355

1356

81


16. Lang AE, Lozano AM, Montgomery E, Duff J, Tasker R, Hutchinson W. Posteroventral medial pallidotomy in advanced Parkinson’s disease. N Eng J Med 1998;337:262-3. 17. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of the ventroposterolateral pallidum: a new therapeutic approach for alleviating all parkinsonian symptoms. Neurosurg 1994;35:1126-30. 18. Pahwa R, Wilkinson S, Smith D, Lyons K, Miyawaki E, Koller WC. High-frequency stimulation of the globus pallidus for the treatment of Parkinson’s disease. Neurol 1997;49:249-53. 19. Gross C, Rougier A, Guehl D, Boraud T, Julien J, Bioulac B. High-frequency stimulation of the globus pallidus internalis in Parkinson’s disease: a study of seven cases. J Neurosurg 1997;87:491-8. 20. Kondziolka D, Firlik AD, Lunsford LD. Complications of stereotactic brain surgery. Neurol Clin North Am 1998;16(1):35-54. 21. Iacono RP, Shima F, Lonser RR, Kuniyoshi S, Maeda G, Yamada S. The results, indications, and physiology of posteroventral pallidotomy for patients with Parkinson’s disease. Neurosurg 1995;36:1118-25. 22. de Bie RMA, de Haan RJ, Schuurman PR, Esselink RAJ, Bosch DA, Speelman JD. Morbidity and mortality following pallidotomy in Parkinson’s disease. Neurol 2002;58:1008-12. 23. Fox MW, Ahlskog EJ, Kelly PJ. Stereotactic ventrolateralis thalamotomy for medically refactory tremor in postlevodopa era Parkinson’s disease patients. J Neurosurg 1991;75:723-30. 24. Diederich N, Goetz CG, Stebbins GT, Klawans HL, Nitter K, Koulosakis A, Sanker P, Sturm V. Blinded evaluation confirms long-term assymmetric effect of unilateral thalamotomy or subthalamotomy on tremor in Parkinson’s disease. Neurol 1992;42:1311-14. 25. Jankovic J, Cardoso F, Grossman RG, Hamilton WJ. Outcome after stereotactic thalamotomy for parkinsonian, essential and other types of tremor. Neurosurg 1995;37:680-7. 26. Zirh AT, Reich SG, Dougherty PM, Lenz FA. Stereotactic thalamotomy in the treatment of essential tremor of the upper extremity: re-assessment including a blinded measure of outcome. J Neurol Neurosurg Psych 1999;66(6):772-5. 27. Benabid AL, Pollak P, Gao D, Hoffman D, Limousin P, Gay E, Payen I, Benazzouz A. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as treatment of movement disorders. J Neurosurg 1996;84:203-14. 28. Koller W, Pahwa R, Busenbark K, Hubble J, Wilkinson S, Lang AE, Tuite P, Sime E, Lozano A, Hauser R, Malapira T, Smith D, Tarsy D, Myawaki E, Norregaard T, Kormos T, Olanow CW. High frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Ann Neurol 1997;42:292-9.

29. Tasker RR, Munz M, Junn FSCK, Kiss ZHT, Davis K, Dostovsky JO, Lozano AM. Deep brain stimulation and thalamotomy for tremor compared. Acta Neurochir Suppl 1997;68:49-53. 30. Favre J, Taha JM, Burchiel KJ. An analysis of the respective risks of hematoma formation in 361 consecutive morphological and functional stereotactic procedures. Neurosurg 2002;50:48-56; discussion 56-7. 31. Kumar R, Lozano AM, Kim YJ, Hutchinson WD, Sime E, Halket E, Lang AE. Double blind evaluation of subthalamic nucleus deep brain stimulation in advanced Parkinson’s disease. Neurol 1998;51:850-5. 32. Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann D, Benabid AL. Electrical stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 1998;339:1105-11. 33. Hosobuchi Y. Subcortical electrical stimulation for control of intractable pain in humans. J Neurosurg 1986;64:543-53. 34. Levy RM, Lamb S, Adams JE. Treatment of chronic pain by deep brain stimulation: Long term follow-up and review of the literature. Neurosurg 1987;21:885-93. 35. Kondziolka D, Ong JG, Lee JY, Moore RY, Flickinger JC, Lunsford LD. Gamma knife thalamotomy for essential tremor. J Neurosurg 2008;108:111-17. 36. Okun MS, Stover NP, Subramanian T, Gearing M, Wainer BH, Holder CA, Watts RL, Juncos JL, Freeman A, Evatt ML, Schuele SU, Vitek JL, DeLong MR. Complications of gamma knife surgery for Parkinson disease. Arch Neurol 2001;58:1995-2002. 37. Friedman DP, Goldman HW, Flanders AE, Gollomp SM, Curran WJ. Stereotactic radiosurgical pallidotomy and thalamotomy with the gamma knife: MR imaging findings with clinical correlation – Preliminary experience. Radiology 1999;212:143-50. 38. Ochsner KN, Kosslyn SM, Cosgrove CR, Cassem EH, Price BH, Nierenberg AA, Rauch SL. Deficits in visual cognition and attention following bilateral anterior cingulotomy. Neuropsych 2001;39:219-30. 39. Jankovic J. Editorial: surgery for Parkinson disease and other movement disorders. Arch Neurol 2001;58:1970-2. 40. Wilkinson HA, Davidson KM, Davidson RI. Bilateral anterior cingulotomy for chronic noncancer pain. Neurosurg 1999;45:1129-34. 41. Ballantine HT, Giriunas IE. Treatment of intractable psychiatric illness and chronic pain by stereotactic cingulotomy. In: Schmidek HH, Sweet WH, editors. Operative neurosurgical techniques. Indications, methods, and results. Philadelphia, PA: W.B. Saunders; 1988. p. 1069-75. 42. Cohen RA, Kaplan RF, Zuffante P, Moser DJ, Jenkins MA, Salloway S, Wilkinson H. Alteration of intention and self-initiated action associated with bilateral anterior cingulotomy. J Neuropsych Clin Neurosci 1999; 11:444-53.


43. Cohen RA, Paul R, Zawacki TM, Moser DJ, Sweet L, Wilkinson H. Emotional and personality changes following cingulotomy. Emotion 2001;1:38-50. 44. Greenberg BD, Malone DA, Friehs GM, Rezai AR, Kubu CS, Malloy PF, Salloway SP, Okun MS, Goodman WK, Rasmussen SA. Three-year outcomes in deep brain stimulation for highly resistant obsessive-compulsive disorder. Neuropsychopharm 2006;31:2384-93.

81

45. Nuttin BJ, Gabrie¨ls LA, Cosyns PR, Meyerson BA, Andreéwitch S, Sunaert SG, Maes AF, Dupont PJ, Gybels JM, Gielen F, Demeulemeester HG. Long-term electrical capsular stimulation in patients with obsessivecompulsive disorder. Neurosurg 2003;52:1263-72.

1357

78 Microelectrode Recording in Functional Neurosurgery W. D. Hutchison . J. O. Dostrovsky . M. Hodaie . K. D. Davis . A. M. Lozano . R. R. Tasker

Introduction Whenever a neurosurgical procedure involves a target or endangers an important neighboring structure that cannot be seen or distinctly imaged intraoperatively, some form of invasive physiological localization is required to assure accuracy and safety. Sometimes a whole structure cannot be visualized clearly (such as STN); on other occasions, the gross structure can be imaged but its important functional subdivisions cannot (such as the case with motor thalamus), even with the highest-quality magnetic resonance imaging (MRI). In both circumstances, the structure must be penetrated and its identity must be established by functional means. The first part of this chapter is concerned with invasive physiological localization of deep brain structures with microor macro-electrodes prior to surgery. The second part of this chapter is devoted to imaging techniques and their use in cortical and subcortical localization; localization relevant to the superficial cerebral cortex regions and to surgery for epilepsy is discussed in the section dealing with epilepsy (see chapters 153 and 157). There is still a need for invasive physiological localization despite any shortcomings since functional stereotactic imaging still does not allow accurate visualization of all stereotactic targets that the surgeon can manipulate. The current consensus is that imaging techniques are still not sufficiently accurate to achieve the best results in most functional stereotactic procedures. If every brain were identical, it would be expected that any target structure would bear a fixed three dimensional relationship in space to brain landmark structures #


such as the anterior and posterior commissures (AC and PC). The fact that this is not so is well known from cortical mapping. For example, a given site on the postcentral gyrus a fixed distance from the midsagittal line may in one patient represent the face, in another the arm, and in yet another the leg [1]. However, it is clear to those who routinely perform subcortical mapping, that there is ongoing variation in initial or image-based targeting due to errors from various sources as revealed by the selection of the final target taking into account the mapping results.

Part 1 – Invasive Physiological Localization Methods of Subcortical Physiological Localization Table 78-1 lists some of the available techniques for invasive physiological localization. The most important are microelectrode recording, and electrical stimulation using micro- and macroelectrodes, which will be covered in some detail first. Electroencephalography (EEG), electrocorticography (ECoG), are used mainly during epilepsy surgery and will not be further described here (see chapters 153 and 157) but results from ‘‘deep EEG’’ studies were at one time used for functional localization of deep brain structures [2–6]. The deep EEG technique can be considered to be reinvented, since more and more centers are also recording LFP activity from micro- and macroelectrodes intraoperatively as well as in the immediate post-operative period from DBS electrode >

1284

78

Microelectrode recording in functional neurosurgery

. Table 78-1 Techniques of invasive physiological localization Microelectrode recordings Electrical stimulation, including microstimulation and macrostimulation Local field potential recordings (‘‘deep EEG’’ electroencephalography and electrocorticography) Noninvasive functional imaging Evoked potential recording Impedance monitoring Semimicroelectrode recordings Neural noise recording Microinjection of test substances (local anesthetic or muscimol) Optical imaging of cerebral cortex

contacts. The characteristic oscillation frequencies such as the STN beta band activity in PD patients off medication indeed helps to confirm STN localization. In a similar way, evoked potential recording has been used as a localization method in the past, and recently our group has used focal microelectrode evoked potentials with some further insight into physiological localization. Finally impedance monitoring is rarely used for localization and the techniques of microinjection of test substances are only occasionally used for specific indications or investigations.

Electrical Stimulation and Microelectrode Recording Currently, stimulation and recording are the most widely used techniques for the physiological localization of subcortical structures. Recording can be done with fine tipped microelectrodes capable of discriminating single cells [7–16] or with semimicroelectrodes that cannot [17–22]. Stimulation can be done with a large tipped electrode (macrostimulation) [3–6,17,23–29] or a microelectrode (microstimulation) [7,8,12]. Each technique has its advantages and disadvantages (> Table 78-2). In its simplest form, physiological localization can be reduced to observing the effects of macrostimulation on motor and sensory function.

In peripheral nerves, roots, and certain long tracts, low-frequency (often 2 Hz) stimulation is used to search for motor twitches and higher-frequency (30–300 Hz) stimulation is used for both sensory and motor effects, the latter in the form of tetanization. A correlation between the effect observed and the parameters used allows, with experience, a reasonable estimation of the distance of the stimulation probe from the target structure. Somatotopographic features also can be obtained, as in exploration of the trigeminal nerve, in which selective manipulation is desired. Similar macrostimulation techniques can be used for selective lesioning of the lateral spinothalamic tract. However, in most subcortical explorations, matters may be a little more complicated. Fritsch and Hitzig are said to have been the first to elicit motor responses by stimulating the motor cortex in experimental animals; Bartholomew was first to do so in humans, while Cushing was the first to elicit sensory effects in humans [19,30]. Macrostimulation was employed for confirmation of probe position in early functional stereotactic procedures [6] only after a probe was thought to have reached the appropriate target site. However, macrostimulation also can be done systematically at fixed intervals as the probe is passed into the brain, allowing the results to be plotted in ‘‘figurine charts’’ of the type used by Woolsey in the laboratory [31]. Such mapping provides more comprehensive localization data that are more easily assessed visually; in addition, it provides information about normal and abnormal brain organization, especially if each electrode trajectory is contained in the same sagittal plane. Wetzel and Snider [32] are said to have been the first to use microelectrode recording in humans in 1958 during a pallidotomy, but the technique did not come into common use until after the introduction of thalamotomy for Parkinson’s disease [2,9–11,18,19,21,22,30,33–51]. The earlier work in the thalamus was done with semimicroelectrodes, while later on true microelectrodes were employed. Semimicroelectrodes can more readily monitor

78


. Table 78-2 Comparison of microelectrode recording with stimulation for physiological localization in stereotactic surgery

Safety Speed of exploration Radius of current spread* Dependence on patient cooperation Ability to record and stimulate with same tip Ruggedness for repeated use Sophistication of backup equipment and activities Ability to detect somatotopy Variety of structures identified

Ability to functionally compartmentalize identified structures such as ventrobasal complex Ability to identify nearby structures Risk of technical problems, cost

Electrical Stimulation

Microelectrode Recording

Macrostimulation

Microstimulation

Low Impedance

High Impedance

High Rapid To 3 mm High No

High Medium 0.02–1 mm High Yes

High Medium Figure 78-1). Microstimulation is done every 1.0 mm along the way. Currents of up to 100 mA do not damage the electrodes,. Physiological data and a voice channel are recorded for off-line analysis using a commercially available digital interface (CED 1401 mk II) that saves two channels of electrode data, 4 EMGs, 2 accelerometer traces and EOG or any other stimulus triggers as required. These are collected continuously into digital files. Similarly a digital video camera saving direct to DVD is used to monitor events and patient’s movements during the entire recording session. Responses elicited by micro- and macrostimulation are similar in quality, but the threshold and current spread differ. The size of the projected fields at threshold current (the minimal current needed to evoke a sensation) usually is smaller with microelectrodes, but in some cases macrostimulation also results in small projected fields (> Figure 78-2). Frequency of stimulation has a variable effect. Sensory effects may not be perceived with very low (1–5 Hz) or high (over 1,000 Hz) rates of repetition, and for the patient to perceive a sensory effect, trains of

several pulses are required rather than single shocks (> Figure 78-3). Frequency of stimulation in the motor system determines whether individual muscle jerks or tetanization will be produced, and in the extrapyramidal system it determines whether involuntary movements will be driven or inhibited. The main part of this chapter reviews physiological mapping with recording and stimulation in structures in which functional stereotactic procedures are regularly done, describing first the responses considered to be normal and then certain pathological ones.

The Thalamus Tactile Area The main sensory nucleus of the thalamus is the ventrocaudal nucleus (Vc), receiving input from the medial lemniscus and sending relay output to sensory cortex. In this region a microelectrode will record very unmistakable high background noise and the presence of large-amplitude units with so-called tactile’ neurons that respond to superficial cutaneous stimuli such as air puffs, (hair bending), and light brushing with gauze [14,47,66–70]. These neurons frequently respond faithfully to stimulation repetitions of up to at least 50 Hz, and those which adapt slowly and rapidly appear to be intermingled in a medially to laterally oriented homunculus that represents the contralateral half of the body; there may be a few neurons with ipsilateral representation on the lips, but otherwise the homunculus is contralateral. Intraoral responses are located in medial Vc at about 11–12 mm from the midline adjacent to the medial thalamic nuclei, and foot responses are located in lateral Vc at 18–20 mm from the midline next to the internal capsule. At least the labial and manual neurons are organized into parasagittal laminae often slightly concave medially so that a parasagittal electrode trajectory


78

. Figure 78-1 Reconstruction of an electrode trajectory through the sensory thalamus (Vc) 15 mm lateral to the midline in a patient with essential tremor. The receptive fields (RFs) of the cells encountered are shown to the left of the vertical line. The stimulation currents (Int.) and the location and quality of each sensation evoked by stimulation at sites throughout the trajectory are shown to the right of the vertical line. The shaded bar indicates the presumed tactile region of Vc, based on the presence of cells responding to light tactile stimuli. Note also the close correspondence of the projected fields and the RFs. Of particular note is the site marked with an asterix, at which stimulation at 3 mA evoked pain similar in quality and location to appendicitis pain previously experienced by the patient. Note also that 1 mm farther inferior stimulation evoked burning pain on the leg, and 0.5 mm below that it evoked a sensation of warmth. PF = projected field; P = paresthesia; O = other sensation; B = burning; W = warm; N = pain; Au = auditory; Ki = kinesthetic. (Reprinted from Davis et al, [72] with permission.)

1289

1290

78


. Figure 78-2 Projected fields and effect of stimulus intensity on evoked sensation. Examples taken from mapping in two patients with deafferentation pain. Figures show perceived location and spread of paresthesia evoked during microstimulation (1-s train, 300 Hz through a tungsten microelectrode) (a) or macrostimulation (with an electrode with a 0.5-mm exposed tip: 1-s train, 300 Hz) (b) in the ventrocaudal nucleus at various intensities. The patient’s ratings of the intensity of tingling (on a scale from 0 to 10) at various stimulation intensities is shown in each graph. Note the increased size and intensity of paresthesias evoked by increasing stimulus intensity. In A, the figure on the left shows the location of the receptive fields of the neurons recorded at the site of the stimulation. (Reprinted from Dostrovsky et al, [66] with permission.)


78

. Figure 78-3 Dependence of sensory threshold on thalamic Vc microstimulation parameters. Each stimulus-response curve shows the stimulus current required (threshold) to evoke a sensation when the train frequencies (left panel), pulse width (middle panel), or train length (right panel) is varied. (Reprinted from Dostrovsky et al, [66] with permission.)

will record neurons with similar receptive fields (RFs) throughout, except for the dorsal and ventral extremes of the trajectory. This medial concavity sometimes appears to be replaced by a laterally concave arrangement. RFs may be as small as 1–2 mm in diameter in the lips and manual digits but are much larger in the trunk and proximal limbs. Thebulkoftherepresentationisgivenovertothelips and manual digits.RFs arefound only in theinferior half to two-thirds of the basal part of Vc; we have not recorded identifiable cells more dorsally or in the central and dorsal subdivisions of Vc. The described somatotopy may be seriously deranged in patients who have suffered strokes, spinal cord injury, or other major deafferenting illnesses [71]. When stimulation is carried out in the thalamus in the region where tactile cells reside, paresthesias are produced in parts of the body similar in location to those from which RFs were recorded from neurons in the stimulation area. These are termed projected fields (PFs). With microstimulation, the match between the locations of RFs and PFs may be virtually perfect and the PFs may be induced with currents as low as 1 mA (> Figures 78-1 and > 78-2a). As the tip size of the stimulating electrode increases, the PF usually grows larger because of current spread [8,17,71–73]. What may confuse the stereotactic

surgeon is the occurrence of RF/PF mismatch even in an apparently healthy brain, possibly caused by stimulation of fibers of passage that activate a set of sensory cells distant from the ones recorded in the vicinity of the electrode tip. Mismatch is of course much more striking after strokes and other major deafferenting illnesses [71–73]. Suprathreshold stimulation in the region of tactile cells produces increasingly strong paresthesias but usually not pain (> Figure 78-2a).

The Kinesthetic (‘‘MovementSensing’’) Area Immediately rostral, adjacent to the tactile area, are neurons with high-amplitude spontaneous activity whose discharges change phasically or tonically in response to deep pressure (but not light stimulation) of the contralateral skin, passive joint bending, and deep tissue squeezing. This region doubtlesscontainsthespindleafferentrelay,though that relay cannot be specifically identified [74]. Such responses are probably located in Vim, though this has never been proved in humans. Again, there is a medial-to-lateral somatotopographic arrangement but one that appears to be less exquisite than that in the tactile area; among these

1291

1292

78


various neurons, the deep skin receptors appear to lie most caudally, immediately adjacent to the tactile neurons of Vc [40,42,59,63,68,69,75–84]. Electrical stimulation among these cells usually induces contralateral paresthesias indistinguishable from those produced in the tactile area in more or less the same part of the body as that of the receptors that activate the neurons. However, thresholds are higher (10–20 mA) and RFs are larger than in Vc. Thus, with macrostimulation alone, it is very difficult to define the presumptive border between Vim and Vc [17], that is, the junction between deep and kinesthetic neurons and tactile neurons. Occasionally, sensory motor effects are elicited rather than paresthesias as shown in > Table 78-3 [17] which represents all such responses found in about 10,000 brain sites

. Table 78-3 Macrostimulation-induced contralateral sensorimotor effects other than paresthesia in the human thalamus Effect Wants to, has to move Pulling Drawing Pressure Tightening Grabbing Pinching Squeezing Pushing Choking Teeth loose Jaw pushes in and out Pulsation Jumping Flicking Eyelid twitching Throbbing Jerking Cheek sliding to and fro Tremorous feeling Part continuously moving Weak Tired Heavy Loss of control

No. sites 10 10 5 5 9 3 3 1 1 1 2 1 6 3 2 2 1 1 1 1 5 3 3 3 1

that were macrostimulated. Another type of response occasionally induced with macrostimulation in Vim appears to be vestibular in origin. True vertigo [17], nonspecific dizziness, and faintness have all been elicited, presumably arising from stimulation of the rostral vestibulothalamic path. Microelectrode recording of vestibular cells has apparently not been reported in humans, probably because the head is immobilized in a stereotactic frame.

Voluntary Cells Moving more rostral again, probably into Vop and Voa, one encounters neurons generally considered to lie in the terminal fields of pallidal efferents, respectively. Spontaneous activity is less here, and the amplitude of the recorded units lower than in Vim and Vc. Particularly in Vop, cells are found that alter their firing pattern in response to a specific contralateral voluntary movement [43,68,75,84–91], and some of these so-called voluntary cells have RFs from passive movements that oppose the voluntary action to which they are related [86]. Lenz and associates [86] identified neurons here with increased or decreased firing rates occurring 200 ms before the onset of their related particular contralateral voluntary movements. The somatotopographic arrangement of these cells is again a loosely medial-to-lateral one, as is found in Vim. Raeva and colleagues [87–90] studied voluntary cells extensively not only in Voa and Vop but also in the basal ganglia and the reticular nucleus of the thalamus. They found voluntary cells that respond to one or more contralateral (and sometimes ipsilateral) movements with various patterns of change (decreased firing, short inhibition and then increased or decreased firing, changing in and out of synchronous firing patterns, and more complex patterns). Some cells, especially in the reticular nucleus, respond at the command to prepare to make a movement; others, at various stages throughout the actual movement. The cells were


found to be of two types: type A cells, firing spontaneously but with an irregular pattern at 1–20 Hz and tending to increase their firing during voluntary acts (71%), and type B cells, firing with short 3–5 Hz rhythmic discharges that tend to be suppressed during voluntary acts. Some cells become rhythmic at the onset or end of a voluntary movement, sometimes only when the same movement is repeated several times; such rhythmicity was not, however, de pendent on the presence of tremor. Other cells fired at tremor rates but not in phase with tremor; their activity was suppressed during voluntary movements. Stimulation among voluntary cells may sometimes induce a contralateral muscle contraction about a joint related to the voluntary movement that causes the neurons in the vicinity of stimulation to alter their firing [17]. These contractions are very focal at threshold stimulation, tend to fatigue after the repetition of stimulation, and involve more and more muscle groups as the strength of stimulation is increased; this effect needs to be distinguished from tetanization [92].

Tremor Cells In patients with tremor, kinesthetic cells are entrained to fire faithfully in time with the peripheral tremor, with this rhythmic firing ceasing when the tremor stops [20,21,23,34,37,59,63,75, 81,82,93–103]. These are known as tremor cells, and they constitute a very prominent feature of recording in Vim in patients with parkinsonian rest tremor; for obvious reasons, they are less conspicuous in patients with action tremor. Tremor cells also are found in the globus pallidus internus [104] and units have also been identified in the subthalamic area [39], which may correspond to fiber input of the prelemniscal radiation that inputs into the motor thalamus. Additionally some of these tremor synchronous units may belong to the zona incerta which has similar connectivity to STN. Tremor cells can

78

also be found in thalamus and pallidum of dystonia patients with tremor, and if sensory driven, these cells may fire in time with dystonic movements [105]. ‘‘Voluntary’’ cells also may fire synchronously with peripheral tremor, just as kinesthetic cells do, and may fire in this manner even when tremor is absent. Since some of these cells fire 200 ms in advance of a related contralateral voluntary movement, they have been nominated for the role of tremor pacemakers. Lenz and associates [86] found that the firing patterns of voluntary tremor cells with and without kinesthetic input was particularly tightly linked to the pattern of the peripheral tremor electromyogram (EMG) [98,99]. Attempts to prove whether tremor in, say, Parkinson’s disease is a result of pacing by voluntary cells or of deranged feedback in kinesthetic cells have tended to suggest that both processes may be at work [100,106–108]. Obviously, then, tremor cells have something to do with the tremor seen in Parkinson’s disease and other conditions, yet despite the fact that this has been known for decades, it remains unclear what that relationship is or what the mechanism of thalamotomy or chronic electrical stimulation of the thalamus is in the alleviation of tremor. > Figure 78-4 shows the reduction of tremor produced by electrical simulation and by lidocaine microinjection in the anterior Vim. Most surgeons have claimed that the lesion site for thalamotomy or the site for implanting a chronic stimulating electrode in the control of tremor is located in Vim or caudal Vop among kinesthetic or voluntary tremor cells at sites where acute stimulation most effectively arrests the tremor [20,40,42,81,109]. However, plots of different surgeons’ target sites vary considerably [110]. Moreover, we and others have experienced tremor recurrence and the complication of ataxia when making lesions presumably in Vim guided only by the location of kinesthetic tremor cells just rostral to manual digital tactile cells, where effective stimulation-induced

1293

1294

78


. Figure 78-4 Effect of thalamic microstimulation and microinjection of lidocaine on parkinsonian tremor. a. This figure shows a reconstruction (based on the AC-PC coordinates) of a microelectrode trajectory through the thalamus in the 15-mm lateral plane in a patient with Parkinson’s disease. The location of neurons responding to movement of joints and/ or deep tissues or to touch are shown on the pattern of stipple, and the locations where stimulation reduced tremor or evoked a sensation are shown by the solid/open bars. The site of microstimulation and microinjection is indicated by the filled circle in Vim. b. The effect of microstimulation (2-s train, 0. 1 -ms pulses, 300 Hz, 100 mA) on tremor is apparent from EMG recordings from the biceps (B), triceps (T), and wrist extensor (WE) and flexor (WF) muscles. The solid bar indicates the time of the stimulation. c. Effect of microinjection of 0.5 ml of 2% lidocaine on EMG activity. (Reprinted from Dostrovsky et al, [61] with permission.)

tremor arrest occurs [78,111]. It is our opinion that the target size not only must meet these criteria but also must be located about 2–3 mm above the intercommissural line and a similar distance rostral

to the rostral margin of the tactile digital cells, similar in location to the site proposed by AlbeFessard [112]. Such a site coincides well with that proposed by Lenz and associates [86] on the basis


of the fact that tremor cells here are most tightly linked to the peripheral tremor EMG [113] and fits with current functional imaging studies implicating the cerebellum in tremorogenesis [114,115]. Such a site encompasses both kinesthetic and voluntary cells as well as unidentified tremor cells that do not alter their firing in response to known sensory input or movement and encompasses such a small volume of thalamus that it would not seem possible to differentially manipulate only one type of tremor cell. In Parkinson’s disease, any postoperative tremor recurrence tends to take place in the first 3 months; tremor recurrence is rare after that time. Whereas equally effective tremor suppression is also seen in essential tremor, there is a tendency for recurrence with time; in the case of cerebellar tremor, suppression is never complete and recurrence with time is the rule [116] (see chapters 104 and 105).

Pain Pathways The main pain and temperature pathway for sensations below the head ascends in the lateral spinothalamic tract (STT). The analogous pathway for oral and facial pain and temperature is the trigeminothalamic tract (TrT) that originates in the caudal part of the trigeminal spinal tract nucleus (subnucleus caudalis) [117]. Recent anatomic techniques have indicated that the ascending fibers in these tracts terminate in a number of distinct thalamic nuclei [118–121]. In the lateral thalamus, these include primarily the Vc nucleus and a region ventroposterior to it. The latter region includes a number of relatively ill-defined nuclei, including the ventrocaudal parvocellular nucleus (Vcpc) [122], posterior ventromedial nucleus (VMpo) [120] ventroposterior inferior nucleus (VPI), and suprageniculate/limitans. It is unclear at the present time whether pain is processed in each, some, or only one of these regions. The STT also projects to the medial thalamus, in particular to the centralis lateralis and ventrocaudal (MD)

78

[118–120]. The medial thalamus also receives ascending projections from the reticular formation and other brain stem structures. It has been assumed for many years that at least some of these reticulothalamic neurons relay nociceptive information from the spinal cord and are thus part of a spinoreticulothalamic pathway (also termed in older literature the paleospinothalamic tract), although there is not much direct evidence for the existence of such connections from modern tract tracing techniques [123]. It has proved difficult to trace the pain and temperature pathways physiologically in human CNS. Only a few studies have searched for the presence of nociceptive neurons, and reports of responses to noxious stimuli have been primarily anecdotal and have not been replicated. Hitchcock and Lewin [124] recorded increased cellular activity in cuneate and gracile fasciculus as well as caudal trigeminal nucleus and STT in three patients when they stimulated the ipsilateral face and limbs and applied contralateral noxious stimulation during percutaneous cordotomy. Amano and coauthors [125] recorded neurons with widespread bilateral RFs responding to nociceptive stimuli in the mesencephalon medial to the STT during mesencephalic tractotomy. The responses were of low voltage, and the latencies in response to pinprick fell into three groups: less than 250 ms, 400–800 ms, and more than 1,000 ms. It is not clear exactly from which region these responses may have arisen. This is a region of brain stem from which, in our experience, no sensory responses are elicited in most patients by electrical stimulation (see the discussion below) [17]; stimulating a little more laterally in the STT elicits a contralateral feeling of warmth or coldness similar to the responses elicited by stimulation of the STT in the upper cervical cord and elicited during percutaneous cordotomy [17]. Within the medial thalamic nuclei where reticulothalamic and STT neurons project, Sano and colleagues [64,126,127] identified two types of neurons responding to nociceptive stimuli, one with a 30–90 ms latency and the other with a 100–500 ms latency; RFs were

1295

1296

78


extensive and bilateral. Although there have been many reports of the existence of nociceptive neurons in the medial thalamus in monkeys, cats, and rats, our group has been unsuccessful in finding such neurons in these regions in humans. Electrical stimulation of the medial thalamic structures, which are thought to represent the nonspecific pain relay, in our experience seldom elicits any conscious response except the responses sometimes produced at high intensities and attributed to current spread to ascending STT axons. However, Sano and colleagues [64,126,127] reported eliciting pain by stimulating the internal thalamic lamina. In the lateral thalamus, where there have been many reports of nociceptive neurons in animal studies, there are currently reports only from one group of the existence of some thalamic neurons activated by noxious stimuli. These neurons were recorded within and ventroposterior to Vc [128,129]. Our group was recently able to record a few neurons activated by cooling in the region ventroposterior to VC [130] Also in the lateral thalamus, there have been several reports of stimulation-evoked pain and temperature sensations [17,72,122,130–136], and lesions there have been reported to relieve, chronic pain [17,131,137]. These sensations of contralateral pain usually are evoked when the electrode is ventroposterior to Vc, in the Vcpc region [72,122,132–135]. These sensations are in contrast to the paresthetic responses seen throughout the rest of Vc. They are often quite striking; as an electrode passes ventroposteriorly through Vc, a succession of paresthetic responses are invariably evoked, each of which is referred to a small region of the contralateral body. Near the base of Vc, the threshold response can change from paresthetic to warmth or pain and the somatotopy can change abruptly (> Figure 78-1). For example, after a succession of PFs in the hand, PFs can shift to the contralateral leg. The evoked sensations frequently include a burning component and occasionally are referred to internal sites [17,72,133,135,136]

(> Figure 78-1). Interestingly, these sensations can frequently evoke distressful and aversive sensations, qualities usually attributed to medial thalamic processing of pain. At some sites, warmth rather than pain is evoked. Sensations of cold can be elicited at some sites in this region but are much rarer, possibly, as suggested by our preliminary findings, because they are evoked from a more medial region that usually is not explored [130]. In a small number of cases, in particular in central pain patients, stimulation within the tactile relay nucleus Vc itself can elicit unpleasant and painful or sometimes warm sensations [72]. However, in the vast majority of cases, stimulation within Vc, even at intensities more than four times threshold to evoke paresthesia, is not reported as painful. In summary, in most patients, stimulation of Vc elicits paresthesia at threshold and suprathreshold levels. In contrast, stimulation ventroposterior to Vc, presumably in Vcpc/VMpo, and also in the mesencephalic STT frequently elicits feelings of pain, warmth, or cold. Stimulation medial to the STT in the midbrain and the medial thalamic nuclei normally produces no perceptual responses. In certain patients with chronic pain, however, remarkable aberrant responses are seen. We have found that threshold macrostimulation of mesencephalon medial to the lateral STT and of the above-mentioned medial thalamic nuclei medial to the intraoral portion of Vc may elicit contralateral sensations of burning in certain patients with neuropathic pain but usually not in those with cancer pain, in whom no effect is elicited [14,17,138]. In some patients with neuropathic pain, especially that caused by stroke, such macrostimulation may cause frank pain, often similar in quality to that from which the patient suffers [139,140]. We have had only limited opportunities to repeat these studies with microstimulation and have found that microstimulation in medial thalamus is not effective in inducing pain, possibly because of the limited (approximately 100 mA)


amount of current that can be passed through them. In the lateral thalamus, we recently showed that there is an increased incidence of stimulationevoked pain responses to microstimulation within Vc in poststroke central pain patients [72].

Responses in Vc in Patients with Deafferenting Lesions Recording of neuronal activity in Vc tests the integrity of the pathway from the receptor to the thalamus; microstimulation here examines the pathway from thalamus to, presumably, the cortex [14,139–142]. If the patient’s disease interrupts the sensory afferent pathway, the related RFs will be absent but may be replaced by new RFs coming from undamaged input that can sometimes be unusual. If the deafferentation is selective (as in lateral medullary syndrome), the RFs dependent on the medial lemniscus may persist while those dependent on STT will be absent. If the afferent pathway from the thalamus to the cortex has been destroyed, there may be no PFs though RFs persist, or PFs may be elicited in a portion of the body different from that expected for stimulation at that site. It is possible that in brains damaged by, say, stroke, the effects of stimulation depend on its retrograde rather than anterograde propagation, but this has not been proved. It is a feature of brain central pain that even in patients in whom neither PFs nor RFs can be identified throughout the thalamus, ongoing steady pain on the side of the body contralateral to the stroke (often associated with allodynia and hyperpathia) may still be present and sensory perception, though diminished, is not absent. Such observations suggest ipsilateral transmission of sensation as well as ipsilateral generation of central pain, including allodynia and hyperpathia [143], a situation similar to that seen after hemispherectomy [144]. We saw one patient with a thalamus badly damaged from stroke in whom sensory motor cortical stimulation on the side

78

ipsilateral to the stroke produced both ipsilateral motor and sensory effects as well as partial control of the patient’s central pain. We explored another patient with a neuropathic pain syndrome involving the right upper extremity caused by an intramedullary tumor in whom no RFs related to that limb could be identified in contralateral Vc but in whom the PFs related to it appeared normal; thus, loss of input to the thalamus does not necessarily result in transsynaptic degeneration of the thalamocortical circuits. In patients with small infarcts affecting a portion of Vc, microelectrode exploration frequently reveals a ‘‘hole’’ devoid of RFs and PFs corresponding to that lesion [139], with adjacent surviving Vc neurons assuming the input of some of the neurons lost in the hole with a resulting disruption of the usual somatotopographic organization. Ohye and colleagues made similar observations [73]. These changes obviously result in varying degrees of mismatch between RFs and PFs and complicate the conduct of the functional stereotactic procedure. A second abnormality seen in patients with destructive lesions of the nervous system is the presence of bursting cells. Thalamic cells typically fire in a bursting pattern during sleep but not during wakefulness [145–147]. Bursting cells have long been known in animal models and in humans at various sites in the nervous system upstream from deafferenting lesions and frequently have been linked to the pathophysiology of neuropathic pain [148,149]. Our recent observations, however, suggest that bursting cells observed in the lateral thalamus in pain states are markers of deafferentation and do not necessarily imply a role in mediating neuropathic pain, since we have observed their occurrence after a previous thalamotomy for Parkinson’s disease, in patients with stroke-induced dystonia who do not suffer from pain, and in patients with multiple sclerosis being operated on for tremor who do not have pain [139,150,151]. They are common in the thalamus of patients with spinal cord lesions whether or not those patients have pain [139]. Many of these

1297

1298

78


bursting cells fire in a characteristic pattern showing interspike intervals that lengthen as the burst proceeds with the first interspike interval and varying inversely with the number of spikes in the burst [71,151,152]. This pattern strongly suggests that the underlying mechanism for the bursting is the activation of a low-threshold calcium current that results in a calcium spike [147]. The final pathophysiological feature to be considered is related to the effects elicited by stimulation. As has been mentioned previously, in certain patients with chronic pain caused by stroke, macro- or microstimulation of Vc may elicit contralateral somatotopically organized painful responses rather than paresthesia. Such responses seem to occur in stroke patients in whom allodynia and/or hyperpathia are prominent [14,138,139, 153,154]. Thus, thalamic stimulation in these patients seems to induce a central allodynia, just as stimulation of non-nociceptors in the periphery produces pain through disorganization of signal processing in the dorsal hom. It is intriguing to speculate about the pathophysiology of pain production by stimulating Vc. We studied four stroke patients in whom stimulation of the periventricular gray (PVG) suppressed allodynia and hyperpathia [155,156]. The practical implication of these observations is that the usual type of paresthesiaproducing deep brain stimulation (DBS) usually administered in Vc is impractical for pain control in these patients, because it actually produces pain, and may be dependent on the activation of pain pathways capable of suppression by PVG stimulation.

The Auditory and Vestibular Pathways Auditory and vestibular responses have been noted rarely in the course of stereotactic procedures, the former with both micro- and macrostimulation and the latter only with macrostimulation [6,26,31]. In the lateral lemniscus, the medial

geniculate, and their proximal projections, neurons can be recorded that respond to auditory stimuli addressed to the contralateral ear [17,154,157]. However, micro- or macrostimulation at such recording sites usually induces a beelike buzzing heard chiefly in the contralateral ear, with the pitch of the sound induced apparently unrelated to the frequency of the stimulation. Vestibular effects consisting of true vertigo, a feeling of side-to-side translation, nonspecific faintness, or dizziness also may be elicited by macrostimulation at the same sites where auditory findings occur as well as in Vim, as has already been mentioned [17,24,26,158].

The Visual Pathways The optic tract is most often encountered during pallidotomy, since it lies ventral to the internal globus pallidus target used to treat Parkinson’s disease [12]. Axonal responses may be recorded showing action potentials in response to light flashes delivered to the contralateral visual field, and macro- or microstimulationinornearthetractproduceswhite or colored phosphenes, usually in the form of spots of light or stars in the contralateral visual field [12,17]. The lateral geniculate lies lateral to most stereotactic target sites, so that it is seldom penetrated; scanty evidence suggests that stimulation here produces colored phosphenes in the contralateral visual field [17,21,159,160]. Occasionally in the course of performing mesencephalic tractotomies, visual responses have been elicited with macrostimulation in the tectum and more deeply in the midbrain 14–17 mm from the midline (deep to sites of oculomotor activation), presumably arising from activation of the tectopulvinar pathway, which projects to the second visual cortex. This structure does not support vision in humans, as it does in some other species of mammals [17]. The responses recorded here are of three types: white phosphenes, a blanking out of vision, and the sense


of movement of the visual fields in the absence of any obvious movement of the eye itself, all referred to the contralateral fields.

The Anterior and Dorsomedian Nuclei of the Thalamus The anterior and dorsomedian thalamic nuclei are of historical interest in the field of psychosurgery [6]; the usual bilateral dorsomedian nucleus lesions may induce recent memory loss. There is little information concerning microelectrode recording in these structures, and in our experience their stimulation does not elicit detectable responses. We have noted that as a microelectrode is passed parasagittally from rostrodorsally toward the medial parafascicular nucleus, bursting activity is encountered as the electrode traverses the dorsomedian nucleus, which ceases at the presumed dorsal border of the parafascicular nucleus in a plane 2 mm lateral to the wall of the third ventricle. This bursting activity is similar to that observed in the lateral thalamus in that the firing pattern suggests an underlying calcium spike [151,161]. Similar activity has been reported by Rinaldi and associates [162] and Jeanmonod and coworkers [161] who suggested that the bursting activity was associated with the neuropathic pain syndromes from which their patients suffered. However, since similar recordings have not been obtained from this region in patients who do not have pain, this interpretation must be viewed with caution until it is confirmed in animals or nonpain patients.

The Pulvinar Pulvinarotomy was advocated in the past for the relief of spasticity and intractable pain, but it is uncertain whether the procedure is still used [17]. Martin-Rodriguez and coauthors [163] identified neurons in the pulvinar that fired rhythmically but not in time with any tremor the patient might

78

exhibit, while other neurons responded to visual and auditory stimuli or altered their firing in relationship to contralateral voluntary acts.

Periventricular and Periaqueductal Gray Matter Reynolds [164] demonstrated relief of nociceptive pain in rats when he stimulated neurons in the wall of the ventricular system between the rostral third ventricle and the aqueduct, apparently through the effects of a descending pathway relaying in the nucleus raphe magnus that inhibited nociceptive spinothalamic tract neurons. This led Richardson and Akil [165] to introduce chronic stimulation of PVG, a comparable structure in humans, for the relief of chronic pain. Controversy has arisen about whether such stimulation relieves both nociceptive and neuropathic pain; whether PVG stimulation is to be preferred over PAG stimulation, since the latter is unpleasant according to some researchers [162,165–168]; and the precise site is not known for stimulation. Gybels and Kupers [169] concluded that medial PF was the preferred target, 5 mm rostral to PC on the AC-PC line and 2 mm lateral to the wall of the third ventricle. Levy and coworkers [170] provided an important review of the clinical outcome of such stimulation, while Young and associates [168] correlated pain relief from PVG stimulation with release of the opioid peptide met-enkephalin into the ventricular CSF. We have found that stimulation of the periaqueductal gray matter (PAG) is usually unpleasant, and so we prefer target sites in PVG [171]. In our hands, PVG stimulation is ineffective for the control of the steady element of neuropathic pain [171], though it relieved the allodynia and hyperpathia seen in several patients with stroke-induced central pain [155] (see chapters 121 and 132). The physiological localization of PVG and PAG stimulation sites is, however, not straightforward; they frequently can be located only on anatomical basis. The cessation

1299

1300

78


of bursting at the dorsomedian-parafascicular nucleus border has been mentioned [162] and is of some help in target localization. In certain patients, however, macro- or microstimulation of PVG induces acute pain relief and/or a sense of satiety, warmth, inebriation, or well-being, while that of PAG may induce a feeling of horror, dizziness, or diplopia, helping the surgeon select the stimulation site [165,172]. > Table 78-4 summarizes the clinical neurophysiology of the thalamus.

The Subthalamic Area This area (not the subthalamic nucleus of Luys), lying beneath the thalamus, which has clinical physiological characteristics similar to those of Vim and Vop (including the presence of tremor cells), is an alternative target to Vim-Vop for the relief of involuntary movements [39]. Perhaps . Table 78-4 Effects on mood, consciousness, and autonomic function elicited in human midbrain and thalamus Response Effects on mood Laughing Crying, whimpering, grimacing Effects on consciousness Dreaminess Loss of memory, ‘‘foggy head’’ Sleepiness Autonomic Comfortable abdominal feeling, warm or cool Unpleasant abdominal feeling, faint, cold Something going to stop heart Something in chest speeding up, anxiety Gas coming out of stomach As if to take a deep breath Total

Number of Observations 9 1 8 3 1 1 1 23 11 4 3 1 1 1 35

the fact that it contains densely packed fiber tracts of the prelemniscal radiations projecting into Vim and Vop explains its efficacy, but it also explains the enhanced risk that a lesion here will produce the ‘‘cerebellar’’ complications of dysarthria, ataxia, and gait disturbance [111].

Motor Pathways Among the most striking effects of subcortical stimulation are the motor ones, which consist of isolated motor twitches in time with the stimulation pulses at lower frequencies that fuse into tetanization when higher rates are used. These twitches may arise from (1) lower motor neurons, (2) the corticospinal tract, (3) other upper motor neuron pathways, and (4) autonomic pathways [17]. In our experience, lower motor neuron effects have been confined to the oculomotor nerve and have been elicited during midbrain exploration. Motor effects are produced by stimulation of the corticospinal tract in the internal capsule during pallidotomy [12,21,25,50,173–177] and thalamotomy. During thalamic procedures, vocalization also may be elicited by stimulation in the capsule or cerebral peduncle [17,68]. Motor effects are elicited by stimulation of the corticospinal tract of the lower brain stem and spinal cord. Stereotactic physiological explorations have revealed a functional organization of the motor fibers of the internal capsule that is different from that taught in classic texts. Rather than a discrete somatotopy extending caudally from the representation of the face at the genu, Guiot and colleagues [173] found a less discrete separation of the representation of various body parts all located far posteriorly in the posterior limb. Other observers reported similar findings [178,179]. Hawrylyshyn and associates [180] found that the distance from the midline of the motor fibers in the internal capsule was heavily influenced by the width of the third ventricle, so that in a patient with an anatomically normal brain, say, a young


individual suffering from primary dystonia, a thalamotomy lesion 14 mm from the midline, typical for the relief of tremor in a parkinsonian patient, may damage the internal capsule. Stimulation in the motor thalamus usually does not evoke movements, although stimulation in parkinsonian patients with tremor is frequently very effective in blocking their tremor (> Figure 78-4), especially at high frequencies (> Figure 78-5). Macrostimulation and occasionally microstimulation can produce asterixis-like inhibition [181] (> Figure 78-6). More complex responses are elicited in other upper motor neuron pathways, including the oculomotor tracts [17] and are found in the hypothalamus and apparently in the medial forebrain bundle, tectospinal tract, and medial longitudinal fasciculus.

78

Stimulation-induced head tilting, always to the ipsilateral side, arises from sites 2 to 7 mm from the midline in the midbrain posterior to the red nucleus, possibly arising from activation of the medial longitudinal fasciculus or central tegmental tract dentatorubrothalmic tract or interstitial nucleus of Cajal [17,23,182,183]. These are sometimes associated with ipsilateral or bilateral facial tetanization from stimuli delivered 9 to 11 mm from the midline. Oculomotor effects have been elicited frequently in humans at subcortical sites such as the hypothalamus [126,184], subthalamus [29, 58,185–187], thalamus [165,188] and midbrain [176,189,190]. In our experience, they occurred widely from the level of the hypothalamus through the midbrain between 0 and 9 mm from the midline, grouped in four different

. Figure 78-5 Effect of varying thalamic stimulus parameters on parkinsonian tremor. Surface EMG recordings obtained from the wrist extensors (WE) and flexor (WF) muscles during thalamic stimulation are shown for 0.5-s stimulus trains of varying pulse frequency (top panels) and 300-Hz trains stimuli varying in duration (bottom panels). (Reprinted from Dostrovsky et al, [66] with permission.)

1301

1302

78


. Figure 78-6 Effect of thalamic stimulation on voluntary contraction of various muscles. Each trace is the average rectified surface EMG obtained from 100 sweeps. Peaks and troughs greater or less than ±2 SD (short horizontal lines on left) are indicated by black shading. The stimuli, 1-ms pulses of 1 mA, were delivered at time zero from a chronic stimulating electrode implanted in the motor thalamus. Note the pronounced inhibition produced in the contralateral muscles. (Reprinted from Ashby et al, [181] with permission.)

areas. One group lay between 2.5 and 6.5 mm from the midline in the superior colliculus, consisting chiefly of conjugate eye movement without pupillary change. A second group was located in the posterior hypothalamus 4–7 mm from the midline, which coalesced with a third group in the subthalamus. Among hypothalamic responses, ipsilateral eye movement with uni- or bilateral mydriasis was the most common (33%); ipsilateral eye movement without pupillary change occurred at 14% of sites, and bilateral eye movement with bilateral mydriasis occurred at 23%. The most common eye movement in the hypothalamic area was adduction of both eyes, seen at 81% of sites, whereas isolated ipsilateral eye movement most often consisted of depression (68% of sites); adduction occurred at 27% of sites. The fourth group consisted of various oculomotor effects in the subthalamus and midbrain tegmentum. Autonomic and psychic [23,26,186,188,191– 196] responses are associated with functions such as regulation of pulse rate, blood pressure,

and respiration and have been identified chiefly during stereotactic procedures on the hypothalamus for the control of chronic pain and behavioral disorders [17]. Since hypothalamic stimulation in an awake patient is said to be unpleasant, such operations usually have been performed under general anesthesia, and so only objective observations could be made [126,184,197] (> Table 78-4). The most commonly observed effect in our experience, occurring at 81% of sites, was respiratory suppression and apnea elicited 2–7 mm from the midline in the posterior hypothalamus and the immediately adjacent rubral area. Tachypnea occurred rarely (6% of sites) in the rostral hypothalamus, hypertension at 20%, hypotension at 14%, tachycardia at 13%, and bradycardia at 9%. Hypotension was elicited by stimulating throughout the hypothalamus 2–4 mm from the midline, and hypertension by stimulating the inferior part and adjacent rubral area in the same sagittal planes. Tachycardia was produced in the anterior rubral area, and bradycardia throughout the hypothalamicrubral area.


Speech Although a number of authors have reported effects on speech during subcortical macrostimulation [17,26,198], it is difficult to synthesize those observations. In our experience in stimulating nonbehaving patients, the only effect seen was speech arrest seemingly from tetanization originating in the internal capsule or motor cortex. Ojemann and Van Buren [199,200] by contrast, demonstrated that when they stimulated a patient carrying out a speaking paradigm, they elicited interference with object naming in the anterolateral dominant Vo causing repetition of the same wrong name. Stimulating more medially produced perseveration and anomia. Stimulating Vo during tasks of input to memory improved recall, but during the stage of retrieval from memory it impaired it. The nondominant thalamus appeared to be involved in processing letters and numerals [199–207].

Globus Pallidus Umbach and Ehrhardt [43,91] recorded neurons in the globus pallidus that altered their pattern of firing during a voluntary movement. Their spontaneously rhythmic synchronous activity was suppressed at the command to move as well as at the initiation of an actual movement. Raeva and associates [88–90] recorded dense low-voltage, high-frequency (150 Hz) activity in the globus pallidus, sometimes with silent intervals, from cells that did not change their pattern of firing in response to a voluntary movement. Thirty-five percent of globus pallidus (GP) cells fired rhythmically near but not precisely at the frequency of any tremor present, while 50% lacked high frequency discharges. The latter cells changed their firing pattern over a 300–400 ms period at the onset and end of a voluntary act. Some cells increased and others decreased their firing rates, while others responded in a complex manner.

78

In GP, neurons might alter their firing in response to both contralateral and ipsilateral movements (see chapter 95 and 107). In the anteromedial putamen, spontaneously active low voltage units firing at or above 120 Hz were found that displayed silent periods as well as episodes of group discharges. Other cells fired infrequently. Thirty percent fired rhythmically near tremor frequency. Discharges might be inhibited or excited when a contra- or ipsilateral voluntary movement began or ended, or the responses might be complex. The findings in the caudate nucleus were similar; any rhythmic activity present there was not synchronous with tremor.

Neurophysiological Findings of our Group The types of neurons encountered in anterodorsal trajectories penetrating the external globus pallidus (GPe) and internal globus pallidus (GPi) in six patients with Parkinson’s disease in the ‘‘off’’ state have been described [208,209], and the findings from our group will be summarized here. In general, these neuronal firing patterns are similar to those described in monkeys that have been rendered parkinsonian with MPTP (1-methyl-4phenyl- 1,2,3,6-tetrahydropyridine) [210,211]. Although a quantitative analysis of the various types has not been carried out in the human GP, our observations are as follows. The GP consists of an external (GPe) and an internal segment which is further subdivided anatomically into an external (GPi,e) and an internal division (GPi,i) as depicted in sagittal sections (18.5 or 20 mm lateral to the midline) of the stereotactic atlas (see > Figure 78-7, top). Most of the neuronal activity encoded in spike trains recorded in GP of these patients is grouped in irregularly occurring bursts, as demonstrated by firing rates calculated from the reciprocal of the most common interspike interval. In GPe,

1303

1304

78


. Figure 78-7 Cellular firing rates in the globus pallidus (GP). Data obtained from the neuronal recordings encountered throughout a typical microelectrode trajectory through the human GP. The top panel shows the location of the trajectory based on a sagittal reconstruction from the Schaltenbrand and Wahren atlas in the 20-mm lateral plane. Tick marks on the trajectory (S2) indicate 1-mm spacing. The bottom panel shows the mean firing rates (±SD) of neurons at sites throughout the trajectory as a function of the distance to the optic tract (based on physiological findings). (Reprinted from Hutchison et al, [208] with permission.)

we found that most units fired in one of three characteristic patterns and were termed (1) slowfrequency discharge (SFD) units, whose activity may or may not be punctuated by pauses in firing; (2) low-frequency bursting (LFB) units that fire overall at about 10 Hz but whose activity is punctuated with rapid bursts; (3) occasionally, highfrequency discharge (HFD) units encountered in GPe that fire in an irregular pattern at high frequency, sometimes with pauses (> Figure 78-8). The mean neuronal activity in GPe was 60–86 Hz

(SD, n = 40). In GPi,i, the neuronal activity is significantly higher than in GPe. Many of the units encountered fire at a frequency above 100 Hz. As observed in GPe, most of the neuronal activity occurs in irregular bursts in the spike trains. The mean firing rate for GPi,i was calculated to be 82–92 Hz (SD, n = 89). The elevated mean firing rate of neurons in GPi has been observed in MPTP-treated monkeys [210,211] and has been confirmed in humans with Parkinson’s disease [209].


. Figure 78-8 Firing patterns of cells encountered in the globus pallidus. Each example shows the spontaneous firing pattern of typical cell types. In the external segment of the globus pallidus, cells included the low-frequency discharge burst (LFB) and slow-frequency discharge firing with pauses (SFD-P) types typical of GPe. Regular firing-type cells (border cell – Bor) were found in the laminae and pallidal borders, and irregular, high frequency cells were found in the internal segment of the globus pallidus (GPi). (Reprinted from Hutchison et al, [208] with permission.)

Border neurons have been described in nonhuman primates and human GP [208,210] and are readily distinguished from other neurons in human GP by a regular pattern of firing and a lower overall rate of 44 17 Hz (SD, n = 17). They occur at or near the borders of GP and also in the internal laminae. Some of these neurons have been observed to switch into a repetitive bursting activity for several tens of seconds and then switch back to the regular firing pattern mode, which we have tentatively termed ‘‘border-burst’’ cells.

78

With the passage of the electrode tip out of the lower border of GP, the background noise in recordings diminishes and few units are encountered. Occasionally, axons can be recorded. A constant-current stimulator is used to microstimulate through the tip of the electrode. At a distance 1–2 mm from the lower border of GPi, patients report visual sensations from electrical excitation of fibers of the optic tract (‘‘phosphenes’’). Usually the patient is requested to close the eyes so that phosphenes can be more readily perceived. Frequently, white or yellow starbursts or flashes are reported with currents as low as 2 mA. Microstimulation parameters do not exceed 100 mA, 0.2 ms, 1 s, 300 Hz. When the patient does not report phosphenes from microstimulation at sites that correspond anatomically to the optic tract, recordings of flash-evoked axonal potentials from the optic tract are made. In isolated instances, recordings of the flash-evoked potentials have enabled confirmation of the location of the optic tract when microstimulation has failed. Lesions are made by radiofrequency electrocoagulation at cell-dense sites in posteroventral GPi that are at least 3 mm from the optic tract and internal capsule. These are the main guidelines that determine final lesion placement. The presence of units with high-frequency firing, units with movement related activity, and units with oscillations in the firing rate tentatively identified to be synchronous with tremor (termed ‘‘tremor cells’’ here) provides additional confirmation of the optimal target.

Pallidal Tremor Cells Although neurons with tremor-related activity have been well studied in the human motor thalamus, there have been only isolated reports in the literature of pallidal tremor cells. Umbach and Ehrhard [43,91], recorded pallidal units that fired 20 ms after the tremor burst in the contralateral limb. Jasper and Bertrand [68] reported tremor cells at the end of long trajectories curving

1305

1306

78


. Figure 78-9 Example showing tremor cell in GPi. The top trace shows the rectified and filtered EMG from contralateral wrist extensors, and the bottom trace shows the firing of a neuron in GPi. Note the close correspondence between firing pattern and tremor (EMG).

anteriorly from the thalamus that were presumed to be in the medial segment of GP. Raeva [88] also reported pallidal tremor cells but did not find significant correlation with limb tremor. We recently identified and characterized 28 tremor cells in human GP in three patients with tremor [104] II8 (> Figure 78-9). Some pallidal tremor cells have been evaluated with crossspectral analysis and have been shown to have rhythmic neuronal activity that is highly coherent with limb tremor [104]. Interestingly, in one patient in whom limb tremor was different for the upper limb and the lower limb, the tremor cell was coherent with the forearm EMG (4.6 Hz) while the foot EMG was at a higher frequency (5.6 Hz) (> Figure 78-10). Other tremor cells could be found that had rhythmic neuronal activity coherent with the foot tremor at 5.6 Hz. Discordant tremor frequencies in the upper and lower limbs have been known since the early work of Jung [212]. Although further study of these cells is required, there are at least three significant implications of these observations. First, they tend to support the existence of multiple parallel loops at the output level of the human basal ganglia, as demonstrated by primate neuroanatomical tracing techniques [213] and

. Figure 78-10 Example showing the results of spectral analysis (Fourier transform) of the firing of a neuron in GPi and of the EMG of wrist extensors and flexors and of the foot dorsiflexor (tibialis anterior) recorded during the same time interval. Note that the firing of this GPi tremor cell was at the same rate (4.6 Hz) as that of the wrist tremor but not that of the foot tremor.


neurophysiological studies [214,215]. This is in contrast to a pacemaker hypothesis that conceptualizes the pallidum as a functional syncytium where all units have the same frequency and serve as a central generator of an efferent oscillatory drive. Instead, individual tremor cells in the human pallidum appear to show discrete, limbspecific frequencies. Second, they do not support the hypothesis of a peripheral origin of parkinsonian rest tremor via ‘‘long-latency’’ spinal reflex arcs [108], since this hypothesis would predict that tremor in the lower extremity should be at a lower frequency than that in the upper extremity (longer loop time). Third, the frequency of tremor cells in the pallidum is in the same range as the limb tremor: 4–6 Hz, not 12–15 Hz, as predicted by the Llinas and Pare hypothesis [216,217]. This information, along with observations of 4–6 Hz neuronal activity recorded in the sensorimotor cortex of monkeys with tremor [218], tends to support a corticopallidothalamic loop mechanism for parkinsonian tremor generation. Pallidal tremor cells are not uniformly distributed throughout GP but appear to be located mainly or exclusively in GPi. In three patients with idiopathic Parkinson’s disease, 28 tremor cells were localized to the ventral portion of GPi [219]. This suggests that the distribution of tremor cells may be coextensive with the sensorimotor portions of GPi and that the so-called direct pathway from striatum to GPi that is part of a loop including the motor thalamus and motor cortex is probably one neural substrate of parkinsonian tremor. The paucity of tremor cells in GPe suggests that the indirect pathway is not primarily involved in the propagation of neural oscillations of tremor but may play a role in the generation of tremor by means of disinhibition of the subthalamic nucleus and excitation of GPi, as hypothesized in models of basal ganglia dysfunction in Parkinson’s disease [220,221], it is also possible that the elevated firing rate observed in GPi discussed above is related to the tremorogenic mechanism, since we have

78

observed elevated firing rates (95 29 Hz; SD, n = 22) in tremor cells compared to the average firing rate of GPi neurons (67 30 Hz, n = 162) in six patients with Parkinson’s disease [104]. In conclusion, the studies discussed here provide further evidence to support the corticopallidothalamic pathway as a substrate for the central generation of parkinsonian rest tremor.

Cerebral Cortex Observations concerning speech have been mentioned above. Various authors have reported microelectrode recordings of cortical neurons with reference to epileptic activity [38,222–224] and other brain functions [225] but these subjects will not be pursued here.

Cingulum Our group [226] recorded neurons in cingulate cortex that displayed widespread bilateral RFs and responded to noxious stimuli and cooling, complementing a number of published observations implicating this structure in nociception [227]. It was noted that 3 months after cingulotomy, the patient’s assessment of nonnoxious warm stimuli was lessened while that of noxious heat and cold was accentuated [228].

Meningeal Stimulation Referred pain can be induced easily by manipulation of the dura during a wake craniotomies but also during stereotactic subcortical explorations. During mesencephalic tractotomy, when an electrode located near the dorsal surface of the midbrain is stimulated electrically or advanced to distort the meninges, pain is commonly elicited, virtually always referred to the ipsilateral face (40% of sites) or forehead (22%)

1307

1308

78


[17]. Such responses are useful in that they help localize the electrode with respect to the surface of the brain stem, and awareness of such responses avoids confusion.

Miscellaneous Responses Alterations of consciousness and gustatory and olfactory effects are rarely elicited in subcortical exploration [17,23,26,188,191] in contrast to the plethora of experiential effects seen in association cortex. > Table 78-5 lists the rare responses seen by us with macrostimulation at 10,000 sites.

Other Structures Schaltenbrand and Wahren [27] have extensive experience with macrostimulation in the human brain stem that will not be reviewed here.

Applications of Physiological Localization Deep Brain Stimulation This section briefly reviews the use of physiological localization in commonly performed . Table 78-5 Autonomic effects observed under general anesthesia in the hypothalamic region Effect Apnea or depressed rate or amplitude of respiration Increased respiratory rate Elevation of blood pressure Depression of blood pressure Elevation of pulse rate Depression of pulse rate No autonomic effects

functional stereotactic procedures, many directed toward the thalamus. > Table 78-6 summarizes the clinicopathological relationships. The simplest procedure is the implantation of a chronic stimulating electrode to produce paresthesias in the distribution of a patient’s chronic pain syndrome, a procedure that appears to us to be most useful in cases of neuropathic pain [229,230]. DBS has a long history [191], with a rekindling of interest after the proposition of the Melzack-Wall gate theory of pain [231] as a cephalad extension of dorsal column stimulation [232,233]. The usual stereotactic approach is used. A micro- or macroelectrode is directed toward the sites in tactile Vc that are related to the painful part of the body. The use of microelectrodes allows accurate confirmation of location by recording responses to touch. The use of macroelectrodes to identify low-threshold sites for evoking paresthetic responses is less accurate because of the large size of the electrode tip and the resulting greater spread of stimulus current away from the tip. Once the appropriate tactile cells have been suitably mapped, a chronic electrode is introduced into their midst whose stimulation produces paresthesias in the patient’s area of pain. If large portions of the contralateral body are painful, the implant may have to be made in the medial lemniscus or internal capsule, which can be located by extrapolation from the position of tactile cells in Vc. Placement of DBS electrodes to chronically stimulate PVG was mentioned above.

No. Patients 131* 9 33 23 21 14 86

*More than one type of response occurred at some sites

Thalamotomy for Pain Destructive lesions for the relief of pain are no longer made in the lateral thalamus [28,234] except possibly in Vcpc, the identification of which has been reviewed sufficiently [17,64, 126,131,137,235].


78

. Table 78-6 The clinical neurophysiology of the thalamus Method of study Microelectrode Nucleus

Stimulation

Recording

Application

Ventrocaudal (Vc)

Paresthesia

Tactile neurons

Parvocellular ventrocaudal (Vcpcfvmpo) Anterior ventral oral (Voa) Posterior ventral oral (vop) Ventral intermediate (Vim) Anterior (A)

Warm, cool, painful effects and paresthesia

Nociceptors Thermoreceptors

Lesions (historical) and chronic stimulation for pain relief Lesions for pain relief

None recognized

Voluntary cells

Lesions for relief of dyskinesia

Motor ‘‘on’’ effects (see text) Paresthesia, vestibular, sensorimotor effects None recognized

Voluntary cells

Parafascicular (Pf) Periventricular gray (PVG), medial Pf Internal lamina (IL) Pulvinar (Pu)

None Various, acute pain relief, feeling of satiety None or pain or burning None recognized

Nociceptors None recognized

Lesions and chronic stimulation for relief of dyskinesia Lesions and chronic stimulation for control of tremor and dyskinesia Lesions for relief of psychiatric disease (historical) Lesions for pain relief Chronic stimulation for pain relief

Dorsomedian (DM)

None recognized

Bursting cells

Centrum medianum (CM)

None, pain

None recognized

Kinesthetic, and deep sensory neurons None recognized

Nociceptors Various

Medial thalamotomy probably is still widely employed. The plethora of potential medial thalamic targets has been mentioned, along with the difficulty of obtaining useful localizing responses in them [3,28,137,154,156,234,236]. Because of the difficulty in physiologically recognizing medial thalamic structures, we extrapolate the location of our target, parafascicular nucleus from the location of Vc. Frank and coauthors [237] pointed out that though the risks of medial thalamotomy are lower than those of mesencephalic tractotomy, so is the success rate in relieving nociceptive pain usually caused by cancer. Medial thalamotomy has not proved very useful in the treatment of steady neuropathic pain [138,154,156] and has been abandoned for the treatment of movement disorders [238].

Lesions for pain relief Lesions for relief of pain and spasticity (historical) Lesions for relief of psychiatric disorders (historical) Lesions for relief of pain and movement disorders (historical)

Mesencephalic Tractotomy Mesencephalic tractotomy, first by open and then by stereotactic means [31], has long been employed to treat chronic pain, functioning as a rostral extension of spinothalamic tractotomy in the spinal cord [6,17,31,125,165,176,190,228]. We have used macro- or microstimulation to localize lesions for mesencephalic tractotomy. First, the expected position of the medial lemniscus is determined from a stereotactic CT or MRI with the usual stereotactic technique. Stimulation is then carried out, searching for the low-threshold contralateral paresthetic effects typical of medial lemniscus in a series of parasagittal trajectories 2–3 mm apart and approximately 12 mm from the midline. Once the

1309

1310

78


medial lemniscus has been mapped out, further trajectories are made 2–3 mm medial and a few millimeters dorsal to it until a series of points are found where contralateral warm or cold effects, like those seen during cordotomy, are elicited, usually 8–10 mm from the midline. The lesion is planned to destroy the loci of these responses, sparing the medial lemniscus but extending medially almost to the aqueduct to destroy the normally physiologically ‘‘silent’’ region of the spinoreticular pathway as well.

Thalamotomy and Thalamic DBS for Tremor It would appear that exactly the same target is used for the implantation of a chronic stimulating electrode for the relief of tremor as was used in the past for thalamotomy [13,16,40,42, 109,113,239]. Our technique in either procedure is to first locate the tactile representation of the manual digits (about 15 mm from the midline), defining their anterior and inferior boundaries. We then search 2 mm more anterior for tremor cells and for neurons that alter their firing in response to passive or voluntary movement of the tremorous part of the body. Among these tremor cells we locate the sites where stimulation most completely abolishes tremor at the lowest threshold without inducing disturbing paresthesias. The final target must meet these criteria but also must lie 2–3 mm above the AC-PC line (see chapter 92).

the posteroventral portion of the GPi. Leksell considered pallidotomy more beneficial than thalamotomy for bradykinesia and rigidity whereas thalamotomy proved superior for the treatment of tremor. Recently, the effectiveness of posteroventral pallidotomy has been confirmed by clinical assessment of movement disorders carried out by our group [241]. We have summarized (see above) our results with microelectrode recording of single units in GP and microstimulation of adjacent neural structures to be avoided during pallidotomy (see chapter 92). The coordinates of GPi are determined from the stereotactic MRI [12] with reference to an atlas in the 20-mm sagittal plane. The microelectrode is lowered through GPe and GPi and on into the optic tract, recording continuously and stimulating every millimeter near the ventral portion of the pallidum. Cells typical of GPe and GPi are sought as well as any tremor cells or cells that alter their firing in response to contralateral and/or ipsilateral movement (> Figures 78-7 and > 78-8) (see above). The anterior, posterior, and inferior margins of GPi are readily defined by the cessation of background activity, the optic tract by stimulation-induced phosphenes and recording of responses to light and the internal capsule by the induction of tetanizing contralateral facial and upper limb responses. A lesion is then placed to destroy the maximum volume of GPi without encroaching on the motor capsule or optic tract.

Part 2 – Imaging in Structural Imaging Pallidotomy The current interest in stereotactic pallidotomy for the treatment of Parkinson’s disease (PD) arose out of the report of Laitinen and associates [240] of the beneficial effect on bradykinesia and rigidity obtained with lesions directed to

The way of the future may revolve around imaging. Just as it is now possible with imaging to recognize a pathological structure for stereotactic biopsy, the hope is that functional targets eventually will be visualized similarly. Thus, intraoperative functional imaging may play a


major role in functional neurosurgery in the future. Indeed, there is currently controversy about whether functional stereotactic surgery can already be done accurately and safely enough using existing MRI techniques with noninvasive radiosurgery [242,243] (see chapter 93). However, current opinion is that the existing imaging techniques are not sufficiently accurate and require corroboration with invasive physiological studies. In fact, there is still controversy about the relative accuracy of CT and MRI compared with ventriculography for functional stereotactic surgery [244], with some surgeons claiming that ventriculography is currently the most accurate imaging technique despite its disadvantages (> Table 78-2). We and others, however, have found CT and MRI to be at least as accurate as ventriculography for functional stereotactic surgery [245,246], and our unpublished experience clearly shows that imaging is more accurate using the Leksell G frame and a 1.5T General Electric Signa scanner than it is with CT [246]. The superiority of MRI probably is at least partly due to the fact that both AC and PC can be readily visualized with MRI, while AC usually cannot be seen well on CT. However, accurate MRI guidance depends on attention to a number of factors [33,247,248] (see chapter 18 and 39). The target structure should be near the center of the field, the chemical in the fiducial tubes should preferably not be petroleum jelly [249], and the frame used must be truly MRI compatible. To address some of these problems when we use MRI to calculate the coordinates of AC and PC (the usual landmarks on which most functional and stereotactic procedures are based), we first do a midsagittal scan to visualize them and then take 1-mm nonoverlapping axial images using the ‘‘spoiled-grass’’ sequence (excitation time, 13 ms; relaxation time, 43 ms). The single axial image best displaying AC and PC is chosen, and their three-dimensional coordinates are determined by using the MRI console’s

78

computer software. These coordinates are selected by bracketing the locations of AC and PC shown in the midsagittal scan. With the proper precautions outlined above, it seems to us that MRI is the structural imaging technique of choice for functional stereotactic surgery. A new type of structural imaging that may help with stereotactic localization is known as diffusion tensor imaging (DTI). This type of imaging can visualize white matter projections associated with (arising from or to) a specific gray matter structure.

Functional Imaging The last decade has seen the emergence of exciting new imaging techniques. Currently, these technologies are being used primarily for basic research. However, some of the limitations to clinical application will be overcome in the near future. Therefore, functional imaging has great potential as a new tool for neurosurgical procedures that rely on the localization of particular brain sites. Modern imaging technologies include functional magnetic resonance imaging (fMRI), positron emission tomography (PET), magnetoencephalography (MEG), and the combination of MEG and MRI techniques known as magnetic source imaging (MSI). A brief overview of the limitations and applicability of each technique for neurosurgical candidates follows.

Positron Emission Tomography PET is an imaging technique that relies on the ability to detect the emission of positively charged electrons. The source of this emission is typically an injectable molecule containing a positron-emitting isotope such as O15 incorporated into water molecules. The emissions (gamma radiation) from the injected material are captured by a ring of radiation detectors contained

1311

1312

78


in a doughnut-shaped PET device. Simultaneous acquisition of many parallel brain slices can be obtained using these emissions. The involvement of cortical and subcortical areas in somatosensory [33,35–38,250,251], pain [227,250,252,253], motor [254], and complex cognitive functions such as memory [255,256] has been investigated with PET technology. PET is particularly useful in patients who cannot be placed in an MR scanner (e.g., patients who are claustrophobic, have difficulty holding their head very still or who have metal in their body). PET can also be used to identify locations of receptor (dys)function [257,258]. However, PETspatial resolution (several millimeters) limits its use in identifying changes that affect small regions of the brain. Also, to achieve a high level of statistical significance, there is a need to average several images. This usually is achieved by pooling data from several subjects, since a limited amount of radioactivity can safely be injected in a single subject. These technical limitations currently limit the use of PET as a presurgical tool, although newer generation PET scanners with increased sensitivity allow for many more scans to be acquired in individual subjects.

Functional Magnetic Resonance Imaging FMRI is an imaging technique that is sensitive to the oxygen saturation of blood and blood flow. The ability of the microvasculature to supply oxygen by means of increased blood flow exceeds the ability of active neuronal tissue to extract oxygen. As a result, oxygen saturation of postcapillary blood from active neuronal areas is higher than that of inactive areas because of a normal flow ‘‘oversupply’’ to those areas. FMRI is capable of measuring and localizing differences in oxy- and deoxyhemoglobin. Since these two forms of oxygen act differently when exposed to a magnetic field (it is actually the deoxyhemoglobin that is

being detected), brain maps can be generated to show areas of active tissue [259]. Since fMRI is a noninvasive technique that can acquire images quickly (seconds per scan) with fine spatial resolution (1–5 mm of tissue) in a single subject [260]. it can be used with some precision to visualize sites activated by brief stimuli in the active human brain. These features are well suited for the examination of somatosensory and other cognitive processes. The ability to develop protocols that can test brain function has clinical applications such as the localization of functional tissue in tumor patients and presurgical mapping of active zones. The standard method of acquiring fMRI data today is with echo planar imaging (EPI), which can simultaneously image multiple brain regions very rapidly (tens of milliseconds per image). Since EPI has nearly real-time functional imaging capabilities, it is especially well suited for studies of functional systems that change with prolonged or repetitive activation and/or involve multiple spatially segregated brain sites. Thus, FMRI has been used successfully to image motor-related basal ganglia activation [261] and cortical activation associated with visual [262,263], motor [263–266], and somatosensory [267–269] stimuli in normal subjects. Other, more complex functions, such as language [270,271] and pain [269] also have been investigated. More recently, the technique of real-time fMRI has been introduced and made available on most clinical scanners. These ‘‘in-house’’ packages can be quite useful to track brain activations during simple tasks (e.g., motor), but should be used with caution for more complex situations due to the simplicity of the statistical analysis being used ‘‘on the fly’’ to generate activation maps in real time [272]. The ability to map functional cortical tissue physiologically has clinical relevance to presurgical localization. Functional imaging has been used to investigate motor, somatosensory, visual, and language function in patients before neurosurgery. Mapping with FMRI has been shown to corroborate the results of intraoperative mapping with


macrostimulation [268] [273] and thus may become quite useful for procedures that do not have to resolve fine subnuclear borders. One study [271] demonstrated activation of Broca’s area during internal speech generation. Hemispheric laterality was also examined in a study of motor cortex activation in left- versus right-handed subjects [266]. These examples illustrate the usefulness of presurgical mapping with fMRI, although undoubtedly there would still be the need for verification of these data during surgery. Hemispheric dominance as revealed by FMRI could replace the Wada test used before certain neurosurgical procedures to determine the laterality of functions. In the future, intraoperative MRI may provide a means to perform noninvasive functional imaging at the time of surgery. Despite great advances in MRI, there are still some limitations to its application with the current technology. First, the spatial resolution of fMRI is not yet fine enough to distinguish very small regions of the brain such as subcortical nuclei. Therefore, FMRI cannot replace intraoperative electrophysiological mapping as is done before procedures such as thalamotomy and pallidotomy. Second, movement artifact resulting from head movements and motion artifact caused by natural internal oscillations (e.g., spinal cord movement) preclude imaging of some central nervous system (CNS) regions. Third, it may not be possible for some patients to be placed in an MR scanner for the 20–40 min necessary for a FMRI series and MRIs cannot be done in patients with implanted metal devices.

Magnetic Source Imaging MEG is a technique similar to EEG that measures the electric currents in the brain, but with greater spatial accuracy. MSI combines MEG with MRI for anatomic localization of the magnetic signals in a functional setting. The exciting features of MSI are that it is easy to use, is noninvasive, and has

78

good spatial resolution (millimeter range) and very fast acquisition time, allowing for real-time imaging [274,275]. MSI has been used to study many cortical regions involved in audition [276] and vibratory/touch sensation [276–279]. The high spatial resolution of MSI has demonstrated distinct representations for different body regions (e.g., digits, hand, arm, face) in the somatosensory cortex [276–279]. Cortical plasticity resulting from injury has also been demonstrated [280]. MSI can also be combined with magnetic resonance angiography (MRA) to provide visual demarcation of important functional regions before neurosurgery. Some studies have demonstrated very good agreement of MSI data with operative results [275]. Therefore, MRA and MSI are two techniques that may provide useful information before neurosurgical procedures involving the cortex. At this time, these techniques are not applicable to subcortical studies. Also, biomagnetometers are not as widely available as are MRI and PET facilities.

Prediction for Future Protocols These new imaging technologies will no doubt affect neurosurgery in the future. As these techniques are refined and become more widely available, they may provide safe alternative methods for the physiological mapping of viable tissue. This could reduce the need for timeconsuming and invasive modes of mapping. Although the current tools for mapping, such as electrophysiological recording and stimulation, probably will still be required in certain instances, the advent of intraoperative imaging married to functional imaging techniques may provide alternatives for certain patients.

Conclusion Although the hope is that in the future imaging, particularly functional imaging, will allow

1313

1314

78


a stereotactic surgeon to recognize whatever target he or she wishes to manipulate and allow a direct approach similar to that currently used in stereotactic biopsy, for the present, invasive techniques of physiological monitoring are necessary for optimal results with minimal complications. If carried out methodically in the manner of brain mapping as originally advocated by Woolsey, the physiological monitoring not only will satisfy the surgeon’s needs but also will provide further knowledge about the brain’s normal and pathological states.

References 1. Penfield W, Boldrey W. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937;60:389. 2. Albe-Fessard D. Electrophysiological methods for the identification of thalamic nuclei. Z Neurol 1973;205:15-28. 3. Ervin F, Mark V. Stereotactic thalamotomy in the human. II. Physiological observations in the human thalamus. Arch Neurol 1960;3:368. 4. Fairman D, Perlmutter I. Physiological observations during stereotactic surgery of the basal ganglia. Confin Neurol 1965;26:298-305. 5. Hughes J. The electroencephalogram in Parkinsonism. J Neurosurg 1966;24:369. 6. Spiegel E, Wycis H. Clinical and physiological applications. I Stereoencephalotomy. New York: Grune & Stratton; 1962. 7. Lenz FA, Dostrovsky JO, Kwan HC, Tasker RR, Yamashiro K, Murphy JT. Methods for microstimulation and recording of single neurons and evoked potentials in the human central nervous system. J Neurosurg 1988;68:630-4. 8. Dostrovsky JO, Davis KD, Lee L, Sher GD, Tasker RR. Electrical stimulation-induced effects in the human thalamus. In: Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and spinal cord. New York: Raven Press; 1993. p. 219-29. 9. Bertrand G, Jasper H, Wong A, Mathews G. Microelectrode recording during stereotactic surgery. Clin Neurosurg 1969;16:328-55. 10. Jasper H. Recording from microelectrodes in stereotactic surgery for Parkinson’s disease. J Neurosurg 1966;2 suppl: 219. 11. Jasper H, Bertrand G. Exploration of the human thalamus with microelectrodes. Physiologist 1963;7:167.

12. Lozano AM, Hutchison WD, Kiss ZHT, Davis KD, Dostrovsky JO. Methods for microelectrode-guided posteroventral pallidotomy. J Neurosurg 1996;84:194-202. 13. Tasker RR, Lenz FA, Yamashiro K, Gorecki J, Hirayama T, Dostrovsky JO. Microelectrode techniques in localization of stereotactic targets. Neurol Res 1987;9:105-12. 14. Tasker RR. The use of microelectrodes in the human brain. In: Bromm B, Desmedt EF, editors. Pain and the brain: from nociception to cognition. New York: Raven Press; 1995. p. 143-74. 15. Tasker R, Dostrovsky JO, Yamashiro K. Effets sensitifs et moteurs de la stimulation thalamique chez l’homme: applications clinique. Rev Neurol (Paris) 1986;142:316. 16. Tasker RR, Yamashiro K, Lenz FA. Thalamotomy for Parkinson’s disease: microelectrode techniques. In: Lunsford LD, editor. Modern stereotactic surgery. Boston: Martinus Nijhoff; 1988. p. 297-314. 17. Tasker R, Organ L, Hawrylyshyn P. The thalamus and midbrain of man. A physiological atlas using electrical stimulation. Springfield, IL: Charles C Thomas; 1982. 18. Fukamachi A, Oye C, Narabayashi H. Delineation of the thalamic nuclei with a microelectrode in stereotaxic surgery for parkinsonism and cerebral palsy. J Neurosurg 1973;39:214-25. 19. Albe-Fessard D, Guiot G, Hardy J. Electrophysiological localization and identification of subcortical structures in man by recording spontaneous and evoked activities. Electroencephalogr Clin Neurophysiol. 1963;15:1052. 20. Narabayashi H. Tremor mechanisms. In: Schaltenbrand G, Walker AE, editors. Stereotaxy of the human brain. Stuttgart: George Thieme; 1982. p. 510-14. 21. Bertrand C, Martinez SN, Hardy J. Stereotactic surgery for parkinsonism: microelectrode recording, stimulation, and oriented sections with a leucotome. In: Krayenbuhl H, Maspes PE, Sweet WH, editors. Progress in neurological surgery. Basel: Karger; 1973. p. 79-112. 22. Guiot G, Derome P, Arfel G. Electrophysiological recordings in stereotaxic thalamotomy for parkinsonism. In: Krayenbuhl H, Maspes PE, Sweet WH, editors. Progress in neurological surgery. Basel: Karger; 1962. p. 189-221. 23. Hassler R, Riechert T, Mundinger F, Umbach W, Ganglberger JA. Physiological observations in stereotaxic operations in extrapyramidal motor disturbances. Brain 1960;83:337-50. 24. Andy OJ. Sensory-motor responses from the diencephalon. Electrical stimulation in man. J Neurosurg 1966;24:612-20. 25. Bertrand C. Functional localization with monopolar stimulation. J Neurosurg 1966;24:403. 26. Sem-Jacobsen CW, Styri OB. Electrical stimulation of the human brain: pitfalls and results. In: Somjen GG, editor. Neurophysiology studied in man. Amsterdam: Excerpta Medica; 1972. p. 27-36.


27. Schaltenbrand G, Wahren W. Atlas for stereotaxy of the human brain. Georg Thieme; Stuttgart: 28. Monnier M, Fischer R. [Localization, stimulation and coagulation of thalamus in man]. J Physiol (Paris) 1951;43:818. 29. Ito Z. Stimulation and destruction of the prelemniscal radiation or its adjacent area in various extrapyramidal disorders. Confin Neurol 1975;37:41-8. 30. Albe-Fessard D, Arfel G, Guiot G, Derome P, Guilbaud G. Thalamic unit activity in man. Electroencephalogr Clin Neurophysiol 1967;25:132. 31. Woolsey CN. Cortical localization as defined by evoked potential and electrical stimulation studies. In: Schaltenbrand G, Woolsey CN, editors. Cerebral localization and organization. Madison, WI: University of Wisconsin Press; 1964. p. 17-26. 32. Wetzel N, Snider RS. Neurophysiological correlates in human stereotaxis. Q Bull Northwest Univ Med Sch 1958;32:386-92. 33. Carreras M, Mancia D, Pagni CA. Unit discharges recorded from the human thalamus with microelectrodes. Confin Neurol 1967;29:87-92. 34. Albe-Fessard D, Arfel G, Guiot G. Derivations d’activites spontanees et evoquees dans les structures cerebrales profondes de l’homme. Rev Neurol (Paris) 1962;106:89. 35. Albe-Fessard D, Arfel G, Guiot G. Activites elecctriques caracteristiques de quelques structures cerebrales chez l’homme. Ann Chir 1963;17:1185. 36. Albe-Fessard D, Arfel G, Guiot G, Derome P, Hertzog E, Vourc’h, G, Brown H, Aleonard P, De la HJ, Trigo JC. Electrophysiological studies of some deep cerebral structures in man. J Neurol Sci 1966;3:37-51. 37. Albe-Fessard D. Thalamic activity and Parkinson’s disease. In: Umbach W, Koepchen HP, editors. Central rhythm and regulation. Stuttgart: Hyppocrates Verlag; 1974. p. 353-62. 38. LI CL, Van Buren JM. Micro-electrode recordings in the brain of man with particular reference to epilepsy and dyskinesia. In: Somjen GG, editor. Neurophysiology studied in man. Amsterdam: Excerpta Medica; 1972. p. 49-63. 39. Lucking CH, Struppler A, Erbel F, et al. Stereotactic recording from human subthalamic structures. In: Somjen GG, editor. Neurophysiology studied in man. Excerpta Medica; Amsterdam: p. 95-9. 40. Ohye C, Fukamachi A, Miyazaki M, Isobe I, Nakajima H, Shibazaki T. Physiologically controlled selective thalamotomy for the treatment of abnormal movement by Leksell’s open system. Acta Neurochir (Wien.) 1977;37:93-104. 41. Ohye C. Depth microelectrode studies. In: Schaltenbrand G, Walker AE, editors. Stereotaxy of the human brain. Stuttgart: Thieme; 1982. p. 510-14. 42. Ohye C. Selective thalamotomy for movement disorders: micro-recording stimulation techniques and results. In:

43.

44.

45.

46.

47.

48.

49.

50. 51.

52.

53.

54.

55.

56.

57. 58.

78

Lunsford LD, editor. Modern stereotactic neurosurgery. Boston: Martinus-Nijhoff; 1988. p. 315-31. Umbach W, Ehrhardt KJ. [Microelectrode leads in the basal ganglia in man]. Arch Psychiatr Nervenkr 1965;207:106-13. Arutjunov A, Kadin A, Konovalov A. Experience of applying the microelectrode method in the neurosurgical clinic practice during single-stage stereotaxic operations for parkinsonism. Vopr Neirokhir 1970;6:12. Bates JAV. Electrical recording from the thalamus in human subjects. In: Iggo A, editor. Handbook of sensory physiology: somatosensory system. Berlin: Springer; 1972. p. 561-78. Bertrand G, Jasper H. Microelectrode recording of unit activity in the human thalamus. Confin Neurol 1965;26:205-8. Bertrand G, Jasper H, Wong A. Microelectrode study of the human thalamus: functional organization in the ventro-basal complex. Confin Neurol 1967;29:81-6. Donaldson IM. The properties of some human thalamic units. Some new observations and a critical review of the localization of thalamic nuclei. Brain 1973;96:419-40. Gaze RM, Gillingham FJ, Kalyanaraman S, Porter RW, Donaldson AA, Donaldson IM. Microelectrode recordings from the human thalamus. Brain 1964;87:691-706. Gillingham F. Depth recording and stimulation. J Neurosurg 1966;2 382. Guiot G, Hardy J, Be-Fessard D. [Precise delimitation of the subcortical structures and identification of thalamic nuclei in man by stereotactic electrophysiology]. Neurochirurgia (Stuttg) 1962;5:1-18. Fukamachi A, Ohye C, Saito Y, Narabayashi H. Estimation of the neural noise within the human thalamus. Acta Neurochir (Wien) 1977; 121–36. Nakajima H, Fukamachi A, Isobe I, Miyazaki M, Shibazaki T, Ohye C. Estimation of neural noise. Functional anatomy of the human thalamus. Appl Neurophysiol 1978;41:193-201. Yoshida M. Electrophysiological characterization of human subcortical structures by frequency spectrum analysis of neural noise (field potential) obtained during stereotactic surgery. Preliminary presentation of frequency power spectrum of various subcortical structures. Stereotact Funct Neurosurg 1989;52:157-63. Tasker RR, Rowe IH, Hawrylyshyn P, Organ L. Computer mapping of brain-stem sensory centers in man. J Neurosurg 1976;44:458-64. Tasker RR, Dostrovsky JO. Computers in functional stereotactic surgery. In: Kelly PJ, Kall B, editors. Computers in stereotactic neurosurgery. Boston: Blackwell; 1992. p. 154-64. Schaltenbrand G, Baily P. Introduction to stereotaxis with an atlas of the human brain. Stuttgart: Thieme; 1959. Sugita K, Takaoka Y, Mutsuga N, Takeda A, Hirota T. Correlation between anatomically calculated target point

1315

1316

78 59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.


and physiologically determined point in stereotaxic surgery. Confin Neurol 1972;34:84-93. Ohye C, Nakamura R, Fukamachi A. Recording and stimulation of the ventralis nuclei of the human thalamus. Confin Neurol 1975;37:258. Organ LW, Tasker RR, Moody NF. The impedance profile of the human brain as a localization technique in stereoencephalotomy. Confin Neurol 1967;29:192-6. Dostrovsky JO, Sher GD, Davis KD, Parrent AG, Hutchison WD, Tasker RR. Microinjection of lidocaine into human thalamus: a useful tool in stereotactic surgery. Stereotact Funct Neurosurg 1993;60:168-74. French LA, Story JL, Galicich JH, Schultz EA. Some aspects of stimulation and recording from the basal ganglia in patients with abnormal movements. Confin Neurol 1962;22:265-74. Narabayashi H, Ohye C. Nucleus ventralis intermedius of human thalamus. Trans Am Neurol Assoc 1974;99:232-3. Sano K, Yoshioka M, Sekino H, Mayanagi Y, Yoshimasu N. Functional organization of the internal medullary lamina in man. Confin Neurol 1970;32: 374-80. Yamashiro K, Tasker RR, Iwayama K, Mori K, be-Fessard D, Dostrovsky JO, Chodakiewitz JW. Evoked potentials from the human thalamus: correlation with microstimulation and single unit recording. Stereotact Funct Neurosurg 1989;52:127-35. Hardy TL, Bertrand G, Thompson CJ. Thalamic recordings during stereotactic surgery. II. Location of quickadapting touch-evoked (novelty) cellular responses. Appl Neurophysiol 1979;42:198-202. Hardy TL, Bertrand G, Thompson CJ. Touch-evoked thalamic cellular activity. The variable position of the anterior border of somesthetic SI thalamus and somatotopography. Appl Neurophysiol 1981;44: 302-13. Jasper HH, Bertrand G. Thalamic units involved in somatic sensation and voluntary and involuntary movements in man. In: Purpura DP, YahrMD, editors. The thalamus. New York: Columbia University Press; 1966. p. 365-90. Lenz FA, Dostrovsky JO, Tasker RR, Yamashiro K, Kwan HC, Murphy JT. Single-unit analysis of the human ventral thalamic nuclear group: somatosensory responses. J Neurophysiol 1988;59:299-316. Ohye C, Fukamachi A, Narabayashi H. Spontaneous and evoked activity of sensory neurons and their organization in the human thalamus. Z Neurol 1972;203:219. Lenz FA, Kwan HC, Martin R, Tasker R, Richardson RT, Dostrovsky JO. Characteristics of somatotopic organization and spontaneous neuronal activity in the region of the thalamic principal sensory nucleus in patients with spinal cord transection. J Neurophysiol 1994;72:1570-87. Davis KD, Kiss ZHT, Tasker RR, DostrovskyJO. Thalamic stimulation-evoked sensations in chronic pain

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

patients and in nonpain (movement disorder) patients. J Neurophysiol 1996;75:1026-37. Ohye C, Shibazaki T, Hirai T, Kawashima Y, Hirato M, Matsumura M. Plastic change of thalamic organization in patients with tremor after stroke. Appl Neurophysiol 1985;48:288-92. Goto A, Kosaka K, Kubota K, Nakamura R, Narabayashi H. Thalamic potentials from muscle afferents in the human. Arch Neurol 1968;19:302-9. Crowell RM, Perret E, Siegfried J, Villoz JP. ‘Movement units’ and ‘tremor phasic units’ in the human thalamus. Brain Res 1968;11:481-8. Hardy TL, Bertrand G, Thompson CJ. Position and organization of thalamic cellular activity during diencephalic recording. II. Joint- and muscle-evoked activity. Appl Neurophysiol 1980;43:28-36. Hardy TL, Bertrand G, Thompson CJ. Position and organization of thalamic cellular activity during diencephalic recording. I. Pressure-evoked activity. Appl Neurophysiol 1980;43:18-27. Jones MW, Tasker RR. The relationship of documented destruction of specific cell types to complications and effectiveness in thalamotomy for tremor in Parkinson’s disease. Stereotact Funct Neurosurg 1990; 54/55:207-11. Maeda T. Lateral coordinates of nucleus ventralis intermedius target for tremor alleviation. Stereotact Funct Neurosurg 1989;52:191-9. McClean MD, Dostrovsky JO, Lee L, Tasker RR. Somatosensory neurons in human thalamus respond to speech-induced orofacial movements. Brain Res 1990;513:343-7. Ohye C. Neurons of the thalamic ventralis intermedius nucleus: their special reference to tremor. Adv Neurol Sci 1985;29:224. Ohye C, Shibazaki T, Hirai T, Wada H, Hirato M, Kawashima Y. Further physiological observations on the ventralis intermedius neurons in the human thalamus. J Neurophysiol 1989;61:488-500. Ohye C, Narabayashi H. Physiological study of presumed ventralis intermedius neurons in the human thalamus. J Neurosurg 1979;50:290-7. Hardy TL, Bertrand G, Thompson CJ. Topography of ‘bilateral-movement-evoked’ thalamic cellular activity found during diencephalic recording. Appl Neurophysiol 1980;43:67-74. Hongell A, Wallin G, Hagbarth KE. Unit activity connected with movement initiation and arousal situations recorded from the ventrolateral nucleus of the human thalamus. Acta Neurol Scand 1973;49:681-98. Lenz FA, Kwan HC, Dostrovsky JO, Tasker RR, Murphy JT, Lenz YE. Single unit analysis of the human ventral thalamic nuclear group. Activity correlated with movement. Brain 1990;113(Pt 6):1795-821.


87. Raeva S. Localization in human thalamus of units triggered during ‘verbal commands,’ voluntary movements and tremor. Electroencephalogr Clin Neurophysiol 1986;63:160-73. 88. Raeva S. Unit activity of some deep nuclear structures of the human brain during voluntary movement. In: Somjen G, editor. Neurophysiology studied in man. Amsterdam: Excerpta Medica; 1972. p. 64-78. 89. Raeva SN. Role of neurons of the neostriatum and nonspecific thalamus in the organization of goal-directed activity. Hum Physiol 1977;3:972. 90. Raeva SN, Livanov MN. Microelectrode studies of neuronal mechanisms of human voluntary amnesic activity. Hum Physiol 1975;1:27. 91. Umbach W, Ehrhardt KJ. Micro-electrode recording in the basal ganglia during stereotaxic operations. Confin Neurol 1965;26:315-17. 92. Strick PL. Activity of ventrolateral thalamic neurons during arm movement. J Neurophysiol 1976;39: 1032-44. 93. Kelly PJ, Ahlskog JE, Goerss SJ, Daube JR, Duffy JR, Kall BA. Computer-assisted stereotactic ventralis lateralis thalamotomy with microelectrode recording control in patients with Parkinson’s disease. Mayo Clin Proc 1987;62:655-64. 94. Albe-Fessard D, Guiot G, Lamarre J, et al. Activation of thalamocortical projections related to tremorogenic processes. In: Purpura DP, Yahr MD, editors. The thalamus. New York: Columbia University Press; 1966. p. 237-53. 95. Bates JAV. The significance of tremor phasic units in the human thalamus. In: Gillingham FJ, Donaldson IML, editors. Third symposium on Parkinson’s disease. Edinburgh: Livingstone; 1996. p. 112-18. 96. Hardy J, Bertrand C, Martinez N. Activites cellulaire thalamiques liees au tremblement parkinsonien. Neurochirurgie 1964;10:449. 97. Hardy TL, Bertrand G, Thompson CJ. Thalamic recordings during stereotactic surgery. I. Surgery topography of evoked and nonevoked rhythmic cellular activity. Appl Neurophysiol 1979;42:185-97. 98. Lenz FA, Tasker RR, Kwan HC, Schnider S, Kwong R, Murphy JT. Cross-correlation analysis of thalamic neurons and EMG activity in parkinsonian tremor. Appl Neurophysiol 1985;48:305-8. 99. Lenz FA, Kwan HC, Martin RL, Tasker RR, Dostrovsky JO, Lenz YE. Single unit analysis of the human ventral thalamic nuclear group. Tremor-related activity in functionally identified cells. Brain 1994;117 (Pt 3):531-43. 100. Lenz FA, Schnider S, Tasker RR, Kwong R, Kwan H, Dostrovsky JO, Murphy JT. The role of feedback in the tremor frequency activity of tremor cells in the ventral nuclear group of human thalamus. Acta Neurochir Suppl (Wien) 1987;39:54-6.

78

101. Lenz FA, Tasker RR, Kwan HC, Schnider S, Kwong R, Murayama Y, Dostrovsky JO, Murphy JT. Single unit analysis of the human ventral thalamic nuclear group: correlation of thalamic ‘‘tremor cells’’ with the 3–6 Hz component of parkinsonian tremor. J Neurosci 1988;8:754-64. 102. Ohye C, Albe-Fessard D. Rhytmic discharges related to tremor in humans and monkeys. In: Chalazonitis N, Boisson M, editors. Abnormal neuronal discharges. New York: Raven Press; 1978. p. 37-48. 103. Ohye C, Saito U, Fukamachi A, Narabayashi H. An analysis of the spontaneous rhythmic and non-rhythmic burst discharges in the human thalamus. J Neurol Sci 1974;22:245-59. 104. Hutchison WD, Lozano AM, Kiss ZHT. Tremor-related activity (TRA) in globus pallidus of Parkinson’s disease (PD). Soc Neurosci 1994;20:783. 105. Lenz FA, Martin R, Kwan HC, Tasker RR, Dostrovsky JO. Thalamic single-unit activity occurring in patients with hemidystonia. Stereotact Funct Neurosurg 1990;54/55:159-62. 106. Schnider SM, Kwong RH, Kwan HC, et al. Detection of feedback in the central nervous system of Parkinson’s patients. IEEE Trans Decis Control 1989; 60:203. 107. Schnider SM, Kwong RH, Lenz FA, Kwan HC. Detection of feedback in the central nervous system using system identification techniques. Biol Cybern 1989;60:203-12. 108. Tatton WG, Lee RG. Evidence for abnormal long-loop reflexes in rigid Parkinsonian patients. Brain Res 1975;100:671-6. 109. Ohye C, Hirai T, Miyazaki M, Shibazaki T, Nakajima H. Vim thalamotomy for the treatment of various kinds of tremor. Appl Neurophysiol 1982;45:275-80. 110. Laitinen LV. Brain targets in surgery for Parkinson’s disease. Results of a survey of neurosurgeons. J Neurosurg 1985;62:349-51. 111. Yasui N, Narabayashi H, Kondo T, Ohye C. Slight cerebellar signs in stereotactic thalamotomy and subthalamotomy for parkinsonism. Appl Neurophysiol 1976;39:315-20. 112. Albe-Fessard D. Physio-pathologie de Parkinson. Presented at Le point sur la maladie de Parkinson symposium, Brussels, Belgium; 1973. 113. Lenz FA, Normand SL, Kwan HC, Andrews D, Rowland LH, Jones MW, Seike M, Lin YC, Tasker RR, Dostrovsky JO. Statistical prediction of the optimal site for thalamotomy in parkinsonian tremor. Mov Disord 1995;10:318-28. 114. Deiber MP, Pollak P, Passingham R, Landais P, Gervason C, Cinotti L, Friston K, Frackowiak R, Mauguie`re F, Benabid AL. Thalamic stimulation and suppression of parkinsonian tremor: evidence of

1317

1318

78 115.

116.

117.

118.

119.

120.

121.

122.

123.

124. 125.

126.

127.

128.


a cerebellar deactivation using positron emission tomography. Brain 1993;116:267-79. Jenkins IH, Bain PG, Colebatch JG, Thompson PD, Findley LJ, Frackowiak RS, Marsden CD, Brooks DJ. A positron emission tomography study of essential tremor: evidence for overactivity of cerebellar connections. Ann Neurol 1993;34:82-90. Shahzadi S, Tasker RR, Lozano A. Thalamotomy for essential and cerebellar tremor. Stereotact Funct Neurosurg 1995;65:11-17. Willis WD Jr. The pain system: the neural basis of nociceptive transmission in the mammalian nervous system. In: Gildenberg PL, editor. Pain and headache. Basel: Karger; 1995. p. 1-346. Apkarian AV, Hodge CJ. Primate spinothalamic pathways. III. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways. J Comp Neurol 1989;288:493-511. Boivie J. An anatomical reinvestigation of the termination of the spinothalamic tract in the monkey. J Comp Neurol 1979;186:343-70. Craig AD, Bushnell MC, Zhang E.-T., Blomqvist A. A thalamic nucleus specific for pain and temperature sensation. Nature 1994;372:770-3. Mehler WR. The anatomy of the so-called ‘‘pain tract’’ in man. An analysis of the course and distribution of the ascending fibers of the fasciculus anterolateralis. In: French JD, Porter RW, editors. Basic research in paraplegia. Springfield, IL: Charles C. Thomas; 1962. p. 26-55. Hassler R. Dichotomy of facial pain conduction in the diencephalon. In: Hassler R, Walker AE, editors. Trigeminal neuralgia. Philadelphia: Sanders; 1970. p. 123-138. Blomqvist A, Berkley KJ. A re-examination of the spinoreticulo-diencephalic pathway in the cat. Brain Res 1992;579:17-31. Hitchcock E, Lewin M. Stereotactic recording from the spinal cord of man. Br Med J 1969;4:44-5. Amano K, Tanikawa T, Iseki H, Kawabatake H, Notani M, Kawamura H, Kitamura K. Single neuron analysis of the human midbrain tegmentum. Rostral mecencephalic reticulotomy for pain relief. Appl Neurophysiol 1978;41:66-78. Sano K. Intralaminar thalamotomy (thalamolaminotomy) and posteromedial hypothalamotomy in the treatment of intractable pain. In: Krayenbuhl H, Maspes PE, Sweet WH, editors. Progress in neurological surgery. Basel: Karger; 1977. p. 50-103. Ishijima B, Yoshimasu N, Fukushima T, Hori T, Sekino H, Sano K. Nociceptive neurons in the human thalamus. Confin Neurol 1975;37:99-106. Lenz FA, Seike M, Lin YC, Baker FH, Rowland LH, Gracely RH, Richardson RT. Neurons in the area of human thalamic nucleus ventralis caudalis respond to painful heat stimuli. Brain Res 1993;623:235-40.

129. Lenz FA, Gracely RH, Rowland LH, Dougherty PM. A population of cells in the human thalamic principal sensory nucleus respond to painful mechanical stimuli. Neurosci Lett 1994;180:46-50. 130. Dostrovsky JO, Davis KD, Kiss ZHT, et al. Evidence for a specific temperature relay site in human thalamus. In IASP World Congress on Pain 1996. 131. Halliday AM, Logue V. Painful sensations evoked by electrical stimulation in the thalamus. In: Somjen GG, editor. Neurophysiology studied in man. Amsterdam: ExcerptsMedica; 1972. p. 221-30. 132. Hassler R. The division of pain conduction into systems of pain sensation and pain awareness. In: Janzen R, Keidl WD, Herz A, Steichele C, editors. Pain; Basic Principles-pharmacology-therapy. London: Churchill Livingstone; 1972. p. 98-112. 133. Davis KD, Tasker RR, Kiss ZH, Hutchison WD, Dostrovsky JO. Visceral pain evoked by thalamic microstimulation in humans. Neuroreport 1995; 6:369-74. 134. Dostrovsky JO, Wells FEB, Tasker RR. Pain sensations evoked by stimulation in human thalamus. In: Inoka R, Shigenaga Y, Tohyama M, editors. Processing and inhibition of nociceptive information, international congress series 989. Amsterdam: Excerpta Medica; 1992. p. 115-20. 135. Lenz FA, Seike M, Richardson RT, Lin YC, Baker FH, Khoja I, Jaeger CJ, Gracely RH. Thermal and pain sensations evoked by microstimulation in the area of human ventrocaudal nucleus. J Neurophysiol 1993;70:200-12. 136. Lenz FA, Gracely RH, Hope EJ, Baker FH, Rowland LH, Dougherty PM, Richardson RT. The sensation of angina can be evoked by stimulation of the human thalamus. Pain 1994;59:119-25. 137. Hitchcock ER, Teixeira MJ. A comparison of results from center-median and basal thalamotomies for pain. Surg Neurol 1981;15:341-51. 138. Tasker RR. Deafferentation pain syndromes. Introduction. In: Nashold BS Jr, Ovelmen-Levitt J, editors. Deafferentation pain syndromes: pathophysiology and treatment. Advances in pain research and therapy. New York: Raven Press; 1991. p. 241-57. 139. Gorecki J, Hirayama T, Dostrovsky JO, et al. Thalamic stimulation and recording in patients with deafferentation and central pain. Stereotact Funct Neurosurg 1989;52:220. 140. Lenz FA, Tasker RR, Dostrovsky JO. Abnormal single unit activity and responses to stimulation in the presumed ventrocaudal nucleus of patients with central pain. In: Dubner R, Gebhart GF, Bond MR, editors. Proceedings of the 5th world congress on pain, vol 18. Amsterdam: Elsevier; 1988. p. 157-64. 141. Lenz FA, Tasker RR, Dostrovsky JO, Kwan HC, Gorecki J, Hirayama T, Murphy JT. Abnormal single-unit activity recorded in the somatosensory thalamus of


142.

143.

144. 145.

146.

147. 148.

149.

150.

151.

152.

153.

154.

155.

156.

a quadriplegic patient with central pain. Pain 1987;31:225-36. Luo L, Davis KD, Kiss ZHT, et al. Phantom sensations elicited by stimulation in thalamus of amputees. In Proceedings of the 8th world congress on pain, Vancouver, BC, Canada,1996. p. 306. Parent AG, Lozano AM, Dostrovsky JO, et al. Central pain in the absence of functional sensory thalamus. Stereotact Funct Neurosurg 1992;59:9. Ralston BL. Hemispherectomy and hemithalamectomy in man. J Neurosurg 1962;19:909-12. McCarley RW, Benoit O, Barrionuevo G., Lateral geniculate nucleus unitary discharge in sleep and waking: state and rate specific aspects. J Neurophysiol 1983;50:798-818. Dostrovsky JO, Tsoukatos J, Kiss ZHT, et al. Neuronal firing patterns in human thalamus during sleep and wakefulness. Physiol Can 1995;26:83. Steriade M, Jones EG, Llinas RR. Thalamic oscillations and signalling. New York: Wiley; 1990. Loeser JD, Ward AAJ, White LE Jr.Chronic deafferentation of human spinal cord neurons. J Neurosurg 1968;29:48-50. Yamashiro K, Iwayama K, Kurihara M, Mori K, Niwa M, Tasker RR, be-Fessard D. Neurones with epileptiform discharge in the central nervous system and chronic pain. Experimental and clinical investigations. Acta Neurochir Suppl (Wien) 1991;52:130-2. Tsoukatos J, Davis KD, Loomis S, et al. Is neuronal bursting activity in lateral thalamus of chronic pain patients related to their pain? In Proceedings of the 8th world congress on pain, Vancouver, BC; 1996. p. 305. Tsoukatos J, Kiss ZHT, Tasker RR, et al. Distinct patterns of bursting activity in the human medial and lateral thalamus. Soc Neurosci Abstr 1995;21:109. Lenz FA, Kwan HC, Dostrovsky JO, Tasker RR. Characteristics of the bursting pattern of action potentials that occurs in the thalamus of patients with central pain. Brain Res 1989;496:357-60. Davis KD, Kiss ZH, Tasker RR, Dostrovsky JO. Thalamic stimulation-evoked sensations in chronic pain patients and in nonpain (movement disorder) patients. J Neurophysiol 1996;75:1026-37. Tasker RR, Dostrovsky JO. Deafferentation and central pain. In: Melzack R, Wall PD, editors. Textbook of pain. Edinburgh: Churchill Livingstone; 1989. p. 154-80. Parent AG, Lozano AM, Tasker RR, et al. Periventricular gray stimulation suppresses allodynia and hyperpathia in man. Stereotact Funct Neurosurg 1992;59:82. Tasker RR. Pain resulting from central nervous system pathology (central pain). In: Bonica J, editor. Management of pain. Philadelphia: Lea & Febiger; 1989. p. 264-83.

78

157. Tasker RR, Organ LW. Stimulation-mapping of the upper human auditory pathway. J Neurosurg 1973;38:320-25. 158. Hawrylyshyn PA, Rubin AM, Tasker RR, Organ LW, Fredrickson JM. Vestibulothalamic projections in man– a sixth primary sensory pathway. J Neurophysiol 1978;41:394-401. 159. Marg E, Dierssen G. Reported visual percepts from stimulation of the human brain with microelectrodes during therapeutic surgery. Confin Neurol 1965;26:57-75. 160. Boethius J, Lindblom U, Meyerson BA, et al. Effects of multifocal brain stimulation on pain and somatosensory functions. In: Zotterman Y, editor. Sensory function of the skin in primates. Oxford/New York: Pergamon; 1976. p. 531-46. 161. Jeanmonod D, Magnin M, Morel A. Thalamus and neurogenic pain: physiological, anatomical and clinical data. Neuroreport 1993;4:475-8. 162. Rinaldi PC, Young RF, Albe-Fessard D, Chodakiewitz J. Spontaneous neuronal hyperactivity in the medial and intralaminar thalamic nuclei of patients with deafferentation pain. J Neurosurg 1991;74:415-21. 163. Martin-Rodriguez JG, Buno W Jr, Garcia-Austt E. Human pulvinar units, spontaneous activity and sensory-motor influences. Electroencephalogr Clin Neurophysiol 1982;54:388-98. 164. Reynolds DV. Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 1969;164:444-5. 165. Richardson DE, Akil H. Pain reduction by electrical brain stimulation in man. Part 1: Acute administration in periaqueductal and periventricular sites. J Neurosurg 1977;47:178-83. 166. Barbaro NM. Studies of PAG/PVG stimulation for pain relief in humans. In: Fields HL, Besson J-M, editors. Progress in brain research. Pain modulation, vol 77. Amsterdam: Elsevier; 1988. p. 165-73. 167. Young RF, Kroening R, Fulton W, Feldman RA, Chambi I. Electrical stimulation of the brain in treatment of chronic pain. Experience over 5 years. J Neurosurg 1985;62:389-96. 168. Young RF, Bach FW, Van Norman AS, Yaksh TL. Release of beta-endorphin and 0methionine-enkephalin into cerebrospinal fluid during deep brain stimulation for chronic pain. Effects of stimulation locus and site of sampling. J Neurosurg 1993;79:816-25. 169. Gybels J, Kupers R. Deep brain stimulation in the treatment of chronic pain in man: where and why? Neurophysiol Clin 1990;20:389-98. 170. Levy RM, Lamb S, Adams JE. Treatment of chronic pain by deep brain stimulation: long term followup and review of the literature. Neurosurgery 1987;21:885-93. 171. Tasker RR, Vilela O. Deep brain stimulation for the control of intractable pain. In: Youmans JR, editor.

1319

1320

78 172.

173.

174.

175.

176. 177.

178.

179.

180.

181.

182.

183.

184.

185.

186.


Neurological surgery. Philadelphia: Saunders; 1995. p. 3512-27. Tasker RR, Yoshida M, Sima A, et al. Stimulation mapping of the periventricular-periaqueteductal gray matter. In: Samii M, editor. Surgery in and around the brain stem and third ventricle. Heidelberg: Springer; 1986. p. 161-7. Guiot G, Sachs M, Hertzog E, Brion S, Rougerie J, Dalloz JC, Napoleone F. [Electrostimulation & surgical lesions of the internal capsule; anatomical & physiology data.]. Neurochirurgie 1959;5:17-42. Guiot G, Hertzog E, Rondot P, Molina P. Arrest or acceleration of speech evoked by thalamic stimulation in the course of stereotaxic procedures for Parkinsonism. Brain 1961;84:363-79. Martinez SN, Bertrand C, Botana-Lopez C. Motor fiber distribution within the cerebral peduncle. Results of unipolar stimulation. Confin Neurol 1967;29:117-22. Nashold BS. Ocular reactions from brain stimulation in conscious man. Neuroophtalmology 1970;5:92. Velasco FC, Molina-Negro P, Bertrand C, Hardy J. Further definition of the subthalamic target for arrest of tremor. J Neurosurg 1972;36:184-91. Hirayama K, Tsubaki T, Toyokura Y, Okinaka S. The representation of the pyramidal tract in the internal capsule and basis pedunculi. A study based on three cases of amyotrophic lateral sclerosis. Neurology 1962;12:337-42. Bertrand G, Blundell J, Musella R. Electrical exploration of the internal capsule and neighbouring structures during stereotaxic procedures. J Neurosurg 1965;22:333-43. Hawrylyshyn PA, Tasker RR, Organ LW. Third ventricular width and the thalamocapsular border. Appl Neurophysiol 1976;39:34-42. Ashby P, Lang AE, Lozano AM, Dostrovsky JO. Motor effects of stimulating the human cerebellar thalamus. J Physiol 1995;489(Pt 1):287-98. Sano K, Sekino H, Tsukamoto Y, Yoshimasu N, Ishijima B. Stimulation and destruction of the region of the interstitial nucleus in cases of torticollis and see-saw nystagmus. Confin. Neurol 1972;34:331-8. Sano K, Yoshioka M, Ogashiwa M, Ishijima B, Ohye C, Sekino H, Mayanagi Y. Central mechanisms of neck movements in the human brain stem. Confin Neurol 1967;29:107-11. Sano K, Mayanagi Y, Sekino H, Ogashiwa M, Ishijima B. Results of stimulation and destruction of the posterior hypothalamus in man. J Neurosurg 1970; 33:689-707. Kalyanaraman S. Some observations during stimulation of the human hypothalamus. Confin Neurol 1975; 37:189-92. Spiegel EA, Wycis HT, Szekely EG, Dams DJ, Flanagan M, Baird HW, III. Campotomy in

187.

188.

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

various extrapyramidal disorders. J Neurosurg 1963; 20:871-84. Spiegel EA, Wycis HT, Szekely EG, Soloff L, Dams J, Gildenberg P, Zanes C. Stimulation of forel’s field during stereotaxic operations in the human brain. Electroencephalogr Clin Neurophysiol 1964;16:537-48. Schaltenbrand G, Spuler H, Wahren W, Wilhelmi A. Vegetative and emotional reactions during electrical stimulation of deep structures of the brain during stereotactic procedures. Z Neurol 1973;205:91-113. Wilson WP, Nashold BS. Evoked photic responses from the human thalamus and midbrain. Confin Neurol 1973;35:338-45. Nashold BS Jr, Gills JP Jr. Ocular signs from brain stimulation and lesions. Arch Ophthalmol 1967;77: 609-18. Heath RG, Mickle WA. Evaluation of seven years’s experience with depth electrode studies in human patients. In: Ramey A, O’Doherty C, editors. Electrical studies on unanaesthetized brain. New York: Hoeber; 1961. p. 214-47. Bondartchuk AN, Nikitina LI, Shmatkov YV. The problem of dominant central mechanisms controlling vegetative functions in man. In IV International Congress of Neurosurgery and IX International Congress Neurology, New York, Sept. 20–27,1969. Obrador S, Dierssen G. Mental changes induced by subcortical stimulation and therapeutic lesions. Confin Neurol 1967;29:168. Siegfried J, Wiesendanger M. Respiratory alterations produced by thalamic stimulation during stereotaxic operations. Confin Neurol 1967;29:220-3. Blum B, Israeli J, Askenasy HM. Changes in coronary vessels induced by repeated hypothalamic stimulation. Microcirculation 1976;1:295. Blum B, Ben-Ari W, Yashin T, et al. Regional changes in pial blood supply induced by hypothalamic stimulation. Microcirculation 1976;1:395. Fairman D. Hypothalamotomy as a new perspective for alleviation of intractable pain and regression of metastatic malignant tumors. In: Fusek K, editor. Present limits of neurosurgery. Prague: Avicenum; 1972. p. 525-8. Schaltenbrand G. The effects on speech and language of stereotactical stimulation in thalamus and corpus callosum. Brain Lang 1975;2:70-7. Ojemann G, Fedio P, Van Buren JM. Evidence for dominance within the human lateral thalamus. In IV Interational Congress of Neurosurgery and IX Interational Congress of Neurology, New York, Sept. 20–27, 1969. Ojemann GA. Language and the thalamus: object naming and recall during and after thalamic stimulation. Brain Lang 1975;2:101-20.


201. Creutzfeldt O, Ojemann G, Lettich E. Neuronal activity in the human lateral temporal lobe. I. Responses to speech. Exp Brain Res 1989;77:451-75. 202. Creutzfeldt O, Ojemann G, Lettich E. Neuronal activity in the human lateral temporal lobe. II. Responses to the subjects own voice. Exp Brain Res 1989;77:476-89. 203. Creutzfeldt O, Ojemann G. Neuronal activity in the human lateral temporal lobe. III. Activity changes during music. Exp Brain Res 1989;77:490-98. 204. Ojemann JG, Ojemann GA, Lettich E. Neuronal activity related to faces and matching in human right nondominant temporal cortex. Brain 1992;115 (Pt 1): 1-13. 205. Ojemann GA. Organization of language cortex derived from investigation during neurosurgery. Semin Neurosci 2:297. 206. Fedio P, Van Buren JM. Memory and perceptual deficits during electrical stimulation in the left and right thalamus and parietal subcortex. Brain Lang 1975;2:78-100. 207. Haglund MM, Ojemann GA, Lettich E, Bellugi U, Corina D. Dissociation of cortical and single unit activity in spoken and signed languages. Brain Lang 1993;44:19-27. 208. Hutchison WD, Lozano AM, Davis KD, Saint-Cyr JA, Lang AE, Dostrovsky JO. Differential neuronal activity in segments of globus pallidus in Parkinson’s disease patients. Neuroreport 1994;5:1533-7. 209. Sterio D, Beric A, Dogali M, Fazzini E, Alfaro G, Devinsky O. Neurophysiological properties of pallidal neurons in Parkinson’s disease. Ann Neurol 1994;35:586-91. 210. Filion M, Tremblay L, Bedard PJ. Effects of dopamine agonists on the spontaneous activity of globus pallidus neurons in monkeys with MPTP-induced parkinsonism. Brain Res 1991;547:152-61. 211. Filion M, Tremblay L. Abnormal spontaneous activity of globus pallidus neurons in monkeys with MPTP-induced parkinsonism. Brain Res 1991;547:142-51. 212. Jung R. Untersuchungen uber den Parkinson tremor und andere Zitterformen beim Menschen. Z Ges Neurol Psychiatr 1941;173:263. 213. Hoover JE, Strick PL. Multiple output channels in the basal ganglia. Science 1993;259:819-21. 214. Alexander GE, Crutcher MD, DeLong MR. Basal ganglia-thalamocortical circuits: parallel substrates for motor, oculomotor, ‘‘prefrontal’’ and ‘‘limbic’’ functions. Prog Brain Res 1990;85:119-46. 215. Alexander GE, Crutcher MD. Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends Neurosci 1990;13:266-71. 216. Pare D, Curro’Dossi R, Steriade M. Neuronal basis of the parkinsonian resting tremor: a hypothesis and its implications for treatment. Neuroscience 1990;35: 217-26.

78

217. Llinas R, Pare D. Role of intrinsic neuronal oscillations and network ensembles in the genesis of normal and pathological tremors. In Findley LJ, Koller WC, editors. Handbook of tremor disorders. New York: Marcel Dekker; 1995. p. 7-36. 218. Microelectrode studies of unit discharges in the sensory motor cortex of monkeys. In: Gybels JM, editor. The neural mechanism of Parkinson tremor. Brussels: Editions Arsia; 1963. p. 83–122. 219. Hutchison WD, Lozano AM, Tasker RR, Lang AE, Dostrovsky JO. Identification and characterization of neurons with tremor-frequency activity in human globus pallidus. Exp Brain Res 1997;113:557-63. 220. Albin RL, Young AB, Penney JB. The functional anatomy of basal ganglia disorders. Trends Neurosci 1989;12:366-75. 221. DeLong MR. Primate models of movement disorders of basal ganglia origin. Trends Neurosci 1990;13:281-5. 222. Calvin WH. Comments on human epileptic neurons. In: Somjen GG, editor. Neurophysiology studied in man. Excerpta Medica; Amsterdam: p. 110-11. 223. Rayport M. Single neuron studies in human epilepsy. In: Somjen GG, editor. Neurophysiology studied in man. Amsterdam: Excerpta Medica; 1972. p. 100-09. 224. Li CL, Friauf W, Cohen G, Tew JM Jr. A method of recording single cell discharges in the cerebral cortex of man. Electroencephalogr Clin Neurophysiol 1965;18:187-90. 225. Li CL, Tew JM Jr. Reciprocal activation and inhibition of cortical neurones and voluntary movements in man: cortical cell activity and muscle movement. Nature 1964;203:264-5. 226. Hutchison WD, Dostrovsky JO, Davis KD, et al. Single unit responses and microstimulation effects in cingulate cortex of an awake patient. IASP world congress on pain, 1993. 227. Talbot JD, Marrett S, Evans AC, Meyer E, Bushnell MC, Duncan GH. Multiple representations of pain in human cerebral cortex. Science 1991;251:1355-8. 228. Davis KD, Hutchison WD, Lozano AM, Dostrovsky JO. Altered pain and temperature perception following cingulotomy and capsulotomy in a patient with schizoaffective disorder. Pain 1994;59:189-99. 229. Hosobuchi Y, Adams JE, Fields HL. Chronic thalamic and internal capsule stimulation for the control of facial anesthesia dolorosa and the dysesthesia of thalamic syndrome. Adv Neurol 1974;4:783. 230. Mazars G, Merienne L, Cioloca C. [Treatment of certain types of pain with implantable thalamic stimulators]. Neurochirurgie 1974;20:117-24. 231. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971-9. 232. Shealy CN, Mortimer JT, Hagfors NR. Dorsal column electroanalgesia. J Neurosurg 1970;32;560-4.

1321

1322

78


233. Wall PD, Sweet WH. Temporary abolition of pain in man. Science 1967;155:108-9. 234. Mark VH, Ervin FR. Role of thalamotomy in treatment of chronic severe pain. Postgrad Med 1965; 37:563-71. 235. Sano K, Yoshioka M, Ogashiwa M, Ishijima B, Ohye C. Thalamolaminotomy. A new operation for relief of intractable pain. Confin Neurol 1966;27:63-6. 236. Hecaen H, Talairach J, David M, et al. Coagulation limitees du thalamus dans les algies du syndrome thalamique. Rev Neurol (Paris) 1949;81:917. 237. Frank F, Fabrizi AP, Gaist G, Weigel K, mundinger F. Stereotactic mesencephalotomy versus multiple thalamotomies in the treatment of chronic cancer pain syndromes. Appl Neurophysiol 50:314-8. 238. Adams JE, Rutkin BB. Lesions of the centrum medianum in the treatment of movement disorders. Confin Neurol 1965;26:231-45. 239. Benabid AL, Pollak P, Gao DM, Hoffmann D, Limousin P, Gay E, Payen I, Benazzouz A. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as a treatment of movement disorders. J Neurosurg 1996;84:203-14. 240. Laitinen LV, Bergenheim AT, Hariz MI. Ventroposterolateral pallidotomy can abolish all parkinsonian symptoms. Stereotact Funct Neurosurg 1992;58: 14-21. 241. Lozano AM, Lang AE, Galvez-Jimenez N, Miyasaki J, Duff J, Hutchinson WD, Dostrovsky JO. Effect of GPi pallidotomy on motor function in Parkinson’s disease. Lancet 1995;346:1383-7. 242. Desalles A, Hariz M, Functional Radiosurgery. In: Desalles A, Goelsch S, editors. Stereotactic surgery and radiosurgery. Madison, WI: Medical Physics Publishing; 1993. p. 389-406. 243. Rand RW, Jacques DB, Melbye RW, Copcutt BG, Fisher MR, Levenick MN. Gamma Knife thalamotomy and pallidotomy in patients with movement disorders: preliminary results. Stereotact Funct Neurosurg 1993;61 Suppl 1:65-92. 244. Alterman RL, Kall BA, Cohen H, Kelly PJ. Stereotactic ventrolateral thalamotomy: is ventriculography necessary? Neurosurgery 1995;37:717-21. 245. Kondziolka D, Dempsey PK, Lunsford LD, Kestle JR, Dolan EJ, Kanal E, Tasker RR. A comparison between magnetic resonance imaging and computed tomography for stereotactic coordinate determination. Neurosurgery 1992;30:402-6. 246. Tasker RR, Dostrovsky JO, Dolan EJ. Computerized tomography (CT) is just as accurate as ventriculography for functional stereotactic thalamotomy. Stereotact Funct Neurosurg 1991;57:157-66. 247. Taha JM, Lamba MA, Samaratunga C, et al. A method to reduce systemic spatial shift associated with magnetic resonance imaging for stereotactic surgery. American Society for Stereotactic and Functional Neurosurgery, March 8-11 1995. 1995.

248. Burchiel KJ, Coombs BD, Nquyen TT, et al. Stereotactic neurosurgical magnetic resonance imaging and geometric distortion. [American Society for Stereotactic and Functional Neurosurgery, March 8-11, 1995]. 1995. 249. Chen DY, Bradley WG, Ali-Ali F, et al. Stereotactic MRI localization in radiosurgery. [American Society for Stereotactic and Functional Neurosurgery, March 8-11,1995]. 1995. 250. Coghill RC, Talbot JD, Evans AC, Meyer E, Gjedde A, Bushnell MC, Duncan GH. Distributed processing of pain and vibration by the human brain. J Neurosci 1994;14:4095-108. 251. Burton H, Videen TO, Raichle ME. Tactile-vibrationactivated foci in insular and parietal-opercular cortex studied with positron emission tomography: mapping the second somatosensory area in humans. Somatosens Mot Res 1993;10:297-308. 252. Jones AK, Brown WD, Friston KJ, Qi LY, Frackowiak RS. Cortical and subcortical localization of response to pain in man using positron emission tomography. Proc Biol Sci 1991;244:39-44. 253. Carlson M, Nystrom P. Tactile discrimination capacity in relation to size and organization of somatic sensory cortex in primates: I. Old-World prosimian, Galago; II. New-World anthropoids, Saimiri and Cebus. J Neurosci 1994;14:1516-41. 254. Shibasaki H, Sadato N, Lyshkow H, Yonekura Y, Honda M, Nagamine T, Suwazono S, Magata Y, Ikeda A, Miyazaki M. Both primary motor cortex and supplementary motor area play an important role in complex finger movement. Brain 1993;116 (Pt 6):1387-98. 255. Fletcher PC, Frith CD, Grasby PM, Shallice T, Frackowiak RS, Dolan RJ. Brain systems for encoding and retrieval of auditory-verbal memory. An in vivo study in humans. Brain 1995;118 (Pt 2):401-16. 256. Swartz BE, Halgren E, Fuster J, Mandelkern M. An 18FDG-PET study of cortical activation during a short-term visual memory task in humans. Neuroreport 1994;5:925-8. 257. Kapur S, Meyer J, Wilson AA, Houle S, Brown GM. Activation of specific cortical regions by apomorphine: an [15O]H2O PET study in humans. Neurosci Lett 1994;176:21-4. 258. Jones AK, Qi LY, Fujirawa T, Luthra SK, Ashburner J, Bloomfield P, Cunningham VJ, Itoh M, Fukuda H, Jones T. In vivo distribution of opioid receptors in man in relation to the cortical projections of the medial and lateral pain systems measured with positron emission tomography. Neurosci Lett, 1991;126:25-8. 259. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci USA 1990;87:9868-72. 260. Cohen MS, Bookheimer SY. Localization of brain function using magnetic resonance imaging. Trends Neurosci 1994;17:268-77.


261. Bucher SF, Seelos KC, Stehling M, Oertel WH, Paulus W, Reiser M. High-resolution activation mapping of basal ganglia with functional magnetic resonance imaging. Neurology 1995;45:180-2. 262. Ogawa S, Tank DW, Menon R, Ellermann JM, Kim SG, Merkle H, Ugurbil K. Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci USA 1992;89:5951-55. 263. Kwong KK, Belliveau JW, Chesler DA, Goldberg IE, Weisskoff RM, Poncelet BP, Kennedy DN, Hoppel BE, Cohen MS, Turner R. Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc Natl Acad Sci USA 1992;89:5675-79. 264. Rao SM, Binder JR, Hammeke TA, Bandettini PA, Bobholz JA, Frost JA, Myklebust BM, Jacobson RD, Hyde JS. Somatotopic mapping of the human primary motor cortex with functional magnetic resonance imaging. Neurology 1995;45:919-24. 265. Kim SG, Ashe J, Georgopoulos AP, Merkle H, Ellermann JM, Menon RS, Ogawa S, Ugurbil K. Functional imaging of human motor cortex at high magnetic field. J Neurophysiol 1993;69:297-302. 266. Kim SG, Ashe J, Hendrich K, Ellermann JM, Merkle H, Ugurbil K, Georgopoulos AP. Functional magnetic resonance imaging of motor cortex: hemispheric asymmetry and handedness. Science 1993;261:615-17. 267. Hammeke TA, Yetkin FZ, Mueller WM, Morris GL, Haughton VM, Rao SM, Binder JR. Functional magnetic resonance imaging of somatosensory stimulation. Neurosurgery 1994;35:677-81. 268. Puce A, Constable RT, Luby ML, McCarthy G, Nobre AC, Spencer DD, Gore JC, Allison T. Functional magnetic resonance imaging of sensory and motor cortex: comparison with electrophysiological localization. J Neurosurg 1995;83:262-70. 269. Davis KD, Wood ML, Crawley AP, Mikulis DJ. fMRI of human somatosensory and cingulate cortex during painful electrical nerve stimulation. Neuroreport 1995;7:321-5. 270. Shaywitz BA, Shaywitz SE, Pugh KR, Constable RT, Skudlarski P, Fulbright RK, Bronen RA, Fletcher JM,

271.

272.

273.

274.

275. 276.

277.

278.

279.

280.

78

Shankweiler DP, Katz L. Sex differences in the functional organization of the brain for language. Nature 1995;373:607-9. Hinke RM, Hu X, Stillman AE, Kim SG, Merkle H, Salmi R, Ugurbil K. Functional magnetic resonance imaging of Broca’s area during internal speech. Neuroreport 1993;4:675-8. Davis KD. Recent advances and future prospects in neuroimaging of acute and chronic pain. Future Neurol 2006;1:203-13. Jack CR Jr, Thompson RM, Butts RK, Sharbrough FW, Kelly PJ, Hanson DP, Riederer SJ, Ehman RL, Hangiandreou NJ, Cascino GD. Sensory motor cortex: correlation of presurgical mapping with functional MR imaging and invasive cortical mapping. Radiology 1994;190:85-92. Naatanen R, Ilmoniemi RJ, Alho K. Magnetoencephalography in studies of human cognitive brain function. Trends Neurosci 1994;17:389-95. Gallen CC, Bloom FE. Mapping the brain with MSI. Curr Biol 1993;3:522-4. Gallen CC, Sobel DF, Lewine JD, Sanders JA, Hart BL, Davis LE, Orrison WW Jr. Neuromagnetic mapping of brain function. Radiology 1993;187: 863-7. Forss N, Hari R, Salmelin R, Ahonen A, Hamalainen M, Kajola M, Knuutila J, Simola J. Activation of the human posterior parietal cortex by median nerve stimulation. Exp Brain Res 1994;99:309-15. Mogilner A, Nomura M, Ribary U, Jagow R, Lado F, Rusinek H, Llinas R. Neuromagnetic studies of the lip area of primary somatosensory cortex in humans: evidence for an oscillotopic organization. Exp Brain Res 1994;99:137-47. Yang TT, Gallen CC, Schwartz BJ, Bloom FE. Noninvasive somatosensory homunculus mapping in humans by using a large-array biomagnetometer. Proc Natl Acad Sci USA 1993;90:3098-102. Yang TT, Gallen CC, Ramachandran VS, Cobb S, Schwartz BJ, Bloom FE. Noninvasive detection of cerebral plasticity in adult human somatosensory cortex. Neuroreport 1994;5:701-4.

1323

82 Radiofrequency Lesions E. R. Cosman Sr. . E. R. Cosman Jr.

The radiofrequency (RF) technique was first put into clinical practice for making controlled therapeutic lesions in the nervous system in the early 1950s, and it have remained an important methodology to this day. This chapter summarizes the physical principles of the RF lesion method and describes the accepted rules for the RF lesion generation that have been verified over the years by clinical experience. Guidelines for making RF lesions of a specific size are stated, and a variety of specific electrode configurations used for various procedures are illustrated. Practical aspects of monitoring the progress of a safe and effective RF lesion are described.

A Brief History of Radiofrequency – Lesion Making Over the years, many physical principles have been utilized to make therapeutic lesions, targeted neurolysis, in the central and the peripheral nervous system. To name a few, these physical principles include: RF heating, direct current coagulation, cryogenics, focused ultrasound, induction heating, chemical destruction, ionizing radiation, stereotactic radiosurgery, mechanical methods such as the leucotomy, focused electromagnetic waves, and lasers. The RF lesion method has remained popular and successful among these techniques because of certain technical advantages that are intrinsic to its application, and these are summarized in the > Table 82-1. Although there have been many contributors to the progress made in the use of the RF technique, there are two names that stand out as particularly significant for their early work in the #


field and for the historic magnitude of their contributions: William H. Sweet and Bernard J. Cosman (> Figure 82-1). Sweet wrote a landmark paper [1] in 1953, coauthored by Vernon Mark, which showed that the use of very high frequency current (in the radiofrequency range) for lesion production has decisive advantages over the then established direct current lesion methods that had been used for decades. Sweet and Mark demonstrated that the lesions made by RF are characterized by a well-circumscribed lesion borders, and that there is better control of size and shape of the RF lesion than there is with the direct current method. Sweet made a second major advance to the field a few years later when he performed the first temperature-controlled RF lesions in the trigeminal ganglion for the treatment of trigeminal neuralgia [2–4]. His technique was rapidly adopted by numerous pain clinicians around the world and helped to underscore the power of the RF method. Bernard J. Cosman (> Figure 82-1) made parallel pioneering contributions to the design and engineering of RF lesion generators and electrodes. Prototype RF generators were made by Cosman and Aranow in collaboration with the neurosurgical group at the Massachusetts General Hospital in Boston in the early 1950s. Shortly thereafter, Cosman introduced the first commercial RF lesion generator, the Radionics Model RFG-2, for sale on the United States. For the next 40 years Bernard Cosman, with one of the present authors, his son, Eric Cosman, Sr., continued to make major advances in RF generator and electrode technology, and their designs dominated the field for decades, being offered by their company Radionics, Inc.

1360

82

Radiofrequency lesions

The modern RF lesion generators and electrode designs of today are elaborate and sophisticated. > Figure 82-2 shows two of the latest RF generators, the Model RFG-1A and the Model G4 . Table 82-1 Advantageous features of the radiofrequency lesion method 1. Well-circumscribed lesions 2. Temperature control allows: TQuantifiable lesions Non-charring, sticking, or boiling Differential selection 3. Excellent target control with: Stimulation Impedance monitoring Recording 4. Easily adapted to stereotactic or fixation devices 5. Versatile, robust electrode configurations for numerous clinical indications and target sites 6. Continuous and pulsed RF provides thermal and electric field modification of neural structures that suit specific indications

Graphics-Four (Cosman Medical, Inc., Burlington, Massachusetts). They both incorporate multiple functions and circuitry including an impedance monitor, wide-range stimulator, recording outputs and input connections, temperature monitors, both continuous and pulsed RF generator circuits, automatic temperature control, and lesion timers. They have built-in microprocessors and computers that can provides automatic function of the lesion process as well as safely checks to guard against untoward effects during the procedure. The Model G4 ‘‘GraphicsFour’’ unit enables use of up to four RF electrodes being applied to the patient at the same time, which has utility, for example, in treating multiple level medial branches for spinal pain. It has full color graphic display of the electrode functions with touch-screen features such as built-in standard procedure menus, recording memory, and printout records of procedure parameters, which are helpful, for example, in a busy pain

. Figure 82-l Two pioneers of radiofrequency surgery. William H. Sweet (left) of the Massachusetts General Hospital wrote the landmark paper in 1953, describing the decisive advantage of the radiofrequency lesion. Bernard J. Cosman (right) was the electrical engineer who built and perfected the earliest commercial radiofrequency lesion generators in the early 1950s, working with Dr. Sweet and Dr. Thomas Ballantine


82

. Figure 82-2 Two modern RF lesion generators. the Cosman Model RFG-1A (left) incorporates multiple functions and controls, digital readouts, and computer controls for effective and safe lesion making. The Cosman Model G4 ‘‘Graphics – Four’’ (right) has full computer graphic screen display, touch screen controls, stored menus and procedure records, and both digital and graphic screen representations for the function processes. It also has four electrode output jacks so that as many as four electrodes can be used on the patient at during the same procedure. Multiple electrode usage provides time-efficiency in some pain treatments, for example, performing multiple segmental medial branch RF heat lesions during the same patient treatment session

clinic where the throughput of patients being treated for spinal pain is high. The modern RF electrodes are also far more elaborate and sophisticate than in the early days of RF technology. Because the RF electrode can be made in a wide variety of shapes and configurations, a large range of styles have been developed for use at specific target sites. Many examples with be shown in this chapter. Today the use of the RF method, which might be referred to as ‘‘radiofrequency surgery,’’ spans the treatment of numerous disease states and spans a wide range of target sites from the brain to nearly every extremity of the body. In this chapter, we will focus on the RF applications in neurosurgery. However, the RF technique, though having its beginnings in neurosurgery with the work of Sweet and Cosman, has today found far wider application in fields primarily outside of neurosurgery, including, to name a few examples: anesthesiology and the peripheral pain therapy, interventive radiology and cancerous tumor ablation, urology and the treatment of BPH, cardiology and the treatment of WPW syndrome, podiatry, and the treatment of vascular disease.

Physical Principles of Radiofrequency Heat Lesion Generation The proper use of the RF lesion method requires a basic understanding of the physical processes involved [5–10]. > Figure 82-3 shows the basic RF lesioning circuit. The RF generator is the source of an RF voltage that is impressed across its output terminals. The terminals are in turn connected by cables to the so-called active electrode and the dispersive electrode (often called a reference electrode). The active electrode produces the heat lesion, and the dispersive electrode is typically a large-area electrode that is not intended to produce heating, but serves as a return pathway for the RF current. Electrical metering can monitor the RF current and voltage. Appropriate metering can also monitor the impedance and temperature at the active electrode tip. > Figure 82-4 illustrates how the RF current flows through the body between the active and dispersive electrodes. The total RF current IRF in the cables reaches the electrodes and spreads out

1361

1362

82


. Figure 82-3 The basic RF circuit

. Figure 82-4 The RF current patterns in the tissue between the active lesion electrode and the dispersive electrodes

in space in the electrolytic medium of the body. This is illustrated by lines of RF current density j, whose patterns are governed by the laws of electricity. The greatest heating takes place in the region of highest current density, which is near the tip of the active electrode. If the dispersive electrode has a significantly large area, the current density near it is low and heating nearby is minimal. For this reason, large-area dispersive electrodes with an area of at least 150 mm2 and an ample amount of conductive gel

between the electrode and the patient’s skin are recommended to avoid any chance of skin burns. The dispersive, or reference, electrode is typically placed on the skin surface over an area of musculature, not far from the site of the active electrode. For example, an active electrode used in the brain may be accompanied by a reference electrode attached and taped to the patient’s shoulder. Galvanic potentials can arise if active and reference electrodes are made from different metals. Thus, when the active lesion tip is made of stainless steel, which is a common type of electrode material, the base material of the reference electrode should also be made of stainless steel to prevent any transient potentials arising between the active and reference electrodes when the electrodes are first connected to the patient and the RF generator. > Figure 82-5 shows the pattern of the electric field and current density near the uninsulated active electrode tip. The mechanism for heating near the tip arises from the RF voltage that is created on the active electrode by the RF generator output. This voltage creates a distribution of the electric field E in the space around the electrode, shown by the E-field lines in > Figure 82-5. The electric field at any given point in space oscillates with the RF


82

. Figure 82-5 The electric field line patterns in tissue around the RF electrode, and the associated isotherm surfaces (dashed lines). The RF current lines follow the same pattern and the electric field lines

frequency and causes the nearby charged ions in the electrolyte to move back and forth in space at the same high frequency, which is typically about 500,000 cycles per second (Hertz) for most modern generators. This ionic oscillation is called the ionic current density j, and the pattern the j-field lines in space around the electrode tip follows the same pattern as the E-field lines shown in > Figure 82-5. It is the frictional heating within the tissue resulting from the RF ionic oscillation, i.e., the current density j, which is the basic mechanism by which the tissue heats up, and accordingly, by which the RF heat lesion is made. The ionic oscillations and current density are strongest near the electrode tip because the electric field is strongest there. Thus, the power deposition and tissue temperatures are highest adjacent to the electrode tip. The tissue that is heated by this mechanism in turn heats up the RF electrode tip, and so, by monitoring the tip temperature, an accurate measure of the nearby tissue temperature, and thus the progress of the RF lesioning, is obtained. Temperature monitoring during the

RF lesion making is an essential for control, consistency, and safety of the procedure. The dashed lines in > Figure 82-5 illustrate the surfaces of constant temperature, which are referred to as isotherms. The 45–50 C isothermal surface is critical, since within that surface the tissue will he hotter than 45–50 C and will be permanently killed. That then defines the heat lesion volume. It can be shown that at equilibrium, in a homogeneous medium, and for a specific tip size and tip temperature, the isotherm dimensions are roughly independent of the electrical and thermal conductivities. For a fixed tip size, a higher tip temperature produces a larger 45 C isotherm surface; that is, the lesion volume is larger. However, for a given tip temperature, the lesion size is also dependent on the size of the electrode tip, with larger tips producing, larger lesions. > Figure 82-6 is a schematic diagram of the temperature falloff in the tissue as a function of nominal distance from the electrode tip. This curve can now be calculated by modern

1363

1364

82


. Figure 82-6 A schematic diagram of temperature falloff as a function of nominal distance from the RF electrode tip

computer methods. Predicted lesion sizes agree with experimentally and clinically observed data. Certain qualitative features are notable. The highest temperatures occur in the tissue near the tip. The specific shape of the curve, and thus the lesion volume, depend on specific parameters such as: the electrode tip dimensions; the temperature that is maintained in the tissue near the electrode tip (by control of the RF voltage on the electrode); and, the electrical conductivity, thermal conductivity, and convection of the surrounding tissue. As shown in the > Figure 82-6, the higher the measured tip temperature, the larger the distance associated with the 45 C isotherm. It is believed that in the range between 42 and 44 C, neural tissue can sustain “reversible” damage by heating [11]. Thus, there may be a zone of reversibility for a given tip temperature corresponding to the shell of tissue near the 45 C isotherm that can be stunned, but not killed, by thermal lesioning. Measurement of the electrode tip temperature not only quantifies the lesion size but also avoids the tissue near the electrode tip reaching 100 C point, and thus avoids the catastrophic effects of

boiling, explosive gas formation, and searing and charring of the tissue. This dangerous condition can be avoided through careful elevation and monitoring of the tissue temperature to prevent temperature runaway to boiling. > Figure 82-7 illustrates another important aspect the lesion making. That is the time progression toward an equilibrium lesion size. This figure shows the increase of the transverse dimension of the lesion as a function of the time during which the electrode tip temperature is held at a fixed value. The size of the lesion increases until it reaches its asymptotic value. Once this value has been approached, no substantial increase of the lesion size occurs. For consistent lesion making, it is desirable to achieve this equilibrium situation, since it avoids some of the uncertainties associated with variability in tissue impedance, vascularity, and proximity to cerebrospinal fluid (CSF), bone, and other heat sinks. The typical time to reach the asymptotic temperature is 30–60 s for electrode tip sizes that are typically used in neurosurgery and pain therapy and for relatively uniform soft tissue environments [7].


. Figure 82-7 DREZ lesion width versus time. The equilibrium lesion size is achieved after 35–45 s for an approximately constant temperature at the electrode tip

82

. Figure 82-8 Irregularities of RF heating can be caused by inhomogenieties in nearby tissue structures and fluid bodies

Figure 82-8 schematically illustrates examples. The proximity of the electrode to CSF bodies such as the ventricles can provide a low-impedance shunt pathway for the RF current density j, and thus sink the heat away and cause irregular lesion shapes and sizes. An example of this situation is RF lesioning in the trigeminal ganglion. The proximity of large blood vessels also has an inhomogeneous cooling effect on the tissue near them. The placement of the RF electrode near a bony structure may have the opposite effect, since a bony mass is an insulator with lower blood circulation. Examples of this can be the RF heating of an intervertebral disk or of the space between other joints in the body. Despite the uncertainties caused by tissue in homogeneities however, the above rules for homogeneous tissue and for nominal conditions have worked well over the past decades, as attested by a mass of successful clinical data using the RF heating method. >

The above discussion can be summarized by some simple rules associated with RF lesion making: Rule 1: The RF current heats the tissue, and the tissue in turn heats the RF electrode tip. Rule 2: Temperature is the basic lesioning parameter, and it should be measured and controlled for consistent and safe RF lesioning. The measurement of electrode tip temperature is directly related to tissue temperature and lesion size. Rule 3: Achievement an equilibrium lesion as a function of time for a fixed tip temperature produces more consistent lesion size (than, for example, the time-buildup lesions). It is desirable to hold the proper tip temperature for 30–60 s to achieve the equilibrium lesion size. Rule 4: The proper electrode size and tip temperature should be selected for a given target site to achieve a consistent and desired lesion size. Specific factors that are difficult to predict or quantify can give rise to variations in the simplified picture of the RF lesion process from one patient to the next. Among these factors are the inhomogenieties in the tissue medium itself.

Typical Lesion Sizes as a Function of Electrode Geometries and Tip Temperatures It is important to have a sense of the size of the lesion that is being made for a given electrode lip geometry and tip temperature. An extensive

1365

1366

82


. Table 82-2 Postmortem lesion sizes versus heating parameters Electrode tipa

Lesion size Thalamus Thalamus Cingulum Cingulum Cingulum

Lesion size, mm

Temperature and times

Diameter, mm

Length, mm

Temperature, C

Time, seconds

Time, Postmortem

A

B

1.1 1.2 1.6 1.6 1.6

5 3 5 10 10

72 65 70 80–90 80

360 120 60 60 60

2 years ? 5 months 6 months 6 months

3 2 8 10 10

7 4 8 10–12 12

a

Straight stereotactic electrodes as shown in > Figure 82-17

discussion of this issue is given in papers by Cosman, et al [5–9], and only summary information will be given here. > Table 82-2 shows lesion sizes in the human brain reported by several stereotactic neurosurgeons using conventional electrode of straight, cylindrical tip geometry. These data for the most part were accumulated in postmortem studies at varying times after the lesion was made, and some variation in the lesion size might be expected as a function of time. As an example from > Table 82-2, it is seen that for thalamotomies performed using electrode tips with a diameter of 1.1 mm and a tip length of 3–5 mm and tip temperatures of 65–75 C, lesions sizes of approximately 3 mm minor in diameter and 4–7 mm in length can be achieved. Lesions for cylindrical electrode tip shapes are typically prolate ellipsoids of revolution. It is also seen that larger lesions can be made in the brain with larger electrode tips and higher temperatures. For example, lesions made in the cingulum with electrodes having 1.6 mm diameter and 10 mm tip length at 80–90 C have dimensions of typically 10 mm in minor diameter and 12 mm in length [12]. Electrodes of greater diameter and length for equivalently high temperatures are accordingly larger. > Table 82-2 shows the range of parameters used by several stereotactic neurosurgeons to produce thalamic lesions. This suggests that some variation in lesion size has been used successfully

in RF thalamotomies, but also that there is a reasonable norm for acceptable lesion parameters. As is described further below, electrodes with more complex geometries, such as with having side-outlet electrode tip extensions, have also been used for more precise or more asymmetric targeting in the region of the thalamus, in cases where physiological testing indicates that the initial electrode placement is not ideal. Information on lesion size for very small RF electrodes, such as those used in the spinal cord, is rare but does exist. Cosman and coworkers [7] have reported that electrodes with tip diameters of 0.25 mm and tip lengths of 2 mm, raised to a temperature of about 80 C and for a lesion time of approximately 15 s, will produce lesion sizes of about 2 mm in width and 2 mm in length. A summary of these data is shown in > Figure 82-9. Transverse lesion diameter, which is equivalent to the width of the prolate ellipsoidal lesion, is plotted as a function of the tip temperature for various electrode diameters. The length of the lesion can be assumed to be approximately 1–3 mm longer than the exposed length of the RF electrode, assuming substantially cylindrical electrode tip geometry with either a hemispherical or a sharpened tip end shape. Although this graph may supply canonical curves for lesion size, it must be recognized that the unpredictable factors of tissue inhomogeneity,


. Figure 82-9 The canonical curves of lesion diameter (A) versus tip temperature as a function of the electrode tip diameter (top). The dimensions of the heat lesion and the electrode tip (bottom)

3.

82

will not occur from encroachment in nontarget volumes. Proper configuration of the electrode as adapted for a particular target region.

The reader is referred to many articles and procedure technique monographs for specific recommendations on these points for the various RF lesion sites. A brief description is given below of successful RF electrode geometries for specific procedures to illustrate the great range of such designs.

Spinal Cord Electrodes

inhomogeneous conductivities and convectivities, and proximity to bony structures all can produce significant variations in the shape and size of lesions in a given situation.

Radiofrequency Electrode Configurations for Specific Clinical Procedures In addition to the choice of the proper RF electrode size and tip temperature, safe and effective lesion making is dependent on the following conditions: 1. 2.

The proper placement of the electrode in the target region. Adequate physiological testing using stimulation and recording to assess that the target is correct and assure that untoward effects

Targets in the spinal cord are discrete and critical. Associated RF electrodes must be appropriately small, with temperature monitoring, despite the fact that their historical forerunners, such as the cordotomy electrodes of Rosomoff [13], Lin and associates [14], Mullan [15,16], and others, were of relatively larger size and of a nontemperature-measuring variety. > Figure 82-10 shows the pointed tip configuration for the LCE Levin-Cosman Cordotomy Electrode Kit (Cosman Medical), which has an approximately 0.25 mm tip diameter and 2.0 mm tip exposure with an indwelling surfacemounted thermocouple [17]. The electrode can penetrate the pia and make very precise and discrete lesions in the spinothalamic tract for the treatment of intractable pain. The unique design of a surface thermocouple sensor within the sharpened tip allows rapid and faithful lesion temperature readings, which are essential in such tight geometries. For such tiny tips, any small thermal perturbation toward boiling during the temperature raising phase of the procedure, would cause explosive gas and steam formation. Thus, superfast temperature monitoring and fine RF power verniation are essential. This electrode is used through a 2l-gauge special spinal needle that is inserted percutaneously.

1367

1368

82


. Figure 82-10 The LCE Levin Cordotomy Electrode (top) and its tip geometry (bottom)

shouldered insulation, staggered tip insulation, and angled-tip configuration are designed for proper electrode placement and target localization in the very tight geometry of the spinal cord that is exposed during laminectomy. In these critical procedures, safe and reliable lesion placement and control require such special tip geometries and rapid surface-mounted thermocouple sensing in the electrode tip.

RF Electrodes for Lesioning in the Trigeminal Ganglion

Figure 82-11 illustrates the tip geometries for the KCTE Kanpolat CT Radionics electrodes, which have both straight and bent-tip geometry [18]. These electrodes are used for percutaneous cordotomy and spinal tractotomies and embody the important innovation of being CT-compatible. Kanpolat and associates [18–22] demonstrated that, with the use of proper materials, a computed tomography (CT) image of the electrode in place in the spinal cord can be made for direct visualization of the positioning of the electrode tip in the lateral spino-thalamic tract. Thus, in addition to the essential physiological testing, which is done through the Levin and Kanpolat electrodes before the lesion is made in percutaneous cordotomies, a new dimension of target position confirmation is achieved by doing the procedure in the CT scanner. > Figure 82-11 shows a confirmational CT slice made with the Kanpolat electrode. > Figure 82-12 shows the Nashold DREZ Electrode tip geometry and the El-Naggar-Nashold Nuclear Caudalis Electrode tip geometry, both from Cosman Medical, Inc. These electrodes have been used by Nashold and coworkers [23–25] to make lesions in the dorsal root entry zone (DREZ) lesion and the nucleus caudalis, respectively. The >

W. H. Sweet and coworkers at the Massachusetts General Hospital in Boston revolutionized the treatment of facial pain from tic douloureux through the use of temperature-monitoring RF electrodes inserted percutaneously into the trigeminal ganglion and posterior rootlets through the foramen ovale [1–4,26,27]. > Figure 82-13 shows the earliest electrode system for this procedure; the Cosman TIC Kit. It evolved from Sweet’s original concepts and was designed by B. J. Cosman, and is still offered by Cosman Medical and used by many neurosurgeons. The set has four insulated cannulas with a removable obdurating stylet. The cannulas having exposed tip lengths 4, 2, 5, 7, and 10 mm. In practice, a cannula, with a stylet in place, is inserted into the trigeminal ganglion, the stylet can be removed, and a thermocouple (TC) temperature electrode can be inserted. The TC electrode is connected to the RF generator, and physiologic testing is done using the stimulation signal output from the generator’s stimulator circuitry. Once the proper target position is thereby confirmed, the RF signal output from the RF generator is applied to the electrode, and the heat lesion is made using while monitoring electrode tip temperature on the generator’s readout display. In this way, Sweet and other workers [28–43] have reported excellent results for the relief of trigeminal pain since the early 1970s.


82

. Figure 82-11 The KCTE Kanpolat Cordotomy Electrode (top) The KCTE tip variations (middle figure) include a straight tip and an angled tip to enable adjustment of the tip position in the spinal cord. The diagrams (bottom) illustrate the approach to the lateral spinothalamic tract under CT control. A CT scan shows the KCTE Electrode in the spinal cord

The straight electrodes of the TIC Kit, which have been adequate for most trigeminal neuralgia procedures, are sometimes inadequate to reach selectively the first division of the trigeminal nerve. Therefore, a modification of the TIC Kit was built to overcome this limitation. > Figure 82-14 shows

the Tew Kit that was developed by John M. Tew, MD and Eric Cosman Sr. [33] The Tew Kit includes an insulated cannula through which either an obdurating stylet, a straight emergent RF electrode, or a flexible side-outlet emergent RF electrode can be passed. The tip configurations

1369

1370

82


. Figure 82-12 Nashold Thermocouple DREZ Electrode (top) is shown with its holding cannula and sizing clamp. The El-Naggar/ Nashold Angled ENA Electrode (second from the top) with its gripping handle and connection leader wire. The tip detail of the DREZ electrode (lower left, top) shows its shouldered insulation for limited penetration of the dorsal root entry zone region of the spinal cord. The ENA electrode tip (lower left, bottom) that has an angled distal shaft end to facilitate penetration of the nucleus caudalis. The surgical approach for the ENA Electrode shown in the lower right drawing

for the straight and curved Tew electrodes are also shown in > Figure 82-14. Both electrodes have TC thermocouple temperature sensors built into their tip ends. The cannula, with the obdurating stylet in it, is percutaneously inserted into the trigeminal ganglion through the foramen ovale. The stylet is removed, and either the

straight TC electrode is inserted, for making axial tip extensions beyond the end of the cannula, or the curved spring electrode is inserted, for making off-axis tip extensions. This gives the clinician the capability to search in the space of the trigeminal rootlets with the electrode to find the desired trigeminal division corresponding to


82

. Figure 82-13 This shows the classical TIC Kit electrodes for trigeminal nerve RF lesioning. The upper figure shows the temperature sensor probe inserted into a cannula. The lower left figure shows the different available cannula tip lengths. The lower right figure shows the electrode cannula’s tip within the trigeminal rootlets

. Figure 82-14 The TEW Electrode accommodates straight and curved RF lesioning tips for use in the trigeminal ganglion. The upper picture shows the TEW cannula with a curved temperature sensor inserted within it. The lower left figure shows the curved spring electrode tip emerging from the cannula’s distal end. The lower right figure shows the curved electrode tip within the posterior rootlets of the trigeminal nerve. The curved tip can be angle upwards or downwards to better reach the painful trigeminal division

the trigeminal pain as determined by sensory stimulation. Each electrode can be connected to the impedance, stimulation, and RF heating functions of the lesion generator with full temperature control. The off-axis electrode illustrates the flexibility that is possible in an RF electrode design to accommodate a specific anatomical target.

RF Lesion Electrodes for Spine to Relieve Neck and Back Pain In the early 1970s, Dr. C. N. Shealy proposed a technique of percutaneous RF Rhizotomy of the medial branch to relieve mechanical lower back pain associated with the lumbar facet joints

1371

1372

82


[44–47]. The original SRK Shealy Rhizotomy Kit was developed by E. R. Cosman, Sr. and B. J. Cosman. It had a set of 14-gauge spinal needles that could be inserted percutaneously to approach the area proximate the suspected pain facets. A 16-gauge RF electrode, insulated except for an exposed 7 mm distal tip, could then be passed through the 14 gauge spinal needle, and could be advanced so that the electrode tip contacts the medial branch innervating the selected facet joint. This electrode was designed to produce a heat lesion with temperature control near the medial branch. There was encouraging success in the early days of the technique, but the results were not always consistent. A variant of the SRK electrode referred to as the RRE Ray Rhizotomy Electrode Kit was designed by E. R. Cosman and produced by Radionics Inc. in association with Dr. C. Ray [48,49]. This electrode has a self-penetrating cannula with an insulated tip and temperature monitoring and eliminated the need for the spinal needles. The RRE electrode is

still offered by Cosman Medical, Inc, and it is still preferred by some pain clinicians because of the large lesions it produces and because of its especially rigid shaft. In the late 1970s, Dr. M. Sluijter, and E. R. Cosman developed a more discrete set of electrodes, referred to as the SMK Kit, having finer gauge cannulas for the purpose of treating spinal facet pain in a similar manner to that proposed by Shealey. Sluijter’s approach utilized somewhat different target positions and electrode trajectories [50–55]. Because of its finer gauge electrodes, Sluijter used the system to treat pain in the lumbar as well and the cervical areas. A modern-day version of this electrode system is the CSK Kit produced (Cosman, Medical, Inc), and is shown in > Figure 82-15. The CSK Kit (Cosman Spinal Kit) contains, in one of its versions, 22-gauge (approximately 0.7 mm diameter) disposable cannulas which are offered with uninsulated tip lengths of 2, 5, and 10 mm tip exposures. Other versions of the CSK

. Figure 82-15 The CSK Cosman Spinal Kit for procedures of the spine such as RF medial branch or DRG procedures. Top figure shows the CSK cannula with the temperature sensing TC probe inserted in it. The middle figure shows the range of shaft lengths to accommodate cervical to sacral approaches in the spine for all size patients. The lower figure shows the straight and curved tips and the sharp and blunt points available to suit clinicians’ approaches, insertion techniques, and injection criteria


Kit are offered with 20 gauge or with 18 gauge cannulas, shaft lengths of 15 and 20 cm, and a variety of tip lengths to accommodate different target sites and patient sizes. The finer cannulas have the great advantage, over the Shealy and Ray designs, of minimizing the discomfort and problems of percutaneous insertion. The larger gauge CSK cannulas (20 and 18 Gauge) are preferred by some clinicians over the 22 gauge types because they make larger heat lesions in the lumbar and the sacral regions or because they can be used to produce multiple heat lesions for sacroiliac (SI) joint denervations. The cannulas have plastic hubs that have the further advantage of enabling the clinician to visualize the target site under C-arm X-rays a view along the direction of the cannula insertion. That is referred to as the “needle view” approach, and it greatly simplifies the speed and accuracy of placing the electrode tip on the medial branch nerve. The hubs have luer tapers that allow injection of anesthetic and contrast media before the RF lesion is made. The obdurating stylet can be removed, and the CSK-TC thermocouple-sensing electrode can be inserted. This thermocouple electrode has a surface-mounted thermocouple that makes its temperature response very rapid and its temperature-reading accuracy extremely precise. Such fine-gauge cannulas can be inserted without a local anesthetic, and the patient remains fully awake and alert. Once the cannula has been percutaneously inserted, it can be manipulated under simple C-arm fluoroscopy, which is essentially a freehand stereotactic technique. The C-arm is first aligned to the anatomy in the direction desired for the needle insertion; i.e., the needle view direction. The needle is then inserted and manipulated until its radio-opaque metal shaft appears essentially as a dot on the C-arm fluoroscopic image. The cannula is then parallel to the direction of the C-arm view, and the tip is positioned at the proper target location relative to the bony structure, as described by Sluijter [50–55]. > Figure 82-16 shows fluoroscopic views using the needle-view method and the CSK cannulas [36].

82

. Figure 82-16 The RRE Ray Rhizotomy Electrode for RF lesioning of medial branch to treat mechanical back pain. The lower figure shows the tip geometry

Stereotactic Radiofrequency Electrodes The earliest RF electrodes produced by B.J. Cosman in the 1950s were for stereotactic intracranial procedures to treat mood disorders. They comprised straight tubular stainless steel shafts that were insulated over their entire length except for an exposed conductive tip. In the earliest intracranial procedures, the stereotactic electrodes were placed either freehand or with crude guidance and stabilization means. However, as sophisticated stereotactic guidance systems became available by the mid-to-late 1950s, the diameters and lengths of the stereotactic electrodes were precisely tailored to fit the apparatus and their associated microdrive accessories. Straight stereotactic RF electrodes are still available (Cosman Medical, Inc.), and used in intracranial stereotaxy, primarily to treat Parkinson’s disease, pain, and mood disorders. A range of diameters and tip lengths, and shaft lengths are available to suit different clinical needs and modern stereotactic systems. Examples are shown in > Figure 82-17. The smooth hemispherical tips and insulation minimize disruption to brain

1373

1374

82


. Figure 82-17 Straight TC Stereotactic electrodes for intracranial RF lesioning, designed for use with a stereotactic guidance system. Tip geometry shown in the lower figure

tissue, and built in thermocouple temperature sensors in the tip enable accurate lesion control. More elaborate stereotactic electrodes have been produced for special stereotactic procedures, and they demonstrate the versatility that is possible in RF electrode design For example, off-axis lesion making to enlarge or to asymmetrically shape the lesion volume lead to electrodes with side extension tips. One of the earliest side-outlet electrodes was designed by N. T. Zervas and built by E. R. Cosman for RF ablation of the pituitary gland [56]. Its tip configuration is shown in the lower right portion of > Figure 82-18. It enabled asymmetrical lesions to be made to ablate all or part of the pituitary gland. An interesting historical note is that the Zervas hypophysectomy was, to the authors’ knowledge, the first example of an RF ablation of a cancerous tumor in humans. Today, RF ablation of tumors is commonly performed on a wide range of target sites throughout the body. > Figure 82-18 also shows a universal, frontfacing, open-lumen cannula that accommodates different inner elements to be passed through, including micro stimulation and/or recording

. Figure 82-18 A range of electrode configurations for straight and off-axis RF lesioning as well as stimulation and recording, after the designs of Zervas, Guildenberg, and Cosman

electrodes and both straight and side projecting RF electrodes. In the upper right is the Gildenberg mini-electrode system, which has a combined recording tip and lesion electrode for more specific targetry during thalamotomy and pallidotomy. J. Siegfried designed an elegant side-issue stereotactic electrode. The SSE Siegfried Stereotactic Electrode is shown in > Figure 82-19. The electrode has a 1.6-mm central shaft with an exposed RF electrode tip of 4 mm for making axial lesions. For stereotactic thalamotomies, it is often necessary to extend the lesion laterally or explore the space around the central electrode tip to determine the optimal target. This is accomplished by a side-emerging electrode with an exposed tip 0.5 mm in diameter and 2.0 mm in length. The tip can be extended a variable amount from the central cannula in any direction and can be used for stimulation, recording, or lesion making. The tiny side-issue tip has an indwelling, surface-mounted thermocouple electrode for very rapid, accurate temperature monitoring if it is desirable to make a lesion from that off-axis tip.

Pulsed Radiofrequency Technique In the modern RF generators, such as shown in > Figure 82-2, there are two modes of RF output that are commonly used for pain therapy. The


output waveforms for these two modes are shown schematically in > Figure 82-20. The first mode is the conventional, thermal (or heat) RF mode which has been described so far in this chapter. It uses a continuous sinusoidal RF waveform output, commonly referred to as continuous RF, or CRF. The second mode, which was first reported on by M. Sluijter, E. R. Cosman, and coworkers in 1998 [57], uses a series of pulsed bursts of RF signal, referred to as pulsed RF, or PRF. To date, PRF has been used primarily . Figure 82-19 The SSE Siegfried Side-Outlet Stereotactic Electrode tip geometry for RF lesioning in the basal ganglia, viz. thalamotomies

82

to treat peripheral nerves and the DRG’s, and most commonly it has been applied to treat back and neck pain and neuropathies. The results have been very good and have now been the subject of a large published literature and several clinical trials. One poplar advantage of PRF over CRF is that it can be done with little or no pain to the patient as the PRF output is being delivered. This is in contrast to some CRF applications in which there is considerable pain and discomfort to the patient during the RF heating of the neural tissue. To date there have been limited attempts to use PRF in the central nervous system, although if it were to have success there, it could have significantly expended potential. In both of the continuous and the pulsed RF modes, the amplitude of the RF voltage, V(RF), of the RF waveforms as shown in > Figure 82-20, is measured in units of Volts (V). As describes above, for the continuous RF waveform, a heat lesion is produced by the action of ionic friction of the RF currents in the tissue caused by the voltage V(RF) on the electrode. This means that

. Figure 82-20 A schematic diagram of the RF waveforms for continuous RF mode (CRF) (top) and for pulsed RF mode (bottom). Axes are not to scale

1375

1376

82


the neural tissue near the uninsulated, metal electrode tip is heated continuously to destructive temperatures (greater than 45–50 C). Thus, the CRF lesion volume includes all tissue within the 45–50 C isotherm boundary, which tends to have an ellipsoidal shape that encompasses the electrode tip. Within this lesion volume, all cell structures are macroscopically destroyed by heat. The action of pulsed RF on neural tissue is different from continuous RF [9]. Because the RF output is delivered in bursts of short duration relative to the intervening quiescent periods, the average temperature of the tissue near the electrode is not raised continuously or as high as for continuous RF for the same RF voltage V(RF). Since the PRF voltage is typically regulated to keep the average tip temperature in a nondestructive range, other mechanisms must produce the clinically-observed pain relieving effects. The electric field, E, is the fundamental physical quantity that governs all the actions of RF output on neural tissue, both for pulsed RF and for continuous RF modes. The electric field is created in space around an RF electrode that is connected to the output voltage V(RF) from an RF generator. His was illustrated in > Figure 82-5, and also is shown in > Figure 82-21 for a pointed electrode that is commonly used for percutaneous pain procedures. E is represented by an arrow (vector) at every point in space around the electrode tip, indicative of the magnitude and the direction the force it will produce on charged structures and ions in the tissue. The E-field produces various effects on tissue including: oscillations of charges, ionic currents, charge polarizations, membrane voltages, and structuremodifying forces. For continuous RF mode, the dominant consequence of these effects is the production of heat in the tissue caused by frictional energy loss due to the ionic currents that are driven by the E-field. However, for pulsed RF, the effects of E-field are more complex and varied, and range from heat flashes, to modification of neuron ultrastructure, to neural excitation phenomena.

. Figure 82-21 Schematic E-field patterns around a pointed RF electrode (top); and the calculated E-field strength distribution (bottom) in tissue for a 22 gauge electrode at V(RF) = 45 V

All of these effects can play a role in neuronal modification, though exactly how they produce antinociception in PRF treatments is a subject of active scientific investigation. Two consequences of theoretical predictions of the electric field in tissue during PRF are supported by experimental and clinical observations [9]. The first is that, as a consequence of the very high E-fields at the electrode tip, there are hot flashes at the electrode tip that can be thermally destructive to neurons. The second is that there are significant non-thermal effects of the E-field on neurons at positions away from the point of the tip that are certainly related to the pain-relieving effects of PRF. During the brief RF pulse, a hot spot occurs at the tip which can be 15–20 C above the average tissue temperature of the tissue [9] that remains


near body temperature of 37–42 C as shown in > Figure 82-22, top. This has been confirmed by ex vivo measurements, > Figure 82-22, bottom, and by finite-element calculations. The intense E-field and hot flashes could be expected to have destructive effects on neural tissue very near the tip point. Evidence for such destruction has been observed in vitro (Cahana et al. [58]). This may play a role in PRF’s clinical effect when electrode point is in the nerve or pressing against the nerve. However, it is unlikely that such focal effects can account for all of PRF pain relief, since the region of extremely high E-fields and T hot flashes are likely confined to less than about 0.2 mm radius from the electrode point (> Figure 82-23). There is evidence that direct, non-thermal effects are important in PRF. It is known that pain relief can be achieved when the side of the electrode tip, not the tip point, is next to an axon or dorsal root ganglion (DRG). While the hot flash fluctuations are less than 1 C at 0.5 mm

82

from the tip in any direction for typical PRF voltages, at lateral distances of greater than 1 mm, the magnitude of the electric field is still large in biological terms. For example, finite-element computation of the E-field for V (RF) = 45 V predict [9] that the E is 20,000 V/m at 0.5 mm, and 12,000 V/m at 1.0 mm laterally (> Figure 82-23). Thus, neuronal modifications in this E-field range should be significant. Comparison of E and T strengths between typical CRF and PRF waveforms show striking differences between these RF modes (> Figure 82-24). Calculations predict that after 60 s of CRF at V(RF) = 20 V, E = 21,000 V/m and T = 60–65 C at the lateral tip surface, and E = 2,750 V/m and T = 50 C at 1.8 mm away. In contrast, after 60 s of PRF with V(RF) = 45 V, E = 46,740 V/m and T = 42 C at the lateral tip surface, and E = 6,100 V/m and T = 38 C at 1.8 mm away. In other words, in PRF, the direct electric field effects are more prominent, whereas in CRF, the

. Figure 82-22 Electric field strength and temperature field strength distributions for the first PRF pulse around a 22 gauge pointed electrode at V(RF) = 45 V and a pulse width of 20 ms

1377

1378

82


. Figure 82-23 Hot flashes during a PRF pulse

. Figure 82-24 E-fields dominate over T-fields in PRF. The opposite is true for CRF


thermal fields are more prominent and largely mask the E-field effects. Combined with the understanding that PRF has a clinical effect even when the electrode is not placed on the nerve directly, these physical observations suggest that the E-field is directly involved in the analgesic effect of PRF. It is known that PRF Efields produce significant trans-membrane potentials on the neuron membrane and organelles. The E-field can also penetrate the membranes of axon and the DRG soma to disrupt essential cellular substructures and functions. For example, PRF done on the DRG of rabbits causes pronounced neuron ultrastructural modifications that are seen only under electron microscopy [59] and that are likely to modify or disable the cell’s function. This would suggest that PRF can produce sub-cellular, microscopic lesions on neurons in a volume around the electrode, possibly resulting in reduction of afferent pain signals. PRF membrane potentials are also capable of neural excitations (action potentials) by a process called membrane rectification. Because the PRF pulse rate is similar to that of classical conditioning stimulation (1–2 Hz), it has been proposed that PRF may have a similar action [9]. Conditioning stimulation is capable of suppressing synaptic efficiency of A-delta and C-fiber afferent nociception signals [60], a phenomenon know as Long Term Depression (LTD). Therefore, the PRF might be reducing transmission of pain information by LTD of synaptic connections in the dorsal horn. The appropriate exposure of PRF for a given pain syndrome and anatomical target, either for microscopic or LTD mechanisms, should be governed by the PRF “E-dose.” E-dose provides a parametric measure of E-field strength and integral pulse/time exposure [9].

Practical Tips on RF Lesion Making Proper lesion parameters, target control, and clinical judgment are paramount to successful RF lesion making. There are also practical

82

considerations in performing the lesion that can make a decisive difference in regard to a successful result. Although the latest RF generator systems have many automatic features, such as automatic temperature control and automatic ramp up of output levels, it should be kept in mind that, although automatic features may be acceptable for uses at lower risk target sites such as the medial branch nerves in the lower back, their use may not be indicated for very critical target site, such as in the central nervous system. It is for the critical targets that the control of the RF generator and of the progress of the RF lesioning itself should be managed with utmost attention by the hands and eyes of the clinician. One of the most common difficulties that can interrupt an RF procedure is failure of the electrode cables. These cables are the most manipulated, cleaned, sterilized, and mishandled objects in the process and thus are the most vulnerable. Spare cables should be on hand at all times. The impedance monitor on the RF generator should be watched for intermittent open or short circuits. The continuous impedance monitor of the RFG1A and G4 models allows monitoring of circuit integrity before, during, and after the lesion, which is a help to detect any fluctuations or abnormal readings that would indicate faulty connections, untoward shunting of current such as at insulation breaks, or incipient temperature instabilities that could, for example, signal focal boiling. A check of the electrode insulation before and after each case should be carefully done, since breaks in the insulation can cause RF current leakage in tissue regions where lesions are not desired and thus can lead to potentially harmful effects. The use of large-area dispersive electrodes of at least 150-cm2 area is essential to avoid skin burns at the reference electrode contact. The use of needles as a reference electrode should be avoided. Before beginning the lesioning making phase of the procedure and when the electrode is in the patient’s body, a tip temperature of approximately body temperature, i.e., 37 2 C, should always be

1379

1380

82


observed on the generator readout. If not, then most likely either the temperature sensing electrode or the connecting cable is faulty, and they should be checked or replaced with spare ones as required. Raising the RF power smoothly and by hand is traditionally good practice, especially when small lesion electrodes that are used in critical anatomical areas such as the spinal cord, brain, and trigeminal ganglion. Runaway temperature to the boiling point can lead to severe and uncontrolled effects, and thus a watchful eye on the temperature meter is required. For smaller electrode tips, the heating process is more sensitive and finicky, because small changes in RF output voltage levels can lead to quick temperature rises and higher risk of a runaway to instability and boiling at the tip. Unusually sluggish or jumpy temperature rise as you raise the RF output control level can be a sign of trouble, and should prompt extra vigilance and even a check of the electrode and cables. While the temperature is the fundamental lesioning parameter and should be measured and carefully watched throughout the procedure, it is good practice to be observant of the voltage, current, and impedance readings. Fluctuations of the voltage and current often signal an erratic effect such as cable interrupt or focal boiling, and under such conditions, the lesion process should be terminated and the system should be checked. It is good practice for a technician to record the lesion temperature, the impedance, and the RF voltage and current during the procedure so that if any question arises, those parameters can be reviewed later.

References 1. Sweet WH, Mark VH. Unipolar anodal electrolyte lesions in the brain of man and cat: report of five human cases with electrically produced bulbar or mesencephalic tractotomies. Arch Neural Psychiatry 1953;70:224-34. 2. Sweet WH, Poletti CE, Roberts JT. Dangerous rises in blood pressure upon heating of trigeminal rootlets; Increased bleeding time in patients with trigeminal neuralgia. Neurosurgery 1985;17:843.

3. Sweet WH, Wepsic JG. Controlled thermocoagulation of trigeminal ganglion and rootlets for differential destruction of pain fibers: I. Trigeminal neuralgia. J Neurosurg 1974;39:143-56. 4. Sweet WH, Wepsic JG. Controlled thermocoagulation of trigeminal ganglion and rootlets for differential destruction of pain fibers. J Neurosurg 1974;40:143-56. 5. Cosman BJ, Cosman EG. Guide to radiofrequency lesion generation in neurosurgery: radionics procedure technique series monographs. Burlington, MA: Radionics; 1974. 6. Cosman ER, Cosman BJ. Methods of making nervous system lesions. In: Wilkins RH, Rengachary SS, editors. Neurosurgery, vol. 3. New York: McGraw-Hill; 1984. p. 2490-8. 7. Cosman ER, Nashold BS, Ovelman-Levitt J. Theoretical aspects of radiofrequency lesions in the dorsal root entry zone. Neurosurgery 1984;15:945-50. 8. Cosman ER, Rittman WJ, Nashold BS, Makachinas TT. Radiofrequency lesion generation and its effect on tissue impedance. Appl Neurophysio1 1988;51:230-42. 9. Cosman ER Jr, Cosman ER, Sr. Electric and thermal field effects in tissue around radiofrequency electrodes. Pain Med 2005;6:(6)405-24. 10. Dieckmann G, Gabriel E, Hassler R. Size, form, and structural peculiarities of experimental brain lesions obtained by thermo controlled radiofrequency. Confin Neurol 1965;26:134-42. 11. Brodkey J, Miyazaki Y, Ervin FR, Mark VH. Reversible heat lesions, a method of stereotactic localization. J Neurosurg 1964;21:49. 12. Hurt RW, Ballantine HT Jr. Stereotactic anterior cingulate lesions for persistent pain: a report on 68 cases. Clin Neurosurg 1974;21:334-51. 13. Rosomoff HL, Carroll F, Brown J, Sheptak T. Percutaneous radiofrequency cervical cordotomy: technique. J Neurosurg 1965;23:639-44. 14. Lin PM, Gildenberg PL, Polakoff PP. An anterior approach to percutaneous lower cervical cordotomy. J Neurosurg 1966;25:553-60. 15. Mullan S. Percutaneous cordotomy. J Neurosurg 1971;35:360-6. 16. Mullan SF, Lichtor T. Percutaneous microcompression of the trigeminal ganglion for trigeminal neuralgia. J Neurosurg 1983;59:1007-12. 17. Levin AB, Cosman ER. Thermocouple-monitored cordotomy electrode. J Neurosurg 1980;53:266-8. 18. Kanpolat Y, Cosman ER. Special radio frequency electrode system for computed tomography-guided painrelieving procedures. Neurosurgery 1996;38:600-3. 19. Kanpolat Y, Savas A, Akyar S, Cosman E. Percutaneous computed tomography-guided spinal destructive procedures for pain control. Neurosurg Q 2004;14(4):229-38. 20. Kanpolat Y, Deda H, Akyar S, Bilgic S. CT guided percutaneous cordotomy. Acta Neurochir Suppl (Wien) 1989;46:67-8.


21. Kanpolat Y, Deda H, Akyar S, Bilgic S. CT guided trigeminal tractotomy. Acta Neurochir (Wien) 1989;100: 112-4. 22. Kanpolat Y. The surgical treatment of chronic pain: destructive therapies in the spinal cord. Neurosurg Clin N Am 2004;15:307-17. 23. Friedman AH, Nashold B, Ovelmen-Levitt J. Dorsal root entry zone lesions for the treatment of post-herpetic neuralgia. J Neurosur 1984;60:1258-62. 24. Nashold BS, El-Naggar A, Abdulhak M. New RF lesion DREZ electrodes for relief of facial pin based on a neuroanatomical study in man of the trigeminal nucleus caudalis at the cervicomedullary junction. American Society for Stereotactic and Functional Neurosurgery June 1991. 25. Sampson JH, Nashold BS. Facial pain due to vascular lesions of the brain stem relieved by dorsal root entry zone lesions in the nucleus caudalis. J Neurosurg 1992;77: 473-5. 26. Sweet WH. The treatment of trigeminal neuralgia (tic douloureux). N Engl J Med 1986;315:174-7. 27. White JC, Sweet WH. Pain and the neurosurgeon: a fortyyear experience. Springfield, IL: Charles C Thomas; 1969. pp 169-97, 607–609. 28. Tew JM. Percutaneous electrocoagulation of the trigeminal nerve in the treatment of trigeminal neuralgia. Radionics procedure technique series. Burlington, MA: Radionics; 1974. 29. Tew JM, Keller JT. The treatment of trigeminal neuralgia by percutaneous radiofrequency technique. Clin Neurosurg 1977;24:557-78. 30. Tew JM, Mayfield FH. Trigeminal neuralgia: a new surgical approach (percutaneous electrocoagulation of the trigeminal nerve). Laryngoscope 1973;83:1096. 31. Tew JM, Jr, van Loveren H, Percutaneous rhizotomy in the treatment of intractable facial pain (trigeminal, glossopharyngeal and facial nerves. In: Schmidek HH, Sweet WH, editors. Operative neurosurgica! technique. Orlando, FL: Grune & Stratton; 1988. p. 1111-23. 32. Tew JM, van Loveren HR, Caputi F. Percutaneous stereotactic radiofrequency rhizoromv for trigeminal neuralgia. Radionics procedure technique series. Burlington, MA: Radionics; 1990. 33. Tobler WD, Tew JM, Cosman ER, et al. Improved outcome in the treatment of trigeminal neuralgia by percutaneous stereotactic rhizotomy with a new, curved tip electrode. Neurosurgery 1983;12(3):313-7. 34. Van Loveren H, Tew JM, Keller JT, et al. A ten year experience in the treatment of trigeminal neuralgia: a comparison of percutaneous stereotaxic rhizotomy and posterior fossa exploration. J Neurosurg 1982;57:757. 35. Siegfried J. 500 percutaneous thermocoagulations of the gasserian ganglion for trigeminal pain. Surg Neurol 1977;3:126-31. 36. Siegfried J. Percutaneous controlled thermocoagulation of gasserian ganglion in trigeminal neuralgia: experiences

37.

38.

39.

40.

41.

42.

43.

44. 45.

46. 47.

48.

49.

50.

51.

52.

82

with 1,000 cases. In: Samii M, Jannetta PJ, editors. The cranial nerves. Berlin: Springer; 1981. p. 322-30. Siegfried J, Broggi G. Percutaneous thermocoagulation of the gasserian ganglion in the treatment of pain advanced cancer. In: Bonica JJ, Ventafridda V, editors. Advances in pain research and therapy. New York: Raven Press; 1979. p. 463-8. Siegfried J, Vosmansky M. Technique of the controlled thermocoagulation of trigeminal ganglion and spinal roots. In: Krayenbuhl H, editor. Advanced and technical stan-dards in neurosurgery, vol. 2. Berlin: Springer. p. 199-209. Broggi G, Franzini A. Radiofrequency trigeminal rhizotomy in treatment of symptomatic, non neoplastic facial pain. J Neurosurg 1982;57:483-6. Broggi G, Franzini A, Lasio G, et al. Long-term results of percutaneous retrogasserian thermorhizotomy for “essential” trigeminal neuralgia: considerations in 1000 consecutive patients. Neurosurgery 1990;26:783-7. Broggi G, Siegfried J. The effect of graded thermocoagulation on trigeminal evoked potentials in the cat. Acta Neurochir (Wien) 1977;24:175-8. Lazorthes Y, Verdie JC, Lagarrigue J. Thermocoagulation percutanee des nerfs rashidens a vise analgesique. Neurochirurgie 1976;22:445-53. Lazorthes Y, Verdie JC, Bouyssen M. Interet de 10 υtilisation d’un cadre stereotaxique dans la thermocoagulation selective du ganglion de Gasser. Neurochirugie 1976;22:77-83. Shealy CN. The role of the spinal facets in back and sciatic pain. Headache 1974;14:101-4. Shealy CN. Percutaneous radiofrequency denervation of spinal facets and treatment for chronic back pain and sciatica. J Neurosurg 1975;43:448-51. Shealy CN. Facet denervation in the management of back and sciatic pain. Clin Orthop 1976;115:157-64. Shealy CN. Technique for percutaneous spinal facet rhizotomy. Radionics procedure technique series. Burlington, MA: Radionics; 1973. Ray CD. Percutaneous radio-frequency facet nerve blocks: treatment of the mechanical low-back syndrome. Radionics, procedure technique series. Burlington, MA: Radionics; 1982. Ray CD. Your low-back pain and facet nerve blocks (audiovisual presentation for patient education). Minneapolis, MN: Institute for Low Back Care, Sister Kenny Institute; 1982. Sluijter ME. Percutaneous thermal lesions in the treatment of back and neck pain. Radionics procedure technique series. Burlington, MA: Radionics; 1981. Sluijter ME. Radiofrequency lesions in the treatment of cervical pain syndromes. Radionics procedure technique series. Burlington, MA: Radionics; 1990. Sluijter ME. The use of radiofrequency lesions of the communicating ramus in the treatment of low back pain. Techniques of neurolysis. Boston: Kluwer; 1989.

1381

1382

82


53. Sluijter ME. The use of radiofrequency lesions for pain relief in failed back patients. International disability studies. Basel: Eular; 1989. 54. Sluijter ME, Koetsveld-Baart CC. Interruption of pain pathways in the treatment of the cervical syndrome. Anaesthesia 1980;35:302-7. 55. Sluijter ME, Mehta M. Treatment of chronic back and neck pain by percutaneous thermal lesions. In: Lipton S, Miles J, editors. Persistent pain, modern methods of treatment. London: Academic Press; 1981. p. 141-79. 56. Zervas NT. Stereotaxic thermal hypophysectomy. Current techniques in operative neurosurgery. Grune & Stratton; New York: 57. Sluijter ME, Cosman ER, Rittman WJ, Kleef M. The effects of pulsed radiofrequency fields applied to the dorsal root ganglion- a preliminary report. Pain Clin 1998;11: (2)109-18.

58. Cahana A, Vutskits L, Muller D. Acute differential modification of synaptic transmission and cell survival during exposure target position pulsed and continuous radiofrequency energy. J Pain 2003;4:(4)197-202. 59. Erdine S, Yucel A, Cunan A, et al. Effects of pulsed versus conventional radiofrequency current in rabbit dorsal root ganglion morphology. Eur J Pain 2005;9:(3)251-6. 60. Sandkuhler J, Chen JG, Cheng G, Randic M. Low frequency stimulation of the afferent A-delta fibers induces long-term depression at the primary afferent synapses with substantia gelatinosa neurons in the rat. J Neurosci 1997;17:6483-91. 61. Fox JL. Experimental relationship of radio-frequency electrical current and lesion size for application to percutaneous cordotomy. J Neurosurg 1970;33:415-21. 62. Kline MT. Stereotactic radiofrequency lesions as part of the management of pain. Orlando, FL: Paul M. Deutsch Press; 1992.

83 Stimulation Physiology in Functional Neurosurgery A. W. Laxton . J. O. Dostrovsky . A. M. Lozano

Despite substantial progress in the clinical application of deep brain stimulation (DBS) [1–5], the specific mechanisms underlying its effects have yet to be fully determined. In this chapter, following a brief description of relevant anatomy, we discuss some of the physiological principles of electrical neurostimulation, and review research investigating the mechanisms of DBS.

DBS anatomy Because DBS is most commonly used for the treatment of movement disorders, the majority of studies that have explored the mechanisms of DBS have investigated stimulation of the principal movement disorder surgery targets: the subthalamic nucleus (STN), the internal segment of the globus pallidus (GPi), and the ventral intermediate nucleus of the thalamus (Vim). It is therefore worthwhile to briefly outline the relevant neuroanatomy of these targets. More detailed descriptions of this anatomy are to be found in other chapters of this textbook.

output to the GPi and the substantia nigra pars reticulata (SNr) as well as the GPe [6,7].

GPi The GPi receives excitatory glutamatergic input from the STN, inhibitory GABA-ergic input from the striatum and GPe, and dopaminergic input from SNc [8]. The GPi sends inhibitory GABAergic projections to the ventral and intralaminar thalamus, and the pedunculopontine region [6,7]. The rodent homologue of the GPi is the entopeduncular nucleus, embedded within the internal capsule [7].

Vim The Vim receives excitatory glutamatergic input from the cerebral cortex and deep cerebellar nuclei, and inhibitory GABA-ergic input from the reticular nucleus of the thalamus [6,8,9]. The Vim sends glutamatergic efferents to the cortical motor regions and the striatum [6,8,9].

STN The STN receives excitatory glutamatergic input from the frontal cerebral cortex [6,7]. It also receives inhibitory GABA-ergic input from the external segment of the globus pallidus (GPe) [6,7]. The parafascicular nucleus of the thalamus, pedunculopontine nucleus (PPN), and substantia nigra pars compacta (SNc) also project onto the STN [7]. The STN sends excitatory glutamatergic #


Basic Physiology of Electrical Neurostimulation DBS is typically delivered in biphasic square wave pulses via cathodal monopolar or bipolar electrodes. The physiological properties of monopolar stimulation are better characterized than they are for bipolar stimulation (see > Table 83-1) [10]. Monopolar stimulation spreads more diffusely

1384

83

Stimulation physiology in functional neurosurgery

. Table 83-1 Basic physiology of electrical neurostimulation* 1. 2. 3. 4. 5. 6.

7.

Axons are more responsive than cell bodies Large axons are more responsive than small axons Heavily myelinated axons are more responsive than less myelinated axons Neural elements are more responsive to cathodal stimulation than anodal High currents (8 times threshold) block action potentials The effect of stimulation on a neural element depends on the electrode’s distance from that neural element Current flow that is parallel to an axon is more likely to produce excitation than current flow that is transverse to an axon

*This table is adapted from Ranck [10]

and can therefore influence neural elements over a greater distance than bipolar stimulation [11]. Compared to monopolar stimulation, the focused current of bipolar stimulation may increase the risk of damage to stimulated regions [12]. Electrical stimulation affects not only the neurons near the electrode but can also excite the axons of neurons at a distance from the stimulation electrode that are passing near by and can also spread to affect neural elements outside of the immediate brain region stimulated. Therefore, although it is common to label stimulation as STN or GPi stimulation, that is really an abbreviated way of saying that electrical impulses are delivered by an electrode placed in the STN or GPi which affect an array of neural elements in the surrounding area [10]. For example, the STN and GPi are surrounded by other important structures, such as the zona incerta and pallidothalamic fiber bundles including the lenticular fasciculus and ansa lenticularis. The effects of DBS may actually result from stimulation of these structures [13]. Thus, the particular brain region, the proximity to surrounding pathways, and even the location within a nucleus will all influence what neural elements are stimulated and therefore what the effect of stimulation will be [14].

The various neural elements differ in their responsiveness to electrical stimulation. Less current is required to excite axons, in particular large myelinated axons, than neuronal cell bodies [10,13]. Furthermore, as the distance of a neural element increases from the stimulating electrode, less current reaches it [3]. The region within which stimulation may influence neural elements increases with stimulation amplitude [1]. This is not a straightforward effect, however, as afferent, efferent, and interneuronal axons may be activated, and this activation may be antidromic or orthodromic [14]. The lowest current (of a theoretically infinite duration) which will generate an action potential in a stimulated neuron is called the rheobase. The time it takes from the onset of electrical stimulation (at twice the rheobase) to the onset of an action potential in the stimulated neuron is termed the chronaxie. A fundamental relationship exists between the amplitude and duration or pulsewidth of stimulation. To maintain a constant effect, the stimulation amplitude must be increased as the duration is decreased, and conversely the stimulation duration must be increased as the amplitude is reduced. This relationship is represented in the following equation: Ith ¼ Irh ð1 þ tad =PWÞ where Ith is the threshold current, Irh is the rheobase, tad is the chronaxie, and PW is the pulse width or duration [8,15]. Different neural elements have different chronaxies. Large myelinated axons have the shortest chronaxies (30–200 ms); smaller axons have longer chronaxies (200–700 ms), and unmyelinated axons, dendrites and neuronal cell bodies have much larger chronaxies (1–10 ms)[8,10].Thismeansthatlargemyelinated axons are more easily activated by electrical stimulationthanotherneuralelements[10,13].Neural tissues also differ in their relative resistivity to electrical impulses. This too affects how stimulation current propagates. White matter generally has 2–3 times greater resistivity than gray matter


which has 4–6 times greater resistivity than cerebrospinal fluid [10]. White matter is also anisotropic which means that the conduction of electrical impulses is directed by the orientation of axons [10]. By comparing the chronaxies for a given DBS effect to known chronaxies, it is possible to infer which neural elements are being stimulated with DBS. Following this reasoning Nowak and Bullier [16] demonstrated that electrical stimulation activated axons and not cell bodies in the cortical gray matter of rodents [8]. This finding was corroborated by a follow-up experiment by Nowak and Bullier [17] in which the orthodromic responses to extracellular electrical stimulation were only slightly lower (15–20%) following N-methyl-D-aspartate (NMDA) block of neuronal cell bodies. Because 80–85% of the normal stimulation response was maintained, the results provide confirmatory evidence that in most cases the neural element most responsible for the effects of electrical brain stimulation is the myelinated axon [1,8]. Holsheimer et al. [18] have extended this work into the clinical realm. They measured the latency of DBS-induced tremor suppression among patients with Parkinson’s disease and found that chronaxies for thalamic and internal pallidal stimulation (129–151 ms) fell within the range of chronaxies of large myelinated axons (30–200 ms). As this is much shorter than the chronaxie of neuronal cell bodies (1–10 ms), these results corroborate that the primary target of DBS mediating the therapeutic benefits of stimulation in those two regions is the myelinated axon [1,8,10]. The distance of the stimulating electrode to the neural elements influences its effect. The rheobase and chronaxie increase as the distance from the stimulating electrode to the neural element increases [8,10]. When the applied current exceeds the threshold by a factor of 8 or more, it can paradoxically block excitation (by inducing a depolarization block) [8]. Therefore, neural elements nearest an electrode may be blocked and distant elements may not receive sufficient

83

stimulation to respond, whereas those elements within an appropriate intermediate perimeter will be excited [8,10]. This range of influence also depends on the properties of each neural element. For example, small axons may be activated close to the stimulating electrode, whereas larger axons may be blocked [10]. The orientation of the stimulating electrode to the axon is also known to influence the effect of stimulation. Current flow that is parallel to an axon is more likely to produce excitation than current flow that is transverse to an axon [10,19].

DBS Mechanisms of Action DBS, surgical lesions, and inhibitory drugs such as the sodium channel blocker lidocaine and the GABAA agonist muscimol, generally produce similar clinical effects when applied to the same neuroanatomical targets [2,20–22]. It is, therefore, reasonable to suppose that DBS and lesions work through a similar mechanism: the inhibition of neuronal activity [3]. While parsimonious and intuitively appealing, this conclusion is difficult to reconcile with the basic physiology of electrical neurostimulation as just described. To better account for the mechanisms of DBS, several explanations have been proposed. It has been suggested that DBS (i) has direct effects onneuronalmembrane properties and ionconductances,(ii)stimulatessynapticactivityleadingtothe release or even depletion of neurotransmitters and postsynaptic desensitization of neurotransmitter receptors, and (iii) alters the frequency or pattern of pathological neuronal activity [3–5,14,22–24].

Direct Cellular and Membrane Effects The ability of high frequency stimulation (HFS) to alter intrinsic membrane properties has been posited as a potential mechanism of DBS [25]. There is evidence that STN HFS can act directly

1385

1386

83


on neuronal membrane sodium and calcium channels [26]. Some in vitro studies suggest that HFS can decrease the excitability of neurons through the inactivation of voltage-gated sodium and calcium currents [27]. By transiently depressing voltage-gated Ca2+ channels, HFS has been shown to block depolarization [27]. HFS may also directly affect Na + channels [28,29]. Using patch-clamp techniques in rat STN slices, Beurrier et al. [27] found that 1 min of high frequency (100–250 Hz) bipolar STN stimulation blocked ongoing STN neuronal firing, and that this blockade lasted for around 6 min after the end of stimulation. Because this neuronal silencing occurred even in the presence of glutamate and GABA antagonists, or cobalt which blocks voltage-gated Ca2+ channels and neurotransmitter release, they concluded that this stimulation effect was not synaptically mediated. Instead, stimulation directly altered neuronal voltage-gated currents leading to a depression of spontaneous neuronal activity. On the other hand, STN HFS (>100 Hz stimulation) in rat brain in vitro slice preparations has been shown to induce depolarization and early rapid firing, then prolonged inhibition [28]. This silencing effect has been attributed to the inactivation of Na+-mediated action potentials. Furthermore, Do & Bean [29] found that the inherent pacemaking rhythmicity of STN neurons is due to persistent Na+ currents flowing at subthreshold voltages. HFS (70 Hz) altered the cells’ inherent rhythmicity through a slow inactivation of Na+ currents. The authors suggest that this stimulation- induced inactivation of sodium currents likely contributes to the clinical effects of DBS at a cellular level. There is also evidence that HFS can cause increased extracellular K+ levels which hyperpolarizes the neuron rendering it less excitable, an effect which may involve the stimulation of glial cells [30]. In rat brain slice preparations, STN HFS (100 Hz) has also been shown to produce long-term alterations in synaptic plasticity [31].

Anderson et al. [32] have identified two types of membrane responses to high frequency extracellular microelectrode stimulation of neurons in rodent thalamic slice preparations using intracellular recording techniques. Although all recorded neurons initially exhibited depolarization and rapid spike activity, some neurons quickly repolarized and ceased to fire (type 1), while others maintained their post-stimulation level of membrane depolarization with or without ongoing spiking (type 2). The authors were able to block the initial depolarization with the sodium channel blocker tetrodotoxin (TTX), the Ca2+ channel antagonist Cd2+, and various glutamate antagonists, including kynurenate. In contrast, the GABAA antagonist picrotoxin did not affect the neuronal depolarization or response types. Overall, their results suggest that the stimulation-induced depolarization is due to the presynaptic release of glutamate, and subsequent activation of postsynaptic glutamate receptors. Furthermore, they found that increasing the applied stimulation current increased the frequency and probability of neuronal firing. Using intraoperative microelectrodes in essential tremor patients undergoing thalamic DBS, Anderson et al. 2006 [33] found that suppression of tremor cell activity occurred in regions beyond the areas of direct current spread from the electrodes. Their results suggest that stimulation produces a functional deafferentation of afferent axons leading to a reversible synaptic depression which thus prevents tremor cell firing in the thalamus. Does DBS inhibit or excite neurons? Studies on the direct cellular effects of HFS suggest that stimulation inhibits neuronal activity [8]. Extracellular bipolar STN stimulation of in vitro rat brain slices can cause an initial increase in STN action potential firing, followed by a longer period of inhibition. In anesthetized rats administered STN stimulation (50 Hz; 300 mA), Lee et al. [34] found an initial increase in STN neuronal firing followed by a longer period of neuronal quiescence.


Boraud et al. [35] have reported that GPi neuronal activity is significantly increased in rhesus monkeys following treatment with MPTP, and that this increased activity is reduced to baseline firing rates with GPi HFS. Bennazouz et al. [36] have also found decreased neuronal activity in the STN and SNr of rats following STN HFS. Anderson et al. [37] recorded ventral thalamic neurons following short trains of GPi HFS stimulation in awake non-Parkinsonian monkeys. Thirty-three of 73 thalamic neurons were inhibited and seven were excited following stimulation. STN HFS (130 Hz; 500 mA; 90 ms) in rats induces c-fos expression but decreases cytochrome oxidase subunit I mRNA levels in STN neurons [38]. This overall reduction in metabolic activity could be compatible with a stimulationinduced inhibition of STN neuronal activity. It should be noted, however, that the intensity of the applied current was quite high (500 mA). In patients with Parkinson’s disease (PD), high frequency microstimulation in STN can decrease firing rates in STN neurons 600 microns from the stimulation site [39,40]. Inhibition of GPi neurons has also been found after GPi microstimulation in PD patients [41]. Pralong et al. [42] showed decreased neuronal firing of tonically active neurons in the pallidal-receiving thalamus of a patient with dystonia following GPi DBS. Other studies show that the effect of stimulation on neuronal firing is variable and can be excitatory. Benazzouz et al. [36] showed that in anesthetized rats, STN HFS led to decreased firing of STN and SNr neurons and increased firing in ventrolateral thalamic neurons in the poststimulation period. The activity of dopaminergic neurons in the SNc of anesthetized rats with and without globus pallidus lesions increased following STN HFS (130 Hz, PW 60 ms, 300 mA) [43]. Based on their previous work showing SNr inhibition with STN HFS [23,36], the authors conclude that STN HFS likely affects SNc neuronal activity by blocking the tonic inhibition of the SNr on the SNc.

83

In human patients the effect of HFS has been more variable. Brief macroelectrode STN HFS (140 Hz; 2 mA; 60 ms) in awake PD patients undergoing DBS surgery caused a decrease in the firing rate and duration of bursting activity of SNr neurons [44]. Low frequency stimulation (LFS; 14 Hz) did not affect SNr neuronal activity. Garcia et al. [45] argue that when stimulation parameters more closely mimic clinical DBS, stimulation excites STN neurons. When electromyographic responses to high frequency (100 Hz) STN stimulation were analyzed in 14 patients with parkinsonism, there was no evidence that stimulation blocked neuronal activity [46]. Instead, the ability of high frequency STN stimulation to attenuate contralateral tremor resulted from the activation of largediameter axonal fibers. Similarly, in 6 PD patients, evoked scalp potentials were recorded following high frequency STN stimulation [47]. The short chronaxie (50 ms) of the stimulated elements suggested that these elements were myelinated axons. The applicability of in vitro findings toward understanding the mechanisms of clinical DBS is unclear [25,27]. The current densities produced in animal studies are often much higher than those used in human DBS, and the slice preparations lack many of the connections, spontaneous activity, and pathological patterns of activity present in human patients, so the results may not be directly applicable to clinical DBS [8,14]. Caution must also be applied when interpreting studies involving anesthetized animals because it is unknown whether the stimulation parameters used produce the same beneficial effects seen with clinical DBS [25]. For example, when stimulation parameters shown to reduce dopamine antagonist induced catalepsy in rats are used, STN HFS increases SNr neuronal firing [48,49]. Anesthetic agents may also alter neuronal responses to stimulation. Furthermore, local inhibitory neuronal effects could still be coupled with the excitation of efferent axons passing, for example, through the STN region from the GPe to the SNr/GPi [50]. Computer modeling of HFS

1387

1388

83


suggests that subthreshold stimulation (relative to a single neuron at a given distance from the stimulating electrode) suppresses intrinsic firing by activating inhibitory presynaptic terminals (depending on the nucleus stimulated), whereas suprathreshold stimulation inhibits the intrinisic firing of the cell body, but generates efferent output at the stimulation frequency via direct activation of the axon arising from the neuron being modeled [24,51]. Finally, even if HFS is excitatory, it may predominantly excite GABAergic axons (e.g., from the thalamic reticular nucleus, putamen, GPe, GPi) and thus reduce neuronal firing rates in the STN, SNr, GPi or Vim [3,35,41,52]. The influence of DBS on neurotransmission will be considered below.

Effects on Neurotransmission STN region stimulation influences glutamatergic and GABA-ergic neurotransmission (see > Figure 83-1) [25,48,53]. In anesthetized rats, Maurice et al. [49] showed that STN HFS decreased firing in the majority of recorded SNr neurons. This inhibition was blocked by the GABA antagonist bicuculline, suggesting that HFS activated inhibitory GABA-ergic projections from the striatum or GPe to the SNr. In freely moving rats, HFS (124 Hz) of the caudate nucleus was associated with increased levels of extracellular GABA in adjacent areas of the caudate [54]. By lesioning the globus pallidus in rats, it is possible to abolish the STN HFS-induced rise in extracellular GABA in the SNr, suggesting that STN-HFS causes GABA-release in the SNr through the activation of pallidonigral axons and/or GPe neurons [55]. When STN HFS was administered to intact and hemi-Parkinsonian rats at an amplitude that produced contralateral limb dyskinesia, increased extracellular glutamate levels were obtained in the ipsilateral SNr [56]. When stimulation was administered below the dyskinesia-inducing

threshold, SNr glutamate levels were unaffected, but ipsilateral extracellular SNr GABA levels increased. Windels [55,57] found that STN HFS in rats leads to increased extracellular glutamate and GABA in the SNr and GPi/entopeduncular nucleus; these results suggest that STN HFS excites STN glutamatergic efferent projections to SNr and GPi as well as GABAergic projections (probably indirectly via GPe). Studies by MacKinnon et al. also suggest that STN stimulation in humans activates pallido-thalamic axons near the dorsal STN [58]. The effects of HFS were also examined in ferret brain slice preparations, in which thalamocortical relay neurons exhibited spontaneous spindle oscillations, and picrotoxin-induced 3–4 Hz absence seizure-like activity [59]. HFS (100 Hz; 10–1,000 mA; 100 ms) generated inhibitory and excitatory postsynaptic potentials, membrane depolarization, and eliminated the spontaneous spindle oscillations and 3–4 Hz absence seizure-like activity. These results suggest that HFS can disrupt abnormal neuronal activity through synaptic neurotransmitter release. Further supporting the synaptic transmission proposal, studies have demonstrated increased firing rates in the GPi following STN HFS [60], and alterations in firing rates in the thalamus following GPi HFS [37,61]. Similarly, glutamate release in the STN [62] and in the downstream GPi and SNr [56,57], has been found to be increased following STN HFS. Windels et al. [55] found that a small proportion of SNr neurons increased firing with STN HFS as stimulation frequency increased from 50 to 130 Hz. This excitation likely represents the activation of glutamatergic projections from the STN to the SNr. In murine thalamic slice preparations, HFS (125–200 Hz; 50 mA) increased the release of adenosine triphosphate (ATP), leading to increased concentrations of the ATP metabolite adenosine [63]. Adenosine A1 receptor activation and adenosine agonists in the thalamus depress excitatory neurotransmission and reduce tremor in mice. Adenosine A1 receptor-null mice experience


83

. Figure 83-1 Summary of potential effects of STN region DBS on neurotransmitter release in GPi/SNr. (a) STN stimulation excites GABA-ergic axons from GPe or striatum; (b) STN stimulation blocks STN neurons and axons; (c) STN stimulation excites STN axons leading to glutamate release

seizures when exposed to even low, subtherapeutic stimulation. These adenosine deficient mice also exhibit involuntary movements without stimulation. These findings provide another plausible mechanism by which thalamic HFS reduces

tremor, and suggest that the effect of HFS in other brain regions may also be due, at least in part, to the stimulation-induced accumulation of adenosine. It is important to note that caffeine, well-known to exacerbate tremor, is an adenosine antagonist.

1389

1390

83


In PD patients, single electrical pulses in the vicinity of GPi neurons produce a cessation of spontaneous activity for 15–25 ms [41]. Because of the latency and duration of this effect, it is likely that stimulation near the GPi causes GABA release from either GPe or striatal axons projecting onto GPi neurons, or local dendritic release of GABA resulting in GPi neuronal inhibition. Although the GPi also receives glutamatergic afferents from the STN, the more numerous GABAergic inputs are believed to overcome these excitatory signals. In patients with dystonia, neuronal firing rates in adjacent regions of GPi and in the downstream Vop decrease during GPi DBS (120 Hz; 3–4 V; 90 ms) suggesting stimulationinduced activation of pre-synaptic GABA-ergic inputs as well as GABA-ergic pallido-thalamic projections [52,61]. Computer modeling suggests that DBS can lead to the propagation of orthodromic and antidromic action potentials [64]. Experimental work in humans is compatible with this conclusion. Following single pulse stimulation of the STN in PD patients, the latency of TMS-induced MEP decreased. This facilitation could be due to orthodromic activation via the subthalamonigropallidal- thalamo-cortical circuit or antidromically via cortico-subthalamic projections [65]. STN HFS in anesthetized rats has also been found to antidromically influence corticosubthalamic projections [49]. Does DBS affect dopaminergic activity? Nigrostriatal axons abut the dorsal STN. Stimulation of the STN may excite these dopaminergic projections, and lead to increased striatal dopamine release. Microdialysis studies in rats have shown increased release of extracellular dopamine and its metabolites following STN HFS [34,66–68]. Using STN HFS in anesthetized rats, Lee et al. [34] found an increased release of striatal dopamine. The authors conclude that STN stimulation excites dopaminergic nigrostriatal projections and the striatal release of dopamine may be responsible for the beneficial clinical effects of STN DBS in patients with movement disorders.

The effects of STN HFS (130 Hz; 80 mA; 80 ms) were studied in a 6-hydroxydopamine-lesioned rat model [69]. Following chronic L-dopa treatment, the rats displayed L-dopa induced dyskinesias (LID). When STN HFS was applied to rats on L-dopa, their LIDs were exacerbated. Off L-dopa, STN HFS did not induce dyskinesias. These results contrast those described above, and suggest that STN HFS may influence L-dopa-modulated transduction pathways, but does not directly increase extracellular dopamine release. Electrophysiological studies in PD patients suggest that STN DBS may normalize pathways adversely affected by L-dopa [70]. It is unlikely, however, that the effect of clinical DBS is mediated through alterations in dopamine neurotransmission [25]. Abosh et al. investigated whether STN DBS increases the release of striatal dopamine in five patients with Parkinson’s disease [71]. Twelve hours after their last dose of L-dopa and 9 h after their DBS systems had been turned off, each patient underwent [11C]raclopride positron emission tomography (PET). Raclopride is a low affinity D2 receptor antagonist. As extracellular dopamine levels increase, raclopride is displaced from the D2 receptor which is reflected in decreased [11C]raclopride binding. Halfway through a 90 min raclopride infusion, the patients’ right STN stimulators were turned on, while their left-sided stimulators were left off. No difference in striatal [11C]raclopride binding was found between the right and left sides. Because unilateral stimulation could cause bilateral dopamine release, bilateral striatal [11C]raclopride binding prior to stimulation was compared to post-stimulation binding, but again no changes were found. With stimulation, the patients’ Unified Parkinson’s Disease Rating Scale (UPDRS) motor scores improved. Hilker et al. [72] obtained similar findings using the same paradigm. These results suggest that the primary mechanism of action of STN DBS is not striatal dopamine release. Does DBS lead to neurotransmitter depletion? Another possibility is that because DBS stimulates


axons, it could produce ongoing neurotransmitter release which could quickly lead to neurotransmitter depletion [13]. Subcortical HFS (125 Hz) in rat brain slices can produce initial depolarization then prolonged depression of excitation in primary motor cortex [73]. This depression was unrelated to GABA-ergic transmission and did not result from a complete block of action potentials. Instead, prolonged HFS reduced excitatory synaptic currents produced by the stimulated pathways, in keeping with a stimulation-induced depletion of excitatory neurotransmitter [50]. Conversely, HFS has been shown to replenish vesicle stores via calcium dependent mechanisms [74]. In a murine brainstem slice preparation, HFS (300 Hz) of presynaptic terminals has been shown to replenish vesicle pools through its influence on voltage-gated Ca2 + channels [74]. It is unlikely that neurotransmitter depletion is a primary mechanism of DBS. Does the effect of DBS vary with stimulation parameters? The effect of DBS depends on stimulation amplitude, pulse width, and frequency as described above. Amplitude. Increasing the current intensity results in an increase in the number of axons/ neurons activated since the effective current spread increases, but can also lead to depolarization block of neurons close to the electrode. The downstream effects can be complex since increased output should increase the magnitude of the stimulation effects on the target nucleus but if the stimulated structure is not homogeneous and/or if the stimulus spreads to affect a different nucleus or axons of passage then additional and/or opposite effects can be induced. In anesthetized rats, STN HFS (50–200 Hz; 20–300 mA; 60 ms) decreased SNr activity at low intensities, but increased it when delivered at high intensity [49]. Because the low current inhibition was blocked by the GABA antagonist bicuculline, it likely resulted from the stimulation-induced release of GABA. In thalamic slices, higher (50 mA) but not lower (10–25 mA) amplitude

83

stimulation led to the increased release of ATP, and only when delivered at frequencies in the 125–200 Hz range [63]. Pulse Width. The effect of variations in stimulation parameters to reduce contralateral wrist rigidity were tested in 10 PD patients with STN DBS [75]. When frequency was held constant, stimulation intensity could be reduced while still achieving clinical effectiveness if the pulse width was increased from 60 to 210 or 450 ms. Similarly, for any particular pulse width, increasing the frequency from 90 to 130 or 170 Hz allowed for a reduction of stimulation intensity while maintaining the desired therapeutic effect. Not surprisingly, the current intensity necessary to induce contralateral limb dyskinesia is higher when shorter pulse widths are used [56]. Frequency. When amplitude and pulse width are held constant, the magnitude of the downstream stimulation effect generally increases with increased stimulation frequency [3]. For example, researchers have found that in some situations the relation between frequency and the clinical or neuronal effects of stimulation are linear [76]. However, at higher frequencies, generally over 100 Hz, the effect of stimulation can frequently change dramatically. For example, abrupt threshold effects can be seen whereby frequency has no effect until a specific threshold of about 80–100 Hz is reached [52]. Similarly, the effect of stimulation may be abruptly blocked once an upper threshold, such as 200 Hz, is reached [30,50]. Moreover, in some situations, low frequency stimulation can have an opposite effect to high frequency stimulation [3]. For clinical applications, the beneficial effects of DBS have generally been achieved with stimulation frequencies within the range of 100–200 Hz [3,77–85]. In rats, STN stimulation induced release of glutamate in GPi and SNr is maximal at frequencies above 130 Hz, and GABA release is increased above frequencies of 60 Hz and only in the SNr [86]. Low frequency stimulation (10 Hz) of STN neurons in naive and dopamine-depleted rat

1391

1392

83


brain slices evoked single 10 Hz spikes, but did not significantly alter the overall ongoing neuronal activity [26]. In contrast, HFS (80–185 Hz) produced a dual effect, completely suppressing previous STN activity and evoking a recurring pattern of spike bursts that were time-locked to stimulation. Because these effects were altered by Na+ (TTX) and Ca2+ (nifedipine) channel blockers, but unaffected by glutamate or GABA antagonists, the authors conclude that STN HFS acts directly on the neuronal membrane, rather than through neurotransmission. They suggest that stimulation directly activates target structures with the effect dependent on the particular characteristics of the stimulated membranes and synapses. Low frequency (5–50 Hz) single pulse microstimulation in the GPi and SNr produces local inhibition [3,41]. This inhibition is not usually seen in the STN, and in the thalamus a brief excitatory response can occur [3]. In the thalamus, activation of local glutamatergic afferents may account for the stimulation-induced excitation. It has been suggested that the balance of glutamatergic and GABA-ergic axons in the STN cancels out any overall inhibitory or excitatory response to stimulation. High frequency (200 Hz) single pulse microstimulation in the GPi and SNr also produces local inhibition [3,41]. When the duration of HFS is increased, the post-stimulation inhibition is reduced, possibly due to the desensitization of GABA receptors [3]. STN microelectrode HFS (100–300 Hz) trains in patients with PD produces, following termination of the train, early inhibition followed by rebound excitation, and then further inhibition of some neurons in STN [39]. This prolonged inhibitory effect is thought to be due to hyperpolarization, possibly due to GABA release from GPe terminals [3,39]. In the thalamus, HFS (100–333 Hz) can also inhibit neuronal firing, particularly in neurons that exhibit spontaneous low threshold spike (LTS) bursting activity, and is likely the result of hyperpolarization.

Highlighting the relevance of frequency for the effects of DBS, in patients with ET, thalamic stimulation above 90 Hz reduces tremor, whereas stimulation below 60 Hz can exacerbate it [87]. Higher frequency stimulation produces a more regular firing pattern which correlates with tremor suppression [87]. Using stimulation parameters that mimic those used in human DBS, Kiss et al. [88] found that the response to stimulation in rodent in vitro thalamic slice preparations began when applied at a frequency of 20 Hz and then increased to a maximum responsiveness at 200 Hz, similar to the response characteristics of human Vim DBS [76,89]. In patients with essential tremor, reductions in tremor have been shown to be maximal at 100 Hz, with increasing tremor reduction between 45 and 100 Hz, and no additional reduction at frequencies above 100 Hz [76,89]. Lee et al. [50] found that the frequency of stimulation significantly altered its effect. Initial excitation of local STN neurons was maximal at 100–140 Hz. At 200 Hz, this activity was entirely blocked, producing poststimulation inhibition alone. Furthermore, the longer the stimulation was applied the longer the poststimulation inhibition persisted. Low frequency stimulation of the STN has been shown to exacerbate b frequency oscillatory activity in the GPi [90]. Using computer models, Grill et al. [91] determined that DBS below 100 Hz is unable to block intrinsic oscillatory neuronal firing patterns, whereas stimulation above 100 Hz completely eliminates it, resulting in a new regular firing pattern.

Effects on Patterns of Neuronal Activity The rate model of basal ganglia function posits that parkinsonism results from reduced rates of neuronal firing in thalamic, cortical, and brainstem components of the motor circuit [6]. The shortcomings of this model are apparent when one considers the clinical features of PD.


The inhibition of motor output seems to fit bradykinesia and rigidity, but not tremor. Moreover, lesions of the thalamus do not result in bradykinesia, and lesions of the GPi do not cause dyskinesias [6]. Rather than rate alone, the underlying pathophysiology of movement disorders likely involves dysfunctional patterns of activity. One notable alteration in the pattern of neuronal activity seen in movement disorders is the presence of oscillatory activity in the STN, SNr, and GPi, particularly in the b frequency range (15–30 Hz) [6,28,92–94]. Additionally, basal ganglia and cortical neurons in patients with parkinsonism exhibit excessive synchronization [6,48]. Successful pharmacological and surgical PD treatments have been associated with a reduction of these anomalous neuronal firing patterns [6,92,95]. It may be the ability of stimulation to reduce burst-firing patterns and de-synchronize pathologic oscillatory activity that is most relevant in explaining its effect [25,96–98]. By interfering with this dysfunctional activity, DBS may allow downstream areas of the motor circuit to perform more normally [6,13,14]. Dopamine alters the oscillatory activity of the STN in PD patients, causing a reduction in b frequency oscillations and an increase in gamma range oscillations (75 Hz) [13,98]. Voluntary movement causes desynchronization of oscillatory activity at 20 Hz and synchronization at 75 Hz [98]. It has been proposed that HFS (STN DBS) may promote higher frequency oscillations similar to dopamine [13]. STN DBS and L-dopa decrease the latency to desynchronization of neuronal activity in the primary sensorimotor and premotor cortices [95]. In a recent rat study of STN DBS, the animals were injected with D1 (SCH-23,390) and D2 (raclopride) receptor blockers to mimic the dopamine-depleted state of PD, and this produced catalepsy [48]. Extracellular microelectrode recordings of SNr neurons were conducted before and after

83

the neuroleptic injections. The activity of SNr neurons changed from regular tonic firing to irregular bursts of firing. STN HFS (130 Hz, 40–100 ms, 2–5 V) was able to reliably abolish the rats’ catalepsy and the abnormal bursting activity in the SNr. Based on these results, the authors conclude that the beneficial effects of STN DBS in PD patients are due to the stimulation’s ability to regulate the pathologic bursting activity in basal ganglia output structures and to restore the balance between the trans-striatal and trans-subthalamic circuits. In an MPTP-treated primate model of parkinsonism, Meissner et al. [99] found that STN HFS (130 Hz; 100 mA; 60 ms) reduced oscillatory activity in the STN. In a different study also in MPTP-treated monkeys, STN HFS (136–185 Hz; 3.5 V) produced regular short latency excitatory responses in GPe and GPi neurons following each stimulation pulse, suggesting the activation of glutamatergic STN projections [60]. This stimulation-locked pattern of increased activity persisted even when stimulation was maintained for more than 5 min. It was also associated with decreased rigidity and increased spontaneous movement in the monkeys. These results further support the theory that DBS effects depend on a disruption or alteration of pre-existing pathological patterns of neuronal activity. DBS produces widespread alterations in cerebral blood flow and metabolism [13]. Computer modeling has demonstrated that STN HFS regularizes GPi firing and restores thalamocortical responsiveness [100]. The effects of STN HFS may be polysynaptic, and effect a variety of downstream targets [48,101]. GPi DBS in patients with dystonia has been associated with alterations in cerebral blood flow in the cerebellum, anterior cingulate cortex, lentiform nucleus, thalamus, pons, and midbrain [102]. The effect of GPi DBS on the neuronal activity in the ventral oralis anterior nucleus of the thalamus (Voa) was investigated in a patient with dystonia [42]. Two types of neuronal firing patterns were

1393

1394

83


observed in the Voa: a low firing rate/high bursting activity type, and a high firing rate/low bursting activity type. GPi DBS reduced the firing rate and increased the bursting activity in the second type of Voa neuron. Because stimulation only affected one type of Voa neuron, it is hard to attribute the effect to the general activation of GABA-ergic pallidal efferents. Thus, while the authors do not propose a specific mechanism for the effect, they do conclude that it reflects the ability of GPi DBS to alter the pattern of pathological neuronal activity. In PD patients, STN DBS is associated with increased cerebral blood flow (CBF; measured with H215O PET) in midbrain, globus pallidus, and thalamus, and decreased CBF bilaterally in frontal, parietal, and temporal cortex [103]. These findings suggest that DBS drives STN activity which increases nigro-pallidal inhibition of thalamocortical projections. Parkinsonian patients with either STN or GPi DBS underwent H215O PET imaging while completing a motor task [104]. With therapeutically effective STN stimulation, movement-related cerebral blood flow changes were seen in the supplementary motor area, cingulate cortex, and dorsolateral prefrontal cortex (DLPFC). During effective GPi stimulation, no significant cerebral blood flow changes occurred. GPi DBS in PD patients increased regional cerebral blood flow, as assessed by H215O PET, in ipsilateral premotor cortical areas [105]. These stimulation-induced changes coincided with improvements in the patients’ rigidity and bradykinesia. In another study examining the H215O PET changes associated with GPi DBS in PD patients, stimulation-induced increases in cerebral blood flow occurred in the left sensorimotor cortex, ventrolateral thalamus, and contralateral cerebellum [106]. In addition to stimulation, participants also performed a motor task during scanning. This may explain why the pattern of activation with stimulation in this study is

somewhat different than that found by Davis et al. [105]. The mechanisms underlying these cortical blood flow changes were not investigated. Nevertheless, these results support the view that DBS influences networks of neuronal activity extending well beyond the site of stimulation, and that these alterations correlate with clinical improvement in humans. Asanuma et al. [107] examined the changes in cerebral glucose metabolism associated with STN DBS and L-dopa therapy. Using FDG-PET, nine PD patients were scanned on and off STN stimulation, and nine others were scanned before and after an intravenous infusion of L-dopa. Following the administration of STN DBS and L-dopa, metabolic reductions were found in the putamen, globus pallidus, sensorimotor cortex, and cerebellar vermis, whereas increases were seen in the precuneus. Relative to L-dopa therapy, STN stimulation was associated with metabolic increases in the STN and decreases in the medial prefrontal cortex. The metabolic alterations associated with STN DBS and L-dopa infusion correlated with clinical improvement. Garcia et al. [26,45] propose that STN HFS has both an activating and inactivating effect. While it silences previous (pathological) neuronal activity, STN HFS also establishes a new beneficial discharge pattern in the gamma frequency (60–80 Hz) range. Such an effect provides a common link between HFS and L-dopa treatment, which is also known to obliterate b frequency (20 Hz) oscillatory activity and replace it with spontaneous synchronization above 70 Hz [45,90]. These results again suggest that, regardless of the specific mechanisms, the success of STN DBS and L-dopa therapy depends on similar alterations to the pathological cerebral activity underlying PD. Vim DBS facilitates TMS-induced motor evoked potentials by activating the primary motor cortex via thalamocortical projections [108]. Vim DBS also facilitates the TMS-induced activation of inhibitory cerebellothalamocortical


projections. Similarly, anterior nucleus of thalamus DBS has been found to drive inhibitory thalamocortical circuits [109]. The influence of HFS on neuronal circuits depends on the specific stimulation site. For example, in a PD patient who had undergone bilateral STN DBS, motor symptoms were improved with left DBS, but episodes of dysphoria were elicited with right DBS alone [110]. Structural imaging revealed that the left electrode was in the inferior STN, whereas the right electrode was marginally superolateral to the STN. Functional magnetic resonance imaging showed that left DBS produced blood oxygen level-dependent (BOLD) increased signal in the premotor and motor cortices, ventrolateral thalamus, putamen, and cerebellum, and decreased signal in the supplementary motor cortex. With right DBS, similar but less pronounced signal changes were seen in these motor areas. However, unique signal increases were seen in the superior prefrontal cortex, Brodmann area (BA) 24, anterior thalamus, caudate, and brainstem, and decreases in the medial prefrontal cortex. These results demonstrate the effect HFS can have on widespread neuronal circuits, and how that effect relates to the neuroanatomy of the stimulation target. They also emphasize the potential range of clinical effects that HFS can influence. Do DBS and surgical lesions produce the same pattern of neuronal activity? Fukuda et al. [111] have identified a Parkinson’s disease related pattern (PDRP) of brain activity using [18F] fluorodeoxyglucose (FDG) PET, characterized by pallidothalamic and pontine hypermetabolism as well as cortical motor hypometabolism. Following unilateral subthalamotomy in PD patients, the PDRP is altered relative to the unlesioned hemisphere. Reduced glucose metabolism is also seen in the ipsilateral midbrain, GPi, ventral thalamus, and pons [112]. These results suggest that STN lesions have widespread effects on motor circuitry. The PDRP is also altered

83

following pallidotomy and GPI DBS, and this alteration is correlated with clinical improvement as measured by UPDRS motor scores [113]. GPi DBS and pallidotomy also increase glucose metabolism in the ipsilateral premotor cortex and cerebellum bilaterally. If GPi DBS inactivates GABA-ergic pallidothalamic axons (i.e., mimics a lesion), one would expect a disinhibition of thalamic neurons, and increase in thalamic spike activity. Anderson et al. [37] examined this hypothesis in two Macaca fascicularis monkeys, and found that the activity of many thalamic neurons decreased during high frequency (120 Hz) bipolar stimulation of the GPi, although some showed increased activity. The authors suggest that the therapeutic effect of GPi DBS is by the activation of stimulated neuronal elements, which may interrupt the pathophysiological corticothalamic circuits producing movement disorders such as PD. Thus, while there may be some similarities in the pattern of neuronal activity associated with stimulation and lesions, the specific mechanisms underlying these effects are not the same. Does DBS promote neurogenesis or neuroprotection? Some researchers have begun to explore the potential neurogenesis-inducing and neuroprotective effects of HFS. For example, increased hippocampal neurogenesis, as measured by the presence of 5’-bromo-2’-deoxyuridine (BrdU) neurons (i.e., new cells) has been shown in rats who had anterior nucleus of thalamus (AN) HFS (130 Hz; 2.5 V; 90 ms) compared with rats who had undergone sham surgery [114]. Furthermore, the reduction of BrdU-positive cells that followed the systemic administration of the neurogenesis suppressor corticosterone was reversed with AN HFS. These results suggest that HFS may promote neurogenesis. Although similar experiments have not been performed with basal ganglia HFS, there is some evidence that STN DBS may promote the survival of midbrain dopamine cells in MPTPtreated monkeys [115]. The mechanism underlying

1395

1396

83


this potential neuroprotective effect may involve stimulation-induced reductions in glutamatemediated excitotoxicty.

Summary and Conclusions DBS affects neuronal function in many ways. The effects of DBS depend on the specific composition of neural elements (axons, neuronal cell bodies, and glia) in the stimulation target, and what set of afferent and efferent pathways are associated with that target. The ways in which DBS influences its target also depend on the amplitude, pulse width, and frequency of the applied stimulation. DBS affects membrane potentials and ion conductances, the synaptic release of neurotransmitters, and general patterns of neuronal activity over widespread networks. Although DBS may have direct inhibitory effects on neuronal cell bodies, it can also excite axons [8,116]. Imaging and electrotrophysiological studies provide evidence that DBS can at least in some situations excite the output pathways from the region stimulated. The ultimate effect of DBS may not be to simply excite or inhibit a specific nucleus, however. Rather, through its combination of excitatory and inhibitory cellular, monosynaptic, and polysynaptic effects, DBS alters the pathological neuronal activity that underlies neurological conditions, such as the synchronized oscillatory patterns in movement disorders, and thereby permits the system to function more normally [3,4,8,14,90,94,116–118].

References 1. Kringelbach ML, et al. Translational principles of deep brain stimulation. Nat Rev Neurosci 2007;8(8):623-35. 2. Benabid AL, Benazzouz A, Pollak P. Mechanisms of deep brain stimulation. Mov Disord 2002;17 Suppl 3:S73-4. 3. Dostrovsky JO, Lozano AM. Mechanisms of deep brain stimulation. Mov Disord 2002;17 Suppl 3:S63-8. 4. Vitek JL. Mechanisms of deep brain stimulation: excitation or inhibition. Mov Disord 2002;17 Suppl 3:S69-72.

5. McIntyre CC, et al. How does deep brain stimulation work? Present understanding and future questions. J Clin Neurophysiol 2004;21(1):40-50. 6. DeLong MR, Wichmann T. Circuits and circuit disorders of the basal ganglia. Arch Neurol 2007;64(1):20-4. 7. Parent A, Hazrati L-N. Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidium in basal ganglia circuitry. Brain Res Rev 1995;20(1):128-54. 8. Perlmutter JS, Mink JW. Deep brain stimulation. Annu Rev Neurosci 2006;29:229-57. 9. Parent A, Hazrati L-N. Functional anatomy of the basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop. Brain Res Rev 1995;20(1):91-127. 10. Ranck JB. Jr., Which elements are excited in electrical stimulation of mammalian central nervous system: A review. Brain Res 1975;98(3):417-40. 11. Follett KA, Mann MD. Effective stimulation distance for current from macroelectrodes. Exp Neurol 1986;92(1):75-91. 12. Temel Y, et al. Monopolar versus bipolar high frequency stimulation in the rat subthalamic nucleus: differences in histological damage. Neurosci Lett 2004;367(1):92-6. 13. Lozano AM, Mahant N. Deep brain stimulation surgery for Parkinson’s disease: mechanisms and consequences. Parkinsonism Relat Disord 2004;10 Suppl 1:S49-S57. 14. Lozano AM, et al. Deep brain stimulation for Parkinson’s disease: disrupting the disruption. Lancet Neurol 2002;1(4):225-31. 15. Weiss G. The law of electrical stimulation in the nerves. Comptes Rendus Des Seances De La Societe De Biologie Et De Ses Filiales 1901;53:466-8. 16. Nowak LG, Bullier J. Axons, but not cell bodies, are activated by electrical stimulation in cortical gray matter I. Evidence from chronaxie measurements. Exp Brain Res 1998;118(4):477-88. 17. Nowak LG, Bullier J. Axons, but not cell bodies, are activated by electrical stimulation in cortical gray matter II. Evidence from selective inactivation of cell bodies and axon initial segments. Exp Brain Res 1998;118(4):489-500. 18. Holsheimer J, et al. Identification of the target neuronal elements in electrical deep brain stimulation. Eur J Neurosci 2000;12(12):4573-7. 19. Rushton WAH. The effect upon the threshold for nervous excitation of the length of nerve exposed, and the angle between current and nerve. J Physiol (Lond) 1927;63(4):357-77. 20. Dostrovsky JO, et al. Microinjection of lidocaine into human thalamus: a useful tool in stereotactic surgery. Stereotact Funct Neurosurg 1993;60(4):168-74. 21. Levy R, et al. Lidocaine and muscimol microinjections in subthalamic nucleus reverse Parkinsonian symptoms. Brain 2001;124(Pt 10):2105-18. 22. Breit S, Schulz JB. Benabid AL. Deep brain stimulation. Cell Tissue Res 2004;318(1):275-88.


23. Benazzouz A, et al. Responses of substantia nigra pars reticulata and globus pallidus complex to high frequency stimulation of the subthalamic nucleus in rats: electrophysiological data. Neurosci Lett 1995;189(2): 77-80. 24. McIntyre CC, et al. Cellular effects of deep brain stimulation: model-based analysis of activation and inhibition. J Neurophysiol 2004;91(4):1457-69. 25. Chang J-Y, et al. Studies of the neural mechanisms of deep brain stimulation in rodent models of Parkinson’s disease. Neurosci Biobehav Rev 2007;31(5):643-57. 26. Garcia L, et al. Dual effect of high-frequency stimulation on subthalamic neuron activity. J. Neurosci 2003;23(25):8743-51. 27. Beurrier C, et al. High-frequency stimulation produces a transient blockade of voltage-gated currents in subthalamic neurons. J Neurophysiol 2001;85(4): 1351-6. 28. Magarinos-Ascone C, et al. High-frequency stimulation of the subthalamic nucleus silences subthalamic neurons: a possible cellular mechanism in Parkinson’s disease. Neuroscience 2002;115(4):1109-17. 29. Do MTH, Bean BP. Subthreshold sodium currents and pacemaking of subthalamic neurons: modulation by slow inactivation. Neuron 2003;39(1):109-20. 30. Bikson M, et al. Suppression of epileptiform activity by high frequency sinusoidal fields in rat hippocampal slices. J Physiol 2001;531(Pt 1):181-91. 31. Shen K-Z, et al. Synaptic plasticity in rat subthalamic nucleus induced by high-frequency stimulation. Synapse 2003;50(4):314-19. 32. Anderson T, et al. Mechanisms of deep brain stimulation: an intracellular study in rat thalamus. J Physiol 2004;559(1):301-13. 33. Anderson TR, et al. Selective attenuation of afferent synaptic transmission as a mechanism of thalamic deep brain stimulation-induced tremor arrest. J Neurosci 2006;26(3):841-50. 34. Lee KH, et al. Dopamine efflux in the rat striatum evoked by electrical stimulation of the subthalamic nucleus: potential mechanism of action in Parkinson’s disease. Eur J Neurosci 2006;23(4):1005-14. 35. Boraud T, et al. High frequency stimulation of the internal Globus Pallidus (GPi) simultaneously improves parkinsonian symptoms and reduces the firing frequency of GPi neurons in the MPTP-treated monkey. Neurosci Lett 1996;215(1):17-20. 36. Benazzouz A, et al. Effect of high-frequency stimulation of the subthalamic nucleus on the neuronal activities of the substantia nigra pars reticulata and ventrolateral nucleus of the thalamus in the rat. Neuroscience 2000;99(2):289-95. 37. Anderson ME, Postupna N, Ruffo M. Effects of highfrequency stimulation in the internal globus pallidus on the activity of thalamic neurons in the awake monkey. J Neurophysiol 2003;89(2):1150-60.

83

38. Salin P, et al. High-frequency stimulation of the subthalamic nucleus selectively reverses dopamine denervationinduced cellular defects in the output structures of the basal ganglia in the rat. J Neurosci 2002;22(12):5137-48. 39. Filali M, et al. Stimulation-induced inhibition of neuronal firing in human subthalamic nucleus. Exp Brain Res 2004;156(3):274-81. 40. Welter M-L, et al. Effects of high-frequency stimulation on subthalamic neuronal activity in Parkinsonian patients. Arch Neurol 2004;61(1):89-96. 41. Dostrovsky JO, et al. Microstimulation-induced inhibition of neuronal firing in human globus pallidus. J Neurophysiol 2000;84(1):570-4. 42. Pralong E, et al. Effect of deep brain stimulation of GPI on neuronal activity of the thalamic nucleus ventralis oralis in a dystonic patient. Neurophysiol Clin 2003;33(4):169-73. 43. Benazzouz A, et al. High frequency stimulation of the STN influences the activity of dopamine neurons in the rat. Neuroreport 2000;11(7):1593-6. 44. Maltete D, et al. Subthalamic stimulation and neuronal activity in the substantia nigra in Parkinson’s disease. J Neurophysiol 2007;97(6):4017-22. 45. Garcia L, et al. High-frequency stimulation in Parkinson’s disease: more or less? Trends Neurosci 2005;28(4):209-16. 46. Ashby P, et al. Neurophysiological effects of stimulation through electrodes in the human subthalamic nucleus. Brain 1999;122(10):1919-31. 47. Ashby P, et al. Potentials recorded at the scalp by stimulation near the human subthalamic nucleus. Clin Neurophysiol 2001;112(3):431-7. 48. Degos B, et al. Neuroleptic-induced catalepsy: electrophysiological mechanisms of functional recovery induced by high-frequency stimulation of the subthalamic nucleus J Neurosci 2005;25(33):7687-96. 49. Maurice N, et al. Spontaneous and evoked activity of Substantia Nigra Pars Reticulata Neurons during high-frequency stimulation of the subthalamic nucleus. J. Neurosci 2003;23(30):9929-36. 50. Lee KH, Roberts DW, Kim U. Effect of high-frequency stimulation of the subthalamic nucleus on subthalamic neurons: an intracellular study. Stereotact Funct Neurosurg 2003;80(1–4):32-6. 51. Miocinovic S, et al. Computational analysis of subthalamic nucleus and lenticular fasciculus activation during therapeutic deep brain stimulation. J Neurophysiol 2006;96(3):1569-80. 52. Wu YR, et al. Does stimulation of the GPi control dyskinesia by activating inhibitory axons? Mov Disord 2001;16(2):208-16. 53. Li T, Qadri F, Moser A. Neuronal electrical high frequency stimulation modulates presynaptic GABAergic physiology. Neurosci Lett 2004;371(2–3):117-21. 54. Hiller A, et al. Electrical high frequency stimulation of the caudate nucleus induces local GABA outflow in freely moving rats. J Neurosci Meth 2007;159(2):286-90.

1397

1398

83


55. Windels F, et al. Pallidal Origin of GABA release within the substantia nigra pars reticulata during high-frequency stimulation of the subthalamic nucleus. J Neurosci 2005;25(20):5079-86. 56. Boulet S, et al. Subthalamic stimulation-induced forelimb dyskinesias are linked to an increase in glutamate levels in the substantia nigra pars Reticulata. J Neurosci 2006;26(42):10768-76. 57. Windels F, et al. Effects of high frequency stimulation of subthalamic nucleus on extracellular glutamate and GABA in substantia nigra and globus pallidus in the normal rat. Eur J Neurosci 2000;12(11):4141-6. 58. MacKinnon CD, et al. Stimulation through electrodes implanted near the subthalamic nucleus activates projections to motor areas of cerebral cortex in patients with Parkinson’s disease. Eur J Neurosci 2005;21(5):1394-402. 59. Lee KH, et al. Abolition of spindle oscillations and 3-Hz absence seizurelike activity in the thalamus by using highfrequency stimulation: potential mechanism of action. J Neurosurg 2005;103(3):538-45. 60. Hashimoto T, et al. Stimulation of the subthalamic nucleus changes the firing pattern of pallidal neurons J Neurosci 2003;23(5):1916-23. 61. Montgomery J, Erwin B. Effects of GPi stimulation on human thalamic neuronal activity. Clin Neurophysiol 2006;117(12):2691-702. 62. Lee KH, et al. High-frequency stimulation of the subthalamic nucleus increases glutamate in the subthalamic nucleus of rats as demonstrated by in vivo enzyme-linked glutamate sensor. Brain Res 2007;1162:121-9. 63. Bekar L, et al. Adenosine is crucial for deep brain stimulation-mediated attenuation of tremor. Nat Med 2008;14(1):75-80. 64. Grill WM, Cantrell MB, Robertson MS. Antidromic propagation of action potentials in branched axons: implications for the mechanisms of action of deep brain stimulation. J Comput Neurosci 2008;24:81-93. 65. Hanajima R, et al. Single pulse stimulation of the human subthalamic nucleus facilitates the motor cortex at short intervals. J Neurophysiol 2004;92(3):1937-43. 66. Paul G, et al. High frequency stimulation of the subthalamic nucleus influences striatal dopaminergic metabolism in the naive rat. Neuroreport 2000;11(3):441-4. 67. Meissner W, et al. High-frequency stimulation of the subthalamic nucleus enhances striatal dopamine release and metabolism in rats. J Neurochem 2003;85 (3):601-9. 68. Bruet N, et al. Neurochemical mechanisms induced by high frequency stimulation of the subthalamic nucleus: increase of extracellular striatal glutamate and GABA in normal and hemiparkinsonian rats. J Neuropathol Exp Neurol 2003;62(12):1228-40. 69. Oueslati A, et al. High-frequency stimulation of the subthalamic nucleus potentiates L-DOPA-induced neurochemical changes in thestriatum in a rat model of Parkinson’s disease J Neurosci 2007;27(9):2377-86.

70. Sailer A, et al. Subthalamic nucleus stimulation modulates afferent inhibition in Parkinson’s disease. Neurology 2007;68(5):356-63. 71. Abosch A, et al. Stimulation of the subthalamic nucleus in Parkinson’s disease does not produce striatal dopamine release. Neurosurgery 2003;53(5):1095-102; discussion 1102-5. 72. Hilker R, et al. Deep brain stimulation of the subthalamic nucleus does not increase the striatal dopamine concentration in parkinsonian humans. Mov Disord 2003;18(1):41-8. 73. Iremonger KJ, et al. Cellular mechanisms preventing sustained activation of cortex during subcortical highfrequency stimulation. J Neurophysiol 2006;96(2): 613-21. 74. Wang L-Y, Kaczmarek LK. High-frequency firing helps replenish the readily releasable pool of synaptic vesicles. Nature 1998;394(6691):384-8. 75. Rizzone M, et al. Deep brain stimulation of the subthalamic nucleus in Parkinson’s disease: effects of variation in stimulation parameters. J Neurol Neurosurg Psychiatry 2001;71(2):215-19. 76. Ushe M, et al. Effect of stimulation frequency on tremor suppression in essential tremor. Mov Disord 2004;19(10):1163-8. 77. Vidailhet M, et al. Bilateral deep-brain stimulation of the globus pallidus in primary generalized dystonia. N Engl J Med 2005;352(5):459-67. 78. Rodriguez-Oroz MC, et al. Bilateral deep brain stimulation in Parkinson’s disease: a multicentre study with 4 years follow-up. Brain 2005;128(Pt 10):2240-9. 79. Deuschl G, et al. A randomized trial of deep-brain stimulation for Parkinson’s disease. N Engl J Med 2006;355(9):896-908. 80. Plaha P, et al. Stimulation of the caudal zona incerta is superior to stimulation of the subthalamic nucleus in improving contralateral parkinsonism. Brain 2006; 129(Pt 7):1732-47. 81. Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine nucleus for Parkinson’s disease. Neuroreport 2005;16(17):1883-7. 82. Stefani A, et al. Bilateral deep brain stimulation of the pedunculopontine and subthalamic nuclei in severe Parkinson’s disease. Brain 2007;130(Pt 6):1596-607. 83. Mayberg HS, et al. Deep brain stimulation for treatmentresistant depression. Neuron 2005;45(5):651-60. 84. Putzke JD, et al. Thalamic deep brain stimulation for tremor-predominant Parkinson’s disease. Parkinsonism Relat Disord 2003;10(2):81-8. 85. Hariz MI, et al. Multicentre European study of thalamic stimulation for parkinsonian tremor: a 6 year follow-up. J Neurol Neurosurg Psychiatry 2008;79(6): 694-9. 86. Windels F, et al. Influence of the frequency parameter on extracellular glutamate and gamma-aminobutyric acid in substantia nigra and globus pallidus during electrical


stimulation of subthalamic nucleus in rats. J Neurosci Res 2003;72(2):259-67. 87. Kuncel AM, et al. Amplitude- and frequency-dependent changes in neuronal regularity parallel changes in tremor with thalamic deep brain stimulation. IEEE Trans Neural Syst Rehabil Eng 2007;15(2):190-7. 88. Kiss ZHT, et al. Neuronal response to local electrical stimulation in rat thalamus: physiological implications for mechanisms of deep brain stimulation. Neuroscience 2002;113(1):137-43. 89. Ushe M, et al. Postural tremor suppression is dependent on thalamic stimulation frequency. Mov Disord 2006;21(8):1290-2. 90. Brown P, et al. Effects of stimulation of the subthalamic area on oscillatory pallidal activity in Parkinson’s disease. Exp Neurol 2004;188(2):480-90. 91. Grill WM, Snyder AN, Miocinovic S. Deep brain stimulation creates an informational lesion of the stimulated nucleus. Neuroreport 2004;15(7):1137-40. 92. Mallet N, et al. Disrupted dopamine transmission and the emergence of exaggerated beta oscillations in subthalamic nucleus and cerebral cortex. J Neurosci 2008;28(18):4795-806. 93. Hutchison WD, et al. Neuronal oscillations in the basal ganglia and movement disorders: Evidence from whole animal and human recordings. J Neurosci 2004;24 (42):9240-3. 94. Brown P. Oscillatory nature of human basal ganglia activity: relationship to the pathophysiology of Parkinson’s disease. Mov Disord 2003;18(4):357-63. 95. Devos D, et al. Subthalamic nucleus stimulation modulates motor cortex oscillatory activity in Parkinson’s disease. Brain 2004;127(2):408-19. 96. Plenz D, Kital ST. A basal ganglia pacemaker formed by the subthalamic nucleus and external globus pallidus. Nature 1999;400(6745):677-82. 97. Levy R, et al. Synchronized neuronal discharge in the basal ganglia of Parkinsonian patients is limited to oscillatory activity J Neurosci 2002;22(7):2855-61. 98. Levy R, et al. Dependence of subthalamic nucleus oscillations on movement and dopamine in Parkinson’s disease. Brain 2002;125(6):1196-209. 99. Meissner W, et al. Subthalamic high frequency stimulation resets subthalamic firing and reduces abnormal oscillations. Brain 2005;128(10):2372-82. 100. Rubin JE, Terman D. High frequency stimulation of the subthalamic nucleus eliminates pathological thalamic rhythmicity in a computational model. J Comput Neurosci 2004;16(3):211-35. 101. Kita H, et al. Balance of monosynaptic excitatory and disynaptic inhibitory responses of the globus pallidus induced after stimulation of the subthalamic nucleus in the monkey. J Neurosci 2005;25(38):8611-19.

83

102. Yianni J, et al. Effect of GPi DBS on functional imaging of the brain in dystonia. J Clin Neurosci 2005;12(2):137-41. 103. Hershey T, et al. Cortical and subcortical blood flow effects of subthalamic nucleus stimulation in PD. Neurology 2003;61(6):816-21. 104. Limousin P, et al. Changes in cerebral activity pattern due to subthalamic nucleus or internal pallidum stimulation in Parkinson’s disease. Ann Neurol 1997;42(3):283-91. 105. Davis KD, et al. Globus pallidus stimulation activates the cortical motor system during alleviation of parkinsonian symptoms. 1997;3(6):671-4. 106. Fukuda M, et al. Functional correlates of pallidal stimulation for Parkinson’s disease. Ann Neurol 2001;49(2):155-64. 107. Asanuma K, et al. Network modulation in the treatment of Parkinson’s disease. Brain 2006;129(10):2667-78. 108. Molnar GF, et al. Changes in cortical excitability with thalamic deep brain stimulation. Neurology 2005;64(11):1913-9. 109. Molnar GF, et al. Changes in motor cortex excitability with stimulation of anterior thalamus in epilepsy. Neurology 2006;66(4):566-71. 110. Stefurak T, et al. Deep brain stimulation for Parkinson’s disease dissociates mood and motor circuits: a functional MRI case study. Mov Disord 2003;18(12): 1508-16. 111. Fukuda M, et al. Networks mediating the clinical effects of pallidal brain stimulation for Parkinson’s disease: a PET study of resting-state glucose metabolism. Brain 2001;124(8):1601-9. 112. Su PC, et al. Metabolic changes following subthalamotomy for advanced Parkinson’s disease. Ann Neurol 2001;50(4):514-20. 113. Eidelberg D, et al. Regional metabolic correlates of surgical outcome following unilateral pallidotomy for Parkinson’s disease. Ann Neurol 1996;39(4):450-9. 114. Toda H, et al. The regulation of adult rodent hippocampal neurogenesis by deep brain stimulation. J Neurosurg 2008;108(1):132-8. 115. Wallace BA, et al. Survival of midbrain dopaminergic cells after lesion or deep brain stimulation of the subthalamic nucleus in MPTP-treated monkeys. Brain 2007;130(8):2129-45. 116. McIntyre CC, et al. Uncovering the mechanism(s) of action of deep brain stimulation: activation, inhibition, or both. Clin Neurophysiol 2004;115(6):1239-48. 117. Lang AE, Lozano AM. Parkinson’s disease. Second of two parts. N Engl J Med 1998;339(16):1130-43. 118. Liu Y, et al. High frequency deep brain stimulation: what are the therapeutic mechanisms? Neurosci Biobehav Rev 2008;32(3):343-51.

1399

84 Stimulation Technology in Functional Neurosurgery B. H. Kopell . A. Machado . C. Butson

Most of the physician quotes regarding neurostimulation leads of the era [1970’s and 1980’s] are unprintable. (Medtronic Archival Document 1988) The evolution of technology in functional neurosurgery is an interesting mix of rapid development and frustrating stagnation, a constant stream of ideas inspired by serendipity and knowledge from eras long past. Functional neurosurgery has always been intimately associated with cutting-edge technology. For example, the vital technique of stereotactic surgery whose principles today underlie the imagingguided surgical advances of general cranial and spinal surgery was itself first applied as a functional procedure in 1947 [1]. Perhaps the most important arrow in today’s functional neurosurgical quiver is the neurostimulation implant designed to manipulate the electrical signals in a targeted neuronal network in order to alleviate neurological symptoms. Electricity was known to have a potentially therapeutic effect as far back as the Roman Empire. Early medical documents record the experience of a physician by the name of Scribonius Largus who noticed the reduction of pain in his patients suffering from gout after they accidently stepped on an electrical-field generating torpedo fish (> Figure 84‐1) [2]. Despite this initial insight, electricity would not be used therapeutically in the nervous system again for many centuries. Instead, efforts focused on making controlled lesions in the nervous system in order to bring about permanent neuromodulatory effects. Initially, such lesions relied on physical force as Russ Meyers pioneering open-craniotomy approach to disrupting neural #


pathways in the 1930s ultimately led to the development of the leukotome in the 1950s [3]. Other means of lesion-generation evolved over the next three decades and included injections of alcohol/wax/oil [4,5], thermal and cryo-energy [6,7], and ultimately radiofrequency energy [8]. The overarching advancement that each technique had was a more precise and better controlled lesion that resulted in decreased patient morbidity. Electricity, however, remained a vital component of these lesional procedures. In the very first human stereotactic procedure, Siegel and Wycis describe the method of stimulating through the lesion-generating needle to confirm that they were not within the internal capsule [1,2]. While most use of electricity during these procedures focused mainly on its usefulness to decrease patient morbidity, eventually clinical effects of electricity became more and more evident. In 1960 Hassler first reported the clinical effects of electrical current on tremor during Vim lesion procedures [2,9]. Soon after, Ron Tasker and others were the first to document carefully the parameters used and the dichotomy between ‘‘low’’ and ‘‘high’’ frequency stimulation [10]. There was a general feeling that low frequency stimulation might drive or increase involuntary movements, especially tremor, and high frequency stimulation might mimic the therapeutic lesional effect, but such observations were inconsistent [2]. Based on these observations of the ability of electrical stimulation to mimic the effects of a lesion, investigators began the era of the chronically implanted electrode. Because there was no

1402

84

Stimulation technology in functional neurosurgery

. Figure 84‐1 An electrical torpedo fish or electric ray. The name comes from the Latin ‘‘torpere,’’ to be stiffened or paralyzed, referring to the effect on someone who handles or steps on a living electric ray

way to provide power to these systems chronically, these electrodes were intermittently powered by direct percutaneous connections to an external power source and stimulus generator. J. L. Pool was the first to employ this technique by implanting silver electrodes in the head of the caudate of a patient suffering from chronic depression [11]. Others soon followed including septal stimulation for chronic pain by Heath [12] and thalamic/pallidal stimulation for movement disorders by Bechtereva [13]. The origin of modern day neurostimulation is generally traced to the publication by Ron Melzack, Ph.D. and Pat Wall, Ph.D., in Science magazine in 1965 (> Figure 84‐2). Their publication ‘‘Pain mechanisms: a new theory,’’ is more commonly known as the Gate Control Theory of Pain [14]. This theory led C. Norman Shealy, M.D., working with his colleagues at Case Institute of Technology, to implant the first Dorsal Column Stimulator (DCS) in 1967. Medtronic became involved at that time in working with Dr. Shealy to develop the first commercially

available unit for clinical investigation. The Medtronic unit was known as the Myelostat Dorsal Column Stimulator (DCS) (> Figure 84‐3) [15]. Sweet and Wespic at the same year developed an implantable system for the treatment of peripheral neuropathic pain that was ultimately commercialized by Roger Avery [16]. Avery Laboratories entered the neurostimulation area several years later, with their design of a DCS system [15]. Spinal cord stimulation (SCS) eventually was utilized beyond neuropathic extremity pain for peripheral vascular disease, angina, spasticity, tremor, and dystonia [17–20]. Shealy, in attempting to better evaluate his patients, began applying external surface stimulation by using the ElecTreat stimulator. He began using this device in the early 1970s as a screening tool for selecting patients for implant. The device had been used extensively in pre-FDA regulation days and was sold with claims that were far in excess of what the device itself could do. Medtronic designed the first external stimulator for use as a screening tool in about 1971.


. Figure 84‐2 Ron Melzack, Ph.D. and Pat Wall, Ph.D., in Science magazine in 1965. (Picture courtesy of Medtronic and used with permission)

. Figure 84‐3 The Myelostat DCS. The first commercial spinal cord stimulation (SCS) system manufactured by Medtronic (Picture courtesy of Medtronic and used with permission)

84

. Figure 84‐4 The first commercially available deep brain stimulation (DBS) electrode manufactured by Medtronic (Picture courtesy of Medtronic and used with permission)

This device was known as a Cutaneous Stimulation Device (CSD). It was designed strictly for use by the physician or the patient in a hospital setting to screen for a later implant. Shortly thereafter, it was recognized that such devices had a therapeutic value in their own right, and could be used by the patient in a home setting. Several companies began producing such patient devices in the early 1970s. Medtronic introduced its first TENS device in 1973. This was known as the Neuromod Transcutaneous Nerve Stimulator (TNS). The terminology TNS was soon modified to Transcutaneous Electrical Nerve Stimulator (TENS) [15]. Paralleling the development of spinal cord and peripheral nerve stimulation was the development of brain stimulation. Brain stimulation began in 1969 at the University of California, San Francisco Medical Center. In that year, Dr. Yoshio Hosobuchi implanted a depth electrode in the sensory thalamus for a pain patient (> Figure 84‐4) [21]. During this time, the discovery of the endogenous opiate, endorphin, led to investigation of brain stimulation in the periventricular gray substance. Richardson and Akil reported the first use of this technique in humans [22].

1403

1404

84


In 1976, the Medtronic Deep Brain Stimulation (DBS) system was commercially released (> Figure 84‐5) [15]. Soon DBS-type systems were being used in clinical situations for epilepsy, spasticity, and psychiatric disease [23–25]. By the late 1970s there were three companies in the US that manufactured brain stimulation technologies for chronic pain: Avery, Neuromed, and Medtronic. At this time the FDA held a joint commission including the AANS and CNS to examine the technologies [26]. Based on this meeting, it was determined that DBS systems for chronic pain (and other conditions) needed to be proven with regard to safety and efficacy and that studies to date fell short of this bar. Each of the manufacturers were given time to perform studies and provide the documentation for safety and efficacy, but two of the companies felt that such a complex study would not be cost-effective when compared to potential sales, so did not submit such a report. Only Avery documented that DBS met the requirements to be approved for pain management, but it was very shortly after, in 1983, that that Roger Avery retired

and sold the company to Bill Dobelle, who concentrated on visual cortex stimulation for blindness and phrenic nerve stimulation for use in paralyzed patients, and no longer supported DBS. Consequently, DBS for pain management was deapproved and has not yet recovered [2]. This seeming death-knell for brain stimulation implants, ironically, heralded the rebirth of DBS technology, this time in the form of brain stimulation for movement disorders. In 1980 Brice and McLellan reported the first use of DBS for movement disorders with their exploration of thalamic stimulation for MS-type tremor [27]. Soon after this, Alim Benabid’s pioneering work in exploring Vim DBS for tremor and STN DBS for PD led ultimately to an FDA approval of this technology and sparked a renaissance in the neuromodulation field and market [28–30]. There have been several seminal advances in neurostimulation technology in the past three decades. The early stimulators came in two parts. These systems utilize an external transmitter which contains the pulse generating circuitry and power source (battery). This energy is

. Figure 84‐5 The first commercially available deep brain stimulation (DBS) system manufactured by Medtronic (Picture courtesy of Medtronic and used with permission)


transmitted through the skin using a radio frequency coupled link. Inside the body is a passive (no battery) receiver which decodes the radio frequency signal and delivers it to the electrode and, therefore, the nervous tissue. Cordis market released the first totally implantable neurostimulator (IPG, or implantable pulse generator) in 1982. Medtronic introduced a fully programmable neurological pulse generator later that year [15]. Other landmark improvements have included increased number of electrode contacts for flexibility in targeting and programming these systems, IPGs capable of controlling more than one electrode array, and transcutaneously rechargeable power sources, extending the time between IPG replacement surgeries.

The Neurostimulation System Today FDA approved neurostimulation systems are the follows: Spinal Cord Stimulation for neuropathic extremity pain and failed back syndrome (1984 510(k)); Peripheral Nerve Stimulation for neuropathic extremity pain (1984 510(k)); Vagus Nerve Stimulation for refractory epilepsy (1997); Vim thalamic DBS for Parkinsonian and Essential tremor (1997); STN/ GPi DBS for Parkinson’s disease (2002); GPi DBS for primary generalized dystonia (limited FDA approval under a Humanitarian Device Exemption); Vagus Nerve Stimulation for treatment resistant Major Depression (2005). In this section, we discuss the currently available types of electrodes and implantable pulse generators. This section reflects the available technology at the first quarter of 2008 and will unquestionably become outdated soon. Nevertheless, some principles are expected to influence the future generations of devices and the reader will be able to apply the inform

Textbook of Stereotactic and Functional Neurosurgery (2nd Edition, Vol.s 1 & 2)

Handbook of Stereotactic and Functional Neurosurgery

Handbook of Stereotactic and Functional Neurosurgery (Neurological Disease and Therapy)

Handbook of Stereotactic and Functional Neurosurgery (Neurological Disease and Therapy)

Stereotactic and Functional Neurosurgery: Movement Disorders and Pain: Medical Management and Functional Neurosurgery

Fretboard Logic Vols 1 2 & 3

Textbook of Endodontology 2nd Edition

Textbook of Psoriasis, 2nd Edition

The LispKit manual, vols.1 and 2

Litauisches Etymologisches Woerterbuch, Bande 1 & 2 - Lithuanian Etymological Dictionary, Vols. 1 & 2

A Course in Applied Mathematics, Vols 1 & 2

Textbook of Clinical Trials, 2nd Edition

Textbook of Palliative Nursing 2nd Edition

Textbook of Natural Medicine (2-Volume Set) 2nd edition

Textbook of Erectile Dysfunction, 2nd Edition

Textbook of Clinical Neuropsychiatry, 2nd edition

Textbook of Anxiety Disorders, 2nd Edition

Functional Fractional Calculus, 2nd Edition

A Textbook of General Practice, 2nd edition

Advances in Functional and Reparative Neurosurgery

Biological Psychiatry [2 Vols]

Field's Virology [2 vols]

Hepatology a clinical textbook 2010 - 2nd edition

Neurosurgery

Primary Care Ophthalmology: Textbook 2nd Edition

History of Indian Philosophy, 2 Vols

sed & awk (2nd Edition)

Neurosurgery

Neurosurgery

Sed & awk (2nd Edition)

Sed & awk (2nd Edition)

Textbook of Stereotactic and Functional Neurosurgery (2nd Edition, Vol.s 1 & 2)

Handbook of Stereotactic and Functional Neurosurgery

Handbook of Stereotactic and Functional Neurosurgery (Neurological Disease and Therapy)

Handbook of Stereotactic and Functional Neurosurgery (Neurological Disease and Therapy)

Stereotactic and Functional Neurosurgery: Movement Disorders and Pain: Medical Management and Functional Neurosurgery

Fretboard Logic Vols 1 2 & 3

Textbook of Endodontology 2nd Edition

Textbook of Psoriasis, 2nd Edition

The LispKit manual, vols.1 and 2

Litauisches Etymologisches Woerterbuch, Bande 1 & 2 - Lithuanian Etymological Dictionary, Vols. 1 & 2

A Course in Applied Mathematics, Vols 1 & 2

Textbook of Clinical Trials, 2nd Edition

Textbook of Palliative Nursing 2nd Edition

Textbook of Natural Medicine (2-Volume Set) 2nd edition

Textbook of Erectile Dysfunction, 2nd Edition

Textbook of Clinical Neuropsychiatry, 2nd edition

Textbook of Anxiety Disorders, 2nd Edition

Functional Fractional Calculus, 2nd Edition

A Textbook of General Practice, 2nd edition

Advances in Functional and Reparative Neurosurgery

Biological Psychiatry [2 Vols]

Field's Virology [2 vols]

Hepatology a clinical textbook 2010 - 2nd edition

Neurosurgery

Primary Care Ophthalmology: Textbook 2nd Edition

History of Indian Philosophy, 2 Vols

sed & awk (2nd Edition)

Neurosurgery

Neurosurgery

Sed & awk (2nd Edition)

Sed & awk (2nd Edition)

Recommend Documents